27th PSI

17 May 2026, 10:00 → 22 May 2026, 17:00 Europe/Berlin

marinaforum REGENSBURG

Rudolf Neu (MPI for Plasma Physics)

27th International Conference on Plasma-Surface Interactions in Controlled Fusion Devices (PSI)

Hotels in Regensburg

PSI27 Flyer.pdf

Sunday 17 May

1 T1 S. Brezinsek: An Introduction to Plasma Wall Interactions

Speaker: Sebastijan Brezinsek (Forschungszentrum Jülich GmbH, Institute of Fusion Energy and Nuclear Waste Management – Plasma Physics, 52425 Jülich, Germany. Faculty of Mathematics and Natural Sciences, Heinrich Heine University Düsseldorf, 40225 Düsseldorf, Germany.)

2 T2 P. Manz: Introduction to Scrape Off Layer Turbulent Transport

We begin with an introduction to instabilities in the near and far SOL and in the divertor region. Special attention is paid to plasma blobs or filaments. We then discuss saturation mechanisms. The special role of turbulence spreading is highlighted. Possibilities for reduced modelling are outlined.

Speaker: Peter Manz (University Greifswald)

15:00 Coffee Break

3 T3 M. Rubel: Dust in Fusion Devices: 40 years of Research

This contribution is based on over 40 years of dust research in several devices with consecutive modifications of plasma-facing materials (PFM):
• JET from full metal wall in 1983-1984, via JET-C with different fuelling schemes (including full D-T in 1997) and three consecutive divertor models, to JET-ILW with Be and W wall components;
• TEXTOR from full metal wall (1983-1984) to the boronised (also siliconized) carbon wall and bulk W test limiters;
• EXTRAP-T2 reversed field pinch with graphite PFM.
Holistic approach to dust studies in fusion devices comprises a broad range of plasma diagnostics (spectroscopy, imaging tools, magnetic sensors), tracer techniques for material migration, development and selection of collection and retrieval methods for secure material transfer to analytical laboratories where over 35 techniques for ex-situ material characterisation were used. A crucial element of the programme was on fuel retention analysis from total amounts of H, D, T to depth profiling, elemental micro-mapping, determination of specific tritium activity and fuel (D, T) quantification in individual dust grains, e.g. distinction of retention between C and metals in mixed samples of dust.
Such integrated efforts have allowed for the determination/ identification of several generation mechanisms including not only disintegration of co-deposits, arcing, brittle destruction and metal splashing but also plasma-chemical effects and metal flaking. In addition, dust-related safety aspects have been experimentally addressed, e.g. particle mobility during in-vessel operations with remote handling, the impact of hot water and/or water vapour on dust generation from affected wall materials, and dust accumulation in filters of the vessel ventilation system
Particular emphasis is given to: (i) boron species, its chemical form, role of boronisation as a marker in the growth of co-deposits; (ii) metal-based or metal-containing dust W, steel, Be in the form of splashes and droplets; (iii) metal oxides, WO$_2$, BeO, whose in-situ formation has been proven. The latter process may lead to quite severe material erosion and transport. The investigation of W presence in dust was often concentrated on searching for loose droplets. No solid evidence for such objects has been obtained either in JET-ILW or in TEXTOR. Very detailed microscopy data clearly indicate that splitting and then flaking of thin metal layers (<0.5 micrometre) occurs in overheated surfaces of bulk W components, e.g. Langmuir probes. This pathway of material loss should not be neglected in the estimates of global erosion.

Speaker: Marek Rubel (Uppsala University, Sweden)

4 T4 T. Looby: Tutorial on plasma power exhaust physics modeling with HEAT

Tokamaks such as SPARC or ITER will operate at high fusion performance to achieve energy gains on the order of Q~10. These high performance burning plasmas will generate extreme plasma power exhaust due to the narrowing of the heat flux channel as plasma current increases. In SPARC, for example, the steady state parallel heat fluxes are projected to reach 10 GW/m$^2$. Operating a tokamak safely under such loads is a key challenge for burning plasma machines and fusion power plants. This tutorial introduces the physics of tokamak power exhaust and a suite of codes used to predict the thermal loads onto PFCs.

The HEAT code bridges the gap between plasma power exhaust physics and engineering CAD geometry. HEAT integrates a variety of plasma physics modules including the optical approximation, gyro orbit tracing, ELM filament tracing, runaway electron transport, photon transport, and 3D field heat loads using M3DC1. These physics solvers are coupled to FEM/FVM modules for calculating the PFC temperature, thermal stress, and tungsten recrystallization kinetics.

This tutorial session will guide attendees through descriptions of the physics behind each of the HEAT modules, and practical examples of validation and verification will be provided from across the tokamak fleet.
An analysis of gyro orbit effects will be provided for NSTX-U. Differentiating between SOL transport channels will be demonstrated with results from ASDEX Upgrade. The effects of 3D fields on the heat flux footprint will be discussed for DIII-D. And leveraging the knowledge of power exhaust physics to optimize PFCs for power handling will be
shown for SPARC.

Speaker: Tom Looby (Commonwealth Fusion Systems)

Monday 18 May

5 Opening / Welcome

Review Talk: Morning session

R1

6 R1 Modeling first wall erosion and impurity migration: from layer formation to W-core ingress

Erosion of the first wall and transport of impurities critically affect the operation of magnetic confinement fusion devices. Erosion limits component lifetimes, while high-Z impurity penetration into the core increases radiation and can degrade confinement. For low-Z materials, repeated transport and re-deposition lead to long-range migration and formation of co-deposited layers that can retain tritium. Driving the ITER material choices, the community’s focus has evolved from carbon migration and remote co-deposition, to beryllium transport and co-deposition, and now to tungsten erosion and ingress into the edge, pedestal, and core. This review traces that evolution and assesses predictive capability.

On the erosion side, energetic ions and charge-exchange neutrals cause physical (and in some systems chemical) sputtering. Quantitative prediction remains limited by evolving surface roughness and microstructure, complex carbon chemistry, and the distinction between gross and net erosion under local re-deposition.

Impurity transport is often treated in trace approximation using codes with varying fidelity, from gyrocenter-averaged solvers (e.g., DIVIMP) to gyro-orbit–resolved tools (e.g., ERO, IMPGYRO). Regardless of fidelity, predictive capability is constrained by the accuracy of the background scrape-off-layer (SOL) plasma solution. Numerous impurity seeding experiments (JET, AUG, W7‑X, TEXTOR, DIII‑D) have been used for validation, but the tight coupling of surface processes and plasma transport complicates isolating transport alone. Layer formation modifies source distributions and the impurity influx, necessitating self-consistent modeling of surface composition evolution. WallDYN addresses this by coupling plasma impurity fluxes with dynamic wall composition, capturing multi-step (re-)erosion and (re-)deposition and long-range migration.

ITER and future devices will primarily use tungsten first walls, mitigating low-Z co-deposition with fuel. However, boronization may be needed for startup conditioning, requiring validated predictions on boron transport and co-deposition for tritium management. For tungsten, core ingress of high-Z impurities is a central concern. AUG and JET experiments demonstrate heating scenarios that suppress core accumulation and show strong divertor compression of high-Z species. Impurity ingress reflects competition between parallel SOL transport from main-chamber sources toward the divertor sink and radial transport into the core. This shifts the emphasis from divertor to main-chamber SOL impurity transport studies. This also demands improved, validated main-chamber plasma models, especially for parallel flow patterns.

This presentation will review the attempts to validate the erosion/migration code packages, point out the main sources of uncertainty and outline the required experimental data and model refinements needed, to improve the predictive capabilities for future fusion devices.

Speaker: Klaus Schmid (MPPL)

Oral: Morning session

7 O1 Integrated modeling of wall material evolution and the boundary plasma transport for EAST long-pulse discharges

A SOLPS-DIVIMP-SDTRIM.SP integrative modeling framework has been established to simulate time-dependent material erosion, impurity transport, and boundary plasma transport for > 100s long-pulse discharges in EAST for the first time. The erosion and redeposition of plasma-facing materials (PFMs) during long-pulse discharges not only affect the boundary plasma conditions but also determine the lifetime of PFMs. The iterative simulation of plasma distribution and impurity migration with grids extended to the first wall are realized by sequential SOLPS and DIVIMP simulations. Then the material evolution is calculated by the SDTRIM.SP with the particle flux derived from SOLPS and DIVIMP. The simulation results reveal that there exist periodic formation and diminishment of tungsten-lithium (W-Li) multilayers at some locations near the strike point on the divertor. With an initial low-Z Li coating on the W amour due to wall conditioning, plasma impact effectively erodes the coating layer on the divertor surface, leaving a W-enriched top layer that protects the underlying low-Z Li from erosion and thus forms a multilayered structure. The multilayered structure is verified by dedicated post-mortem EDS analysis of the EAST divertor target [1]. As plasma impinges, the W enriched surface layer gradually diminishes, leading to a transition from the W-dominant surface material to underlying Li. The periodic formation and diminishment of the multilayered structure significantly influence local particle recycling processes through changing the reflection, reemission and deposition rates of D on PFMs [2]. Unlike the D reemission, which emits D at the thermal energy, the D reflection can lead to energetic D neutral emission with a much higher averaged energy even above 100 eV, which results in stronger neutral penetration. The dynamic evolution of the D recycling reshapes the boundary plasma distribution and can even affect the sustainment of divertor detachment in EAST long-pulse discharges. This phenomenon is evidenced by the corresponding variation of Da and Langmuir probe data.

Comparative studies of material evolution under Li and boron (B) coatings show that B coatings have over twice the lifetime of Li coatings under typical EAST high-recycling divertor conditions. Periodic W-B multilayer formation also occurs, but its impact on global particle recycling is less significant than that of Li coatings. Modeling advances in this work highlight the crucial role of wall material evolution in predicting dynamic boundary plasma behavior, especially during long-pulse discharges.

[1] G. Xu, et al. IAEA-FEC, 2025, Chengdu, China.
[2] Q. Long, et al. NME 41 (2024): 101840.

Speaker: Guoliang Xu (ASIPP)

8 O2 Understanding the sputtering phenomena of complex surfaces

The plasma facing materials of fusion reactors are subjected to harsh environments, and especially to possible synergetic effects of several different factors at once. To understand both the effects of single factors and their synergistic effects, we
need to understand the phenomena at the atomistic level. Molecular dynamics (MD) simulations have been utilized to understand these phenomena, and with the increasing computational power, these simulations can be done on large and complex enough systems, to ultimately understand the sputtering effects. In the last decade, not only flat low index surfaces have been investigated, but also the effects of random and amorphous surfaces, and surface features. We showed that amorphous surfaces, usually used as a proxy for a random orientation, do not work. However, random surface orientations were studied, and they were shown to follow the Sigmund theory at high
energies [1]. Our combined experimental and computational study showed that surface structures can dramatically affect sputtering and its mechanisms [2].

In addition to surface structures and orientation, we can have impurities at the surface or just below the surface, e.g. deuterium or tritium decoration and boron from boronization. The study of the effects of hydrogen isotopes and boron on tungsten surfaces is enabled by newly developed and extremely accurate Machine Learning interatomic potentials. We studied the effects of decoration of the surface and found that not only is the tungsten sputtering affected, but also the fuel was seen to easily be recycled into the plasma. The sputtering of different elements followed their own trends, additionally affected by ion type. Sputtering of different molecules was also observed. Alloy surfaces showed preferential sputtering, and that at the steady state the surface composition is clearly different compared to the initial one. The preferential sputtering was not dictated by the elemental sputtering yields, but depending on other factors, such as mass. For all cases, decoration, boronization or alloying, the underlying mechanisms
were determined by the atomistic insight enabled by MD simulations.

Our results will not only give greater insight into the sputtering phenomena, but can be used to enable more accurate larger scale simulations. Additionally, the data obtained can be used to train surrogate models, which would permit having a fast model directly
integrated into the larger scale models, to enable very accurate simulations.

[1] Schlueter. et al. Physical Review Letters 125 (2020) 225502
[2] Cupak. et al. Physical Review Materials 7 (2023) 065406

Speaker: Fredric Granberg

10:10 Coffee Break

Invited Talk: Morning session

9 I1 Erosion and Screening of Tungsten in the JET divertor during inter/intra-ELM Periods in DT Ne-seeded plasma

Tungsten (W), chosen as the plasma-facing material for ITER, offers high thermal robustness and low tritium retention, but its erosion and penetration into the plasma core can degrade the confinement through dilution and radiation. The net W content depends on the interplay of erosion, divertor screening and transport, which vary with plasma species. To resolve these dependencies, we investigate W sources in DT plasmas at JET-ILW using optical emission spectroscopy, focusing on intra-ELM erosion and the effect of impurity seeding.
The identification of tungsten (W) atomic sources in deuterium-tritium (DT) plasmas has been improved through optical emission spectroscopy performed at JET-ILW. Intra-ELM erosion dominate the total W gross erosion source in DT plasmas. Notably, the sputtering yield (Y_W) is found to be roughly twice as large in H-mode neon (Ne)-seeded pulses, with concentrations of n_Ne⁄n_e ≈1.0-1.6%, compared to unseeded pulses. This increase is due to (a) the additional sputtering caused by Ne ions and (b) the higher T_(e,ped), which leads to a corresponding increase in impact energy. Although the sputtering yield is lower in the inner divertor compared to the outer divertor, W erosion is still dominated there. Lower Y_Wduring the ELMs in the inner divertor aligns well with predictions from the free-streaming model, which assumes that electrons transfer their parallel energy E_(e,∥)to ions as they stream toward the divertor target during ELMs, causing the resulting sputtering to depend mainly on n_ped and T_ped.
In unseeded plasmas, T ions are primarily responsible for W erosion during the ELM phases, whereas in Ne-seeded plasmas, both T and Ne ions contribute significantly to intra-ELM W erosion, with marginal contributions from D and Be ions. The contribution of W self-sputtering to the overall W erosion is limited to a maximum of ≈10%, consistent with predictions from the PIC/MC code BIT1 and from Monte Carlo simulations that account for multiple ionizations and electric fields in the Magnetic Pre-Sheath (MPS).
The W source in Ne-seeded plasmas is larger compared to unseeded reference DT plasmas. On the other hand, Ne-seeded plasmas show a decreased W density in the confinement region, indicating significantly improved W divertor screening. This can be explained by the fact that in Ne-seeded H-mode plasmas, increased Ti and reduced ne in the pedestal cause the drift direction to reverse, pointing outward as predicted by neoclassical transport (|R/Ln|<|R/2LT|, where Ln and LT are the density and temperature gradient scale lengths in the pedestal)

Speaker: Alexander Huber (Institute of Fusion Energy and Nuclear Waste Management–Plasma Physics, Forschungszentrum Jülich GmbH)

Oral: Morning session

10 O3 Time-resolved simulation of erosion on tungsten-coated divertor plate in the Large Helical Device using ERO2.0

Time-resolved simulations of erosion on tungsten-coated divertor plates in the Large Helical Device (LHD) were performed using ERO2.0, a three-dimensional Monte Carlo code for plasma–surface interaction analysis [1]. The temporal evolution of the erosion areas during the experimental campaign was investigated using a new modeling approach that accounts for time-dependent changes in surface composition on the plates. This approach enables more accurate lifetime estimations for tungsten layers.
Before the experimental campaign in FY2019, conventional carbon divertor plates in the closed helical divertor region of one of the ten helical sections were replaced with tungsten-coated plates to evaluate their compatibility with plasma discharges [2]. During the experimental campaign, the tungsten-coated plates successfully suppressed deposition layers, thereby reducing dust particle generation in the divertor region without significant degradation of plasma performance. However, post-campaign inspections of the plasma-facing components revealed erosion of the tungsten coatings, with depletion at strike points on about half of the tungsten-coated divertor plates. This finding poses a critical challenge for maintaining the lifetime of plasma-facing components and highlights the need for further investigation into material migration.
To elucidate the physical mechanism underlying tungsten depletion, ERO2.0 simulations were performed, revealing that the calculated tungsten-erosion profiles matched observations and demonstrated the progressive expansion of depleted areas throughout the experimental campaign. The evolution of depleted areas was driven by tungsten self-sputtering and interactions with intrinsic carbon ions in the peripheral plasma. This analysis identified two impurity transport processes contributing to tungsten erosion:
1. Long-range transport of carbon sputtered from strike points on carbon divertor plates in other divertor regions,
2. Short-range transport of carbon sputtered from depleted areas on the tungsten-coated divertor plates.
Post-mortem surface analysis of a tungsten-coated divertor plate using Glow Discharge Optical Emission Spectrometry (GD-OES) revealed the formation of three distinct areas characterized by different material composition depth profiles:
1. Carbon deposition areas, where carbon is deposited on the tungsten-coated layers,
2. Tungsten-coated areas, where tungsten layers remain on the carbon substrate,
3. Depleted areas, where tungsten layers were removed entirely, exposing the carbon substrate.
To elucidate the mechanisms of formation of the three areas, impurity migration on the tungsten-coated divertor plate is investigated using ERO2.0. It provides valuable insights into the physical processes governing material migration in the tungsten-coated divertor region.

[1] J. Romazanov, et al., Nucl. Mater. Energy 18, 331 (2019)
[2] G. Motojima, et al., Proc. 28th IAEA FEC 2020, IAEA-CN-286/EX/P6-22, Vienna, Austria, May 2021

Speaker: Mamoru Shoji (National Institute for Fusion Science)

11 O4 Refining High-Z Impurity Sourcing and Transport Models for Mixed Material Plasma-Facing Component Operations at DIII-D

The transport of tungsten (W) impurities from W-coated plasma-facing components (PFCs) in the largely carbon (C) environment of the DIII-D tokamak was assessed leading to refinements of the existing modeling and theory framework that will be needed as the fusion community advances to high-Z PFC operations while maintaining a mixed-material environment in transition. Recent progress at DIII-D is presented - including the effects of mixed-material environments and surface morphology on impurity sourcing, as well as the role of kinetic effects on high-Z impurities transiting the Scrape-Off Layer (SOL) plasma and how that impacts divertor impurity leakage.

W impurity sourcing was measured from a W-coated divertor spectroscopically during H-mode discharges with Edge Localized Modes (ELMs) and compared with predicted erosion yields from the Free Stream Recycling Model (FSRM) [1]. The FSRM was found to over-predict W erosion yields, leading to the investigation of C-W mixed-material layer effects on sputtering yields and reflection coefficients and PFC surface roughness effects on C deposition. The effect of PFC surface morphology on re-deposition was further investigated in L-mode plasmas using the PYEAD-RustBCA-GITR modeling framework [2]. Net erosion can be significantly underestimated if the magnetic field B-to-PFC pitch angle (α) is larger than expected due to surface roughness or localized melting events. Sensitivity analysis suggests this geometric effect can compete with local plasma density and temperature in setting re-deposition rates [2].

W transport from a closed-divertor geometry was studied with regard to the ion Bx∇B drift direction while impurity leakage from the divertor region was assessed with impurity collector probes (CPs) to determine transport in the far-SOL [3] as well as soft x-ray tomography (SXR) and vacuum ultraviolet spectroscopy (SPRED) for W contamination of the core. Core W content was found to increase by up to ~3x when the ion Bx∇B drift was directed towards the divertor (favorable field direction for H-mode) relative to discharges with the unfavorable direction. However, CP measurements taken in the outer midplane far-SOL showed ~75% lower W deposition for discharges with favorable Bx∇B drift. A SOLPS-DIVIMP workflow was used to assess these results and determine the plasma conditions under which kinetic effects become dominant in high-Z parallel impurity transport in plasma regimes of both low impurity ion charge states and low collisionality.

Work supported by US DOE under DE-SC0023378.

[1] Cacheris et-al. PPCF, 65 (2023) 085010
[2] Easley et-al. PPCF, 67 (2025) 035023
[3] Messer et-al. NME, 38 (2024) 101566

Speaker: Prof. David Donovan (University of Tennessee-Knoxville)

Invited Talk: Morning session

12 I2 Tungsten neoclassical temperature screening in the periphery of ASDEX Upgrade plasmas

High-Z impurities such as tungsten must be kept out of the plasma, to avoid excessive radiation or plasma collapse. Previous assessments of the neoclassical impurity transport expected at the edge of ITER plasmas has indicated favorable outward convection of tungsten, due to the expected high pedestal ion temperatures and lower density gradients [R. Dux et al, PPCF 2014; NME 2017] leading to temperature screening. While the edge plasma conditions expected on ITER cannot be reproduced in current devices, this effect has been experimentally observed at the periphery of optimized, high performance hybrid scenario pulses at JET [J. Hobirk et al, NF 2023; A.R. Field et al, NF 2023].

In the full W-wall ASDEX Upgrade, experiments to study the neoclassical temperature screening of tungsten in the plasma periphery were undertaken. The required plasma conditions are strong edge $\mathrm{T_i}$ gradient, low edge $\mathrm{n_e}$ gradient, as well as low collisionality and strong rotation [D. Fajardo et al, PPCF 2023]. To this end, the “improved H-mode scenario” [J. Hobirk et al, IAEA-FEC 2012] was optimised by tuning the heating and gas waveforms. High plasma performance was achieved $\mathrm{(H_{98(y,2)} \sim 1.3}$, $\mathrm{β_N \gt 3.0}$, $\mathrm{W_{MHD} \sim 1.3\;MJ)}$, while the radiation was maintained at low levels $\mathrm{(f_{rad}=P_{rad,tot}/P_{in} \sim 0.3)}$. The tungsten concentration remains low, and bolometry and soft X-ray measurements indicate a significant reduction of impurities in the ‘mantle’ region, indicating that tungsten is effectively kept out at the plasma edge. Transport modelling with FACIT [P. Maget et al, PPCF 2020; D. Fajardo et al, PPCF 2022] indicates outward convection of tungsten in the region just inside the pedestal top.

To assess this scenario in detail, the tungsten density profiles inferred from bolometry and soft X-rays are compared with the modelling and the tungsten sources are qualitatively assessed. The analysis shows a strong poloidal asymmetry of the tungsten density in the pedestal. The impact of the ELMs on the tungsten density, as well as its inter-ELM evolution are discussed. Furthermore, small amounts of neon were injected in selected pulses to study the neon transport using CXRS following the method developed in [T. Gleiter et al, NF 2025]. This aims to compare the transport of impurities with different charge, and to assess the effectiveness of impurity seeding for power exhaust in these conditions.

Therefore, this work enables a comprehensive investigation of tungsten transport, demonstrating experimentally its neoclassical temperature screening in the plasma periphery, relevant to ITER.

Speaker: Athina Kappatou (Max-Planck-Institut für Plasmaphysik, Garching, Germany)

12:20 Lunch

Invited Talk: Afternoon session

13 I3 Physics basis and status of the ITER tungsten First Wall

The ITER 2024 re-baseline defines a revised approach to reach the main burning plasma objective. A key component of this strategy is the switch from Be to W main wall armour, making ITER a full-W device from the beginning of operations. In fact, the new wall becomes two walls, with a Temporary First Wall (TFW) in place for the “Start of Research Operations” campaign (SRO) in the revised ITER Research Plan, followed by the final, actively cooled FW in place for DT operations. Both are being designed in parallel, but the lengthy DT wall procurement means that the second cannot benefit from experience gained with the first. Both must interface with the same set of Blanket shield blocks, but the TFW is an inertially cooled component with a mandate to permit achievement of the principal SRO objectives of 15 MA / 5.3 T hydrogen L-modes, 7.5 MA / 2.65 T deuterium H-modes. These two factors are driving TFW design choices, e.g.: thermal envelope defined by an entire pulse rather than a single (burning plasma) operating heat flux for the DT wall; standardizing FW panel tiles, simplified shaping, deployment of armour variants in different locations (e.g. bulk W, W heavy alloy).

The new walls benefit from improvements in physics understanding of both stationary and transient heat loads over the past decade since the final design review of the original Be variant. Disruption transients constitute the main rationale for the TFW, which will allow mitigation and avoidance techniques to be developed during SRO without fear of water leaks. In turn, this relaxes the need for the DT wall design to account for the worst-case disruption current quenches since it is assumed that the SRO Research Plan will guide improved operation. Nevertheless, the spectre of runaway electrons (RE) can never be completely alleviated by experience gained in SRO; Compton and tritium decay beta seed electrons will appear for the first time in the DT phase. Much improved understanding of RE impact, from incoming energy distribution to deposition in the material and subsequent thermal transfer, have driven an increase in design armour thickness in areas of the DT wall expected to be critical for RE interactions.

The paper presents the physics basis for the loads which are determining the design of the new ITER W walls. It will also provide an update on the wall designs at the time of the conference.

Speaker: Richard Pitts (ITER Organization)

Oral: Afternoon session

14 O5 Study of Charge-exchange Neutral Distribution and its Impact on First wall Erosion in EAST

The energy distribution of charge-exchange neutral (CXN) measured on EAST is interpreted by SOLPS-ITER modeling with the wide grid version, and the CXN-induced all erosion near the inner midplane and baffles is found to be significant especially under the enhanced SOL transport regime. Simulations for ITER predict that plasma and CXN fluxes onto the first wall are very sensitive to the cross-field transport in the far-SOL [1,2], while the CXN transport has not been well investigated. Recently, self-consistent simulations for edge plasma and neutrals are conducted and compared with the measurements of CXN distribution by the low-energy neutral particle analyzer (LENPA) in EAST. The CXN energy spectrum obtained from the SOLPS-ITER simulations with realistic wall geometry matches well with LENPA measurements. A new EIRENE interpretive module has been developed to analyze the generation and transport mechanisms of CXNs with different energies.

The formation mechanism of the “turning point”, which is a typical characteristic of the CXN energy spectrum in H-mode plasmas that separates the low-energy part and the high-energy part with two slopes, is found to be caused by the steep Ti gradient in pedestal region. The modeling is verified by dedicated L-H transition experiments on EAST. Meanwhile, more than one half of CXNs generated in the plasma undergo collisional losses including charge exchange and ionization during their flight pathway to the first wall. The baffles experience a high integrated flux (Γ_{int}) but a low mean energy (Emean) of CXN, while the midplane region exhibits the opposite characteristics. The particle anomalous diffusion coefficient (D_{⊥}) is scanned in different locations in the SOL. Extending the grid to the full wall is found to significantly decrease the sensitivity of the solution with far-SOL D⊥ because of the less recycling sources from the wall surface. Whereas the CXN flux at the first wall increases with D_{⊥} mainly due to the enhanced recycling source at the baffles. The impact of CXN on material erosion is also analyzed for both boron and tungsten. The boron erosion mainly depends on the Γ_{int} and thereby peaks at the baffle regions near divertors. The tungsten erosion is sensitive to the incident particle energy so that it peaks near the inner midplane where the maximum Emean exists.

[1] J. Romazanov et al., Nuclear Materials and Energy, 26 (2021) 100904.
[2] N. Rivals et al., Nuclear Fusion, 65 (2025) 026038.

Speaker: Jin Guo (ASIPP)

15 O6 Impact of turbulent SOL fluctuations on tungsten sources in ASDEX Upgrade

The physical sputtering rate of tungsten (W) and its sensitivity to scrape-off-layer (SOL) fluctuations is characterised as the first application of time-dependent ERO2.0 [1] full-orbit Monte Carlo simulations in fluctuating turbulent plasma conditions. A time-dependent 3D description of global edge-SOL turbulence is obtained from validated GRILLIX [2] simulations. The studied plasma scenario is a diverted attached L-mode plasma in the ASDEX Upgrade tokamak. The time evolution of the plasma profiles is extracted from GRILLIX into ERO2.0 with a resolution of 5 µs in time and ~1 mm in space. The simulated W sources are validated by comparing W I 400.9 nm line emission spectrometer measurements with synthetic W I diagnostics at the low-field side divertor target.

The GRILLIX plasma solution has density fluctuation amplitudes up to 150%, and ion temperature fluctuation amplitudes up to 250% of the mean-field in the low-field side SOL [2]. Plasma filaments propagating radially outwards from the separatrix cause intermittent erosion of the W first wall in the main chamber. ERO2.0 simulations predict that the W net erosion rate in the far-SOL varies in time by more than an order of magnitude, mainly due to the intermittent impurity ion (B, O, W) fluxes.

In contrast to the strong fluctuations in the low-field side far-SOL, cross-field transport is low near the high-field side mid-plane and at the near-SOL outer divertor target in the studied conditions. A time-averaged mean-field approach enabling the use of a steady-state plasma solution is therefore a fair approximation at the outer strike line, but causes an underestimation of the W erosion sources in the far-SOL divertor and main chamber. While the highest W gross and net erosion source in the ERO2.0 simulations is located near the strike line, the W sources in the main chamber are less efficiently screened and contribute more to the predicted W density in the confined region.
[1] J. Romazanov et al., Nucl. Mater. Energy 18 (2019) 331-8
[2] W. Zholobenko et al., Nucl. Mater. Energy 34 (2023) 101351

Speaker: Henri Kumpulainen (FZJ)

16 O7 3D numerical modeling of W erosion and migration in WEST plasma: impact of magnetic ripple on erosion patterns and core contamination

With the transition to full tungsten (W) plasma-facing components in ITER, the modeling of plasma wall interaction and the prediction of W erosion, transport and screening it has become a crucial issue for preparing ITER operation. In this context the development of improved modeling tools able to predict W behavior and its impact on plasma performances as well as the transport of seeded impurities is decisive. In this contribution we will report recent results obtained with 3D simulations using the SOLEDGE3X-ERO2.0 workflow, with application and validation on WEST plasma experiments. We have considered realistic 3D magnetic equilibria in WEST where the impact of magnetic ripple requires a fully 3D treatment of the problem, with relatively important variation of the magnetic field in the toroidal direction. Computing W erosion and migration with ERO2.0 on the 3D SOLEDGE3X plasma backgrounds, we are able to reproduce the 3D patterns of erosion on the WEST divertor, that follow the heat flux patterns with maxima and minima periodically spaced on the two sides of the divertor. Using such 3D plasma backgrounds and considering Lower Single Null configuration, the results previously obtained using 2D axisymmetric simulations are revisited. The lower divertor is an important source of W but well screened, even if the 3D cases seem to suggest that the screening is less efficient than what previously obtained with axisymmetric assumption. Interestingly very low erosion rate on the upper divertor is found accordingly with experiments and contrary to what obtained with 2D axisymmetric simulations. Other important contributors to W migration toward core region are the baffle, as expected and predicted also with axisymmetric assumption, and the antenna limiter, as shown in recent 3D studies. These numerical results are compared with analytical models for erosion rates as well as experimental findings from visible spectroscopy and recent post mortem analyses, showing very good agreement and indicating the relevance of having fully 3D geometry description in order to catch W erosion patterns and migration paths. The validation of SOLEDGE3X-ERO2.0 workflow on WEST experiments is a crucial step in order to apply such workflow to ITER relevant scenarios and predict the impact of wall erosion on plasma operation.

Speaker: Dr Guido Ciraolo (IRFM CEA)

Invited Talk: Afternoon session

17 I4 Evaluating the dominant tungsten erosion source during limiter start- up on ASDEX Upgrade

The ITER Research Plan re-baseline [1], especially the change of first wall (FW) material from beryllium to tungsten (W) has required the re-validation of the initial ramp-up phase in limiter configuration. The direct plasma contact on W surfaces, without the screening of eroded atoms provided by diverted operation, is expected to lead to high radiative ractions of about f rad = 70-80% as found in simulations for ITER with the SOLPS-ITER code [2, 3]. The W is generated by self-sputtering and self-regulated due to the high radiative power, leading to moderate last closed flux surface (LCFS) temperatures. These findings are in contrast with earlier W limiter start-upexperiments on ASDEX Upgrade (AUG) [4], where W sputtering was found to be mainly by lowand medium Z impurities, e.g. oxygen (O), with W self-sputtering only a minor contributor.

New W limiter experiments in support of the ITER re-baseline were performed in 2024-5 on EAST, AUG and WEST [2,5]. Here we report on further, complementary studies on AUG in which low density, strongly ECRH heated and well diagnosed plasmas were executed on W outboard limiters following a fresh boronization. This allows easier initial start-up, but, because the boron (B) layer erodes rapidly in the limiter contact area, is equivalent to limiting on a W surface. The primary aim was to access ITER-like conditions with high temperatures at the LCFS. In these plasmas the W source on the limiter is observed to increase even though the plasma density, the heating and also the low Z impurity sources and concentrations are rather constant. Radiation losses increase together with the W source and slowly saturate on timescales of a few 100 ms. This is an indication of a regime where the W self-sputtering dominates the radiation build up, a supposition supported by dedicated, time dependent SOLPS-ITER simulations. The latter, well constrained by the multiple experimental measurements, predict that these AUG plasmas should have a high W self-sputtering source and moreover find that the prompt re-deposition fraction is the principal mechanism allowing for good agreement between code and experiment regarding the 2D radiation distribution.
References
[1] A. Loarte et al., PPCF 67 (2025) 065023
[2] R.A. Pitts et al., NME 42 (2025), 101854
[3] Y. Zhang et al., NF 65 (2025) 056035
[4] A. Kallenbach, et al., NF 49 (2009) 045007
[5] J. Hobirk et al., 2025 paper presented at 34th IAEA FEC (Chengdu, China), EX-35811, https://conferences.iaea.org/event/392/papers/35811/files/13761-IAEAPaper25v3.pdf

Speaker: Jörg Hobirk (MPPL)

18 I5 Kinetic effects on tungsten impurity transport and divertor leakage mechanisms under different divertor conditions

Kinetic effects are found to significantly diminish the thermal force on W relative to conventional fluid modeling, underscoring the necessity of kinetic approaches for modeling W transport in boundary plasmas. A recently developed kinetic impurity transport model in DIVIMP has been used to investigate kinetic effects on W transport and screening across various divertor conditions. Results indicate that the kinetic correction to the W thermal force is significant not only in low-collisionality regimes (ν* < 20), but also under conditions of low ion-temperature (Ti < 20 eV) and high effective-charge (Zeff). In pure D discharges, the kinetic correction to the W thermal force is weaker in high-density H-mode plasmas than in low-density L-mode plasmas. However, the strong friction force in high-density H-mode cases makes the influence of kinetic effects on W screening more significant. These modeling results align with experimental observations on EAST, where increasing boundary plasma collisionality via D₂ puffing reduces core W density by more than 20%. With Ne-seeding, the higher plasma temperature in the main-SOL leads to a higher Zeff than in the divertor, resulting in a more pronounced kinetic correction on the W thermal force in the main-SOL. As divertor conditions transition from the high-recycling to detached regimes, the decrease in Ti and increase in Zeff further enhance kinetic effects on W screening. To further validate the role of Zeff-dependent kinetic effects, dedicated He-plasma experiments were conducted. As expected, superior W screening is observed in He plasmas, particularly under detached conditions, due to increased Zeff and the corresponding reduction in W thermal force.
The synergy between kinetic and E×B drift effects on W transport has also been delineated. With the reduction of the W thermal force by kinetic effects, E×B drifts are demonstrated to have a more significant effect on W leakage through a reversed near-SOL flow on EAST. In contrast, E×B drifts are weak in the large-size and high-Bt devices like CFETR, and W edge transport is governed by thermal and friction forces. Under Ne-seeding detached divertor conditions on CFETR, kinetic effects can even reverse the direction of the W thermal force from pointing upstream to pointing divertor target, thereby further enhancing divertor W screening. However, a clockwise plasma flow in the low-field-side SOL is observed as a potential W leakage pathway for CFETR. By increasing Ne injection rates, the reverse flow is effectively suppressed, which narrows the near-SOL leakage path and enhances W screening.

Speaker: Dr Hui Wang (Institute of Plasma Physics, Hefei Institutes of Physical Science, Chinese Academy of Sciences)

15:50 Coffee Break

Postersession 1: Tracks A, B, C, D, E, F and G

19 1.014 Helium bubbles migration characterized by in-operando GISAXS and in-situ TEM.

In fusion devices like WEST, ASDEX Upgrade, EAST and ITER, tungsten (W) has been chosen as the plasma facing material for the divertor, where the heat and particles fluxes are the most intense. In particular, helium (He) irradiation leads to the formation of nano-sized bubbles in the subsurface area, which increase hydrogen isotopes retention. More dramatically, W-fuzz may form, presumably caused by the repeated bursting of bubbles reaching the surface. Understanding bubble migration and diffusion is therefore of prime importance.

We have used a nanoscience approach to characterize the bubbles in order to control separately the multiple factors met in a tokamak environment such as surface conditions, temperature evolution, microstructure, pre-existing defects and He irradiation. Well prepared single crystals were implanted below and above the displacement threshold of W using 400 eV or 2 keV He ions. Bubbles were characterized by Grazing Incidence Small Angle X-ray Scattering (GISAXS) at the ESRF synchrotron. This technique provides statistical information on the sample morphology. At ITER relevant temperature (1273 K), GISAXS measurements have revealed that bubbles are facetted. The equilibrium shape has been precisely described and consists of a truncated rhombic dodecahedron composed of {100} and {110} facets [1]. The growth kinetics of bubbles has been measured in real time during implantation by in-operando measurements. The growth mechanism has been identified as migration-coalescence [2].

We will present the latest results addressing the migration of bubbles. In complementary to in-operando GISAXS measurements, in-situ Transmission Electron Microscopy (TEM) measurements were performed. Thin lamellae extracted from implanted samples were heated from room temperature to 1100 °C, enabling direct visualization of bubble evolution events. Brownian diffusion of bubbles through the W matrix, coalescence leading to bubble growth and bursting events at the surface have been captured. The diffusion coefficient of bubbles was measured between 900°C and 1100°C for bubbles ranging from 1 to 4 nm in diameter by in-situ TEM and from 4 to 10 nm in diameter by in-operando GISAXS. In parallel, kinetic Monte Carlo simulations were conducted to confront experimental results. The diffusion coefficient decreases exponentially with the bubble diameter, which is consistent with a diffusion process hindered by the nucleation of a ledge on the bubble facets.

[1] L. Corso et al., Nucl. Mater. Energy 37, 101533 (2023).
[2] L. Corso et al., Nucl. Mater. Energy 42, 101894 (2025).

Speaker: Loic Corso (CINaM - Aix Marseille University)

20 1.036 Simulation of Nano-tendril Bundle Growth on Tungsten Surface under Helium/Neon Plasma Irradiation

Abstract
Island-like nano-tendril bundles (NTBs) composed of fuzzy nanofibers can form on a tungsten (W) surface under simultaneous sputtering by helium (He) as well as seeded impurity like neon (Ne) and argon (Ar). In future fusion devices like ITER, a high-energy ion flux that consists of He and seeded impurity Ne/Ar may facilitate the growth of NTBs through the sputtering and deposition of W. Although the exact growth mechanism of NTBs remains unclear, the formation conditions has already been identified. These conditions require a net sputtering yield (Ynet) of∼10-3-10-2, a W surface temperature range of 900–1600 K, and the presence of a pre-existing fuzzy structure on the W surface [1].
A series of experiments have observed that changes in Ynet can lead to significant structural variations in NTBs, and ion fluence can affect the growth characteristics of NTBs. This indicates that the sputtering-deposition process plays an important role in the growth of NTBs. Dedicated modeling has been conducted to explore NTB growth under He/Ne plasma irradiation by using the three-dimensional kinetic Monte Carlo code SURO-FUZZ [2-4]. The growth characteristics of NTBs under different ion fluences have been simulated by SURO-FUZZ. It is found that the height of NTBs increases as the ion fluence rises. Moreover, the impacts of impurity concentration and impinging energy on the changes in NTB growth have also been investigated.

Reference
[1] D. Hwangbo et al., Nucl. Mater. Energy 18 250-257 (2019).
[2] D. H. Liu et al., Nucl. Fusion 60 56018 (2020).
[3] K. R. Yang et al., Nucl. Fusion 62 096019 (2022).
[4] J. Y. Chen et al., Nucl. Fusion 64 056006 (2024).

Speaker: Shuyu Dai (Dalian University of Technology)

21 1.050 Modeling the Impact of Boron Powder Injection on Runaway Electron Dynamics Using JOREK

Abstract:
The EXL-50U device is specifically designed for hydrogen-boron (p-¹¹B) fusion research, with boron powder injection employed as one of its boron fueling schemes [1]. However, plasma perturbations induced by boron injection may trigger magnetohydrodynamic (MHD) activities and then facilitate the generation of runaway electron beams [2]. This poses challenges for the safe operation of the EXL-50U during boron powder fueling processes.
Comparative experiments on similar discharges with and without boron powder injection were conducted on the EXL-50U device. Infrared imaging results reveal that boron powder injection facilitates the growth of runaway electron avalanches. To elucidate the underlying physical mechanism, the three-dimensional (3D) nonlinear code JOREK is used to simulate the MHD instabilities and runaway electrons [3,4]. The simulation results demonstrate that boron powder injection induces the growth of nonlinear MHD modes and current perturbations, which in turn enhance the core toroidal electric field. This results in faster electron energy growth and a higher proportion of runaway electrons compared to the case without boron powder injection. These results indicate that boron powder injection facilitates the generation and acceleration of runaway electrons by enhancing the toroidal electric field via MHD perturbations.
[1] Shi Y. J. et al., 2025 Nucl. Fusion 65 092004
[2] T. Onchi et al., 2021 Phys.Plasmas 28 022505
[3] M. Hoelzl et al., 2021 Nucl. Fusion 61 065001
[4] Hu D. et al., 2024 Nucl. Fusion 64 086005

Speaker: Mr yang Feng (Key Laboratory of Materials Modification by Laser, Ion and Electron Beams (Ministry of Education), School of Physics, Dalian University of Technology, Dalian 116024, People’s Republic of China)

22 1.070 Quantifying Fluid Model Discrepancies for Low-Charge State Tungsten in Non-Maxwellian Edge Plasmas

Tungsten (W) impurity transport in fusion edge plasma conditions is significantly influenced by the characteristics of its velocity distribution function, limiting the accuracy of standard fluid models and the ability to accurately predict W sources in a future reactor. In particular, there is a stark disagreement between the fluid transport of low-charge state heavy impurities and the quasi-kinetic treatment of the same. Fluid descriptions inherently assume Maxwellian impurity distribution functions, but low-collisionality plasmas do not allow for the necessary equilibration times to achieve them for heavy impurities [1].
We introduce a quantifiable metric, the Root Mean Square Error Percentage (RMSEP), to assess the fidelity of the Maxwellian fluid assumption against the non-Maxwellian distribution functions calculated by the quasi-kinetic Monte Carlo trace impurity transport code, DIVIMP [2]. DIVIMP tracks impurity ions as discrete particles, modeling collisions, atomic processes, and drifts based on finite acceleration and thermalization times of ions, which explicitly permits non-Maxwellian behavior. In contrast, fluid models inherently assume immediate acceleration and thermalization of impurities, corresponding to local Maxwellian distributions.
DIVIMP is used to inject tungsten (W) into SOLPS-ITER plasma backgrounds spanning the detachment threshold. The full parallel fluid impurity momentum balance is calculated using local parameters from DIVIMP. The RMSEP then quantifies the deviation (or error due to non-Maxwellian impurity distributions) by comparing the DIVIMP-derived momentum to the calculated fluid force.
Results show that for low-charge states (e.g., W1+), the non-Maxwellian effects are intense, with RMSEP values reaching 250%, indicating a clear failure of the Maxwellian fluid approximation in these regimes. Using a fluid-only approach could result in unrealistic W leakage profiles. Conversely, the high-charge states (W6+ and above) most commonly found in hot plasmas are accurately described by the full fluid model, exhibiting an RMSEP of less than 50% across all modeled scenarios. Finally, an assessment of common fluid approximations demonstrates that simplified balances – such as those considering only friction and thermal forces – can yield an error more than double that of the full fluid balance, highlighting the need for complete momentum conservation in fluid modeling of these regimes. Informed application of various approximations in appropriate regimes could significantly reduce computational time for future impurity transport simulations.
1. Schmid, K. and T. Lunt, 3D global impurity transport modeling with WallDYN and EMC3-Eirene. Nuclear materials and energy, 2018. 17: p. 200-205.
2. P.C.S. and Elder J.D., A GUIDE TO THE DIVIMP CODE. 1995.

Speaker: Seth Messer (University of Tennessee)

23 1.018 Development of Machine Learned Interatomic Potentials for Simulating Transmutation Products in First Wall Materials

First wall materials in fusion reactors will be subject to extreme conditions including high fluxes of a variety of plasma species including hydrogen, helium, and other impurities as well as high heat loads of 10-20 MW/m2. In addition, neutrons will both alter the microstructure and composition through the creation of defects and transmutation elements. In particular, these transmutation products will influence defect evolution and thermomechanical properties of the material. Understanding the evolution of first wall materials under neutron irradiation will be critical for determining lifetime limits of these materials. Atomistic modeling like molecular dynamics (MD) can play a key role in determining the mechanisms at the atomic scale that drives the radiation response of the material. However, the accuracy of the simulations will be determined by the interatomic potential (IAP) used. Many classical potentials cannot reproduce the relevant physics for these types of extreme conditions at the plasma-material interface and typically are poor at extrapolating from the parameters they were fit to. A new class of IAPs using machine learning (ML-IAPs) have been shown to have increased accuracy compared to classical IAPs and may be more flexible in modeling plasma-material interactions.

In this work, we will describe the development of ML-IAPs for studying transmutation products in first-wall materials including W-Re, SiC-Mg, and Fe-Cr-Mn. Each potential was fit to a wide variety of training data that includes defect structures for the transmutation elements. In addition, an Fe-Cr-Mn potential for studying ⍺’ phase precipitation and potential trapping of Mn at those interfaces. We have developed a W-Re potential that accurately represents the interaction and clustering of Re with point defects. Density functional theory (DFT) predicts that Re should only bind with other Re atoms if there are vacancies present which our potential reproduces using MC-MD. In addition, we have developed an ML-IAP for SiC-Mg that reproduces the Mg point defect energetics as calculated using DFT. We plan to discuss current work on Fe-Cr-Mn ML-IAP development and the use of these potentials in studying defect evolution using MD.

Speaker: Mary Alice Cusentino (Sandia National Laboratories)

24 1.041 Modelling of tungsten prompt redeposition at the inner wall of ITER during ramp-up

Tungsten as high-Z material has a relatively low physical sputtering. Chemical erosion – if occurring at all – is negligibly small. In addition, tungsten has a very large melting point of about 3400°C. However, the core plasma tungsten concentration in a fusion device has to be kept at extremely low values around 3E-5 to minimize plasma dilution and in particular plasma cooling. Therefore, it is of great importance to control and minimize the net source of tungsten from the various wall elements. In this context prompt redeposition, which primarily happens for high-Z elements of large mass like tungsten, is an important process, which significantly can reduce the gross erosion to much smaller net erosion.
The prompt redeposition of sputtered tungsten at the inner wall of ITER during current ramp-up in limiter configuration has been simulated with the ERO code. Plasma parameters from SOLPS-ITER simulations in deuterium for a medium-density (with a peak electron density of 4E12 $cm^{-3}$ at the inner wall) and a high-density (1E13 $cm^{-3}$) case have been used as input for ERO. CX neutrals and impurities like oxygen or seed species hitting the wall will change the tungsten gross erosion source but not the fraction of prompt redeposition and are therefore not of relevance for the present study. Simulations without anomalous cross-field diffusion for sputtered tungsten ions reveal peaked prompt redeposition profiles in poloidal direction. At the tangency point with largest electron temperature ($T_{e}$ = 65 eV for medium and 48 eV for high-density case) and density, maximum prompt redeposition fractions of about 60% for the medium-density and 80% for the high- density case occur. At a distance of 50 cm away from the tangency point, prompt redeposition decreases to 10% (medium-density) and 20% (high-density case). The simulations without anomalous cross-field diffusion show that the overall redeposition is the same as the prompt redeposition thus the overall redeposition is only due to prompt redeposition. An anomalous cross-field diffusion of 1 $m^{2}/s$ leads to slightly increased prompt redeposition, however, for both medium and high-density case there is now also a significant amount of non-prompt redeposition. The modelled profiles of prompt redeposition can be used as input for plasma simulation codes like SOLPS-ITER to improve the assumptions of net tungsten wall sources.

Speaker: Dr Andreas Kirschner (FZJ)

25 1.049 Deuterium retention properties of advanced neutron multipliers for fusion applications

Beryllium intermetallic compounds, known as beryllides, such as Be12Ti and Be12V, are highly promising advanced neutron multipliers for JA demonstration (DEMO) power reactors due to their low swelling and high stability at high temperatures. Advanced neutron multipliers are being developed by Japan and the EU as part of their Broader Approach (BA) activities within the International Fusion Energy Research Center (IFERC) project.
The release and retention of tritium in beryllium and beryllides are extremely important properties since tritium can be generated in beryllium/beryllides by reactions between beryllium and neutrons. However, since database on tritium release and retention in beryllides fabricated by plasma sintering (only suggested by QST, Japan) is insufficient and unsatisfactory, it is difficult to clarify the mechanism of tritium retention and release. In addition, it is not easy to establish the database on tritium retention in beryllides owing to reactor restriction and high cost for neutron irradiation examination,
Accordingly, the authors' group has conducted research and development on deuterium release and retention in beryllium and beryllides using deuterium ion implantation. It was clear that the desorption properties and retention of hydrogen isotopes (deuterium) using the newly developed beryllide are better than those of beryllium. This is evidenced by lower starting-up temperature, total retention, and activation energy for deuterium desorption. Moreover, through comparison with transmission electron microscopy, it has been clarified that no bubbles exist at high temperatures due to the structure complexity of beryllides.
In this presentation, we will address an overview of R&Ds on deuterium desorption and retention properties of beryllides, focusing on temperature dependence, and compares them with tritium desorption and retention properties of beryllium and beryllides.

Speaker: Jae-Hwan Kim (QST)

26 1.015 Boron erosion and deposition with quartz crystal microbalances in EAST

Boronization is anticipated to serve as a routine wall-conditioning method for the fully tungsten (W) first wall under ITER’s new baseline [1]. To assess the performance of boron (B) coatings under ITER-relevant metallic wall conditions, experimental campaigns involving carborane-based boronization was conducted in the EAST from 2023 to 2024 [2]. B deposition rate during boronization and material erosion/deposition rate during subsequent plasma operations have been studied using quartz crystal microbalances installed at ports C (C-QMB) and J (J-QMB).
During boronization, B deposition rates ranged from 0.31 to 3.40 μg·cm⁻²·h⁻¹ at C-QMB and from 0.04 to 16.95 μg·cm⁻²·h⁻¹ at J-QMB, respectively. The B deposition behavior is influenced by the locations of sublimation oven and ICRF antenna, oven and substrate temperatures, and substrate material. Higher deposition rates were measured closer to the ICRF antenna, in regions with denser gaseous carborane near the oven, and at higher oven temperatures, whereas reduced deposition was observed on high-Z substrates (W) and at elevated wall temperatures.
During plasma discharges, material erosion was detected in the majority (>50%) of discharges, regardless of whether they were completed normally or terminated due to disruptions. Erosion rate increased with rising plasma density, which was attributed to enhanced flux of neutrals in 20–500 eV range. Heating configuration also played a significant role on material erosion/deposition. ECRH discharges resulted in 1.20–1.95 times higher erosion rate than that in discharges using combined LHW and ECRH heating. While NBI discharges tended to enhance deposition. Higher NBI power leads to greater deposition due to interactions between shine-through neutrals and charge-exchange neutrals with the first wall.
B-based coating on C-QMB exhibited a lifetime of ~10⁴ s, as measured following a single boronization using 10 g of carborane. Subsequent post-mortem analyses revealed that approximately 30 nm of a B-carbon film persisted on the C-QMB after entire campaign. Notably, this film exhibited strong oxygen gettering capability and only minor contamination by iron and copper. Furthermore, deuterium retention in the film was measured at 2.12 × 10²⁰ m⁻²—over eight times higher than that in pure tungsten (W)—highlighting the pronounced deuterium trapping capacity of B-containing films in low-flux regions. These findings thus provide valuable insights into the application of boron in next-step devices such as ITER.

[1] J. Winter, H.G. Esser, L. Könen, et al., J. Nucl. Mater. 162–164 (1989) 713–723.
[2] Y.H. Guan, G.Z. Zuo, W. Xu, et al., Nucl. Fusion 65 (2025) 096020 (12pp).

Speaker: Rong Yan (ASIPP)

27 1.016 Exposure of advanced tungsten materials and ultra-high temperature ceramics to divertor plasmas in DIII-D: Analysis of erosion, surface morphology changes, recrystallization, and surface cracking

We examined the effects of divertor plasmas on 14 distinct tungsten and ultra-high temperature ceramic (UHTC) materials, providing insight into how combined high heat and particle fluxes affect their surface composition and structure. The experiments were carried out using the Divertor Materials Evaluation System (DiMES) at DIII-D. The test matrix included commercially available tungsten alloys doped with (20 ppm and 30 ppm) K and (10 %) Re, along with UHTC materials produced at Stony Brook University, including NbC, (Nb+Ta)C, ZrC, WC, (W+Si)C, and SiC. In each case, the samples were exposed to 6-7 H-mode plasma shots, with an average steady-state perpendicular heat flux of 2.4 MW m$^{-2}$, an ELM heat flux of 6 MW m$^{-2}$, and an ELM frequency of 75 Hz. Select samples were cut to a 10° angled geometry to increase the intercepted heat flux to > 10 MW m$^{-2}$.

Postmortem characterization revealed that all tungsten alloys survived exposure to divertor plasmas well, with modest surface morphology changes, near-surface cracking, and leading-edge melting observed as the main effects of the plasma exposure. Spectroscopic ellipsometry, obtained before and after plasma exposure over a wavelength range of 245 – 1000 nm, was consistent with nm-scale roughening of the surface. This was confirmed with scanning electron microscopy, which also revealed evidence of grain boundary grooving commonly observed with high-temperature annealing of tungsten. Slight changes in fiducial marker geometry (including rounding of corners) due to erosion were also noted. All samples showed evidence of minor leading-edge melting (over regions spanning 50 – 100 μm in width), highlighting the sensitivity of tungsten materials to slight misalignment. The flush-mounted UHTC specimens also demonstrated promising performance, with minimal surface morphology changes observed following exposure. Additional X-ray and Auger spectroscopies are underway to assess preferential sputtering of the UHTC surfaces, as well as the use of grazing incidence X-ray diffraction and electron microscopy techniques to study grain growth and crystal structure stability. These results are expected to provide insights needed for further optimization of doped W and UHTC materials and provide guidance on materials selection for upcoming materials testing campaigns in DIII-D.

SNL is managed and operated by NTESS under DOE NNSA contract DE-NA0003525. This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Award(s) DE-FC02-04ER54698, DE-FG02-07ER54917, DE-SC0019256, DE-AC05-00OR22725, and DE-AC52-07NA27344.

Speaker: Robert Kolasinski (Sandia National Laboratories)

28 1.017 Thermal and elastic property evolution of displacement damaged tungsten at fusion-relevant annealing temperatures

The ability to quantify irradiation damage and recovery of thermal transport and elastic response in tungsten is essential for understanding and accurately predicting stress evolution, heat management, and even failure modes in fusion plasma-facing components. Transient grating spectroscopy (TGS) non-destructively quantifies material properties in the first few microns [1]. Whereas previous TGS works on tungsten (W) [2-4] have focused on thermal diffusivity (α) alone, this work also determined Poisson’s ratio (𝝂) and Young’s modulus (Ε). In addition to the surface acoustic wave (SAW) typically reported, another higher frequency peak is clearly discerned in the Fourier spectra of the TGS signal. TGS measurements on both Si and W confirm this frequency is due to a surface skimming longitudinal wave, constraining the analytical SAW equations and allowing both 𝝂 and Ε to be extracted.

Polycrystalline W (PCW) and single-crystal W (SCW) samples were partially masked during self-damage to preserve a pristine region, used for normalizing results. PCW samples were self-damaged with W ions at 20 MeV, while the SCW samples were self-damaged with 11 MeV ions. To study recovery, a 0.020 dpa SCW sample was annealed in 100 C increments. X-ray diffuse scattering quantified dislocation loop size and concentration in the SCW samples. TGS measurements were performed ex-situ with varied penetration depths from ~0.6 to 1.9 μm to probe within and beyond the induced-defect region.

Thermal diffusivity was found to decrease for all self-ion damaged samples. Previously reported 𝝂 and Ε values for neutron or self-ion irradiated W at elevated temperatures resulted in no change within the uncertainty of the impulse excitation technique or cantilever measurements [5-6]. For self-ion damage performed at room temperature, we measured a reduction in Ε of ~0.8, 1.4, and 2.0% for 0.008 dpa SCW, 0.020 dpa SCW, and 0.23 dpa PCW, respectively. Annealing the 0.020 dpa SCW sample showed partial recovery of α and E. All samples measured showed 𝝂 remained effectively unchanged within the measurement uncertainty, well below 1%. Additional results for higher dpa PCW samples will also be presented.

Work is supported by US-DOE-FES No. DE-SC0022528.

[1] Käding et al., Applied Physics A 61 (1995) 253–261
[2] F. Hoffmann et al., Sci. Rep. 5 (2015) 1-7
[3] Dennett et al., Appl. Phys. Lett. 110 (2017) 211106
[4] Reza et al., Acta Mater. 232 (2022) 117926
[5] Dellis et al., J. Nucl. Mater. 551 (2021) 152985
[6] Gibson et al., Phys. Scr. T159 (2014) 014056

Speaker: Michael Simmonds (University of California San Diego)

29 1.019 Angular distributions of sputtered particles: influence of surface topography and texture

In fusion devices, the plasma-facing wall is primarily eroded via physical sputtering, which introduces impurities into the plasma core that degrade device performance. Aiming to gain a fundamental understanding of the sputtering processes relevant for plasma-wall interactions with machined plasma-facing components, we investigate angular distributions of sputtered particles in a controlled laboratory environment. Already in 1955 G. K. Wehner [1] discovered that in single crystals, sputtered particles are predominantly ejected along certain symmetry axis. However, plasma-facing materials typically have more complex structural compositions and surface topologies. Surface roughness is also known to significantly influence sputtering; recently, the first moment of the surface inclination angle distribution was found to be a strong predictive parameter [2, 3].

A rough tungsten model system served as a substitute for industrially machined tungsten first-wall components used in modern fusion devices, such as WEST, ASDEX-Upgrade or ITER. To achieve a specific surface roughness, a silicon substrate was etched under controlled conditions, resulting in pyramid-like surface structures on the micrometer scale. A thin tungsten coating was grown onto it by magnetron sputter deposition, preserving the pyramidal substrate topology. Roughness parameters and topography were quantified and investigated via atomic force microscopy (AFM). The sputtering behavior of one flat and two rough samples was studied, their root mean square roughness values being approximately 15 nm, 690 nm and 1.2 µm, respectively. Irradiation was performed with with a low flux ($\Gamma\approx10^{16}Ar^{+}m^{-2}s^{-1}$), energy selected 2 keV argon ion beam. For the angular resolved measurement of the sputtered particles, a precise quartz crystal microbalance (QCM) was used. Mounted on a 5-axis manipulator, the QCM enabled measurement of angular-resolved mass gain rates, corresponding to the local intensity of sputtered particle flux relative to the sample surface normal.

Simulations were carried out using the in-house developed “SPuttering simulation via RAYtracing of particles” – SPRAY code. It utilizes a data repository of sputtering data generated by state of the art binary collision approximation codes, such as SDTrimSP-7, and arbitrary input geometries, including AFM image data. First comparisons between experiment and simulation show that only accounting for the surface roughness in modeling is not sufficient, implying that further characteristics, such as e.g. crystalline texture, have to be considered to accurately model and understand the observed sputtering process.

[1] G.K. Wehner, Journal of Applied Physics, 26 1056 (1955).
[2] C. Cupak et al., Applied Surface Science, 570 (2021)
[3] A. Uccello et al., Nuclear Fusion, 65 (2025)

Speaker: Benjamin Burazor Domazet (TU Wien)

30 1.020 First mirror cleaning and impedance characterization for the ITER Core Plasma Thomson Scattering diagnostic

The Core Plasma Thomson Scattering (CPTS) diagnostic system is designed to precisely measure the electron temperature and density in the ITER core plasma region. During the fusion plasma operation, the metallic first mirror in the CPTS first mirror unit (FMU) is expected to degrade gradually due to plasma exposure, which requires radio-frequency (RF) plasma cleaning for the periodic CPTS mirror recovery. In addition, the water cooling for the FMU inevitably leads to the DC grounding of the mirror, which brings new challenges to the RF cleaning system design. To finally realize efficient contamination removal and to avoid damage to mirror materials and RF components, it is of great necessity to test the plasma cleaning for the FMU to characterize the load impedance as an input parameter to further improve the FMU design and circuit matching.

Argon plasma cleaning tests in the FMU mock-up were first conducted using mirror coupons (coated with an Al2O3 layer) at 13.56, 40.68 and 60 MHz with selected powers and pressures. The surface composition and reflectivity of the coupons before and after cleaning were analysed by X-ray photoelectron spectroscopy and a spectrophotometer. The cleaning result showed that complete contamination removal could be performed at 13.56 MHz for both configurations with and without DC grounding, which was not the case for higher frequencies due to reduced ion energies in this mock-up. Unlike the cleaning without DC grounding, structural material was deposited on both mirrors in the mock-up after cleaning with DC grounding, due to enhanced wall sputtering and weaker mirror cleaning. Furthermore, for all the tested frequencies, the whole RF circuit experienced substantial heating due to the highly power reflective load as confirmed by the standing wave voltage ratio measurement. Finally, the dependencies of the load impedance on power, pressure and frequency were characterized using three different devices: the offline substitution method (i) by a VNA, or the real-time measurement (ii) by an Octiv V/I probe or (iii) a dual directional coupler. This measurement allowed circuit modelling for power distribution analysis and the CPTS FMU pre-matching element design in the future.

The work leading to this publication has been funded by Fusion for Energy under the contract F4E-OMF-0847-01-01. This publication reflects the views of only the authors, and Fusion for Energy cannot be held responsible for any use which may be made with the information contained therein.

Speaker: Youpeng Wang (University of Basel)

31 1.021 Application of the linear plasma device PSI-2 as benchmark of hydrogen plasma chemistry

Hydrogen plasma chemistry plays crucial role for both power and particle exhaust in low temperature detached divertor plasmas as it can give rise to molecular assisted recombination via $H_2^+$, $H^-$, $H^+_3$. In the present contribution the application of the linear plasma device PSI-2 for validation of the corresponding plasma-chemical models will be discussed.

The effect on which the comparison of experiment with model calculations is based is the generation of excited hydrogen atoms in the processes involving molecules, molecular and negative ions. To make use of this effect as model benchmark the absolutely calibrated spatially resolved imaging measurements of Balmer radiation have been performed. In addition, the shapes of hydrogen lines taken with high resolution spectrometer are analysed. These latter help to isolate the contribution of the excitation channels of ionic (rotating with plasma) origin and deduce the translational temperature of emitting particles. Radial profiles of plasma parameters measured by movable Langmuir probe and pressure gauge measurements serve as input for modelling.

The measurements have been compared with local 0D calculations based on the standard model of $H^+_2$ plasma-chemistry applied in tokamak divertor simulations [1, 2, 3, 4]. Preliminary, this model was found to be sufficient to describe the pure hydrogen dischargesat low gas pressure 0.04 Pa, with electron temperature $T_e$ = 2...10 eV, density $n_e\sim$5$\cdot$10$^{17}$ m$^{-3}$. The calculations suggest that in this regime $H^+_2$ channel is likely to be the dominant source of Balmer radiation. At the same time, the model [1, 2, 3] was clearly seen to be insufficient to describe the discharges with pressure elevated up to 0.5 Pa and with lower $T_e$. Current status of comparison of the experiment with plasma-chemical models will be presented.

References
[1] Sawada K, Eriguchi K and Fujimoto T 1993 J. App. Phys. 73 8122
[2] Sawada K and Fujimoto T 1995 J. App. Phys. 78 2913
[3] Reiter D, May C, Baelmans M and Börner P 1997 J. Nuclear Mater. 241-243 342
[4] Kukushkin A S, Pacher H D, Kotov V, Reiter D, Coster D P and Pacher G W 2005 Nuclear Fusion 45 608

Speaker: V Kotov

32 1.022 Vision transformer based model regression for plasma exposed surface structures

Exposing a surface to an ion beam or hot plasma leads to erosion and the development of surface structures on the nanoscale. Such nanostructures have been observed on tungsten samples exposed to plasma in the PSI-2 linear plasma device and the LHD stellarator. Existing studies show that these nanostructures can have an influence on the erosion process of plasma facing components (PFCs). Better understanding those can lead to improved erosion models and potentially refine simulations modelling the plasma wall interaction by including the effects of surface morphology on sputter yields and emission distributions. This could lead to more accurate predictions for the lifetimes of PFCs.

Simulating the evolution of surface structures during ion beam impact quite often suffers from large computational effort. A more convenient description of the evolution of these structures is possible using a Kuramoto-Sivashinsky (KS) type model whose parameters we aim to infer for given experimental data. For real world data, only a single surface profile is available for this time dependent chaotic process, because the surface’s height profile h(x,y) cannot be obtained in-situ, especially for experiments in a magnetic confinement device. There exist some previous approaches to this problem, for example training a regression model on the Fourier transform of the surfaces, or using large pretrained convolutional neural networks, finetuned on the regression task. We propose a different approach using the vision-transformer architecture and including additional physically informed input features to the training process. We show that training such a model on our KS-dataset leads to good predictive performance on unseen test data across different parameter regimes. Furthermore, we investigate the embedding the model creates for the profiles, by visualizing the so-called class-token of the transformer. We will present details of the method, parameter studies and results on our synthetic dataset.

The results show the capability of our architecture to understand and extract information from fusion relevant surface structures. This serves as a starting point for creating models that do not only predict analytical surface models, but that can be used as surrogates for simulations that create important input quantities for PWI codes. Our framework can be applied to this task by changing the target variables while keeping the architecture we have developed, as we have shown that it can learn surface structure dependent processes.

This work is part of the project FusKI, funded by the BMFTR under grant no. 13F1012C.

Speaker: Torben Schmitz (ZJFJ)

33 1.023 an experimental study of interaction between plasma and flowing Sn free-suface

The application of flowing liquid metal for plasma-facing components (PFCs) in fusion reactors has attracted interest due to its potential heat removal capabilities, self-healing surface, and the expectation of improved plasma confinement. However, research on the interaction between liquid metal and plasma in well-controlled laboratory-scale experiments is not sufficient to understand the interaction comprehensively. To assess the feasibility of its application for PFCs, understanding of some key factors are needed, such as droplet ejection mechanisms and their suppression, the influence of local recycling rates on plasma, liquid metal related MHD phenomena, differences between flowing and steady liquid metal and so on.
In this study, our new experimental device with linear plasma and liquid Sn, along with the preliminary results, will be presented. A liquid metal free-surface flow system was integrated into the downstream end of the linear plasma device TPD-II to study plasma-liquid Sn free-surface interaction and droplet ejection. Liquid Sn is injected from the top of the vacuum chamber, forming a thin vertical free-surface flow with a width of about 50 mm. The plasma is incident perpendicular to the Sn free-surface flow. In the experiments, the Sn free-surface flow was exposed to cylindrical pure He plasma (n_e ~ 1 x 10¹⁸ /m³, T_e ~ 5 eV). Plasma at about 12 mm upstream from the exposure surface were measured by a scanning Langmuir probe.
Spectroscopic observation showed no significant peaks from Sn in the range of 300–600 nm wavelength. However, apparent peaks from Zn were observed. The origin of the Zn is likely an impurity in the Sn (99.9 % purity) in the system. When the Sn inlet was stopped while maintaining the He plasma, numerous bubbles of visible size (< 1 mm) were observed on the plasma-exposed surface (the surface of adhered molten Sn) after several tens of seconds. Many adhered Sn droplets were also observed on the stainless steel thin plate installed in the exposure chamber.　Even when bubbles were observed, no peaks from Sn were detected, indicating that our measurement was insufficient to capture the release of Sn droplets. The fact that there were no observable bubbles during Sn flow indicates suppression of foaming and/or growth of bubbles. However, further research is needed to understand the conditions for the bubbling mechanism and to confirm the suppression of bubbling.

Speaker: yukinori hamaji (national institute for fusion science)

34 1.024 Transport of light impurities in Wendelstein 7-X

Since the operational period OP 2.1 the stellarator Wendelstein 7-X (W7 X) operates with a CFC divertor, graphite tiles, and a mostly stainless-steel outer wall. Boronizations are used for wall conditioning in order to reduce the amounts of oxygen and water.
The divertor strike line is a net erosion area [1, 2]; the amount of eroded carbon depends on the concentration of oxygen in the plasma. Some amount of carbon eroded at the strike line is redeposited in the divertor in the vicinity of the strike line, but this redeposited amount is considerably smaller than the eroded amount. The inner wall shows either erosion or deposition, but both effects are relatively small. Erosion at the position of the outer wall was studied by exposing samples to individual discharges using the multi-purpose manipulator (MPM). Erosion was observed, but was also small.
Carbon containing redeposits are observed on a number of first wall components. These were analysed using elastic backscattering spectrometry (EBS) with incident protons at MeV energies, with scanning electron microscopy (SEM), and with focused ion beam cross sectioning (FIB) after the operational periods OP 2.1 and OP 2.3. Thick deposits were found on baffle tiles and adjacent inner wall tiles, the thickest observed deposits exceeded thicknesses of 10 µm. Moderately thick deposits were found on toroidal closure tiles. The surface of deposits can be smooth, but can also show conical structures parallel to the surface. Deposits consist mostly of carbon with 20-40 at.% of hydrogen, 10-20 at.% oxygen, and < 10 at.% of boron. Some deposits flake off. The rupture line can be between deposit and substrate, but can also be inside the layered structure of the deposit. The appearance of flaking could not be directly connected to layer thickness or layer composition.
The erosion rate of boronization layers was considerable higher by about one order of magnitude than that of hydrocarbon layers at the position of the MPM [3]. The expected lifetime of boronization layers on surfaces close to the plasma is small, probably only 1-2 days of plasma operation. However, the effect of boronizations lasts much longer than this expected lifetime. This means, that the gettering effect of boronization layers lasts much longer than the initial lifetime of these layers.
[1] M. Mayer et al., Phys. Scr. T171 (2020) 014035
[2] M. Mayer et al., Nucl. Fusion 62 (2022) 126049
[3] F. Maragkos et al., this conference

Speaker: Matej Mayer (MPPL)

35 1.025 ERO2.0 simulation of the global material erosion and redeposition with multi-species on EAST

Global material erosion, impurity transport, and redeposition processes involving carbon (C), tungsten (W), and molybdenum (Mo) in EAST tokamak have been systematically investigated using the 3D Monte-Carlo code ERO2.0 [1], with the background plasma generated by the SOLPS-ITER [2] wide grid version. The background plasma extended to the first wall is mapped onto the toroidal wall mesh in ERO. Through a stepwise modeling approach, this study explores the erosion and deposition of plasma-facing components (PFCs) and key multi-physics governing impurity dynamics. Initial modeling focuses on a segment simulation for the key PFCs with first wall and upper/lower divertor to establish basic erosion/transport characteristics. Key findings reveal that the eroded C, W, and Mo exhibit distinct transport and redistribution characteristics which are governed by the material mixing effects. Notably, due to the high energy threshold of W and Mo sputtering by deuterium, C impurities eroded from the high field side first wall act as the primary factor for Mo and W erosion. Net erosion of W main exists at both the inner and outer divertor targets, near the strike points, with approximately 90% of the eroded material undergoing short-range redeposition close to its source. In contrast, Mo eroded from the first wall travels farther in the edge plasma with a lower redeposition rate. Despite the first wall being closer to the main plasma than the divertor, the impurity density of eroded Mo is lower than that of W, due to the higher loss rates directed toward the divertor. Then, the simulation model is extended to include different low field side components such as the main limiter and guard limiters of the heating systems with a full toroidal simulation. The impact of varying magnetic connection lengths for different limiters on the impurities transport is quantified by the simulations. Finally, the migration and redeposition of boron (B) have been investigated to clarify its modification of boron layer distribution and subsequent impacts on the erosion behavior of the underlying PFCs materials. The findings quantify key inter-dependences between plasma-material interactions, magnetic geometry, and surface layer evolution, providing a better understanding of wall erosion in the presence of multiple impurities on EAST.
[1] J. Romazanov, et al., Nucl. Fusion 64 (2024) 086016
[2] I. Senichenkov, et al., Contributions to Plasma Physics 64.7-8 (2024) e202300136

Speaker: Hai Xie (Institute of Plasma Physics, Chinese Academy of Sciences)

36 1.026 Testing low-Z PFC slag removal with strike point sweeping on DIII-D tokamak

Build-up of plasma facing component (PFC) debris, a.k.a. “PFC slag”, is a potentially serious concern for the next generation of the magnetic fusion devices. A PFC slag management experiment on DIII-D tokamak tested the ability of strike point sweeps at cleaning low-Z slag from the outer divertor shelf. Layers of enriched boron isotope B10 were deposited on the outer target with the impurity powder dropper (IPD) to simulate low-Z slag accumulation. The strike point was swept over the layers to redeposit B10 onto the Divertor Material Evaluation System (DiMES) head [1] to study the controlled movement of material with sweeping. Initial visual inspection and scanning electron microscopy analysis of the DiMES heads indicate qualitative differences in deposition patterns before and after sweeping as well as the presence of B dust after the sweeps. Isotopic analysis with laser ablation mass spectroscopy (LAMS) will distinguish deposited B10 from the background B content.
Since DIII-D is a short-pulsed device and thick debris layers do not develop naturally, low-Z B slag can be simulated with high powder injection rates from the IPD. During these injections, the B10 powder was observed to agglomerate, leading to increased delivered quantities. Plasma disruptions further contributed to powder deposition on the divertor shelf. Spectroscopy data still indicates over a 3x increase in B erosion on the outer shelf after the deposition discharges, while fast camera imaging suggests most of the additional loose powder had ablated before the beginning of the sweeping stage. The innovative “clam shell” DiMES design utilized then enables direct comparison of the B deposition on DiMES before, during, and after this stage [1].
Strike point sweeping changes areas of net erosion and deposition at the targets, potentially responsible for the successful slag management during the JET DTE1 campaign [2] compared to the WEST high fluence campaign [3]. This experiment investigates this concept further and whether the sweeps erode the material atomistically or as dust. The initial findings suggest some combination of both.
[1] D. L. Rudakov et al., Fus. Eng. Design 124 (2017)
[2] J. P. Coad et al., Fus. Eng. Design 138 (2019)
[3] J. Gaspar et al., Nucl. Mater. Energy 41 (2024)
Work supported by US DOE under: DE-SC0023378, DE-FC02-04ER54698, DE-AC05-00OR22725, DE-AC02-09CH11466

Speaker: Jeremy Mateja (University of Tennessee-Knoxville)

37 1.027 Sputtering behavior of tungsten-boron mixed layers

Plasma-facing components in fusion reactors are exposed to extreme conditions including high heat flux and bombardment by energetic particles. Tungsten, the chosen first wall material in ITER, offers favorable properties such as a high melting point and low sputter yield. To enhance plasma performance and minimize impurity influx such as oxygen, wall conditioning techniques like boronization are applied [1]. Boronization deposits a thin boron layer on the reactor wall, which modifies surface properties and its erosion behavior.

This laboratory study investigates the effect of varying boron concentrations on the sputtering behavior of tungsten-boron (W-B) thin films under ion bombardment. Five samples, ranging from pure tungsten via 3 different W:B mixtures to pure boron, were irradiated with $2\,\mathrm{keV}\,\mathrm{Ar}^+$ and subsequently $2\,\mathrm{keV}\,\mathrm{D}_2^+$ ions to simulate fusion plasma conditions. Argon serves as a seeding gas, while deuterium is a main component of the fusion fuel cycle. The elemental compositions of the samples were characterized ex situ by time-of-flight elastic recoil detection analysis (ToF-ERDA) before and after irradiation [2]. Additional high-resolution Rutherford backscattering spectrometry (HR-RBS) was performed after irradiation.

Sputter yields were quantified using a quartz crystal microbalance (QCM), capable of detecting mass changes down to $10\,\mathrm{pg\,cm^{-2}\,s^{-1}}$ through eigenfrequency shifts of the oscillating quartz [3]. The relationship between frequency and mass change is described by the Sauerbrey equation [4].

The results revealed a clear trend: the total eroded mass decreased with increasing boron content under both $\mathrm{Ar}^+$ and $\mathrm{D}_2^+$ irradiation with the effect more pronounced for $2\,\mathrm{keV}\,\mathrm{Ar}^+$ irradiation.

The measured mass loss was converted into atoms per incident ion using the known film composition from ToF-ERDA and assuming steady state conditions. Our data show that tungsten sputtering decreased and boron sputtering increased with higher boron concentrations.

These findings highlight the beneficial impact of boron enrichment on reducing high-Z impurity sources in fusion environments, supporting boronization as an effective wall conditioning technique.

[1] J. Winter $\textit{et al. J. Nucl. Mater.}$ $\textbf{162}$ 713 (1989).

[2] E. Pitthan $\textit{et al. Surf. Coat. Technol.}$ $\textbf{417}$ 127188 (2021)

[3] R. Stadlmayer $\textit{et al. Rev. Sci. Instrum.}$ $\textbf{91}$ 125104 (2020).

[4] G. Sauerbrey $\textit{Z. Phys.}$ $\textbf{155}$ 206 (1959).

Speaker: Raphael Gurschl (TU Wien)

38 1.028 Overview of runaway electron (RE) interaction with the beryllium inner limiters in JET

Runaway electrons (REs) generated during plasma disruptions in tokamak reactors present a challenge due to their capacity to induce severe damage to the plasma facing components (PFCs) [1]. Investigations on the Joint European Torus (JET), operating with the metallic wall (formerly known as ITER Like Wall – ILW) [2] between 2010 and 2023 demonstrated the damaging effect of REs on beryllium (Be) main chamber limiters [3], raising concerns about reactor integrity and operational reliability. Despite the high risk of these events damaging the PFCs, a limited number of experiments have been carried out and results reported.
In the final operational phase of JET, during December 2023, additional experimental sessions were conducted with the focus on RE production, mitigation, impact on PFCs and implications for reactor operations. For this purpose, REs were deliberately triggered using argon (Ar) massive gas injection (MGI), enabling controlled studies of their triggering mechanisms and impact [4]. In the following intervention, high resolution photography of in-vessel components revealed a new toroidal distribution map of the damage induced to the Be inner limiters by these events. Infrared (IR) wide-angle camera assessments, combined with temperature data from thermocouples embedded in Be limiters, revealed that two inner limiters, located approximately 120° apart in the toroidal direction, were impacted by the REs with surface damages in the range of 30 to 40 cm2. Moreover, the analysis showed that for one limiter the damage was primarily from a single RE event, whereas the other showed cumulative damage, resulting from multiple RE impacts. In-vessel Laser Induced Breakdown Spectroscopy (LIBS) [5] with an average spot size of 150 to 200 µm in diameter was performed on both limiters to assess fuel retention and material composition in the damaged and non-damaged areas. During the JET 2024 intervention, damaged tiles from these two limiters were removed for detailed post-mortem analysis. Consequently, a comparative analysis of single event and multi-event RE damage on the Be limiters will be presented, highlighting the morphological and structural properties modifications, material property changes, molten Be behaviour dynamics and the implication to the fuel retention and release.

[1] S. Ratynskaia et al, Plasma Phys. Control. Fusion (2025) DOI 10.1088/1361-6587/ae1c6c
[2] G. F. Matthews et al, Phys. Scr. (2007) T128 137
[3] I. Jepu et al, Nucl. Fusion 64 (2024) 106047 (15pp)
[4] C. Reux et al, Nucl. Fusion 55 (2015) 093013
[5] J. Likonen et al, Nucl. Mater Energy 45 (2025) 102021

Speaker: Dr Ionut Jepu (UKAEA, Culham Campus, Abingdon, OX14 3DB, UK)

39 1.029 Physical and chemical erosion of thin boron and amorphous boron-hydride films investigated at the linear plasma device PSI-2

Re-baselining ITER to have a full tungsten wall eliminates the strong impurity gettering capabilities of beryllium. Consequently, the intrinsic oxygen level makes it challenging to start-up the plasma in limiter configuration and achieve high-performance operating conditions in divertor configuration. The selected procedure in ITER to getter oxygen is boronization [1] by depositing a thin amorphous BD film (a-B:D, thickness ~100 nm) via glow discharge on the wall [2]. This film rapidly erodes during nominal plasma operation in high-flux areas. The oxygen gettering lasts longer due to coverage of the first wall and recessed areas. Thus, the layer lifetime determines how often boronization is required. Codes simulating the plasma-wall interaction, like ERO 2.0, provide predictions for this lifetime. In this case, physical and chemical erosion primarily determine the lifetime, and the absolute yields are not yet known precisely. This necessitates conducting experiments to measure the absolute erosion rates and benchmark the ERO 2.0 code for predicting the lifetime of thin boron films.
This contribution presents an analysis of thin boron film erosion experiments conducted at the linear plasma device PSI-2. Magnetron sputtering created 100 nm thick boron and a-B:D films covering polished tungsten substrates, simulating a boronized wall in a fusion reactor. Deuterium discharges at PSI-2 provide the flux of low-energy deuterons onto the films for studying their near-threshold erosion behavior. The biasing of the sample enables an impact energy scan from 40 eV to 100 eV. Additionally, limiting the fluence of the deuterons to 3∙10^23 m^−2 avoided the complete removal of the thin films, allowing for post-mortem layer thickness measurements. Comparing the thickness before and after plasma exposure gives the net erosion rates of the boron layers [3].
Spatially resolved emission spectroscopy provides information on gross erosion by measuring the boron 2p-3s transition (249.8 nm). The spatial distribution of the emission, i.e., the decay length, is compared to the emission profile predicted by ERO 2.0 and used to benchmark the code. Finally, varying the surface temperature of a bulk boron sample enables an investigation into the temperature dependency of chemical erosion by measuring the BD A-X transition around 433 nm. The PSI-2 results, in the impact energy range covered, suggest that the chemical erosion is small compared to physical sputtering.
[1] J Winter et al J. Nucl. Mater. 162-164 713 (1989)
[2] R Pitts et al NME 42 101854 (2025)
[3] M Sackers et al NME 45 102003 (2025)

Speaker: Marc Sackers (Forschungszentrum Jülich GmbH, Institute of Fusion Energy and Nuclear Waste Management - Plasma Physics, 52425 Jülich, Germany)

40 1.030 Sputtering Behavior of He–W Co-Deposited Layers: Experiments and Modeling

Helium(He)–tungsten(W) co-deposited layers are expected to form in future magnetic fusion devices as a consequence of simultaneous sputtering and re-deposition of tungsten (W) under mixed hydrogen and He plasma exposure. Understanding their sputtering behavior is essential for predicting long-term material erosion and impurity sources in ITER and DEMO. In this study, He–W co-deposited layers were synthesized in the linear plasma device Co-NAGDIS, and their microstructure and sputtering characteristics were comprehensively investigated. Binary collision approximation simulations were additionally performed to elucidate the microscopic origins of their reduced sputtering yields.
SEM and TEM analyses revealed that the co-deposited layers consist of nanocrystalline W aggregates containing numerous sub-100-nm cavities, with porosities ranging from 30% to 55%. Unlike the well-defined helium bubble structures typically formed in bulk W, these cavities exhibited indistinct boundaries and were uniformly distributed throughout the layer. X-ray diffraction confirmed that the layers retained crystallinity, while peak broadening indicated the presence of nanoscale crystalline domains. Optical emission spectroscopy during Ar ion bombardment showed that the apparent sputtering yield, inferred from the W I / Ar II line ratio, was suppressed to approximately 30% of the bulk W value at the initial stage of irradiation. As the ion fluence increased, the yield gradually recovered and ultimately reached the bulk value once the co-deposited layer was completely removed. The fluence required for full recovery scaled with the layer thickness, demonstrating that the suppressed sputtering yield is an intrinsic property of the co-deposit itself rather than an artifact of transient surface effects.
To interpret these experimental observations, simulations were performed for He–W mixed structures with varying porosities and defect densities. The initial atomic structures were generated through structural optimization using an Embedded-Atom Method (EAM) potential. The simulations revealed that the nanoporous morphology and complex surface topography characteristic of the co-deposited layers strongly dissipate the energy of collision cascades, thereby reducing the effective momentum transfer to surface atoms. The roughened surface was also found to enhance the re-trapping of sputtered W atoms, providing a mechanism consistent with the experimentally observed suppression of net erosion. Moreover, as the co-deposited layer becomes thinner, sputtering induced by backscattered Ar atoms originating from the underlying bulk material becomes increasingly significant.

Speaker: Noriyasu Ohno (Nagoya University)

41 1.031 Impact of energetic fusion products on the first wall of a GenF’s ICF reactor

GenF, a spin-off of the Thales Group, is developing an accelerated pathway toward a direct-drive inertial confinement fusion (ICF) reactor. As part of the TARANIS project, GenF collaborates with French research institutions including CEA and CNRS to address the critical physics problems in ICF physics.

One of the most demanding challenges is the design of the ICF reaction chamber, particularly the first wall (FW) components. The FW material is subjected to intense, pulsed emissions of X-rays, ions, and neutrons resulting from fusion reactions. Following ignition, the energy is distributed among burned and unburned ions from the fuel target— helium (He), deuterium (D), tritium (T), and elements from the target's outer layers such as hydrogen (H) and carbon (C).

This study presents simulations of physical sputtering, bulk damage, and compositional changes in various candidate FW materials induced by energetic ions, taking into account their energy distribution and spectra based on a typical target design proposed by CEA for the TARANIS project. Sputtering, composition change and damage production in the FW is evaluated for tungsten (W) and stainless steel (SS) using SDTrimSP code allowing for the first predictions of the lifetime of the FW.

Coupled numerical simulations using the FESTIM code [2] enabled evaluation of T accumulation and diffusion in W under ICF-relevant thermal conditions, considering the relevant material damage. This work forms part of GenF’s broader initiative to numerically model the FW materials performance under extreme pulsed thermal and particle loads, ensuring sustained performance of the reaction chamber throughout the reactor’s operational life.

References
[1] A.Mutzke et al., SDTrimSP Version 7.00, IPP Report 2024-06
[2] R. Delaporte-Mathurin et al., FESTIM: An open-source code for hydrogen transport simulations, International Journal of Hydrogen Energy, 63 (2024)

Speaker: Mykola Ialovega (GenF)

42 1.032 Experimental analysis of Pt marker erosion in the outer divertor region of ASDEX Upgrade during L- and H-mode discharges

Net erosion of tungsten (W) in the divertor region has been investigated by exposing small platinum (Pt) marker samples, as well as bulk molybdenum (Mo) and W samples, to a series of plasma discharges in the outer strike point (OSP) region during dedicated L- and H-mode experiments in ASDEX Upgrade (AUG). Pt was chosen as a proxy for W due to its comparable erosion characteristics, addressing the previously assessed limitations [1] of Au in earlier experiments [2]. Net erosion of the Pt marker samples was quantified using Rutherford Backscattering Spectrometry (RBS), while the material composition and spectral peak identifications were performed using Nuclear Reaction Analysis (NRA) for low-Z elements and Particle-Induced X-ray Emission (PIXE) for high-Z elements. Measurements were performed on both the sample surfaces and the regions between the Pt markers, allowing characterisation of poloidal variations in erosion/deposition patterns on the markers and deposition of W and other materials between the marker areas. PIXE and NRA analyses were performed at the same points, providing a detailed picture of the material composition along the poloidal coordinate with additional microbeam measurements and microscopy studies to characterise the shape of the re-deposition patterns with high spatial resolution. On the bulk Mo and W samples, Focused Ion Beam (FIB) cross-sectional analysis was performed to investigate net erosion rate as well as the structure and thickness of deposited layers.

Analysis of RBS data shows pronounced net erosion at the OSP, gradually decreasing toward the far scrape-off layer. In H-mode, measurements indicate nearly an order of magnitude stronger net erosion compared to L-mode, potentially reflecting the contribution of ELM-induced erosion. Both experiments reveal, alongside Pt re-deposition, accumulation of W and low-Z elements on the surfaces near the OSP. Additionally, indications of local deposition were observed directly adjacent to the Pt markers near the OSP. If confirmed as deposited Pt, these observations would represent clear experimental evidence of prompt re-deposition in marker erosion experiments, extending earlier studies based on campaign-integrated W marker erosion [3]. This offers valuable insight into material deposition processes in the OSP area and the transport of W and other species in plasma discharges.

[1] S. Saari et al., Nuclear Materials and Energy 45 (2025) 102032
[2] A. Hakola et al., Nuclear Fusion 61 (2021) 116006
[3] D. Naujoks et al., Journal of Nuclear Materials 210 (1–2) (1994) 43–50

Speaker: Samuli Saari (VTT)

43 1.033 Modelling divertor erosion and tungsten core contamination in DTT negative triangularity scenario

The Divertor Tokamak Test (DTT) facility, currently under design and construction at the ENEA Research Centre in Frascati (Italy), aims to assess alternative solutions to the heat and power exhaust challenge in future fusion reactors [1]. A key issue for its operation is plasma-wall interaction (PWI) [2], as material erosion can limit component lifetime while eroded impurities may migrate into the core plasma, causing dilution and radiative cooling. These effects must be carefully monitored and controlled across all DTT-envisioned scenarios, including negative triangularity (NT), which is a promising solution for achieving high confinement performances while working in a regime free from Edge Localised Modes (ELMs).
This work investigates tungsten divertor erosion and core plasma contamination in the DTT NT scenario using ERO2.0, a 3D Monte Carlo impurity transport code [3]. The deuterium background plasma is taken from SOLPS-ITER simulations, together with neon ion fluxes used for detachment control. Oxygen is added in ERO2.0 as a proxy for unseeded light impurities. Local simulations including the full 3D divertor geometry are also performed to identify optimal locations for PWI diagnostics, such as quartz crystal microbalances (QMBs). The results are compared with those obtained for the corresponding positive triangularity (PT) configuration.
In the absence of ELMs, Ne and O impurities dominate tungsten erosion. The influence of the ion incidence angle on the estimated erosion rate is assessed over the 0-70° range. The outer divertor leg operates under significantly hotter conditions than the inner leg, leading to a higher net erosion rate of approximately 0.2-0.4 nm/s near the strike point, substantially larger than in the corresponding PT case. Net deposition regions are observed in proximity to the erosion peak at both strike points. The sensitivity of erosion and migration estimates to the extrapolation of SOLPS-ITER plasma parameters from the computational grid to the material surface is also evaluated. From a migration perspective, the outer target is identified as the main contributor to core contamination, with the private flux region near the outer strike point being the most critical zone for tungsten influx due to weaker screening and relatively high local plasma parameters, in terms of ion flux and impact energy. Optimal QMB installation points in the dome-target gaps are assessed to ensure detectable W deposition.

[1] Romanelli F. et al., NF 64.11 (2024) 112015
[2] Roth J. et al., JNM 390–391 (2009) 1–9
[3] Romazanov J. et al., PS T170 (2017) 014018

Speaker: Gabriele Alberti (Politecnico di Milano)

44 1.034 Surface morphology damage of rolled W-K plates under different high heat load

In fusion reactors, the divertor fulfills two core functions: first, expelling impurities produced by plasma-first wall interactions and helium (a fusion product); second, withstanding high heat flux from the plasma and dissipating plasma energy out of the tokamak device. During operation, the divertor endures extreme thermal loads. As one of the most promising candidate materials for future divertors, tungsten (W) faces considerable challenges. Conventional powder metallurgy-derived tungsten suffers from low recrystallization temperature and high ductile-to-brittle transition temperature (DBTT), failing to meet the requirements for divertor base materials. In contrast, the novel rolled W-K alloy effectively increases recrystallization temperature, enhances mechanical properties, and reduces DBTT. Thermoplastic deformation further optimizes its mechanical performance and ductility while retaining tungsten’s inherent high thermal conductivity, endowing the W-K alloy with excellent comprehensive properties to cope with the harsh service environment of future fusion reactor divertor facing materials.
In this study, large-sized bulk W-K alloy was fabricated via hydrogen sintering, followed by hot rolling with a deformation degree of 60–80%. Tensile tests revealed that the alloy exhibited significant plastic deformation at 50°C, with tensile strain exceeding 5% and room-temperature tensile strength surpassing 1400 MPa. High-heat-flux test mock-ups of various sizes were prepared using the thermoplastically deformed W-K alloy, and tested under different loading conditions via the EMS-60 facility. Different tungsten blocks on the same mock-up underwent combined transient-steady-state tests with varying pulse durations and counts, as well as high-heat-flux tests simulating plasma disruption and vertical displacement events (VDEs). Subsequent systematic characterization of the test mock-ups’ surface morphology and crack depth showed that under combined transient-steady-state conditions, surface crack formation of W-K alloy samples was closely associated with the pulse count and duration of transient thermal loading. When the pulse duration was 0.1 ms, no obvious surface cracks were observed after 50,000 thermal shock cycles; however, when the pulse duration increased to 0.5 ms, a crack network formed on the surface after only 10,000 cycles. In tests simulating plasma disruption and VDEs, the area and morphology of the sample’s heat-affected zone changed significantly with increasing loading power. When obvious surface melting occurred, the W-K alloy samples exhibited distinct surface morphological differences compared to pure tungsten samples.​

Speaker: Fan Feng (Southwestern Institute of Physics)

45 1.035 Thermophysical characterization of surface deposits from the first high fluence campaign of WEST

During the WEST experimental campaigns, various evolutions of the plasma-facing components (PFCs) have been observed. One of these degradations is erosion of material and redeposition on other components. These deposits have unknown but probably degraded thermal properties (low diffusivity and/or low thermal contact with the PFC). Some of these deposits created Unidentified Flying Object (UFO) during the recent experimental campaigns. These UFO could cause the premature termination of the discharge.
The aim of the present work is to improve our understanding of the evolution of the thermal properties of PFCs during their use in the WEST Tokamak. To achieve this, an experimental bench has been developed to measure the thermal properties of ITER-grade components (thermal diffusivity, thermal conductivity, etc.). This method can be applied directly to a component (without cutting or modifying it) to study its ageing and characterization of the deposits will be part of these analyses. As they are thin (from µm to tens of µm) a front face flash method is developed with laser heat source (10ns), or flash lamp (80µs) to identify parameters by inversion of photothermal measurements.
The present work exposes the thermal properties of theses deposits. Due to significant variations of thickness and composition, three zones are chosen on PFCs exposed to plasmas during different campaigns. Namely, a zone of high thickness (several tens of microns), an intermediate deposit zone (a few microns) and an oxidation zone (hundreds of nanometres). The thermal model is taken in consideration the surface roughness and the composition of the deposit. The identified thermal properties offer the possibility to improve existing thermal model used to access the temperature of the top surface of the PFC, but also the gradient of temperature in the deposit thickness.

Speaker: clément Monet-Vidonne (Aix -Marseille Univ., CNRS, IUSTI)

46 1.037 Preliminary Analysis of Chromium erosion and redeposition in DIII-D

Experimental results evaluating chromium (Cr) as a candidate main-chamber plasma-facing material (PFM) for next-generation magnetic fusion devices are presented. Tungsten (W) and W-based alloys are currently the leading plasma facing materials (PFMs) candidates for future magnetic fusion devices such as SPARC, ITER and FPP/DEMO. W resistance to sputtering and high melting temperatures makes it suitable to withstand the extreme thermal and particle fluxes of fusion reactors. However, significant challenges remain due to the potential degradation of fusion performance caused by W contamination of the plasma core and the uncontrolled release of W dust originating from co-deposited layers, which in present devices are shown to trigger disruptions at a critical rate. These issues motivate the search for alternative PFMs that could mitigate such risks while maintaining acceptable thermal and mechanical performance. Chromium (Cr) emerges as a promising alternative to W for the main chamber first wall of fusion power plants. Cr offers several intrinsic advantages: it exhibits low neutron activation, acts as an effective oxygen getter that reduces the concentration of light impurities (and consequently physical erosion of plasma-facing surfaces), and features good radiation properties in the plasma core with Cr cooling factor being at least an order of magnitude lower than that of W for Te greater than 5 KeV.
In this work, we analyze experimental data on the erosion, redeposition, and surface evolution of Cr samples exposed in the lower divertor region of the DIII-D tokamak using the Divertor Material Evaluation System (DiMES) manipulator. Samples were exposed under H-mode ELMy conditions, reaching peak heat fluxes of 10–15 MW/m². Gross erosion was measured in situ using calibrated spectrometers. After exposure, visible evidence of Cr redeposition was observed on the DiMES surface, and post-mortem analysis was conducted via energy-dispersive X-ray spectroscopy (EDS) and Auger electron spectroscopy (AES). Experimentally determined erosion rates are compared with predictions from binary collision approximation (BCA) models (e.g., RustBCA) using characteristic DIII-D divertor plasma parameters as input. Redeposition is modeled using a local transport Monte Carlo (LPTMC) code, and a one-dimensional surface evolution model is applied to estimate long-term surface changes. This work contributes to assessing chromium’s viability as a next-generation plasma-facing material for magnetic fusion applications.

Speaker: Luca Cappelli (ORAU)

47 1.038 Molecular Dynamics Simulation of Sputtering Behavior and Morphology Evolution of Tungsten Surfaces under Neon Plasma Irradiation

Neon (Ne) seeding has emerged as a promising technique for radiative divertor operations in future fusion devices, such as ITER and DEMO, to effectively dissipate high heat fluxes. However, the interaction between Ne impurities and plasma-facing materials, particularly tungsten (W), introduces complex erosion and retention issues that directly impact the lifetime of divertor targets. While macroscopic erosion is well-monitored, the atomic-scale mechanisms initiating surface degradation remain difficult to observe experimentally. In this work, we employ large-scale molecular dynamics (MD) simulations to systematically investigate the sputtering characteristics and microstructural evolution of W surfaces under low-energy Ne irradiation, providing a critical comparison with relevant experimental observations.
The investigation is divided into two phases. Initially, we simulated single ion impact processes on perfect W(100), (110), and (111) crystal surfaces under a wide range of incident energies and angles. The simulations analyzed the effects of crystallographic orientation on sputtering yields, reflection coefficients, and the energy/angular distributions of sputtered atoms. The results reveal that the atomic packing density significantly modulates the penetration depth and energy deposition of incident ions, leading to orientation-dependent sputtering behaviors in the low-energy regime.
Furthermore, to explore the surface response under prolonged plasma exposure, cumulative Ne injection simulations were performed on W surfaces across an energy range of 30-200 eV. These simulations track the dynamic evolution of the surface from a pristine state to a heavily damaged, roughened state. We observed that the accumulation of sub-surface defects and the gradual saturation of trapped Ne atoms drive a transition in surface morphology. Crucially, the simulations demonstrate a feedback mechanism where the evolving surface roughness alters the local incidence angle of subsequent ions, thereby modifying the effective sputtering yield over time. Notably, the specific morphological features observed in the simulations show a strong qualitative correspondence with the micro-pleated structures found on W surfaces exposed to 100 eV Ne irradiation in linear plasma device experiments. This study not only elucidates the atomic-scale origins of Ne-induced damage but also provides essential source-term data for boundary plasma transport codes, offering a microscopic basis for interpreting the formation of macroscopic morphological features observed in fusion environments.

Speaker: Zhongshi YANG (GNOI)

48 1.039 Tungsten erosion and injection investigations in the stellarator Wendelstein 7-X: Results from OP2.1-2.3

Tungsten (W) has emerged as a favourable material for plasma-facing components (PFCs) in nuclear fusion devices. It has been incorporated in several tokamaks, however, in the stellarator, with 3D geometry, its suitability as PFCs, in terms of erosion, redeposition, ionization, transportation and accumulation of impurity particles in the plasma core, is yet to be demonstrated. In the stellarator Wendelstein 7-X (W7-X), tungsten PFCs are being introduced stepwise by increasing the surface area over the last few plasma campaigns [1]. During the recent OP2.2 + OP2.3 campaigns, tungsten PFCs surface areas of about 2 m2 in the heat shield (out of 47 m2) and about 0.8 m2 in the baffle (out of 33 m2), were provided.
Dedicated tungsten sputtering experiments were conducted using island control coils to drive higher heat/particle fluxes onto the tungsten baffle areas while simultaneously introducing impurity gas N2/Ne/Ar seeding. In addition, tungsten samples were exposed to the plasma edge island via a multi-purpose manipulator at defined exposure positions and durations. Tungsten tracer injection was carried out using the Tracer-Encapsulated Solid Pellets (TESPEL) and Laser blow-off (LBO) methods. Plasma discharge conditions were varied during two different magnetic configurations, i.e., standard and high mirror [2].
Observations were made at the edge with UV spectroscopy, Langmuir probes and thermal He-beam diagnostic and in the plasma core with high-resolution X-ray imaging spectroscopy (HR-XIS), a high-efficiency XUV overview spectrometer (HEXOS) and a pulse height analysis system (PHA). W I and W II line signals were detected by the UV edge spectrometers having a line of sight on the tungsten baffle tiles. Tungsten signals from erosion caused by plasma-wall interactions, TESPEL/LBO injection and intrinsic background were followed using multiple core viewing spectrometers and showed variations in the intensities when varying the island geometry and impurity seeding. The Langmuir probe measurements at the divertor surface have shown plasma temperatures of about 10 eV and asymmetry in the edge density for the upper and lower island divertors. A detailed analysis will be presented at the conference.

[1] D. Naujoks et al., Nuclear Material Energy 37 (2023) 101514.
[2] O. Grulke et al., IAEA-FEC (2025) Chengdu.

Speaker: Chandra Prakash Dhard (Max-Planck-Institut für Plasmaphysik, Greifswald, Germany)

49 1.040 Divertor Material Evaluation System (DiMES) at the DIII-D tokamak: 30 years of PMI research

Controlling the Plasma-Material Interactions (PMI) is one of the key issues to be resolved for the success of machines like ITER and SPARC and the following generation of the magnetic fusion devices. The Divertor Material Evaluation System (DiMES) was a workhorse of PMI research at the DIII-D tokamak for about three decades and still remains one of the leading material testing facilities in the world with exceptional diagnostic coverage.
The DiMES program was started in 1992 as a collaborative project between General Atomics, Sandia National Laboratories and Argonne National Laboratory [1]. The DiMES manipulator was fabricated and installed on DIII-D in 1995 [2]. Since then DiMES contributed to over 100 publications in scientific journals and numerous presentations at US and international conferences. DiMES allows inserting material samples up to ~5 cm in diameter in the lower outer divertor of DIII-D, typically flush with the divertor tile surface. Samples elevated about the surface and angled towards the incident plasma fluxes are used to achieve reactor-relevant heat fluxes of over 10 MW/m2. Since DIII-D has all-C (graphite) plasma facing components (PFCs), the initial DiMES studies concentrated on studies of C erosion/deposition [3]. Alternative PFC options were also investigated, e.g. experiments with Li were conducted in 2003. A sample heating capability was added in 2004, allowing studies of the temperature effects on C erosion/deposition and diagnostic mirror coating with C [4]. As the magnetic fusion community moved away from C PFCs and C was excluded from the ITER design, DiMES was used to test high-Z PFC materials including W, V, Mo, Ta and Zr, first in thin coatings [5], then in bulk, including the leading edge effects.
One of the main strengths of the DiMES program lies in a worldwide network of collaborating institutions. In recent years, DiMES has become the primary point for the industry engagement at DIII-D, fueled by a rapid growth in the fusion private industry community. Details of the industry collaborations will be covered in a companion presentation by F. Effenberg et al.

[1] C.P.C. Wong et al, J. Nucl. Mater. 196-198 (1992) 871
[2] C.P.C. Wong et al, J. Nucl. Mater. 258-263 (1998) 433
[3] D.G. Whyte et al., Nucl. Fusion, 41 (2001), 1243
[4] D.L. Rudakov et al., Phys. Scr. T128 (2007) 29
[5] D.L. Rudakov et al., Fusion Eng. Des. 124 (2017) 196

Work supported by the US DOE under DE-FC02-04ER54698

Speaker: D.L. Rudakov (University of California, San Diego)

50 1.042 Quantified depth profiling of deeply located deuterium with ps-laser-induced ablation quadrupole mass spectrometry

Picosecond-Laser-Induced Ablation Quadrupole Mass Spectrometry (ps-LIA-QMS) was employed to investigate deeply located deuterium retention in polycrystalline tungsten (W). As the foreseen primary material for fusion reactors, its high heat conductivity, low sputtering yield, and good fuel retention properties make it a promising choice as first wall material. However, high-energy particles, particularly ~14 MeV neutron fluxes, will induce defect generation that drastically enhances the saturation level of deuterium (D) and tritium (T) retention, affecting the fuel cycle. To address this challenge, ps-LIA-QMS has been developed, combining the strengths of quantification inherent to Laser-Induced Desorption Quadrupole Mass Spectrometry (LID-QMS) and depth-resolved measurements possible with Laser-Induced Breakdown Spectrometry (LIBS). Using ps-LIA-QMS enables depth profiling and quantification of retained fuel, with a compact setup which is supposed to be suitable for remote handling in future fusion reactors. Moreover, it allows simultaneous measurements with LIBS to build a complementary diagnostic setup.
This study addresses two major challenges in the application of ps-LIA-QMS: Understanding the QMS signal composition and possible probing depth. The remarkable depth profiling capabilities of QMS are showcased in 1.4 MeV, 0.05 dpa proton-irradiated and subsequently deuterium-decorated W samples, achieving a depth profile of over 15 μm. Quantification of the released deuterium yields a deuterium content of (0.7±0.2) at%, closely matching $^3$He Nuclear Reaction Analysis (3He NRA) results (1±0.15) at% within the first 4 µm depth. The QMS signal composition was analyzed for the proton-irradiated and for W$^{3+}$ self-damaged samples decorated with D.
Both profiles reveal a peak with exponential decrease at the surface, not observed in NRA measurements. Moreover, up to 50 at% of D are detected in the first 75 nm of the proton-irradiated sample. Both profiles could be attributed to strong thermal influences during the first laser pulses. Simulations of deuterium diffusion and trap annealing under transient heating in application to reference NRA measured D depth profiles qualitatively agree with ps-LIA-QMS results. These simulations suggest that the D signal is dominated by outgassing from regions below the ablated surface on time scales much larger than the laser pulse duration (up to ms).
In conclusion, this study marks a significant progress in the analysis of deuterium retention in tungsten. It showcases the unparalleled probing depth and sensitivity of ps-LIA-QMS and gives new insights about the signal creation during ps-LIA-QMS, making it a suitable candidate for fuel retention diagnostic in reactor-relevant conditions and studies of physical processes in damaged materials.

Speaker: Christoph Kawan (FZJ)

51 1.043 Thermal desorption and retention of deuterium in helium-implanted and displacement-damaged EUROFER97 at reactor-relevant temperatures

A critical challenge in the fusion materials science lies in understanding the influence of radiation damage on hydrogen isotopes (HIs) retention, particularly concerning safety and tritium self-sufficiency. Reduced activation ferritic/martensitic (RAFM) steels are leading candidates for use as structural material in the first wall such as the breeding blanket modules. In deuterium-tritium (D-T) plasmas, structural materials will be exposed to 14 MeV neutrons, leading to atomic displacements and helium (He) production as a result of nuclear reactions. The resulting defects and He bubbles act as trapping sites for HIs, significantly influencing their transport and retention behavior.
This study investigates deuterium (D) retention in displacement-damaged and He-implanted EUROFER97. The objective is to characterize He-related trap sites for D and study their thermal evolution. EUROFER97 samples are irradiated with 20 MeV tungsten ions at temperatures ranging from 290 to 770 K, inducing a primary peak damage dose of 0.6 dpa. Selected samples are additionally implanted by 0.5 MeV He ions to a peak concentration of up to 0.6 at.% to form He bubbles. These samples are then exposed to a low-temperature and low-energy D plasma or to D$_2$ gas to fill the radiation-induced defects with D. Nuclear Reaction Analysis ($^3$He-NRA) is applied to determine D depth profiles. Thermal Desorption Spectroscopy (TDS) is employed to investigate D desorption.
Transmission electron microscopy (TEM) confirmed the formation of nanoscale He bubbles, which were found to significantly enhance local D concentration. Although the density of He-related trapping sites decreases with increasing annealing temperature, they are not completely recovered even after annealing at 720 K, unlike displacement-damage without the presence of He that recovers at 620 K. Irradiation at reactor-operating temperatures leads to lower D retention compared to annealing at the same temperatures after irradiation at 290 K. D is de-trapped from He-related defects and desorbed from EUROFER97 at about 520 K. Notably, TDS spectra differ significantly between plasma-exposed and gas-exposed samples, with the latter showing two additional high-temperature desorption peaks.
Finally, rate-equation simulations employing the TESSIM-X code were conducted to determine trapping and de-trapping energies for both He-related and intrinsic defects, enabling the interpretation of the experimental TDS spectra. This work significantly improves the ability to predict the tritium loss in a DEMO first wall by providing experimentally determined defect parameters that account the impact of He-related traps. This enhances the accuracy of tritium inventory models under fusion-relevant irradiation and thermal cycling conditions.

Speaker: Andreas Theodorou (MPPL)

52 1.044 Beam fueling low recycling discharges with modest lithium impurity fraction in LTX-beta

We show that beam fueling of a low recycling discharge with low energy neutral beams is feasible with modest (~ 5%) lithium impurity concentrations. Low recycling discharges (R ~ 0.5) on LTX-beta have been documented with a near flat electron temperature profile. Edge temperature measurements in the SOL indicate a hot and sparse SOL with collisionality < 0.1 and values as low as 0.01. Low gradients in electron and ion temperature profiles are predicted to suppress temperature gradient driven instabilities and enhance confinement. High confinement discharges on LTX-beta can reach a H-98 of close to 2 in a limited plasma without H-mode, in the complete absence of a pressure pedestal. Fueling a low recycling discharge while maintaining a hot edge is non-trivial and will likely only be tractable either with neutral beams or with pellet injection from the high field side. Gas puffing will lead to a collapse of the hot edge. LTX-beta was recently vented to increase the tangency radius of the neutral beam. Fast ion orbit modeling indicated that injecting the beam at larger tangency radius reduces the first orbit losses and improves beam coupling. LTX-beta now shows beam fueling. Charge exchange cross section of beam hydrogen neutrals on Li ions can be 10-15 times the charge exchange cross section of beam hydrogen neutrals on ionized hydrogen from the plasma. Therefore, the effect has a functional dependence on the core lithium impurity fraction, this effect will be explored. Beyond fueling hot edge, high confinement discharges with low recycling, these results indicate a path to fueling tokamaks and stellarators by introducing a modest lithium impurity fraction; similar results are predicted for a fusion plasma with a modest helium ash fraction. A modest lithium impurity has little effect on Z effective, and would getter common higher Z impurities present on the surfaces of metallic plasma facing such as carbon and oxygen. Results for neutral beam fueling with solid and liquid lithium plasma-facing surfaces will be presented. The dependence of beam fueling on lithium concentration and on surface conditions as monitored by in-situ secondary ion mass spectroscopy and temperature programmed desorption will be discussed.

This work is supported by USDoE contracts DE-AC02-09CH11466, DE-AC52-07NA27344.

Speaker: Anurag Maan (Princeton Plasma Physics Laboratory)

53 1.045 Doppler Shift of Hydrogen Balmer-alpha in picosecond Laser Induced Breakdown Spectroscopy on W substrates: The Art of Timing and Wavelength Tuning

Hydrogen (H) isotopes retention in plasma-facing materials is a critical issue for nuclear safety in fusion devices operating with a deuterium-tritium mixture and tungsten plasma-facing components. Its reliable detection is of great importance for both safety and material characterisation after plasma exposure. However, we observed that in vacuum conditions, the Balmer alpha spectra (Ha) measured by LIBS exhibit complex Doppler shifts, leading to significant spectral distortions.
In this work, we report the first systematic study of Doppler shifts in hydrogen spectra obtained by picosecond Laser Induced Breakdown Spectroscopy (ps-LIBS, pulse width: 35ps, max energy: 25 mJ, wavelength: 355 nm) on tungsten samples loaded with hydrogen. Our results reveal not only Doppler shifts associated with single-velocity H atoms, but also, at different time delays, distinct dual-wavelength shifts arising from the simultaneous presence of fast and slow H atoms in the LIBS plasma. The fast component exhibits a fluence-dependent shift, whereas the slow component shows the opposite behavior. Fast and slow H atoms can be temporally separated due to their different velocities. Moreover, it is demonstrated that, in the near surface layer, fast hydrogen corresponds to surface-adsorbed H, while slow H possibly corresponds to H present in the W lattice. Building on these findings, we demonstrate a new approach for tuning the Hα wavelength over a range of up to nm level (depends on laser fluence), making it possible to temporally isolate fast and slow hydrogen atoms and thereby suppress spectral interference. The Doppler characteristics additionally offer a potential spectroscopic tool using LIBS for distinguishing between hydrogen present at the surface and in the tungsten lattice. Finally, the dependence of the Doppler shift on fluence suggests that it may also serve as a proxy for in-operando monitoring of the laser fluence during LIBS measurements.
This work not only resolves a fundamental spectral distortion mechanism, but also provides guidelines for improving the accuracy of H isotopes detection in fusion-relevant plasma-facing materials (PFMs) in LIBS measurements. Importantly, neglecting this Doppler-induced effects may cause misinterpretation in H isotope analysis and absolute isotope-resolved quantification, since the Doppler-induced wavelength displacements can mislead isotope signatures.
This work was supported by the BMFTR project “SyrVBreTT” (Grant No: 13F1011G).

Speaker: Rongxing yi (FZJ)

54 1.046 Investigating the Soret Effect of Deuterium in Metals – A Novel Approach Using Ion Beam Analysis

The development of suitable materials for structural and plasma-facing components is a decisive challenge on the path towards nuclear fusion as a future energy source. In particular, the transport and retention of hydrogen isotopes is of relevance as they have a significant influence on the tritium inventory. The Soret effect describes diffusion processes driven by a temperature gradient, which are also expected to occur in the actively cooled components of a future fusion reactor. To estimate the resulting hydrogen transport, the Soret coefficient $S_T$ is required, which not only depends on the material combination but also varies with temperature and the temperature gradient. However, it is largely unknown for the fusion-relevant materials tungsten and copper.

In this contribution, the development and validation of a novel approach for determining $S_T$ for deuterium in metals is presented. Rod-shaped samples are gas-loaded with deuterium and subsequently coated with a diffusion barrier layer. After long-term exposure to a constant temperature gradient, the concentration distribution is determined using ion beam analysis. This has the advantage that $S_T$ can be determined for a wide temperature range using a single sample.

A dedicated experimental setup was developed to expose rod-shaped samples to a constant temperature gradient over a prolonged time period. The cold side is maintained at room temperature, while the hot side can reach temperatures of up to 500°C. Depending on the thermal conductivity of the material, a heat flux of up to 600 W is possible. In order to prevent oxidation of the sample, the process is performed under vacuum.

To validate the measurement concept and setup, the Soret coefficient of deuterium in niobium was determined and compared with known literature values. 30 mm long samples were gas-loaded in a deuterium atmosphere, coated with a copper diffusion barrier layer and exposed to a stationary temperature gradient for 21 days. The deuterium concentration was determined both before and after with a lateral resolution of 1 mm using the $\textrm{D}(^3\textrm{He},\textrm{p})^4\textrm{He}$ nuclear reaction. The concentration profile was then used to determine $S_T$, whereby a high degree of consistency with the literature values could be observed.

Speaker: Mr Alexander Feichtmayer (Max Planck Institute for Plasma Physics)

55 1.047 Evaluation of deuterium retention properties on pure tungsten with various working rates

Tungsten (W) is a promising candidate material for the first-wall and divertor armor in fusion reactors, in which since one of the disadvantages of W is inherently hard and brittle, work-hardening (working) processes such as forging or rolling are required to improve its brittleness. Although the working rate may influence not only mechanical properties of W but also its hydrogen isotope retention properties due to the drastic change of the internal microstructures, e.g., grain size/structure, porosity and dislocation density etc., systematic investigations focusing on the correlation between microstructures and hydrogen isotope retention properties almost have never been performed to date. In this study, retention properties of hydrogen isotope (deuterium) in pure W samples as a function of various working rates were examined.

Mirror-polished pure W samples with various working rate of 0%, 30%, 60%, 90%, and 98% were prepared. Each sample was irradiated with 1.5 keV-D⁺ ions to a fluence of 3×10²¹ D/m² at room temperature. Approximately three hours after irradiation, thermal desorption spectroscopy (TDS) measurements were performed with a heating rate of 1 K/s. Microstructural features such as grain size/structure, pore density and accumulated distortion field were obtained by means of scanning electron microscope (SEM) and electron backscatter diffraction (EBSD) analyses. Inverse pole figure (IPF) and kernel average misorientation (KAM) maps were also obtained by EBSD data.

SEM images revealed that size and density of the potentially existing pores decreased with increasing working rate. The grain size was on the order of several tens of micrometers at 0% working rate, and the grain size was decreased towards the few micrometers with increasing the working rate. This fact implies that the density of grain size/boundary increases with increasing the working rate. KAM maps generally reflect information on distortion fields accumulated in W samples during a fabrication process, in which relatively high distortion fields were observed in working rate of 98% compared with others.

Comparing 60% and 90% worked W, those have relatively low pore density, it was found that the higher rolling rate (90%) contributed to release of all retained deuterium at lower temperature (below 550 K). Furthermore, all deuterium was completely desorbed at an extremely low temperature (around 400 K) in the 98% worked W. These results indicate that the working rate significantly affects the deuterium retention properties of W.

Speaker: Shuka Tsuda

56 1.048 The plasma content in hydrogen and its control on the full-metal tokamak WEST, from restart phases to long pulses.

In present-day magnetic fusion devices, hydrogen serves as a minority species for Ion Cyclotron Resonant Heating (ICRH). The efficiency of the heating scenarios depends on the hydrogen fraction in the plasma core that should be controlled within a few percents. Plasma-Facing Components (PFCs) release/trap hydrogen into/from the discharge, thereby complicating this control. Hydrogen increases neither the radiation nor the effective charge, but might dilute the fusion fuel, hence degrading the reactivity of burning plasmas.
This contribution systematically surveys the hydrogen content $n_H/(n_H+n_D)$ in WEST deuterium plasmas over six experimental campaigns and presents some techniques to influence it. Using a high-resolution visible spectrometer, we estimate $n_H/(n_H+n_D)$ in the scrape-off layer from the line ratio $H_\alpha/D_\alpha$ at $\lambda$~656nm. Three lines of sight aim at one inboard limiter, the inner and outer lower divertor. Using a residual gas analyzer, we also monitor the composition of the pumped gas. Although, not surprisingly, the various measurements do not fully agree quantitatively, they generally exhibit similar parametric variations.
In the restart phase after a vent, the plasma always appears naturally rich in hydrogen, while fueled with pure deuterium. Outgassing from the PFCs is suspected: ~1week of baking at 170°C did not empty all their hydrogen reservoirs. In 2024, we identified newly-installed semi-inertial bulk tungsten tiles on the inboard limiters as contributors to the hydrogen influx. Glow Discharge Cleaning (GDC) for 1h or more overnight proves efficient at reducing $n_H/(n_H+n_D)$ and was applied systematically before ICRH sessions early in the campaigns. However, GDC acted only temporarily and $n_H/(n_H+n_D)$ subsequently re-increased gradually.
Several weeks of plasma operation brought $n_H/(n_H+n_D)$ below the 10% suitable for ICRH. One could then control the minority fraction by dosing hydrogen puffs pulse by pulse.
Hydrogen re-appeared during long-pulse operation, when large amounts of energy were injected into the plasma. The hydrogen fraction rose both after some time within a discharge and between successive similar pulses, leading to reduced D-D neutron rates (quantitatively consistent with $D^+$ dilution) and eventually a loss of the density control. The hydrogen release decreased over successive experimental sessions with similar plasma scenarios. Outgassing is suspected from semi-inertial remote objects (not clearly identified) only heated by radiation or charge exchange neutrals.
We finally briefly discuss future reactors. In addition to the D-T mix (needing control), the minority species will be $^3He$, the requested minority fraction will be low and the fuelling efficiency will be lower than in present-day machines.

Speaker: Dr Laurent COLAS (CEA, IRFM, F-13108 St-Paul-Lez-Durance, France)

57 1.051 Near-surface hydrogen inventory response to picosecond laser pulses in tungsten.

Tungsten is the leading candidate for plasma-facing components in fusion devices. In future fusion reactors, neutron irradiation creates material defects that trap hydrogen isotopes, including the fusion fuel tritium and deuterium. Monitoring the hydrogen inventory is mandatory for nuclear safety and important for the efficient use of fuel. Laser-based diagnostic methods such as Laser-Induced Breakdown Spectroscopy (LIBS) and Laser-Induced Ablation with detection of volatile species by Quadrupole Mass Spectrometry (LIA-QMS) provide quantitative, depth-resolved assessments of the local fuel inventory and release. They are offering a controlled measurement of near-surface retention on short time scales for in-situ and ex-situ applications.
To describe the effect of transient material heating on fuel release beyond the ablated layer, a one-dimensional modelling framework is developed in FEniCS/FESTIM for picosecond laser pulse trains. The simulations address depths below 1 mm. Transient heat conduction with temperature-dependent properties is coupled to a minimal near-surface trapping-detrapping scheme. Initial near-surface fuel fractions are in the percent to sub-percent range. Coarse ablation is represented by removing a thin surface layer tens of nanometers thick at the start of each pulse, which updates both the thermal and hydrogen-isotope boundary conditions. Per-pulse release is obtained by integrating the simulated desorption flux. Evolution over fuel release is analyzed for tens to hundreds of pulses.
Here in this modelling approach, the initial fuel distribution is taken from NRA depth profiles measured on proton-irradiated and self-damaged, deuterium-decorated tungsten samples which act as reference.
Results show that heating-induced release dominates over the contribution attributable to the imposed ablation layer release. Early pulses are characterized by rapid mobilization of the near-surface fuel reservoir, leading to pronounced, prompt desorption transients. Consequently, when ablation is applied at the start of subsequent pulses, it largely removes a layer already partially depleted by prior heating. With a simple defect-annealing term, model predictions are compared to LIA-QMS-derived depth reconstructions, yielding reasonable agreement.
The modelling framework supports the interpretation of LIA-QMS data for self- or proton damaged tungsten as proxy for neutron-damaged tungest and clarifies the respective roles of fast heating and material ablation in pulse-train experiments with ps-laser systems.
Work supported by BMFTR project SyrVBreTT.

Speaker: Maria Popova (Forschungszentrum Jülich GmbH)

58 1.052 The effect of boron coatings on deuterium retention in tungsten

Deposition of low-Z films with strong getter properties is widely used to improve plasma performance in modern fusion devices, reduce the influx of heavy impurities from the plasma-facing components (PFCs) and enhance hydrogen recycling control [1]. Boronization is the most common technique involving the deposition of a boron–containing coating on the PFCs. Boronization is a candidate technique for ITER [2,3] in case of problems during plasma operation with a pure W first wall. Numerous studies in fusion devices have confirmed its favorable impact on key discharge parameters [4,5].
This work investigates deuterium retention in polycrystalline tungsten samples coated with boron films deposited by magnetron sputtering. The deposition rate was 75 nm/h, and the film thickness ranged from 50 to 500 nm. The behavior of the boron coating under thermal heating was studied up to 1500 K. The results show that the thin B film gradually disappeared from the tungsten surface when heated to this temperature. The thick (500 nm) film is unstable at the surface at high temperatures, cracks and blisters have been formed only due to heating without plasma exposure.
The samples were irradiated under two different conditions: with a low flux ($10^{18} D/m^2s$) of mass-separated deuterium ions (667 eV/D) at the MEDION facility, and with a high flux ($10^{21} D/m^2s$) of low-energy ions (100 eV/D) at a beam-plasma discharge (BPD) facility. The temperature varied from room temperature to 700 K. The D retention was measured using thermal desorption spectroscopy (TDS). The deuterium retention in the presence of the B film was higher than in bare W. Deuterium release from B coated W observed in a wide temperature range up to 1200 K.
[1]. J. Winter, Plasma Physics and Controlled Fusion 38 (1996) 1503–1542.
[2]. T. Wauters et.al. Nuclear Materials and Energy 42 (2025) 101891.
[3]. J. Snipes et.al. Nuclear Materials and Energy 41 (2024).
[4]. Y. Cheng et al. Nuclear Materials and Energy Т. 41 (2024) 101744.
[5]. V. Sharapov, Problems of Atomic Science and Technology, Series Thermonuclear Fusion 43 (2020) 5–12.

Speaker: Anastasiia Umerenkova (NRNU MEPhI)

59 1.053 Parametric Study of Plasma-Driven Permeation in the Diagnostic First Wall of ITER

While ITER is a full-tungsten (W) tokamak, a substantial surface area of water-cooled stainless steel remains exposed to charge-exchange neutrals (CXN) originating from the plasma. Hydrogen isotopes implanted in steel can permeate rapidly and subsequently be released into the cooling water, where inventory limits are imposed by regulatory requirements. This work presents a sensitivity study of permeation through the stainless-steel (SS) 316L-IG ITER diagnostic first wall (DFW), which comprises ~30 m² of upper and equatorial port surface area.
The analysis employs the FESTIM modelling framework [1], which implements the McNabb–Foster reaction–diffusion model to resolve mobile and trapped hydrogen isotopes, while incorporating additional physics such as the Soret thermo-diffusion phenomenon. CXN wall fluxes are obtained from recent SOLPS-ITER and SOLEDGE3X plasma backgrounds using simulation grids extended to the wall contour, providing uncertainty ranges for particle flux and heat loads at the DFW surfaces. Tritium permeation to the water coolant loops occurs predominantly during baking phases. The influence of baking duration, temperature, and steel thickness is assessed.
Results indicate that permeation rates during baking are most sensitive to the impacting particle fluxes and heat loads during plasma operation. To reduce conservativeness, plasma exposure conditions are set as closely as possible in terms of heating power, pulse durations, and tritium content according to the ITER Research Plan. Depending on material properties, the inventory limit in the torus cooling water circuit may be reached by the end of the first DT operation phase (DT-1).
Finally, the application of permeation barriers at the DFW plasma-facing surface is evaluated. These may be implemented if early measurements during the initial DT-1 campaigns indicate permeation rates exceeding acceptable limits. For tens of micrometre thin coatings, assuming continuity of chemical potential, double layers such as SS/W on SS or SS/alumina on SS are highly effective in mitigating permeation. However, since such layers may not withstand disruption-induced thermal loads, a more robust solution—2 mm W cladding on SS—is proposed, reducing permeation by two orders of magnitude. The potential influence of material interface properties on barrier effectiveness is assessed through parameter scans.
[1] R. Delaporte-Mathurin et al., Int. J. Hydrog. Energy, vol. 63, pp. 786–802, Apr. 2024

Speaker: Tom Wauters (ITER Organization (IO))

60 1.054 Low-Z materials Wall Conditioning using Lithium and Boron for Improved Plasma Performance in EAST with full-metal wall

Fuel recycling and impurity accumulation remain major obstacles to sustaining long-pulse, high-performance plasmas in magnetic confinement fusion devices. In EAST, under full-metal wall conditions, strong impurity radiation and recycling—particularly from heavy impurities—severely limit the duration and stability of H-mode discharges [1]. Addressing this challenge requires advanced wall-conditioning techniques capable of continuously mitigating plasma–wall interactions.
In the EAST tokamak, equipped with an ITER-like tungsten divertor and full-metal first wall, both low-Z coatings and in-situ powder injections have been implemented to enhance wall conditioning. Lithium or boron films are applied using ovens at the O, J, and F ports, assisted by glow discharge cleaning (GDC) or ion cyclotron wall conditioning (ICWC) [2]. Two powder droppers from PPPL enable real-time, actively controlled lithium or boron powder injection during plasma operation [3].
Lithium coatings markedly reduce impurity content and fuel recycling, maintaining tungsten core impurities at 3–15 ppm, H/(H+D) < 5%, and a global recycling coefficient below 1 during >100 s H-mode operation. However, their short lifetime (~300 s) limits long-pulse performance [4]. To overcome this, a dynamic lithium powder injection system with feedforward–feedback control has been developed. Using this technique, EAST achieved a 605 s H-mode discharge with H/(H+D) < 4% and only a 12% rise in total radiation [1], , and later demonstrated a record >1000 s steady-state H-mode operation in 2025.
Boron coatings show comparable impurity suppression and better stability [2], though a >15% confinement reduction occurs, likely due to enhanced hydrogen release from carborane-based films. Real-time boron powder injection has since enabled reproducible 100 s H-mode discharges with sustained impurity control.
In summary, lithium and boron wall-conditioning approaches demonstrate strong capability in impurity suppression and plasma performance enhancement in EAST, providing a solid foundation and valuable insights for steady-state operation in ITER and future fusion reactors.

Speaker: Zhe Wang (GNOI)

61 1.055 Study of ICWC- and GD-Assisted Boronization in EAST with Full Metal Wall

Wall conditioning is critical for improving plasma confinement and stability to reduce impurity influx and ensure long-term reliable operation. Boronization, which forms a dense boron film to capture impurities and enhance plasma performance, has become a new baseline for ITER and thus gained growing global attention in fusion research.
In EAST, ITER like tungsten divertor and molybdenum first wall have been successfully upgraded. In order to remove impurity particles from plasma-facing components, routine wall-conditioning techniques–such as electric and hot N2 baking, ion cyclotron wall conditioning (ICWC) and glow discharge cleaning (GDC) have been successfully established[1]. Novel low-Z wall coatings, including lithiation and boronization, have also been applied to further improve particle control capability[2, 3]. Either ICWC or GDC is typically performed to remove residual impurities and prepare the substrate for boronization using C2B10H12 as working materials. Comparative studies over these two methods assisted boronization show that both methods produce boron films of similar thickness (~163–171 nm). However, ICWC coatings exhibit higher hardness (2.73 vs. 2.27 GPa), a higher H/E ratio (0.018 vs. 0.015), and more-delayed pop-in events. These features suggest a denser film structure and stronger film–substrate adhesion, which are beneficial for providing more effective wall protection and mitigating plasma–wall interactions. As the discharge progresses, ICWC Boronization shows a longer lifetime, as evidenced by the sustained higher level of B II emission. Regarding plasma performance, discharges following ICWC boronization show reduced impurity radiation, lower effective charge (Zeff), and ~10% higher energy confinement (H89) and stored energy relative to those following GDC boronization. In conclusion, ICWC- boronization offers significant advantages over GD-assisted methods, including higher film density, stronger adhesion, better impurity control, and improved plasma confinement. Importantly, ICWC-boronization can be conducted directly under magnetic field conditions. This operational flexibility enables rapid post-discharge wall conditioning without demagnetization, making ICWC-boronization a more efficient and practical approach for routine wall maintenance and steady-state operation in future superconducting fusion devices.
These results indicate that ICWC is an efficient technique for assisting boronization, suggesting that it can play a crucial role in optimizing plasma performance in ITER and other future fusion devices.

[1] G. Z. Zuo, et al., Plasma Phys. Control. Fusion 67, 055011 (2025).
[2] Y. H. Guan, et al., Nucl. Fusion 65, 096020 (2025).
[3] Z. Wang, et al., Nucl. Fusion 65, 104001 (2025).

Speaker: Guizhong Zuo (Institute of Plasma Physics, Hefei Institutes of Physical Science, Chinese Academy of Sciences)

62 1.056 Absorption processes of electron cyclotron waves in electron cyclotron wall conditioning plasma and their influence on plasma expansion in JT-60SA

Electron cyclotron wall conditioning (ECWC) is one candidate as inter-shot wall conditioning tool in tokamak devices with superconducting toroidal coils. Since ECWC plasma is primarily generated at the electron cyclotron resonance layer, poloidal magnetic fields are required to expand the plasma toward the wall where fuels and impurities are necessary to be removed. In JT-60SA, ECWC plasma produced by the fundamental ordinary mode (O1-mode) EC wave expands along the poloidal magnetic field, whereas plasma produced by the second harmonic extraordinary mode (X2-mode) EC wave remains localized. This difference may be attributed to the variation in single-pass absorption rates of the EC waves.
First-pass absorption rates of EC waves were estimated in JT-60SA. Two EC waves were injected at an angle with normal to the toroidal magnetic field: one was an 82 GHz O1-mode EC wave, and the other was a 110 GHz X2-mode EC wave. Temperature increases in carbon tiles facing the EC launchers were measured by thermocouples embedded in the carbon tiles. Single-pass absorption rates of EC waves were also calculated based on the electron temperature ($T_\mathrm{e}$) estimated by the He I line intensity ratio measured by a visible spectrometer, and the electron density ($n_\mathrm{e}$) was estimated by the He I line intensity ratio and a CO$_2$ laser interferometer.
Temperature increases in the carbon tiles by the first pass of the O1-mode and X2-mode in the He-ECWC were consistent with that observed during vacuum injection, indicating that the first-pass absorption rates of the O1 and X2-mode EC waves in the He-ECWC plasma were similarly low within the measurement uncertainty (±20%). These results were confirmed by calculation of the single pass absorption rate of the O1 and X2-mode EC waves: 12% for the O1-mode EC wave at $n_\mathrm{e}$ = $2\times10^{18}$ $\mathrm{m^{-3}}$ and $T_\mathrm{e}$ = 100 eV and 28% for the X2-mode EC wave at $n_\mathrm{e}$ = $10^{19}$ $\mathrm{m^{-3}}$ and $T_\mathrm{e}$ = 16 eV. In contrast, the single pass absorption rate of the X1 and O2-mode EC waves differed significantly: 100% for the X1-mode EC wave and 0.1% for the O2-mode EC wave. These findings suggested that mode conversion from the O1-mode to X1-mode EC wave at the in-vessel wall may play an important role in the expansion of the He-ECWC plasma in JT-60SA.

Speaker: Masakatsu Fukumoto (National Institutes for Quantum Science and Technology)

63 1.057 Effect of boron coatings on hydrogen isotope exchange effect in tungsten

The replacement of the originally planned beryllium first wall with tungsten in the current ITER baseline makes boronization an important strategy for reducing impurities such as oxygen in the plasma fuel. Consequently, the effects of boron on plasma-material interactions during and after plasma operations are of high importance. Although hydrogen isotope retention in tungsten has been widely studied because of tritium inventory and safety concerns, the influence of boron layers on hydrogen isotope retention and removal in tungsten is still not well known.

Hydrogen isotope exchange effect has been shown to successfully decrease near-surface tritium retention in tungsten by replacing tritium with lighter isotopes. The exchange process is the strongest near the material surface, where incoming light isotopes diffuse in and displace heavier isotopes from the trap sites. This near-surface effect makes the role of any surface coating critical for the overall efficiency.

This study experimentally investigates how thin boron layers will affect hydrogen isotope exchange in tungsten. Boron films with fusion-relevant thicknesses of 50 nm and 250 nm were deposited on tungsten substrates by magnetron sputtering. The 50 nm case corresponds to a typical layer formed by a single boronization, while the 250 nm thickness reflects deposition-dominant regions, allowing the influence of coating thickness on isotope exchange to be measured. Samples were implanted with deuterium at two energies (5 keV/D and 20 keV/D) to investigate trapping within the boron layer and deeper in the tungsten. Isotope exchange was promoted by subsequent annealing in H2 gas which was chosen to introduce hydrogen gently without energetic high-flux ion sputtering of the boron film. The results were compared to vacuum-annealed and uncoated tungsten references. The resulting deuterium and hydrogen depth profiles were measured by elastic recoil detection analysis (ERDA) to quantify isotope replacement as a function of depth, coating thickness and implantation energy.

By investigating hydrogen isotope exchange efficiency in the boron layer, across the boron-tungsten interface, and within the underlying tungsten bulk with a boron surface layer, this work clarifies the role of boron in modifying near-surface tritium removal by isotope exchange. A key hypothesis of the study is that thin boron layers may still permit efficient isotope exchange, while thicker layers could increasingly limit hydrogen permeation and thus reduce the exchange efficiency at the tungsten surface. The findings are relevant for tritium inventory management and wall-conditioning strategies in future fusion devices and reactors with tungsten-based first walls.

Speaker: Dr Tomi Vuoriheimo (University of Helsinki)

64 1.058 Boronization in ITER: Study of the hydrogen isotope interaction and isotope exchange in B:D layers

Due to the change of first wall material in ITER from Be to W, a glow discharge boronization (GDB) system is included in the re-baseline in order to guarantee efficient plasma operation (https://doi.org/10.1016/j.nme.2024.101854). Even though the GDB is used in many fusion devices for decades, its conditioning effect and possibility for recovering fuel from the B layers was not studied in detail so far. Furthermore, ITER is expected to have a lower carbon inventory compared to the majority of today’s fusion devices. The presences of C has an influence on the B layer characteristics, which might change the boronization mechanism as well.
The study focuses on plasma exposure conditions relevant to Ion Cyclotron (IC) Wall Conditioning and Glow Discharge (GD) Conditioning in ITER. With this study, the efficacy of these tritium removal techniques for ITER can be assessed and further insight into the boronization mechanism will be gained.
W substrates are coated with B:D layers (160-320 nm) by magnetron sputter deposition. Two kind of samples with different D content (D/B ratio between 0.1 and 0.2) are produced and pre-characterized (e.g. microstructure, composition, morphology and crystal phase). For the study of isotope exchange, samples are exposed to different hydrogen plasmas at the toroidal plasma device TOMAS and the linear plasma device PSI-2 at different sample temperatures. The hydrogen isotope content is measured before and after exposure by thermal desorption spectroscopy and nuclear reaction analysis.
The deuterium content in the layer of a GD plasma exposed sample is significantly reduced by about 35% compared to a non-exposed sample. In opposite to this observation, a first IC plasma exposure of a B:D layer results in a much smaller reduction of the D content in the layer. Further exposures with different plasma conditions are performed and the analysis will show, if it is possible to efficiently clean the B:D layers by isotope exchange and which plasma condition is suitable for the tritium removal.

Speaker: Anne Houben (Forschungszentrum Jülich GmbH, Institute of Fusion Energy and Nuclear Waste Management - Plasma Physics, 52425 Jülich, Germany)

65 1.059 Spectroscopic characterisation of boronisation glow discharges in ASDEX Upgrade

During boronisation glow discharges, the spectral radiance on several lines-of-sight (LOS) through the vessel was measured for the complete visible spectrum in the range from 350-730 nm. The spectroscopic setup was absolutely calibrated by illuminating the respective optical heads inside the vessel with an integrating sphere of known spectral radiance. The glow discharges were run at the standard parameters, i.e. pressure of 0.5 Pa, voltage of 500 V, and current of 1.8 A per anode. Discharges with 4 and 2 anodes were compared. The gas was the standard mixture of 90% He with 10% deuterated diborane.

No signal is seen on the LOS deep in the slits between roof baffle and lower divertor tiles. On the other LOS through the main chamber and in the upper divertor region, the spectrum is dominated by the emission lines from neutral He in the singlet and triplet system with one electron going from n=3-7 to n=2. The strongest line is the 3p-2s transition in the singlet system. The n=4-3 transition of ionised He is weak. The lines of the Balmer series are well visible up to n=7-2. The BD molecule is seen on the A-X transition with the vibrational bands 0-0 and 1-1. A wide region with Fulcher band emission of the deuterium molecule is visible where the diagonal and the non-diagonal vibrational bands are emitting.

We mainly analysed the radiances of the 11 HeI-lines emitted on the transitions n=3-2 and n=4-2 with the help of a collisional radiative model which was adopted from the evaluation of the He-beam diagnostics. Even though the unknown electron energy distribution had to be approximated by a thermal distribution with an effective temperature around 30 eV, the measured radiances can be well fitted when taking the strong optical thickness on the resonance lines into account. Furthermore, we considered that the dwell time of the He atoms before colliding with the vessel walls is much shorter than the equilibration time of the exited states with the ground state. The fitted density of these fast 30eV-electrons is around 10$^{12}$m$^{-3}$ and increases by a factor of 2 when using 4 instead of 2 anodes.

Speaker: Ralph Dux (MPI for Plasmaphysics, Garching, Germany)

66 1.060 Enhanced hydrogen isotope removal by HiPIMS-inspired pulsed GDC with high-peak-current in the EAST tokamak

Tritium retention remains a critical issue for future fusion reactors owing to its impact on plasma performance, radiological safety, and fuel availability. Glow discharge cleaning (GDC) is a simple yet essential technique for hydrogen isotope removal in fusion research devices; however, its efficiency is limited under certain operating conditions. Inspired by the high power impulse magnetron sputtering (HiPIMS) regime, a high-peak-current pulsed GDC approach is proposed for hydrogen isotope removal in fusion devices. Unlike conventional GDC, in which increasing the discharge current often leads to arcing and surface damage, the pulsed operation enables stable access to a previously unexplored high-peak-current regime, thereby enhancing transient plasma-surface interactions and removal efficiency while mitigating thermal and arcing risks.

The high-peak-current pulsed GDC was experimentally investigated in a linear plasma device using helium at 5 Pa, with a pulse frequency of 76.8 kHz. The results show that the deuterium removal efficiency was improved by 53% by increasing the GDC peak current from 1.1 A to 4.0 A, while keeping the same average GDC current of 1 A. Moreover, at the same average current of 5 A, the pulsed GDC with a high peak current of 20 A improved the removal efficiency by 215% compared to continuous GDC. These results demonstrate a strong enhancement in hydrogen isotope removal enabled by high-peak-current operation without arcing.

This technique was successfully demonstrated on the EAST tokamak using a single GDC anode in 1 Pa helium at a pulse frequency of 10 kHz. Despite the more restrictive operational constraints and complex geometry of a tokamak, stable and arc-free pulse GDC operation with high peak current was maintained throughout the experiments. The results show that increasing the peak discharge current from 8 A to 16 A, while maintaining the same average current of 4 A, resulted in 55% enhancement in deuterium removal efficiency. This tokamak demonstration confirms that the high-peak-current pulsed GDC is a robust and effective approach for hydrogen isotope removal under fusion-relevant conditions, thereby providing a new and promising technique for tritium removal in future fusion reactors.

Speaker: Mr Hao Sun (Institute of Plasma Physics, Hefei Institutes of Physical Science, Chinese Academy of Sciences)

67 1.061 Boronization and erosion rates of samples exposed in the W7-X stellarator

Wall conditioning improves plasma performance in fusion devices by reducing the amount of impurities, especially carbon- and oxygen-based impurities, in the plasma [1,2]. Standard wall-conditioning procedures include baking and glow-discharge cleaning (GDC) of the inner walls using hydrogen or helium plasmas for removing impurities contained at the plasma facing components (PFCs). The amount of impurities can be further mitigated by depositing thin boron layers on PFCs surfaces (boronization). These boron layers act as oxygen getters [3, 4].

Wendelstein 7-X (W7-X) [5,6] is an advanced stellarator fusion reactor with major parts of its first wall comprised of carbon components. Since 2022 (operational period OP 2.1) the device has operated with a water-cooled carbon-fibre-composite (CFC) divertor. Boronizations have been incorporated into the wall conditioning procedures since 2018 (operational period OP 1.2b).

In the present work, a series of samples exposed to boronizations or plasmas in W7-X were analyzed. All samples were exposed using the multi-purpose manipulator (MPM) equipped with the MAT1 probe head during operational phases OP 2.2 and 2.3. The characteristics of the samples (material composition, surface roughness) varied. The boronizations were implemented with a B2H6 (90%) – He (10%) gas mixture. Samples were exposed to hydrogen plasmas for multiple discharges of varying duration, with electron densities ranging from 2 to 16 * 1019 m-2. In addition, a number of samples were exposed to He-GDC plasmas. The sample analysis was carried out at the Tandem Accelerator facility of the Max-Planck Institute for Plasma Physics in Garching, Germany using Nuclear Reaction Analysis (NRA) with a 3 MeV 3He beam.

The measured amounts of boron in the deposited layers varied significantly between boronizations, up to a factor of 4. No correlations were observed between the deposited amount of boron and the substrate composition or its surface roughness. Carbon and oxygen concentration profiles did not scale proportionally with boron. For samples exposed to hydrogen plasma, the net erosion rate of boron was substantially higher than that of carbon or oxygen. In some cases, a net oxygen uptake rather than erosion was recorded.

[1] J. Winter, Plasma Phys. Control. Fusion 38 (1996) 1503
[2] A Goriaev et al., Phys. Scr. T171 (2020) 014063
[3] M. Mayer et al., Nuclear Materials and Energy 41 (2024) 101778
[4] S. Sereda et al., Nucl. Fusion 60 (2020) 086007
[5] Beidler et al., Fusion Technology, 17(1) (1990) 148–168.
[6] T.S. Pedersen et al., Phys. Plasmas 24 (5) (2017) 055503.

Speaker: Fotios Maragkos (MPPL)

68 1.062 Investigation of boron distribution on multiple anodes applied in W7-X using ex-situ picosecond laser-induced breakdown spectroscopy

Boronization is an established techniques to condition the first wall of fusion devices. Boronization systems are designed to deposit a ~10-100 nm thin boron (B) layer on PFCs by injecting diborane (B₂H₆/B₂D₆) mixed with a carrier gas (He, H₂, D₂) into a cleaning Glow Discharge (GD). For optimal boronization performance, thus, achieving a homogeneous layer is desirable in order to getter the oxygen (O) and cover recessed areas of the first wall. This ensures a long lifetime of the boronization effect, thus reduction of the impurity influx and controlled fuel recycling resulting in good plasma performance. The burden to use boronization is the need for specific gas inlets for toxic diborane and the need of a homogenous, stable GD. The B distribution can become highly non-uniform, depending strongly on the spatial arrangement of anodes and gas-feed locations. The regions near the anodes and gas inlets tend to develop substantially thicker B layers, whereas more remote areas exhibit much thinner deposition. Subsequent plasma operation can further redistribute the deposited B, making the dynamic behaviour of B layers and their influence on plasma performance..
In the non-axis symmetric stellarator W7-X, the challenge to achieve homogenous boron coverage is even higher than in tokamaks. The success of boronization in W7-X was demonstrated and the lifetime determined empirically, material samples in the multipurpose manipulator revealed thin B layers of less than ten nm, whereas migrated B of µm thickness was found campaign-integrated in the divertor. No systematic study of B deposition on the anodes was yet performed.
The boronization in W7-X uses a GDC with a 90:10-mixture of He and B₂H₆, typically employing 8-10 electrodes at 310-490 V and 4-11×10⁻³ mbar for 2-5 hours. After three boronizations in OP1.2b and five in OP2.1, five toroidally distributed GD anodes were retrieved, providing an opportunity to examine the spatial distribution of B deposition. ps-LIBS was applied to characterize mixed B deposition on multiple anodes. The results reveal that variations in operational duration, anode position, and local discharge conditions produce pronounced differences in B-layer thickness from nm-to-µm. Significant variation is observed among different anodes located toroidally and different distance to the inlets, and individual anodes exhibit strong radial and circumferential asymmetry. The deposits consist primarily of B-O with minor H incorporation. These findings provide essential experimental benchmark to validate models of B deposition during boronization and subsequently input to B migration studies in W7-X and beyond.

Speaker: Sebastijan Brezinsek (ZJFJ)

69 1.063 A heuristic model for the effects of real-time and glow discharge boronization

Boronization, a process involving coating of the plasma facing components (PFCs) with boron (B) either by glow discharge (GDB) or solid boron injection (SBI), is an established way of improving operation of tokamaks and stellarators due to medium and high-Z impurity level reduction. However, a quantitative prediction of the amount of B required on next-step fusion devices such as ITER or a Fusion Power Plant (FPP), while critical, remains elusive. We put forward a 0-D heuristic model to interpret and predict the evolution of wall conditions in response to boronization. The model assumes that B introduced in the vessel passivates with plasma operation, expressed as cumulative energy injected in the machine, as a proxy for particle and energy fluence to the machine PFCs, and that only non-passivated, “active”, B is responsible for wall improvements. The evolution of active B amount is governed by a 0-D continuity equation with an effective passivation rate ε [MJ], representing the effect of passivation mechanisms, such as erosion of deposited layers, migration of B to shaded areas, saturation of retention capacity, oxidation or implantation of carbon. The wall improvements are expressed in terms of indicators such as brightness of specific impurity lines, radiative losses, neutral pressure or wall fueling, whose changes are assumed to depend on the active B through a response function, defined by the parameter μ [mg], the mass of active B required to saturate an indicator’s response to an increase of active B. The model is applied to different experimental DIII-D datasets where wall conditioning indicators are tracked throughout a series of plasmas, including plasmas following GDB or with phases of SBI. With an optimized choice of the parameters ε and μ, the model reproduces the experimental evolution of wall conditioning indicators reasonably well. Notably, different ε and μ are needed to describe indicators associated with different physics mechanisms. This allows us to account for significant differences in the experimentally observed evolution of indicators of wall fueling (plasma density, main chamber pressure, characterized by a prompt, relatively short-lived response to B injection) compared to indicators of impurity retention (oxygen or nitrogen brightness at breakdown, characterized by a saturated, lasting response). The values of ε and μ determined for various indicators and datasets can be used to inform and quantify the applicability of boronization to other conditions and devices.

Work supported by the U.S. DOE under DE-AC02-09CH11466 and DE-FC02-04ER54698.

Speaker: Alessandro Bortolon (Princeton Plasma Physics National Laboratory)

70 1.064 Lithium migration and functional lifetime of wall conditioning in NSTX: linking pre-discharge coatings to real-time wall conditioning

The functional lifetime of wall conditioning (e.g., Li, B) is a key constraint for long‑pulse operation and for operation with metallic (W) plasma‑facing components, where sustained control of recycling and high‑Z impurity sources is required[1]. Despite extensive use of pre‑discharge coatings and, increasingly, real‑time powder delivery, the lifetime of a single conditioning event (in shots or boundary fluence) remains challenging to predict because migration can redistribute low‑Z material.
This contribution uses NSTX as a testbed to quantify how Li migration controls conditioning longevity and to contrast passive pre‑coating with active real‑time delivery. We revisit NSTX sequences following a large Li evaporation dose with no subsequent inter‑shot conditioning over ~100 H‑mode discharges, together with reference cases using smaller inter‑shot Li evaporations and real‑time Li aerosol injection using the “dropper” device. This regime provides a rare opportunity to decouple the initial deposition profile from the real-time conditioning surface. The coating at the high-flux strike point is theoretically eroded/saturated within milliseconds; however, enhanced confinement and reduced recycling persisted for varying durations (0.5 - 1.0s). The new analysis exploits NSTX 2D divertor Li-I and Dα camera data, together with divertor heat‑flux from IR thermography, to track the evolution of Li sources and D recycling mitigation between near‑SOL/strike-point zones and far‑SOL/background regions across a long sequence of shots. These diagnostics enable us to distinguish between simple depletion at the strike point, migration toward cooler regions, and possible “self‑freshening” via re‑erosion and redeposition, while also quantifying the impact on fueling requirements and divertor recycling levels. To interpret these trends and connect to present-day tungsten devices, SOLPS‑ITER simulations with drifts will be used to examine ExB- and flow-driven Li transport between inner/outer targets and between far‑SOL and near‑SOL, scanning the local effective recycling coefficient in a way constrained by the measured Dα response. Crucially, we contrast this passive replenishment with active real-time powder injection. We demonstrate that real-time injection creates a distinct transport pathway that can bypass the migration mechanism depending on the injection location and plasma parameters. By validating these transport pathways with 2D imaging and modeling, this work provides a physics basis for optimizing the duty cycle and injection geometry of real-time B and Li systems in future devices.
*Sponsored in part by U.S. Dept. of Energy under contracts DE-AC02-09CH11466 and DE-AC52-07NA27344.
References
[1] A. Loarte, et al., Plasma Physics and Controlled Fusion 67 (2025) 065023.

Speakers: Dr Zhen Sun (PPPL), Rajesh Maingi (Princeton Plasma Physics Laboratory)

71 1.065 Experimental investigations of impurity enrichment in the divertor of Wendelstein 7-X

The island divertor concept implemented at Wendelstein 7-X (W7-X) is one of the most extensively investigated solutions for power and particle exhaust in future quasi-isodynamic stellarator power plants. In the latest experimental campaign (OP2.3), W7-X demonstrated successful feedback control of radiative detachment via impurity seeding, based on real-time bolometric measurements of the total radiated power. This capability marks an important step toward reactor-relevant power exhaust scenarios, where strong impurity radiation in the Scrape-Off Layer (SOL) is required to protect plasma-facing components. At the same time, future fusion reactors must minimize the impurity content in the core plasma to sustain the fusion reaction. The impurity enrichment factor η = cimp,SOL/cimp,Core quantifies the divertors ability to retain impurities and maintain high divertor concentrations without degrading core performance. This work presents the experimental methods for impurity enrichment studies at W7-X, illustrated with representative discharges for helium, neon and argon.
We measure the impurity concentration in the residual gas with a novel Time of Flight (ToF) mass spectrometer in the Diagnostic Residual Gas Analyzer (DRGA). This measurement relies on the long-lasting, steady state conditions (≈20s) provided by the feedback system to equilibrate the neutral gas components at the DRGA with the pumping gap several meters away.
We complement the residual gas analysis by divertor spectroscopy measurements, which grant direct insight into local plasma parameters of the SOL including impurity concentrations. We investigated line ratio spectroscopy NeII in the divertor plasmas and found transport processes of neutrals to increase the radiance of the spectral lines and complicate quantitative concentration estimates. However, the line ratio spectroscopy remains sensitive to electron density. These values derived from the impurity line ratios as well as from Stark broadening analysis both indicate W7-X operating at significantly lower electron densities in the SOL (ne = 1–5×10¹⁹ m⁻³) compared to tokamaks such as AUG and JET with densities typically exceeding ne = 10²⁰ m⁻³.
To assess the enrichment factor, we measure the impurity concentration in the core plasma via Charge Exchange Recombination Spectroscopy (CXRS) during diagnostic NBI blips.
A first evaluation of neon yields an enrichment of η = cimp,SOL/cimp,Core ≈ 1, largely independent of upstream density and total radiation fraction. The same methodology is now being extended to helium and argon to provide a comprehensive comparison of impurity enrichment across species.

Speaker: Frederik Henke (MPPL)

72 1.066 Neoclassical impurity transport in the tokamak transport barrier and pedestal with SOLPS-ITER

Impurity transport inside the separatrix of a tokamak is considered to combine neoclassical and anomalous properties. If $L_{n_i}$, $L_{T_i}$ are main plasma density and ion temperature radial characteristic lengths and $L_{T_i}$<1/2 $L_{n_i}$ the impurity radial convective flux is outward [1]. However, for strong gradients in the edge transport barrier it’s not applicable [‎4] and the convective flux may be reduced [‎2] or enhanced [‎3] depending on various factors. Ionization sources [‎4], strong radial non-neoclassical transport and non-neoclassical details of momentum balance diverge impurity transport from neoclassical predictions. These conditions can be studied only via modelling with drifts.
We present analysis of impurity transport for H-mode transport barrier for several tokamaks: ASDEX-Upgrade, JET and ITER (“the standard barrier” [5] with $T_i \approx 2.5~\text{keV}$ and “the enhanced barrier” with $T_i \approx 4 ~\text{keV} $ at the barrier top) based on SOLPS-ITER modeling. Poloidal distributions and radial fluxes of seeded impurities (N and Ne) are studied in detail. Significant difference with respect to neoclassical theory predictions is obtained. Poloidal distributions of impurity reveal strong HFS-LFS asymmetry that either coincide with predictions of [‎‎2],[‎3],[‎6] or diverges due to non-neoclassical factors listed above. This leads to specific inward impurity convection.
For ASDEX-Upgrade: in the ETB inward drift convection $|V^{dr}_{I,r}|$ is smaller than 5 m/s, which is up to 10 times smaller than standard neoclassical prediction [1], in the pedestal neoclassical theory is also not applicable and the convection is zero. For JET: in the ETB $|V^{dr}_{I,r}|$ is smaller than 2 m/s, which is more than 10 times smaller than prediction of [1], in the pedestal outward $V^{dr}_{I,r}$ is of the order of [‎1]. For ASDEX-Upgrade and JET convection in the ETB and the prediction of [1] have opposite signs. For ITER with “the standard barrier”: both in the ETB and the pedestal $V^{dr}_{I,r}$ is close to standard neoclassical prediction [‎1], but the value is still small, $V^{dr}_{I,r}$<0.2 m/s. For ITER with “the enhanced barrier”: in the ETB outward drift convection is up to 0.5 m/s, in the pedestal $V^{dr}_{I,r} \approx 0$.
1. S.P. Hirshman, D.J. Sigmar, Nucl. Fusion 21 (1981) 1079.
2. P. Helander, Phys. Plasmas 5 (1998) 3999
3. C Angioni and P Helander, Plasma Phys. Control. Fusion 56 (2014)
4. V. Rozhansky et al Nucl. Fusion 55 (2015) 073017
5. I. Veselova, et al. Nucl. Mater. Energy 26 (2018).
6. E. Kaveeva, et al. 42nd European Physical Society Conference on PlasmaPhysics, EPS, 2015.

Speaker: Prof. Vladimir Rozhansky (SPbPU)

73 1.067 Prompt Redeposition of Divertor and ICRH-Sourced Tungsten in WEST Measured with Ultraviolet W I/W II Spectroscopy

Prompt redeposition of tungsten plays a central role in impurity transport, main-chamber erosion, and net wall evolution in full-metal fusion devices. Yet experimental quantification of prompt redeposition remains limited due to the diagnostic challenges of ultraviolet (UV) tungsten spectroscopy and the need for reliable atomic physics data. This work presents spectroscopic measurements, collisional–radiative (CR) modeling, and tungsten-sourcing experiments aimed at quantifying prompt redeposition in both the WEST divertor and in regions influenced by ion cyclotron resonance heating (ICRH).
High-resolution measurements of the W I 400.9 nm and W II 364.1 nm emission lines were used to constrain the dynamics of neutral and singly ionized tungsten. These transitions were selected because W I 400.9 nm provides a direct indicator of neutral tungsten sourcing, while the W II 364.1 nm line offers a robust measure of the earliest ionized population. Line-integrated fluxes were inferred via the S/XB method, using electron temperature and density from Langmuir probes. CR coefficients were taken from Smyth et al. [1] (W I) and Dunleavy et al. [2] (W II).
Under X-Point Radiator divertor conditions, comparison of W I- and W II-based fluxes indicates that approximately 90–97% of grossly eroded tungsten is promptly redeposited within the divertor region. Spatially resolved analysis reveals systematic asymmetries: prompt redeposition is consistently higher on the high-field side.
A dedicated ICRH tungsten-sourcing experiment was conducted by scanning the plasma vertical position to modify antenna sheath potentials. Resulting variations in W I and W II emission demonstrate clear signatures of ICRH-driven tungsten release and rapid return of ionized particles to nearby surfaces. These measurements confirm that antenna-sourced tungsten largely undergoes short-range prompt redeposition rather than long-range transport to the divertor.
Overall, this work demonstrates that coordinated W I/W II spectroscopy combined with simplified CR-based modeling provides a practical and experimentally grounded method to quantify prompt redeposition in WEST. The diagnostic robustness of the W II 364.1 nm line, in particular, enables clear separation of gross erosion and ionized-flux components. The methods developed here support future efforts to link in-situ redeposition measurements with post-mortem surface analysis under ITER-relevant conditions.

This work was funded under DE-AC05-000R22725.
References

[1] R.T. Smyth, , C.P. Ballance, C.A. Ramsbottom, C.A. Johnson, D.A. Ennis, and S.D. Loch. Physical Review A 97, no. 5 (2018)
[2] N.L. Dunleavy, C.P. Ballance, C.A. Ramsbottom, C.A. Johnson, S.D. Loch, and D.A. Ennis. Physics B: Atomic, Molecular and Optical Physics 55, no. 17 (2022)

Speaker: Curtis Johnson (ORNL)

74 1.068 Simulation study on the effect of boronized wall condition on first wall erosion and core tungsten impurity accumulation in EAST

The first wall plays a vital role in ensuring the safety and stability of the tokamak device. EAST is upgrading its first wall to full tungsten (W); however, W impurities (high-Z) are incompatible with the main plasma. Boron (B) can improve wall conditions by coating (boronization), and reduce impurity level in core region[1]. Schmid et al evaluated the erosion behavior of the all-tungsten first wall in ITER and the influence of boronized layers on fuel retention[2] based on the WallDYN3D model. EAST experiment found that boronization primarily reduces metallic impurity sources rather than significantly altering impurity transport in the core plasma[1]. However, there is a paucity of studies on the effects of boronization of W wall on physical sputtering and impurity transport. Meanwhile, previous simulation studies was limited to the divertor and not extended to the entire first wall [3]. Importantly, the key challenge lies in the accurate characterization of the dynamic evolution of boronized layers and B-W interface interactions, as well as coupled multi-physics modeling. In this work, the new SOLPS-ITER wide-grid code [4] coupled with the kinetic impurity transport code IMPEDGE [5], impurity migration and wall composition dynamics code IMWCD [6] is used to investigate the effects of boronized wall conditioning on wall erosion and core W impurity accumulation in EAST. The plasma background is provided by SOLPS-ITER, while IMPEDGE [5] is applied to calculate the redistribution matrix and simulate W impurity transport and core accumulation. The lifetime and thickness of the boronized layer are evaluated using the IMWCD. The dynamic evolution of boronized walls is evaluated, and their inhibitory effect on W sputtering of the first wall is quantitatively analyzed. Specifically, the difference in first-wall erosion rates with and without boronization is compared, the influence of boronized layer thickness and B/W ratio on erosion inhibition efficiency are investigated, and the regulatory mechanisms of boronized walls on wall erosion and W impurity core accumulation will be clarified.

Keywords: boronized first wall, erosion, impurity core accumulation, impurity transport

References
[1] Cheng Y et al 2024 Nucl. Mater. Energy 41 101744
[2] Schmid K 2022 Nuclear Materials and Energy 33 101230
[3] Effenberg F et al 2025 Nuclear Materials and Energy 42 101832
[4] Dekeyser W et al 2021 Nuclear Materials and Energy 27 100999
[5] Wu Y et al 2022 Nuclear Materials and Energy 33 101297
[6] He Y et al 2025 AIP Adv. 15 055230

Speaker: Chaofeng Sang (GNOI)

75 1.069 Impact of plasma shape on impurity sources and resulting W core contamination in KSTAR and WEST.

A series of experiments with plasma shaping scans were conducted in WEST and KSTAR illuminating the respective role of lower divertor impurity sourcing (high impurity production and screening) compared to the non-divertor regions (lower impurity production and screening) and the non-intuitive resulting impact on W core contamination. Since the new ITER baseline was presented in 2024, W has become the leading candidate Plasma-Facing Component (PFC) in next-step fusion devices. With extended pulse length relevant for ITER scenario development, PFCs will have to withstand massive cumulative particle fluences leading to significant deposition over long time scales not yet experienced on short pulse devices. The lack of understanding in both managing the release of W into the Scrape-Off Layer (SOL) to maintain performance (avoiding radiative collapse via high-Z core accumulation) and controlling material build-up without creating copious redeposition layers are serious uncertainties in qualifying W as a viable PFC material in future devices. A series of experiments were conducted in KSTAR and WEST to characterize impurities by varying strategically the plasma shape in front of hypothetically high impurity sourcing regions. In KSTAR, the lower X-point height was increased in upper (USN) and lower (LSN) single null to see the effect of the screening efficiency of the lower divertor. The impurity sourcing and core contamination increased with the lower X-point height. In WEST, the primary separatrix driven away from the upper divertor in LSN led to the upper divertor source being strongly reduced (-90% deposited power, -40% radiated power, -60% of W and -80 to -90% of light impurity source) while the lower inner target source significantly increased (+70% heat load, +40% radiated power, +55% of W and +180 to +380% of light impurity source) measured from the recently developed integrated multidiagnostics suite. Reducing the expectedly poorly screened upper divertor impurity source contribution but increasing the strongly screened lower inner divertor source should be beneficial for plasma performance by reducing the W core concentration. However, experimental results showed only a slight decrease of the core radiated power (-6%), while central Te decreased (-11%) and W central concentration increased (+27%). Interpretative modelling with the SOLEDGE3X-ERO2.0 codes will be presented to discuss the non-intuitive experimental results from WEST and KSTAR, highlighting the complex interplay between sources and core pollution.
This work is supported by the U.S. DOE under Grant Number DE-SC0020414.

Speaker: Alex GROSJEAN (University of Tennessee - Knoxville (UTK))

76 1.071 Turbulent W core influx in negative triangularity DIII-D with Flan

The turbulent trace impurity transport code Flan is used understand what causes tungsten (W) to cross the last closed flux surface (LCFS) in a limited negative triangularity (NT) DIII-D discharge from a hypothetical W source. A turbulent representation of an NT discharge is simulated with the gyrokinetic solver Gkeyll. Time-dependent fluctuating quantities like the electron density and electric field are passed to Flan. Flan is a Monte Carlo particle-in-cell code that transports impurities via the Lorentz force and uses the Nanbu collision model to handle deflections with the background plasma. It is applicable in all regions of the plasma as long as the background plasma is valid and the impurities exist in a trace amount. Flan is ideal for turbulent simulations where the guiding center approximation may not be satisfied because the full impurity gyro-orbit is resolved. For the discharges studied, W was not actually in the vessel at the time. Therefore, a hypothetical W10+ source 3 cm away from LCFS in the scrape off layer (SOL) is used to motivate future experiments when DIII-D transitions to a full-W wall. Radial W fluxes along the LCFS show that W tends to be expelled from the core at the outboard midplane side of the plasma and generally enters the core from everywhere else with speeds on the order of ~1,000 m/s. This is roughly interpreted as caused by the E×B drift but strictly speaking it is a turbulent phenomenon because the W Larmor radius at the LCFS is comparable to the spatial scale of the fluctuations of the electric field (the guiding center drift approximation is violated). Next, a mockup of a laser blow-off (LBO) experiment is run in Flan to compare the particle confinement time for various impurities in NT. The confinement times (~10 μs) are generally too small compared to experimental expectations (~1 ms) because the entire core is not simulated, but the particle confinement time increases with atomic number. This qualitatively agrees with neoclassical transport theory and motivates dedicated LBO experiments in future campaigns.
Work supported by US DOE under DE-FC02-04ER54698 and DE-AC02-09CH11466.

Speaker: Shawn Zamperini (General Atomics)

77 1.072 Advancing Fusion Pilot Plant Readiness through the Tungsten Wall Transition on DIII-D

The DIII-D National Fusion Facility is preparing for a transition from graphite to tungsten (W) plasma-facing components to enable reactor-relevant studies of plasma scenarios, core-edge integration, and plasma–material interaction for next-step fusion devices.

Leveraging DIII-D’s unique capability for short-pulse, high-performance operation with flexible shaping, advanced actuator control, and comprehensive diagnostics, the new metal-wall environment would also accommodate innovative divertor concepts and iterative design optimization. This upgrade would directly support the mission of future fusion pilot plants (FPPs) by enabling validation of plasma boundary physics, impurity transport, and material performance in high-Z conditions. The Full Wall Change-Out, targeted for 2029, would proceed in stages, prioritizing W installation in the highest heat and particle flux regions while preserving flexibility for future expansion.

A structured Physics Validation Review (PVR) is underway to ensure compatibility of the full-metal configuration with DIII-D’s present scenario portfolio. Preliminary PVR results suggest that key advanced scenarios, including Advanced Tokamak, Negative Triangularity, and Quiescent H-mode, remain viable in high-Z environments, provided core W accumulation is effectively managed. Active mitigation tools under evaluation include strike-point sweeping, ECH impurity flush-out, seeding of radiative impurities, and optimized plasma shaping.

Updated analysis of heat fluxes and fast ion losses identifies the main strike-point regions, upper inner wall, and upper baffle as critical areas requiring either bulk W or tungsten coatings. The remaining wall can be made out of stainless steel or molybdenum without significantly increasing high-Z impurity levels. Erosion estimates confirm that W-based alloys self-stabilize to W-like surfaces within hours of operation. Runaway-electron calculations show that beam currents up to 500–900 kA remain benign for W, assuming melting limits of 3–5 MJ m⁻² and deposition areas of 0.1–0.5 m². Maintaining tile alignment within ≤ 0.5 mm ensures that even leading edges stay within safe thermal limits.

Additional infrared and visible cameras, thermocouples, and interlocks for surface monitoring would be required to enable safe metal-wall operation. The PVR incorporates lessons from JET and ASDEX-Upgrade showing that robust startup, ELM, and impurity control are achievable in full-metal environments with active plasma–surface monitoring. The tungsten wall at DIII-D will provide a versatile platform to advance and validate the physics basis and control strategies for ITER; long-pulse tungsten devices such as EAST, WEST, and KSTAR; and future FPPs.

This work was supported by the US DOE under DE-AC02-09CH11466, DE-FC02-04ER54698, DE-NA0003525, DE-SC0014264, DE-AC05-00OR22725, DE-SC0023378, DE-AC52-07NA27344, and DE-SC0022270.

Speaker: Tyler Abrams (General Atomics)

78 1.073 Kinetic modeling of tungsten transport under electromagnetic perturbation induced by Low-n X-point mode

This study investigates the transport of tungsten impurities induced by the low-n X-point mode (LNXM) in plasma, utilizing the full-orbit particle code PTC[1] and the turbulence code Hermes-3[2]. LNXM, observed during boron and carbon injection experiments in EAST[3,4], enhances plasma transport and shows potential for impurity expulsion in small or no edge-localized mode (ELM) regimes. The model of impurity-induced LNXM is based on this paper[5]. Our findings reveal that LNXM drives both inward and outward radial transport of tungsten impurities, with outward transport being predominant. The radial distribution range of LNXM is relatively narrow and influenced by impurity distribution, suggesting that effective management of impurities could further enhance LNXM's transport efficiency and shielding effectiveness.

We found that LNXM's inward transport may cause short-term accumulation of wall-sputtered tungsten particles in the edge region, but its stronger outward transport keeps these particles confined to the edge, preventing deeper penetration into the core. This study highlights the promise of LNXM as an alternative to ELMs for radiative energy dissipation and impurity control in high-confinement operational regimes. Furthermore, LNXM exhibits a narrower perturbation range and greater sensitivity to impurity distribution compared to ELMs. Future work will focus on characterizing LNXM's impact on tungsten particle velocity distributions, transport coefficient and their temporal evolution, alongside a comparative analysis of ELM effects on tungsten distribution and velocity profiles based on previous work.

References
[1] C. Team, F. Wang, R. Zhao et al, PTC: Full and Drift Particle Orbit Tracing Code for α Particles in Tokamak Plasmas, Chinese Phys. Lett. 38 (2021) 1–8.
[2] B. Dudson, M. Kryjak, H. Muhammed et al, Hermes-3: Multi-component plasma simulations with BOUT++, (2023).
[3] Z. Sun, A. Diallo, R. Maing et al, Suppression of edge localized modes with real-time boron injection using the tungsten divertor in EAST, Nucl. Fusion 61 (2020) 14002.
[4] Y. Ye, G.S. Xu, Y.Q. Tao et al, Sustained edge-localized-modes suppression and radiative divertor with an impurity-driven instability in tokamak plasmas, Nucl. Fusion 61 (2021).
[5] H. Chang, B. Dudson, J. Sun et al, Hermes-3 simulation of the low-n X-point mode driven by impurity in tokamak edge plasmas, Nucl. Mater. Energy 43 (2025) 101913.

Speaker: Huayi Chang (Dalian University of Technology)

79 1.074 Linking simulated ICRF sheath-rectified potentials to measurements in ASDEX Upgrade and JET

Following the re-baselining of ITER aimed at operation with full tungsten wall, a renewed interest in studies of the impurity sources during operation of Ion Cyclotron Range of Frequencies (ICRF) antennas emphasizes the importance of validation of the modelling tools simulating the sheath-rectified DC potentials $V_{DC}$. We use the SSWICH-SW code coupled with RAPLICASOL or TOPICA, in order to establish comparisons between calculations and experiments in ASDEX Upgrade (AUG) and JET, increasing the confidence of the code predictions.

In AUG, SSWICH-SW was successfully used in the past to describe the reduction of impurity production by the 3-strap antenna compared to the 2-strap antenna in L-modes. Using recent measurements by a retarding field analyzer to obtain $V_{DC}$ in H-modes with small ELMs, a good agreement with the experimental data is obtained for spatially resolved cases with various 3-strap antenna feedings: optimized dipole, misbalanced power dipole and -90°. This requires a careful coupling between RAPLICASOL (or TOPICA) which calculates near-field $E_{||}$ and 2D SSWICH-SW which calculates $V_{DC}$ taking into account the slow wave dispersion, the RF sheath boundary condition and DC current transport. Defining the interface plane for coupling between the codes at the radial position of the Faraday Screen (FS), as has been used previously, is insufficient in H-mode, despite the plasma density profiles lying above the lower hybrid density as in L-modes: the important $E_{||}$ contributions from the RF image currents on antenna limiters are not treated adequately due to a different current pattern than in L-modes. To properly take these into account, the interface plane radially in front of the antenna limiters should be chosen, or at least 5 mm in front of the AUG antenna FS.

In JET, direct $V_{DC}$ measurements are not available. The RF sheath behavior induced by the 4-strap A2-antenna D is inferred by measuring BeI line intensity at the magnetically-connected outboard limiters, as a function of the power balance between the central straps ($P_{cen}$) and the outer straps ($P_{out}$) in $(0-\pi-0-\pi)$ and $(0-\pi-\pi-0)$ phasings. Trends of dependencies on $P_{cen}/(P_{cen}+P_{out})$ are well described by spatially averaged $|E_{||}|$ , but show a sharper reaction than the calculations of $V_{DC}$ by SSWICH-SW which compare better with the experiments. A heuristic Be-sputtering model by deuterium with the sputtering yield $Y_{Be}=Y_{Be}(V_{DC})$ provides an improved agreement between the calculations and the experiments, but the results become more sensitive to assumptions on $V_{DC}$ averaging.

Speaker: Dr Volodymyr Bobkov (Max-Planck-Institut für Plasmaphysik, Boltzmannstraße 2, 85748 Garching, Germany)

80 1.075 SOLPS-ITER modeling of neon/argon-seeded EAST plasmas and its effects on tungsten impurity behaviors

The transport of neon (Ne) and argon (Ar) impurities, along with their impact on tungsten (W) impurity behaviors in EAST plasmas, was numerically investigated using the SOLPS-ITER edge modeling package. The key findings are as follows. Results show that E×B drift significantly modifies target profiles of plasma density (ne) and temperature (Te) by regulating divertor particle transport, thereby strongly influencing W sputtering. In the absence of drifts, radial particle transport into the far divertor SOL is minimal, resulting in low ne and high Te—conditions that promote significant W sputtering in this region. When drifts are included, they shift the peak positions of ne and Te at the target and induce pronounced in–out/up–down divertor asymmetries. Notably, in double-null configurations, the in–out asymmetry in Te exhibits a distinct dependence on the Ne/Ar puffing rate (Γpuff) compared to single-null plasmas. W sputtering can be effectively suppressed by increasing Γpuff, but only when E×B drift substantially enhances radial particle transport into the far divertor SOL. In drift-free cases, low Γpuff leads to stagnation points in the poloidal velocity within the divertor, causing considerable impurity leakage through the near SOL. This leakage can be mitigated by raising Γpuff. Conversely, at high Γpuff, Ne/Ar seeding affects the deuterium ionization source in the divertor, redistributing Ne/Ar density among different divertor regions via the main SOL. In full-drift scenarios, E×B drift dominates the poloidal flow of impurity ions, similar to single-null cases. However, in double-null configurations, the stagnation point and divertor retention are less sensitive to variations in Γpuff. Neglecting drifts leads to a substantial overestimation of impurity densities near the core boundary. The spatial distributions of ionization sources from Ne⁺, Ar⁺, and W⁺ neutrals differ, resulting in distinct ion flux patterns for each species. Importantly, increasing Γpuff helps reduce W core leakage. Further studies employing the wide-grid SOLPS-ITER v3.2.0 are underway to investigate far-SOL impurity transport and its effects on edge plasma behavior and core impurity content.

Speaker: Fuqiong Wang (Donghua University)

81 1.076 Operational space exploration of the X-Point Radiator scenario in WEST

Among the potential reactor-relevant divertor regimes of operation, the “X-Point Radiator” (XPR) regime [1] which features a radiating MARFE at the X-Point and strongly reduced target heat loads, is a strong contender. This regime is investigated extensively in the WEST tokamak on the actively cooled ITER-grade tungsten divertor [2], over a wide range of operational parameters (density, power, magnetic configurations), and characterized in a database of over a hundred discharges. XPR’s in WEST are demonstrated to be real time controllable with a simple interferometry signal, and highly repeatable.
Upon sufficient seeding, the divertor plasma eventually condenses within microseconds into a stable, dense and cold (Te < ~5 eV at both targets) regime, followed in the millisecond timescale by the formation of a characteristic radiating MARFE at the X-Point. Divertor heat loads and tungsten erosion are strongly reduced. Nonetheless, a finite ion flux is maintained on both divertor targets, and its deposition pattern changes. Main scrape-off layer profiles are weakly impacted by the XPR appearance, but SOL flows do reorganize, as a consequence to the change in target density and temperature. Core confinement is improved through a combination of ion dilution effects and reduced tungsten contamination.
Statistical analyses are performed on the WEST XPR database, which now includes long and repeated pulses from the new XPR high-fluence campaign. They show the influence of operational parameters (plasma current, density, input power) on divertor particle and heat flux deposition and radiated power patterns. Nitrogen visible spectroscopy signals informing on divertor impurity concentration and sources are also linked to core plasma performance metrics including central ion and electron temperatures and effective charge.
ACKNOWLEDGEMENTS
This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No 101052200 — EUROfusion). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the European Commission can be held responsible for them.
References:
[1] M. Bernert et al. Nucl. Fusion, 61 (2020) 024001
[2] J. Bucalossi et al., Nucl. Fusion, 64 (2024) 112022

Speaker: Nicolas RIVALS (CEA)

82 1.077 The effects of lithium plasma-facing surfaces on the scrape-off layer and core plasma in tokamaks

An overview of the effects of low recycling lithium walls, plasma-limiting surfaces, and divertor targets on the scrape-off layer (SOL) and core confined plasmas is presented. The discussion will primarily reference the tokamak, although consequences for alternative magnetic confinement configurations (primarily stellarators and mirrors) will also be discussed. For the SOL plasma, reduction of recycling to minimal levels drastically lowers the collisionality. Trapped particle populations develop which are mirror confined, producing a confined plasma with finite pressure in the SOL. The disparity between electron and ion pitch angle scattering times leads to the development of an ambipolar or Pastukhov potential in the SOL, which can produce drifts and transport which are significant on the SOL “confinement time”. In addition to a reduction in main (and impurity) ion recycling, a clean lithium surface has the lowest secondary electron emission (SEE) coefficient known, with a SEE coefficient peaking at ~ 0.5, to reduce SOL cooling from electron recycling as well. Although a divertor for a reactor concept may be designed with accurately controlled, near-tangential magnetic field lines to prevent introduction of secondary electrons to the SOL, such precise magnetic control is not always available on medium-scale devices. The large reduction in secondary electron production, with an SEE coefficient of 0.1 – 0.2 at relevant energies, produced by a lithium surface relaxes magnetic field design requirements at the divertor target. For the core plasma, attention has focused on the ability of lithium to generate flat temperature profiles for both electron and ion populations. But low recycling lithium walls can also mitigate charge exchange losses in the core plasma, by strongly reducing the neutral population, even in small scale devices (like LTX-beta). The confinement properties of novel magnetic configurations (e.g. quasisymmetric stellarators, or magnetic mirrors) can thus be accurately assessed at much smaller scale than would otherwise be possible with high recycling walls, where charge exchange losses can reduce ion energy confinement, and mask any reduction in turbulent transport. Although results from LTX-beta will be cited here, the intent is to call attention to the broader advantages offered by the use of low recycling walls on a variety of novel, smaller scale, magnetic confinement configurations.

Work was supported by US DoE contract DE-AC02-09CH11466.

Speaker: Dick Majeski (PPPL)

83 1.078 Increased power and performance at Wendelstein 7-X: effect on the scrape-off layer transport and first attempts of combined core/edge scenarios

The plasma exhaust concept of the Wendelstein 7-X (W7-X) stellarator is based on the island divertor configuration, which exploits the interaction of magnetic islands with ten discrete water-cooled targets. This translates into a scrape-off layer (SOL) with long connection lengths and small field line pitch angles, affecting the balance of the different transport channels, favouring the perpendicular (radial, binormal) over the parallel.
In the most recent campaigns, the operational space of W7-X was extended including input powers above $7$ MW (i.e. $P_{SOL} > 6.5$ MW). Under these conditions, line-averaged densities $\langle n_e \rangle$ of up to $11.3 \cdot 10^{19}$ m$^{-3}$ were sustained while maintaining the plasma attached (i.e. $f_{rad}=P_{rad}/P_{in} < 0.45$). This increased range of densities allowed the observation of features that had not been detected in lower-power/density attached experiments. Bolometer measurements show variations in the toroidal radiation distribution that strongly reduce with increasing $\langle n_e \rangle$ rather than with $f_{rad}$, as implied by low-power observations. At constant $\langle n_e \rangle$, radiation asymmetries are stronger with increasing power. Moreover, higher divertor densities were observed at higher power, indicating a more beneficial, nearly-quadratic, scaling with respect to the last-closed-flux-surface density compared to what previously measured. These increased divertor densities were accompanied by higher neutral pressures and stronger parallel flow velocities. Additionally, the SOL density profiles showed steeper radial gradients inside the magnetic islands, correlated with increased density/temperature fluctuations, therefore possibly driving higher perpendicular (mostly radial) transport. This change in transport resulted in heat loads being deposited on plasma-facing components not designed to withstand them. This prevented safe attached operation ($f_{rad}< 0.45$) within the mid-density range ($6$ to $9 \cdot 10^{19}$ m$^{-3}$) in some of the most common W7-X magnetic configurations for heating powers above $4 – 5$ MW.
Due to the challenges of operating under the described attached conditions, significant effort was invested in developing a feedback-control system on the radiated power. Its successful implementation enabled the first attempts at combined scenarios with both high core performance ($T_i > 1.5$ keV) and acceptable divertor heat loads. However, the improved performance of these experiments is associated with suppressed core turbulence, which reduces the typical diffusive impurity flushing. This results in a strong accumulation of both intrinsic and seeded impurities in the core, ultimately limiting the plasma stored energy via radiative cooling. This highlights the need to further optimise impurity seeding and divertor density in order to minimise the impact on core performance.

Speaker: Valeria Perseo (Max Planck Institute for Plasma Physics - Greifswald)

84 1.079 Physics of the decay length broadening and asymmetry: Insights from SOLPS modelling

Due to its high magnetic field SPARC is expected to have a narrow heat-flux width (λq~0.5mm) in the similar range to ITER, making power exhaust a central challenge. Previous studies explored SPARC’s operational space through density and impurity scans [1, 2], but the diffusivity values (D, χ) used to reproduce λq are not uniquely constrained. Recent experiments reported upstream decay lengths broadening, extending the multi-machine power width scaling by up to a factor of ~2.5 near the density limit. This work examines how much of this broadening can be explained by parallel transport toward the divertor, and to what extent increased radial turbulent transport contributes [3].
For SPARC conditions in SOLPS, λq broadening appears with increasing density even at fixed D and χ, driven by divertor conditions (similar to ITER [4,5]), reaching factors of 2-3, consistent with experiments [3,6]. Additional broadening arises from thermo-electric current contributions, which become significant under SPARC’s high density, and narrow decay lengths condition [2].
We investigate the impact of D and χ (D/χ = 10, 1, 1/3; χ = 0.03-1) across two SPARC scenarios, full-field (12.2 T, 29 MW) and 2/3-field (8 T, 10 MW) [1]. Separatrix density scans from 1/4 to 1/2 Greenwald reproduce the reported lam-T (and thus lam-q) broadening, though not the much larger widening of the density channel. Notably, scans at fixed D and χ capture a substantial portion of the turbulence correlated broadening observed on AUG, supporting the continued use of constant transport coefficients for reproducing first-order broa\dening trends.
An additional requirement for SPARC is to avoid abrupt changes in the in/out power balance. Analysis of parallel flow asymmetry shows that lower D and χ increase power to the divertor while reducing radial losses. The D/χ = 10 cases show the strongest in/out power imbalance, triggering the HICO → HOCI transition [2], while increasing density or choosing different D/χ combinations mitigates this asymmetry. These results identify an operational regime that avoids strongly asymmetric divertor conditions, requiring a minimum collisionality of αt ~ 0.1-0.5 and ~1/4 Greenwald fraction at the separatrix.

[1] Kuang A. Q. et al., Journal of Plasma Phys 86.5 (2020) 865860505.
[2] Lore J. D. et al., Nucl Fusion 64.12 (2024) 126054.
[3] Eich T. et al., Nucl Fusion 60.5 (2020) 056016.
[4] Park J.-S. et al., Nucl Fusion 64.3 (2024) 036002.
[5] Rivals N. et al., Nucl Fusion 65.2 (2025) 026038.
[6] Brown, Goldston, NME (2021) 101002

Speaker: Jae-Sun Park (Oak Ridge National Laboratory)

85 1.080 Detachment of type-I ELMy H-mode TCV plasmas via nitrogen seeding and correlations with scrape-off layer properties

This contribution presents first experimental observations on the effect of nitrogen (N₂) seeding on the scrape-off layer (SOL) properties of type-I ELMy H-mode plasmas in the TCV device. This work is motivated by the significant uncertainties still persisting in the physical understanding of SOL transport processes in seeded regimes, which represents a crucial aspect for reliable extrapolation to reactor scenarios. Indeed, impurity seeding is among the most attractive solutions for mitigating the heat and particle loads traveling in the SOL region towards material surfaces. The addition of extrinsic impurities leads to significant divertor momentum and power dissipation, thereby detaching the divertor targets and reducing the exhaust fluxes reaching the target plates [1]. Further benefits of seeded impurities have been demostrated on improving particle confinement [2] and buffering or suppressing edge-localized modes [3,4].

Herein, two datasets have been acquired in TCV at respectively low density, high power and high density, medium power. The deuterium (D₂) gas puffing and the N₂ seeding rates have been varied independently to disentangle their effect on the SOL profiles and transport characteristics. N₂-seeded phases are associated with an increase in average effective ion charge, core radiation level and, at high density, a more irregular ELM frequency, possibly indicating core impurity accumulation. Spectrally filtered images of the outer leg show detachment of the CIII emission front from the machine floor during N₂ seeding, although limited to inter-ELM phases only. At high density, N₂ injection does not correlate with modifications in either the near SOL density gradient length or far SOL shoulder amplitude, in line with previous TCV L-mode results [5]. At low density, higher D₂ fueling is associated to shallower SOL density gradients, supporting prior TCV results [6]. Conversely, in the presence of N₂ seeding the near SOL density profile becomes considerably steeper and the far SOL shoulder amplitude is distinctly reduced. These findings, as well as the observed differences, make interpretation of the underlying physical mechanisms challenging and call for further experimental investigation, focusing in particular on ionization sources and turbulent fluctuation properties.

[1] P.C. Stangeby (2018) Plasma Phys. Control. Fusion 60 044022
[2] C. Giroud et al (2025) 30^th IAEA-FEC (Chengdu, China)
[3] M. Zurita et al (2026) This conference
[4] M. Bernert et al (2021) Nucl. Fusion 61 024001
[5] O. Février et al (2020) Plasma Phys. Control. Fusion 62 035017
[6] A. Stagni et al (2022) Nucl. Fusion 62 096031

Speaker: Adriano Stagni (Consorzio RFX)

86 1.081 Surface and interface damage of ITER-like W/Cu monoblocks in the lower divertor of EAST

Maintaining the integrity of plasma-facing components (PFCs) under high thermal loads is critical for ITER and other future fusion devices. In 2021, ITER-like W/Cu monoblocks with a large chamfer (1.5 × 17 mm) were installed on EAST’s lower divertor to evaluate their in-service behaviour. Combining high-resolution infrared (IR) thermography, thermal modelling, and post-mortem microstructural analysis, we report both surface and interface degradation that progressed over successive campaigns (spring 2021 → autumn 2021 → spring 2022). IR cameras frequently recorded bright hot spots—affecting ≈41% of modules in the toroidal direction—whose spatial distribution correlated with local heat flux and installation misalignment. Post-mortem inspection revealed extensive surface damage (crust formation, increased roughness, net-like macro/microcracking) and abnormal grain growth near melted regions (grain sizes up to 7.1 mm). High-heat-flux tests and simulations indicate that interface debonding markedly reduced interfacial thermal conductance: from baseline values down to ≈2×10^4 W·m⁻²·K⁻¹ after ~4,800 discharges (yielding an ~18% increase in steady-state surface temperature) and further to ≈5×10^3 W·m⁻²·K⁻¹ with continued operation (≈31% temperature rise). Progressive interface cracking produced local temperature excursions that promoted incipient melting at leading edges; with auxiliary heating above 10 MW, more extensive melting along the 17 mm chamfer is anticipated. Microchemical analysis also identified reactions between molten W and low-Z impurities, including W₂C formation at extreme temperatures. Notably, prior to widespread melting, the observed interface cracks exerted only a negligible influence on plasma performance. These observations quantify critical degradation pathways for ITER-like W/Cu monoblocks and provide reference data for assessing long-term heat-removal capability, refining HHF test protocols, and informing manufacturing/qualification practices (e.g., welding and joint standards) for ITER and future fusion reactors.

Speaker: Yang Wang (Institute of Plasma Physics, Hefei Institutes of Physical Science Chinese Academy of Sciences, Hefei 230031, China)

87 1.082 Passive detachment front stabilization in baffled divertors using upstream pumping

${}$
Scrape-off layer (SOL) transport simulations including magnetic and ExB drifts flows, performed with the multi-fluid edge-plasma code UEDGE, predict passive stabilization of the detachment front along the low-field side (LFS) divertor leg when main-SOL pumping upstream of the LFS divertor target is applied. With the ion $\mathbf{B}\times\nabla\mathrm{B}$ drift directed into the divertor, the $\mathrm{T_e}=1.5~\mathrm{eV}$ detachment front stabilizes at the pumping plenum with increasing main ion gas injection, maintaining $\mathrm{T_e}\gt80~\mathrm{eV}$ at the X-point, when the pump duct entrance is located 7-9 cm upstream of the divertor target. Neutral pressure buildup downstream of the pumping plenum, necessary to remove the injected particles and obtain particle-balance, induces particle, momentum, and power losses through plasma-neutral interactions. These plasma-neutral losses detach the LFS divertor without the radiation front moving upstream to the X-point, as typically observed for DIII-D [1,2]. UEDGE predicts plasma burn-through of the target-shielding neutrals for pump-to-target distances below 7 cm, whereas pump-to-target distances longer than 9 cm are predicted to reduce the interval of gas-injection rates for which $\mathrm{T_e}\gt80~\mathrm{eV}$ at the X-point are sustained.
${}$
The simulations are based on a 12.5 MW DIII-D Stage 2 divertor reference geometry with a 6 cm wide pumping plenum situated 25-31 cm downstream of the X-point [3,4]. Predictive simulations performed with reduced X-point–to–pump distances indicate passive stabilization of the detachment front can be sustained for 17 cm X-point–to–pump separation with similar development of the X-point Te as a function of gas injection rate in the reference geometry. Passive stabilization of the detachment front is also predicted by UEDGE when the divertor is pumped upstream of the target on the private-flux side but at higher gas injection rates. The UEDGE simulations, evaluated for steady-state conditions, consider intrinsic carbon and seeded neon impurities as separate neutral and ion species in the simulations. Experimental measurements from the highly diagnosed dissipation-focused Stage 2 Chimney divertor, expected operational in 2027, will provide critical validation of the predictive UEDGE simulations, used in its design, underpinning FPP-relevant divertor design. Carbon-free UEDGE simulations of a DIII-D–like metallic FPP, assessing detachment-front stabilization with upstream pumping in the absence of intrinsic carbon, will also be presented to guide FPP design.
${}$
[1] A.E. Jaervinen, et al., Nucl. Fusion $\mathbf{60}~(2020)~056021$
[2] A.G. McLean, et al., J. Nucl. Mat. $\mathbf{463}~(2015)~533$
[3] J. Yu et al., Nucl. Mat. Energy $\mathbf{41}~(2024)~101826$
[4] A. Holm et al, Mat. Energy $\mathbf{41}~(2024)~101782$
${}$
$\mathrm{This}\;\mathrm{work}\;\mathrm{was}\;\mathrm{supported}\;\mathrm{in}\;\mathrm{part}\;\mathrm{by}\;\mathrm{the}\;\mathrm{US}\;\mathrm{Department}\;\mathrm{of}\;\mathrm{Energy}\;\mathrm{under}$ $\texttt{DE-FC02-04ER54698}$, $\texttt{DE-AC52-07NA27344}$, and $\texttt{DE-AC05-00OR22725}$. $\texttt{LLNL-ABS-2014170}$

Speaker: Andreas Holm (Lawrence Livermore National Laboratory)

88 1.083 Enhanced Turbulent Transport and Drift Suppression in Divertors at the Detachment Onset

At the onset of divertor detachment, enhanced cross-field transport can modify particle and heat fluxes to plasma-facing components. Using fast reciprocating Langmuir probe measurements in the TCV tokamak, we investigate turbulent transport and profile evolution in single-null (SN) and snowflake (SF) divertors during the transition from attached to onset of detachment conditions.

Measurements show that radial turbulent particle transport increases by a factor of $2$-$4$ upstream of the dissipative divertor region in the near scrape-off layer. This enhanced transport leads to a broadening of density profiles upstream of the dissipative region, with the density decay length increasing by a factor of $4$-$5$, while electron temperature profiles flatten along the divertor leg. The particles therefore enter the dissipative divertor already spread out. The broadened profiles are subsequently convected through the divertor by parallel transport, with a transition of parallel heat transport from conduction- to convection-dominated. Within the dissipative divertor, plasma potential profiles collapse and flatten, reducing radial electric field gradients and suppressing local $E\times B$ drift transport. Radial transport at the detachment onset is therefore driven by enhanced turbulent spreading upstream of the dissipative region rather than local cross-field transport in the near-detached divertor.

These results highlight the importance of upstream turbulent transport in setting divertor density and heat flux profiles in near-detached conditions, with direct implications for footprint broadening and power handling in advanced divertor configurations.

Speaker: Rabel Rizkallah (UCSD)

89 1.084 Effects of triangularity on power asymmetry and detachment in TCV double-null plasmas using SOLPS-ITER with drifts

Double-Null (DN) configurations are a leading candidate for future tokamak power exhaust management, yet the precise impact of the magnetic configuration on divertor asymmetry remains a critical challenge. This work employs SOLPS-ITER simulations with full drifts to investigate the interplay between triangularity ($\delta$) and transport in TCV L-mode DN discharges, focusing on the physics of detachment. The simulations reveal distinct roles for drift mechanisms in determining divertor asymmetry: $E \times B$ drifts are found to primarily drive in-out asymmetry, while diamagnetic drifts significantly contribute to the up-down asymmetry. Crucially, these drift effects are modulated by the presence of port protection tiles, which modify the effective divertor leg lengths and closure, thereby altering the power sharing and detachment access.Regarding detachment, defined as a target electron temperature ($T_e < 5$) eV, Negative Triangularity (NT) configurations are found to be inherently more difficult to detach than Positive Triangularity (PT). Simulations show that core density ramps alone fail to cool the NT outer target below this 5 eV threshold. This resistance to detachment is attributed to a combination of higher heat flux flowing to the Low Field Side (LFS), a physically shorter outer divertor leg which limits the power dissipation volume, and a narrower power fall-off length ($\lambda_q$) accompanied by a lower spreading factor ($S$). To mitigate these geometric and transport limitations, Nitrogen ($N_2$) seeding is investigated. The simulations demonstrate that $N_2$ seeding effectively accesses a high-radiation regime, enabling detachment at the outer target without the upstream profile degradation associated with pure density ramps. Finally, to strictly disentangle the fundamental effects of shaping from topological variations, a specific controlled case with matched divertor leg lengths and closure is analyzed; this confirms that triangularity intrinsically modifies the drift-driven transport patterns even when geometric parameters are fixed. These results indicate that realizing the benefits of DN in reactor designs like DEMO requires optimizing the magnetic shape to manage the specific coupling between triangularity-induced leg length variations and drift-driven plasma transport.

Speaker: Yanjie Zhang (Nanyang Technological University)

90 1.085 Divertor Turbulence Control via Leg Geometry: Experimental Tests on TCV tokamak

Understanding the mechanisms that govern heat and particle transport in the divertor region is critical for the design and operation of future fusion reactors. Turbulent cross-field transport plays a key role in determining the heat flux distribution at divertor targets, affecting both the peak heat load and the overall power exhaust scenario. A key metric for characterizing heat flux spreading is the $S$ parameter in the Eich’s fit. While previous studies have investigated $S$ under various conditions [1, 2, 3, 4], its behavior remains poorly constrained. A new generation of 3D plasma turbulence codes has been developed to tackle edge/SOL turbulence in diverted geometry (e.g. [5,6]), even revealing unexpected results under reactor-relevant conditions [7], highlighting the challenges of extrapolating current tokamak experiments to reactor scenarios in the edge and SOL regions.
In this contribution we present an experimental test for such codes, based on the theoretical ideas proposed in [8]. The key idea is that the relative orientation of the magnetic curvature vector $\kappa$ and the pressure gradient $\nabla p$ determines the turbulence transport into the private flux region of the two divertor legs, depending on their geometry. This alignment can either destabilize or stabilize drift-interchange and drift-wave turbulence. Controlling the turbulence drive and its impact on transport through diverted plasma geometry represents a promising approach in the design of alternative divertor configurations. Leveraging the shaping flexibility of the TCV tokamak, a scan of the outer divertor leg orientation has been performed while maintaining matched upstream plasma parameters and target flux expansion. Initial results from Langmuir probes and infrared analysis do not show any significant difference in the divertor spreading parameter $S$. However, according to [8], the inner divertor leg is expected to be more affected, as the $\kappa$ and $\nabla p$ drifts are parallel when the leg is vertically oriented. A database analysis is currently underway, based on discharges in negative triangularity, to further investigate these effects. Moreover, multifluid plasma simulations of transport and turbulence using the Hermes-3 [9] code are foreseen to independently validate the initial hypothesis of a change in divertor turbulence transport.

Speaker: Marco Cavedon (University of MIlano-Bicocca)

91 1.086 A Numerical Study of the Isotope Effect on the Scrape-Off Layer Density Shoulder Formation in JET-ILW plasmas

nHESEL simulations of L-mode JET ITER-like wall (JET-ILW) plasmas in hydrogen, deuterium and tritium isotopes reveal a non-linear dependence of scrape-off layer (SOL) transport on the isotope mass number A. When upstream profiles are held fixed across isotopes, both the perpendicular turbulent fluxes and the parallel convective fluxes exhibit systematic but non-linear trends with A. The simulations show that increasing the isotope mass reduces the instability-driven cross-field transport while simultaneously altering the balance with parallel losses, leading to modified density and pressure profile gradients in the SOL. These combined effects give rise to a more pronounced density shoulder at the outer midplane for higher-A isotopes.

The formation and strength of the density shoulder are found to depend on the re-distribution between perpendicular and parallel transport channels, as well as their coupling to local ionization and charge-exchange sources from hydrogen isotope neutrals. A variation in the isotope mass number produces a strongly A-dependent response in the relative transport fractions, thereby affecting the onset and radial extent of the shoulder into the far SOL region.

nHESEL is a drift-reduced two-fluid model that evolves electron density, electron and ion pressures together with the generalized vorticity in a two-dimensional cross-field domain applied to the JET outer midplane [1, 2]. Parallel losses are included through parameterized sink terms, while interactions with dynamic neutral species (Franck-Condon atoms, charge-exchanged atoms, and molecules) provide self-consistent volumetric sources.

The simulations demonstrate how isotope-dependent transport re-balancing affects both upstream profile evolution and the parallel heat flux redistribution. The predicted parallel heat flux projected onto the divertor target [3] yield target heat-flux patterns consistent with JET measurements [4] and previous EDGE2D-EIRENE modelling [5]. This agreement supports the use of nHESEL as a synthetic diagnostic for studying isotope effects on SOL transport and density shoulder formation in future devices and for ITER-relevant conditions.

[1] A.S. Thrysøe et al., Physics of Plasmas, Phys. Plasmas 25, 032307 (2018)
[2] A.H. Nielsen et al., Plasma Phys. Control. Fusion 59 025012 (2017)
[3] B. Sieglin et al. Plasma Phys. Control. Fusion 58 055015 (2016)
[4] C.F. Maggi et al. Plasma Phys. Control. Fusion 60 014045 (2018)
[5] M. Groth et al, Nuclear Materials and Energy 34 10134 (2023)

Speaker: Alexander Simon Thrysøe (DTU)

92 1.087 3D detachment in the new upper divertor of ASDEX Upgrade induced by a local gas-puff

Toroidal symmetry is one of the basic design principles of a Tokamak. ASDEX Upgrade (AUG) has recently undergone a major hardware upgrade [Herrmann FED 2017,Herrmann FED 2019,Teschke FED 2019] in order to study alternative divertor configurations [Lunt NME 2017,Pan PPCF 2018]. Apart from a pair of in-vessel coils, a new inner and outer upper target and a new cryo-pump a large set of new edge diagnostics as well as new gas valves located at several toroidal locations have also been installed. Due to the fact that the total flux of ions to all plasma facing components is typically larger than the injected deuterium flux by more than an order of magnitude, the local gas puff is typically regarded as a small perturbation and its exact position to be rather irrelevant. However, in the 2025 AUG campaign substantial toroidal asymmetries were observed in the divertor shunt current measurements [Giannone EPS 2025]. Puffing deuterium with a gas valve located near the strike line in segment 14 the currents in segments 2, 6 and 10 are approximately equal at -10 A per tile, while the one in segment 14 shows a dramatic increase to about +40 A. The phenomenon is explained by a toroidally localized detachment of the plasma in segment 14 accompanied by a reversal of the thermo-currents. The toroidal asymmetry in the thermo currents remains stable for several hundred milliseconds up to the point where the nitrogen seeding is switched-on and the plasma undergoes a transition to a globally detached state with symmetric currents of the order of +10 A per tile. A breaking of the toroidal asymmetry is also observed when going to a low-field side snowflake minus configuration where the gas is puffed into the secondary X-point, i.e. into a region with long connection lengths to the strike lines. If the gas is injected by another valve at the same radial position but in segment 2, the current asymmetry also shifts to segment 2 as expected, confirming the robustness of the observations. In this contribution we will make an attempt to describe the observations by the 3D transport code EMC3-EIRENE and study the possibility of using the localized gas puff system as a re-attachment avoidance system.

Speaker: Tilmann Lunt (Max Planck Institute for Plasma Physics)

93 1.088 Evaluation of light impurity turbulent transport in the edge plasma from 3D multi-fluid turbulence simulations

Mean-field transport coefficients are the main work-horse for the modelling of the edge plasma of magnetic fusion devices and are heavily used in the frame of predictive studies for the design and preparation of operation of future devices. These codes offer a high-fidelity description of plasma-neutrals and multi-species interactions at the cost of a simplified description of transverse turbulent transport in the form of a gradient-diffusion model. In practice, transverse transport is prescribed by 3 ad-hoc diffusion coefficients (for particles, parallel momentum and energy) for every single particle species in the plasma. While experimental Scrape-Off-Layer (SOL) width scaling laws can serve as a guide-line to tune the transport coefficients of the dominant species (in general hydrogenic ions), very little information is available to guide the choice of sensible transport coefficients for impurities.
In this contribution, we report on a first evaluation of the turbulent transport of light impurities from 3D turbulence simulations performed with the SOLEDGE3X code. Of particular interest for this study is the capability of SOLEDGE3X to model plasmas of arbitrary composition thanks to its implementation of the multi-fluid Zhdanov closure. We consider a TCV attached deuterium plasma based on the TCV-X21 reference scenario. Carbon is chosen here as the impurity of interest. Two simulations are considered: one in which carbon is self-consistently generated by erosion processes at the divertor target plates and one in which it is injected in the plasma from the outboard mid-plane. By doing so we can compare whether the resulting transport differs whether the carbon source is downstream or upstream of the main species flow. In both simulations, turbulence is found to be the dominant contributor to the transverse transport of carbon as it is for the deuterium. A key difference with the main species however is that in the concentrations considered here (a few percents) carbon mostly behaves like a passive tracer. This results in a diffusive behavior for carbon independently of it being injected from the mid-plane or at the targets, allowing carbon to radially penetrate inwards against the deuterium flux. Like for the main species, the transport coefficients of carbon are found significantly inhomogeneous in space, with enhanced transport in the outboard mid-plane and along the outer divertor leg. We finally compare the amplitudes of transport coefficients for carbon and deuterium and come up with 0th order recommendations for mean-field models.

Speaker: Patrick Tamain (CEA)

94 1.089 Impact of the toroidal field direction on SOLEDGE3X 3D turbulent simulations of the COMPASS boundary plasma

The magnetic field topology has been observed to systematically affect the power threshold for the L–H transition in several tokamaks. In particular, a lower (higher) power threshold is found when the ion-$\nabla B$ drift points toward (away from) the X-point, indicating a "favorable" ("unfavorable") configuration for the transition. Early explanations for this behavior were proposed by [LaBombard et al.,2005], who experimentally linked topology-dependent scrape-off layer (SOL) parallel flows to changes in core toroidal rotation and, ultimately, to the L–H transition. Later studies suggested a possible contribution from edge magnetic shear–induced Reynolds stresses, dependent on the drift configuration[Fedorczak et al.,2012]. More recent experiments, however, do not fully support these interpretations[Plank et al.2023] and instead show that the observed differences between the two drift configurations (especially those involving the edge radial electric field profile[Vermare et al.,2022) are not consistently captured by the existing explanations. This indicates that the phenomenon is not yet completely understood, and given the central importance of reliably accessing H-mode in future fusion reactors, further investigation is needed.
In this contribution, 3D turbulent simulations of the COMPASS tokamak boundary plasma in both favorable and unfavorable configurations are carried out using the first-principle drift-reduced fluid code SOLEDGE3X. The magnetic equilibrium reproduces the shot #15487 (ohmic L-mode D-shaped discharge[Cavalier et al.,2019]) and the two different drift configurations are obtained by intentionally reversing the direction of the toroidal magnetic field $B_T$. Neutrals are self-consistently included through the coupling with the kinetic code EIRENE.
The study compares edge and SOL plasma dynamics across two otherwise identical L-mode scenarios, differing only in the direction of $B_T$. The upstream Mach number profile indicates a displacement of the stagnation point from below (favorable) to above (unfavorable) the outer mid-plane (OMP), corresponding downstream to the relocation of the denser and colder divertor leg respectively from the inner to the outer target. Near-sonic SOL parallel flows in the high-field side are driven by ballooning-like transport and the edge plasma poloidal rotation reverses direction. Remarkably, in the favorable configuration the OMP radial electric field well becomes noticeably steeper and the positive peak in the SOL nearly doubles, pointing also a link to divertor physics. The analysis of the interplay between these equilibrium changes and turbulence itself will highlight the impact of the toroidal field direction on intensity, structure and propagation of plasma fluctuations.
A detailed discussion on the relation between these results and the existing theoretical explanations is presented.

Speaker: Michele Lambresa (Aix-Marseille Université)

95 1.090 Analysis of non-resonant divertor experiments in W7-X

Divertors are important for particle and heat removal in future stellarator power plants. Currently, three types of stellarator divertors are studied: the island divertor, the helical divertor, and the non-resonant divertor [1]. Where the island divertor and the helical divertor have been studied experimentally in W7-X and LHD, respectively, the non-resonant divertor has mainly been studied theoretically [2].

The non-resonant divertor (NRD) concept makes use of chaotic magnetic structures in the plasma edge. Modelling shows that NRDs are resilient across changes in plasma current and magnetic equilibrium in contrast to the resonant island structure which is typically used in the island divertor approach of the W7-X stellarator [1,3]. Recently, a particular magnetic field configuration in W7-X sharing properties of such a NRD configuration has been identified and experiments in this configuration have been performed. To mitigate coil stresses resulting from the high rotational transform of this configuration, only experiments at reduced magnetic field using purely NBI heating were possible. A first analysis indicates that there are similarities between observed strikelines with infrared cameras and predictions made by EMC3-Lite [4] (which uses an anisotropic heat diffusion model).

The use of NBI heating, however, causes some challenges. It can contribute to strikeline splitting through fast ion heat loads [5] which is not considered in the EMC3-Lite simulation. Furthermore, this NBI causes a steep density profile in the core leading to impurity accumulation. Consequently, the radiated power is dominated by core radiation.

To analyze how the heat and particle transport behaves in the case of edge radiation, EMC3-EIRENE [6] is used. Special attention is placed on grid quality to avoid artificial power leaking and on determination of anomalous transport parameters to obtain agreement between simulations and experiments. This study is a first step towards power and particle exhaust studies in non-resonant divertors.

[1] K.A. Garcia et al., Plasma Phys. Contr. Fusion 67, 035011 (2025)
[2] A.H. Boozer and A. Punjabi, Phys. Plasmas 25, 092505 (2018)
[3] K.A. Garcia et al., Nucl. Fusion 63 (12), 126043 (2023)
[4] Y. Feng, Plasma Phys. Contr. Fusion 64, 125012 (2022)
[5] M.J.H. Cornelissen et al., Plasma Phys. Contr. Fusion 64, 125015 (2022)
[6] Y. Feng, et al., Contr. Plasma Phys. 54, 426-431 (2014)

Speaker: Dr Dieter Boeyaert (University of Wisconsin-Madison)

96 1.091 Variation of Radial Electric Field due to cross-field transport in ADITYA-U edge/SOL region

The cross-field transport in tokamak edge plasma remains one of the major challenges in magnetic confinement fusion research, as it causes heat and particle loss from the main plasma, thereby affecting plasma confinement [1]. This transport modifies density and temperature gradients, which can alter the neo-classical radial electric field. On the other hand, ion-orbit loss mechanisms produce the outward ion flux, which in turn generates an inward radial electric field [2]. The E×B shear, generated by radial electric field is believed to be essential for the suppression of edge turbulence [3], and thus is necessary to the confinement improvement expected in ITER’s H-mode scenario.
Recently, the UEDGE code has been used to investigate the effect of convective transport in ADITYA-U tokamak edge/SOL plasmas [4]. To match the experimentally observed electron density (n$_e$) profile in typical ADITYA-U discharges, an inward convective velocity (v$_{conv}$) of about 1.5 m/s is required in addition to the constant perpendicular diffusion coefficient (D$_{\perp}$) of around 0.2 m$^2$/s. In the present study, the E×B drift has been incorporated into the UEDGE code to analyze its influence on the edge plasma dynamics in the ADITYA-U tokamak. The profiles of the radial electric field and electrostatic potential are evaluated and compared with the available experimental measurements [5]. The radial variation of the electric field in the edge and SOL regions, which is a maximum of 10 kV/m near the limiter surface. In the present study, the cross-field diffusion coefficient is taken to be 0.5 m$^2$/s. Finally, the impact of the transport parameters on the radial electric field has been investigated.

Speaker: Ritu Dey (Indian Institute of Technology Tirupati)

97 1.092 Perpendicular transport by turbulence and drifts in the W7-X island divertor

In the island divertor of Wendelstein 7-X, modular divertor targets intersect a chain of resonant magnetic islands that act as the scrape-off layer (SOL). The island SOL is characterized by long parallel connection lengths of several 100m, resulting in a high efficiency of perpendicular (radial and bi-normal) transport compared to parallel transport. The two main perpendicular transport channels are believed to be turbulent transport and stationary ExB drift flows, where the former is dominant in the radial direction (normal to the last closed flux surface) due to the main pressure gradients. The ExB flows are mostly oriented in the bi-normal direction, which can deviate from the island flux surfaces.

Due to the 3D geometry and the toroidally discontinuous divertor targets, perpendicular pressure gradients are not uniform, resulting in a spatially varying impact of turbulent transport on the SOL plasma. Further, radial electric fields feature multiple shear layers (including sign flips) across the SOL. As a consequence, counter-propagating bi-normal ExB drift flows are observed [1]. The 3D drift and turbulent transport act back on the plasma parameter distribution, resulting in a self-consistent equilibrium.

We employ two key diagnostic systems installed in approximately toroidally symmetric positions of W7-X and give insight into both turbulence and drifts. Reciprocating electric probe arrays provide electron temperature and density as well as estimates of electric fields and turbulent radial particle transport via floating potential measurement. Gas Puff Imaging observes H-alpha emission fluctuations from a localized gas puff with high temporal resolution. From these measurements, turbulence properties and drift flows (via spatio-temporal analysis) are obtained [2].

We find that the velocity and therefore the impact of drift flows decreases for high plasma densities. Accordingly, the asymmetries between upper / lower divertors, which are presumably caused by drift flows, decrease for higher densities, but still persist up to (at least) $n_{av}=7\cdot 10^{19}m^{-3}$. Turbulent transport, however, increases in terms of absolute particle fluxes while moving to higher plasma densities. This is mostly a consequence of steeper SOL density gradients, while diffusion coefficients estimated from particle flux and density gradient measurements remain roughly in line with initial findings in the test divertor operation phase of W7-X [3].

[1] Killer NF 65 056026 (2025)
[2] Baek NME 43 101937 (2025)
[3] Killer NF 61 096038 (2021)

Speaker: Carsten Killer (MPPL)

98 1.093 EDGE2D-EIRENE and EIRENE interpretation of Balmer and Fulcher band emission in JET-ILW Ohmic deuterium plasmas

The agreement in the EDGE2D-EIRENE predicted and the measured Balmer-$\alpha$ line emission and the molecular Fulcher band emission for a set of deuterium Ohmic plasmas in JET-ILW validates the neutral and the target recycling models in EDGE2D-EIRENE. Since the plasma-neutral interactions are the strongest in the divertor, a validated EDGE2D-EIRENE neutral model provides higher fidelity to predicting the divertor performance in future fusion devices. In low-recycling and detached conditions EDGE2D-EIRENE predicted Balmer-$\alpha$ emission is within 10% of the measurements, and in high-recycling higher by a factor of 2. The discrepancy in the predicted Balmer-$\alpha$ emission in high-recycling conditions is, in part, explained by the lack of thermo-electric currents and drifts in the current EDGE2D-EIRENE simulations. The predicted Fulcher band emission is within 20% of the measurements in all divertor conditions. The magnitude of the Fulcher band emission averaged over the low-field side (LFS) divertor is found to be within a factor of 2 from low-recycling to detached plasma conditions, which is explained by the strong dependence of the population density of the $N=3$ molecular excited state on the local electron temperature. Since the target electron temperature decreases, and the molecular density increases, with deepening detachment, the decrease in the relative population density of the upper Fulcher band state is approximately compensated by the increased molecular density.

To calibrate the molecular influx, and to validate the fuel recycling model in EDGE2D-EIRENE, a deuterium gas injection module was set up in JET-ILW Ohmic deuterium plasmas to obtain a localized calibrated gas puff into the divertor. The Fulcher band spectrum is measured using two spectrometers viewing the LFS divertor, with one viewing the injection point. The intensity of the diagonal transitions are determined [1,2], from which the rotational and vibrational temperatures are derived, and the total Fulcher band emission is estimated. The EIRENE predicted Fulcher band emission is within 30% of the measurements at the injection site in high-recycling ($T_{e,t} \approx 5$eV, $n_{e,t} \approx 1.5 \cdot 10^{19} m^{-3}$) and detached conditions ($T_{e,t} \approx 1.0$eV, $n_{e,t} \approx 3.0 \cdot 10^{19} m^{-3}$), when a gas puff is applied on the predicted plasma. Using the inverse photon efficiency [3], the estimated molecular influx is within 20% of the analytically estimated gas-injection rate, validating the inverse photon efficiency technique.

[1] E.M Hollman et al., 2006 Plasma Phys. Control. Fusion 48 1165
[2] G. Sergienko et al., 2013 JNM 438
[3] S. Brezinsek et al., 2003 JNM 313–316

Speaker: Vesa-Pekka Rikala (Aalto University)

99 1.094 Fast processes in the divertors of Wendelstein 7-X and ASDEX Upgrade

Understanding and quantifying particle and energy transport at the boundary, detachment and its stability in the divertor region, is crucial for magnetic confinement fusion, as this determines both plasma performance and target loads. For this reason, several fast helium beam systems, based on the ASDEX Upgrade (AUG) midplane system [1], have been installed in the divertors of AUG [2] and Wendelstein 7-X (W7-X). These diagnostics allow the investigation of edge transport processes, such as modes and filaments, with high spatiotemporal resolution in both $n_\text{e}$ and $T_\text{e}$ [3] as well as detachment phenomena in the divertor [4]. With the new possibility of measuring highly temporally resolved 1D profiles, we observe strong variations (\textasciitilde 50\% relative fluctuation amplitude in $T_\text{e}$ and $n_\text{e}$) and movements of the radiation front in the divertor region of W7-X. These events change transiently the plasma state for \mbox{1--20 ms} with transition durations from \mbox{0.5--2 ms}. Since such a diagnostic is placed in two divertors of W7-X, the toroidal and poloidal correlation of these events is investigated, along with other diagnostics. This contribution explores the physical nature of these events and their influence on the target heat flux due to the strong $T_\text{e}$ and $n_\text{e}$ variations. These investigations are carried out for different magnetic configurations and plasma parameters. Furthermore, the measurements will be compared with similar events observed by the same diagnostic at AUG. This presentation aims at a better understanding of detachment on fast timescales in \mbox{W7-X}, the influence of fast events on the divertor island region, and it classifies the phenomena in relation to the tokamak divertor.

[1] Michael Griener et al. 2018 Rev. Sci. Instrum. 89, 10D102
[2] Sebastian Hörmann et al. 2024 Rev. Sci. Instrum. 95, 113507
[3] Joey Kalis et al. 2024 Nucl. Fusion 64, 016038
[4] Sebastian Hörmann et al. 2025 Plasma Phys. Control. Fusion 67 075016

Speaker: Sebastian Hörmann (Max-Planck-Institut für Plasma Physik (Garching))

100 1.095 Study of the dependence of charge-exchange neutral flux on the operation conditions of EAST

Based on the charge-exchange neutral (CXN) flux database measured by the low-energy neutral particle analyzer (LENPA) in EAST, the dependence of CXN flux on basic plasma parameters of EAST has been summarized, and scaling expressions for integrated flux and mean energy of CXNs are obtained. Simulations for ITER [1] and DEMO [2] predict that CXNs can play an important role in the erosion of the first wall material. It is important to study the CXN flux and energy distribution dependent on global plasma parameters. The LENPA diagnostic on EAST can measure the energy spectrum of CXNs from 20 to 3000 eV [3]. A database of LENPA measurements under different plasma operation conditions in 473 EAST discharges in the year of 2023 are generated for CXN analysis. These discharges contain multiple magnetic configurations (LSN, DN, USN) and auxiliary heating methods (LHW, ECRH, and NBI). It is found that the low-energy (<500 eV) CXN flux ($\Gamma_{20-500}$) increases while the high-energy (500-3000 eV) CXN flux ($\Gamma_{500-3000}$) decreases with the increase of the line-averaged plasma density crossing the core ($n_e$). Whereas a high line-averaged plasma density usually leads to a high edge plasma density, which has a notable mitigation effect on the high-energy CXN flux due to the strong collisional losses and can be indicated by the neutral pressure ($p_n$). Meanwhile, the high-energy CXN flux increases significantly with the increase of the total heating power ($P_{tot}$) and plasma stored energy ($W_{MHD}$). Therefore, regression analyses of the integrated CXN flux and the mean energy ($E_{mean}$) on the basic plasma parameters including $n_e$, $P_{tot}$, $W_{MHD}$, $I_P$, $p_n$ are conducted. The scaling expressions for the low-energy and high-energy CXN flux are: $\Gamma_{\left(20-500\right)}\ =\ n_e^{0.03}W_{MHD}^{0.20}P_{tot}^{0.08}I_p^{-0.57}p_n^{0.77}$, $\Gamma_{\left(500-3000\right)}\ =\ n_e^{0.69}W_{MHD}^{0.51}P_{tot}^{0.13}I_p^{0.54}p_n^{-0.42}$. These expressions are further validated by a LENPA measurement database from EAST experiments in the year of 2024.

References:
[1] A. Eksaeva et al., Physica Scripta, 97 (2022) 014001.
[2] D. Matveev et al., Nuclear Fusion, 64 (2025) 106043.
[3] N.X. Liu et al., Review of Scientific Instruments, 92 (2021) 063507.

Speaker: Prof. Junling Chen (Institute of Plasma Physics, Chinese Academy of Sciences)

101 1.096 Dynamics of Toroidal Radiation Belt and Its Impact in EAST with Impurity Seeding

In future fusion reactors, active impurity seeding is expected to establish a highly radiating region at the plasma edge, enabling efficient power dissipation. Due to radiation condensation triggered by the temperature dependence of impurity radiation, this region typically exhibits strong poloidal localization, manifesting as a toroidally symmetric radiation belt. When stabilized near the X-point, this structure is termed X-point radiator (XPR). If the radiation belt migrates upward, it becomes magnetohydrodynamically unstable, called multifaceted asymmetric radiation from the edge (MARFE), possibly leading to a disruption. To maintain stable power exhaust with high radiative losses while avoiding disruptions, a thorough understanding of the underlying physics and identification of critical parameters governing the radiation belt’s dynamics are essential.

In EAST, experimental observations show that a strong radiation belt initially forms in the divertor region following impurity seeding. After an H-L mode back transition or confinement degradation, the radiation belt moves toward the X-point and the high-field side (HFS) of the plasma boundary. Without feedback control, the belt advances further along the HFS but typically returns to its original position within 1 s. However, this process is occasionally accompanied by plasma shrinkage and subsequent disruption.

The belt’s movement is closely linked to overall edge neutral pressure evolution: Forward motion coincides with rising neutral pressure and retreat occurs as pressure recovers. However, local pressure increases and wall heat load decreases when the belt approaches, suggesting strong radiative cooling and recombination near the belt. Reversing the toroidal magnetic field (BT) direction inverts both the initial position of the radiation belt and its trajectory. The reversal of neutral pressure asymmetry between upper and lower divertors following BT direction change demonstrates the role of B-direction-dependent drift effects.

Furthermore, experimental observations reveal that the movement of the radiation belt is consistently accompanied by localized fluctuations with frequencies up to 10 kHz in both plasma radiation and density signals. The poloidally and toridally distributed Mirnov coil array captures the temporal and spatial evolution of magnetic perturbations. During belt movement, the magnetic oscillations transition from a broadband spectrum to a coherent narrowband emission, identified as an m/n=2/1 magnetohydrodynamic (MHD) instability mode through Mirnov coil phase analysis. Notably, excessive increases in the magnetic oscillation frequency typically precede plasma disruptions, while stable XPR regimes maintain low-frequency oscillations. This frequency-dependent behavior may suggest a critical link between oscillation spectral characteristics, radiation belt stability, and disruption onset mechanisms.

Speaker: Dr Fang Ding (Institute of Plasma Physics, Hefei Institutes of Physical Science, Chinese Academy of Sciences)

102 1.097 Classifying ST40 plasmas through SepOS framework

The electron temperature (T$_{e,sep}$) and density (n$_{e,sep}$) at the separatrix at the outboard midplane (OMP) are the key parameters in mapping the operational space for power exhaust handling in tokamak fusion devices. Here, three methods of determining the Te,sep and ne,sep in ST40 are compared. First is through the reverse two-point modeling [1], wherein the electron temperature (T$_{e,target}$) and density (n$_{e,target}$) at the outer divertor from Langmuir probe measurements are used to predict the upstream parameters at the OMP. Secondly, available T$_{e,sep}$ and n$_{e,sep}$, from Thomson scattering data are used. Both methods will use the separatrix location given by EFIT during the calculation. Lastly, the Heat flux Engineering and Anaysis Toolkit (HEAT) [2] and IR data are used to determine the correction factor of the separatrix location calculated from the difference of the simulated and experimental heat flux peak locations. The upstream parameters are then employed to predict the operational space for power exhaust handling in ST40 using the separatrix operational space framework (SepOS) [3]. Initial results showed that observed periods of enhanced plasma stored energy and reduced D-alpha emissions are consistent with an H-mode phase, given their SepOS parameters. The results will be used to determine whether the parameters from ST40, a compact high-field spherical tokamak, are comparable with published normalized multi-machine SepOS database.

[1] J H Nichols et. al, Plasma Phys. Control. Fusion 66 (2024) 045013.
[2] EJC Tinacba et. al., Nuclear Materials and Energy 41 (2024) 101791.
[3] T Eich et al., Nuclear Materials and Energy 42 (2025) 101896.

Speaker: Erin Joy Tinacba (ORNL)

103 1.098 Study of alternative divertor configurations with first-principles, self-consistent, global turbulence simulations

While the conventional lower single-null (LSN) divertor will be tested in ITER, its extrapolation to power-plant conditions remains uncertain, motivating the exploration of further optimised solutions. Three geometrical parameters can be changed independently in the LSN configuration: poloidal length of the divertor legs, poloidal flux expansion at the target and radial position of the strike points. Additional magnetic nulls can be included in the path towards improved heat exhaust solutions. Each of these variations has an associated alternative divertor configuration (ADC): long-legged divertor (LLD), X-divertor, super-X divertor and X-point target. Different experimental campaigns on the TCV tokamak have investigated the impact of these ADCs on divertor power exhaust, showing varying degrees of improvement depending on the configuration. Modelling efforts, however, remain scarce and focused so far on interpretative modelling with mean-field codes lacking turbulent transport.

In order to improve the physics understanding of the ADCs, we performed turbulence simulations of TCV plasmas with different magnetic geometries, precisely the LSN and the four aforementioned ADCs mentioned. The simulations were carried out by using GBS, a first-principles, 3D, global turbulence code that self-consistently evolves the plasma dynamics with the neutral atoms modelled kinetically. A detailed comparative analysis of the considered configurations highlights the role of turbulence in particle and heat exhaust and its interplay with magnetic geometry, pointing out their relative strengths and weaknesses. All ADCs show the decrease in divertor heat load with respect to the LSN configuration. The LLD achieves this heat flux reduction through its increased divertor volume, while the turbulent dynamics are not affected. In contrast, the other configurations benefit from modifications of the turbulent dynamics. These findings constitute important elements for more reliable predictions of ADC performance in future devices.

Speaker: Sergio Garcia Herreros (Swiss Plasma Centre - EPFL)

104 1.099 Edge localized mode suppression by the low-temperature supersonic molecular beam injection on EAST

To improve fueling efficiency and achieve deeper injection, a low-temperature supersonic molecular beam injection (LT-SMBI) system was successfully developed and implemented on the Experimental Advanced Superconducting Tokamak (EAST). This system utilizes cryogenic cooling with LN2, maintaining the gas at about 120 K to promote cluster formation. Compared with the RT-SMBI, the LT-SMBI results in about 5 – 7 cm deeper injection depth and the fueling gas consumption is decreased by 33% in the feedback control experiments using the LT-SMBI. The LT-SMBI system is also used to mitigate the type-I edge localized modes (ELMs) on the EAST. Before the SMBI experiments, the mixed D2 and Ne is cryogenically cooled to less than 120 K to enhance the clusters in the beam. In our past studies, the particles could be injected into the top pedestal by LT-SMBI. During the high-confinement mode (H-mode) plasma, the high-energy ELMs were mitigated for about 90 ms after LT-SMBI. By performing the LT-SMBI at a frequency of 10 Hz, stable 400 ms ELM suppression was achieved. Furthermore, no large ELM bursts were observed over the subsequent 1.5 s, even though SMBI had stopped. The LT-SMBI will be an efficient method for ELM mitigation and impurity injection in EAST and future fusion devices.

Speaker: Junhao Qiao (Institute of Plasma Physics, Chinese Academy of Sciences, Hefei 230031, China)

105 1.100 ITER Q = 10 divertor operational window with realistic plasma-facing component geometry

The ITER divertor design has been guided by extensive scoping studies focused on baseline burning conditions at $Q_{DT} = 10$. They were conducted with the SOLPS-4.3 plasma boundary code without drifts and currents, assuming fuel injection from the top of the machine and pumping directly underneath the dome umbrella. The resulting simulation database was used to optimize the divertor operational window limited by target heat loads, fuel throughput, helium (He) exhaust and core-edge integration requirements. In 2015 the ITER Organization launched a new SOLPS-ITER version of the code which, in subsequent years, has been systematically upgraded, providing a much more complete edge plasma description in comparison with SOLPS-4.3. Together with switching to gas injection from the sub-divertor volume (now the scheme favoured at ITER), this resulted in a shift of the operational window to much greater divertor pressures ($p_n$) and higher average separatrix neon (Ne) concentrations ($c_{Ne}$), posing questions of compatibility with the throughput requirements and core plasma performance.

However, when also including more complex sub-divertor geometry (accounting for realistic neutral bypass conductances around the divertor cassette body), and a proper description of pumping system capabilities [1], the results again change radically. We present here a completely new SOLPS-ITER $Q_{DT} = 10$ simulation database in which, rather unexpectedly at first sight, the operational window reverts to something rather close to the initial SOLPS-4.3 database. This is identified as due primarily to two factors: 1) a strong influx of neutrals from the sub-divertor space to the far-SOL reducing the peak target heat flux and decreasing the value of $p_n$ required to meet the $10\ MW/m^2$ constraint; 2) the proper assessment of pressure at the pumping duct translating to a wider range of pn compatible with throughput requirements.

Accounting for the neutral bypasses also leads to a reduction of the plasma temperature in the divertor baffle regions. This is extremely beneficial given the results of simulations with the new wide-grid SOLPS-ITER capability, also presented here, which show these locations to be the origin of significant quantities of sputtered W impurity which is rather poorly screened compared with that released from vertical target areas. Finally, taking into account the improved divertor description and the new code capabilities, the important question for reactor divertor designs beyond ITER of the impact of the dome umbrella on the global divertor performance has also been explored.

[1] N. Vasileiadis et al., Fusion Eng. Des. 151 (2020) 111383

Speaker: Andrei Pshenov (ITER Organization)

106 1.101 Toward Active Power Exhaust Control: Integration of Small ELMs, Detachment, and AI Diagnostics on EAST

Small Edge Localized Modes (small ELMs) represent a promising scenario for power exhaust control in future fusion reactors, offering a favorable balance between heat load mitigation and impurity exhaust in high-confinement plasmas [1,2]. Achieving active control of power exhaust will require the integration of stable small-ELM operation, robust detachment sustainment, and real-time diagnostic capabilities—three critical elements that must be co-developed to enable reactor-relevant applications. On EAST, recent experiments have advanced this integration goal by systematically characterizing SOL heat flux in small-ELM regimes, demonstrating long-pulse detached operation compatible with small or suppressed ELMs, and developing AI-enhanced diagnostics for future control applications.
High-resolution infrared thermography has been employed to characterize the scrape-off layer (SOL) heat flux width (λq) across different wall conditioning regimes. In small ELMs, the time-averaged λq is broader than inter-large ELM phases but narrower than during large ELMs [3]. Notably, 20–45% of the ELM energy is deposited in the far-SOL, indicating measurable cross-field transport. Experiments with both lithium- and boron-conditioned walls show that λq in small ELM regimes is strongly influenced by plasma density and heating schemes, underscoring the role of edge turbulence and pedestal dynamics in determining power deposition profiles.
Crucially, under boronized wall conditions, sustained detached operation with small or fully suppressed ELMs has been achieved for over 50 seconds via real-time feedback injection of nitrogen impurities. Modeling suggests that high-frequency broadband turbulence (HFBT) at the plasma edge enhances particle and heat transport, contributing to the stability of this high-performance, detached state.
To enable active control in future devices, two advanced diagnostic frameworks are being developed. First, a hybrid AI-tomography method enables 2D reconstruction of divertor radiation (Dα and impurity emission) from visible-light imaging. Second, a multimodal neural network is under training to provide fast, robust inference of detachment and confinement states by fusing imaging and conventional diagnostics.
These integrated efforts establish a critical pathway toward reliable, long-pulse power exhaust control in reactor-relevant scenarios—advancing both the physics understanding and technological readiness for fusion energy.
[1] X.Q. Xu, N.M. Li, M.L. Zhao et al Nucl. Mater. Energy 42 (2025) 101866
[2] Q.Q. Yang, G.S. Xu, N. Yan et al Nucl. Fusion 60 (2020) 076012
[3] G.T. Chen, Q.Q. Yang*, J.H. Yang et al Nucl. Fusion 65 (2025) 076040

Speaker: Qingquan Yang (ASIPP)

107 1.102 Magnetic configuration effect on X-point radiation in W7-X plasmas

In the stellarator W7-X, three magnetic island-chain configurations with rotational transform values ι = n/m = 5/4, 5/5, and 5/6 (referred to as high-ι, standard, and low-ι configuration) provide a versatile platform for exploring the island divertor concept. As the poloidal mode number m increases, the connection length increases while the internal field-line pitch decreases (Feng et al 2024 Nucl. Fusion), both strongly influencing the scrape-off-layer transport and the resulting plasma–wall interaction (PWI) patterns on the divertor targets. An interesting question is which configuration yields the most favorable divertor performance for power and particle exhaust. This study investigates impurity radiation—particularly the radiation localized near the X-points—across these three magnetic topologies. The primary results are obtained using bolometric tomography at an up/down-symmetric triangular cross-section and are complemented by video and spectroscopic diagnostics.
In the standard configuration, highly-radiative, detached plasma phases (with a radiation fraction frad = 0.6–0.9) are routinely achieved by increasing the plasma density or reducing the heating power. During detachment, impurity radiation exhibits a multi-X-point radiation structure with significant up/down asymmetry. A simplified model, supported by field-reversal experiments, explains the dominant role of the poloidal E×B drift causing an asymmetric modification of the parallel impurity flow due to friction with main ions (Zhang et al 2025 Nucl. Fusion). In high- and low-ι configurations, 2D radiation structures with up/down asymmetry are also observed. However, intense radiation near certain X-points occurs at lower frad (~0.2-0.3) in the attached plasma phases. Analysis has shown that in the high-ι cases, this phenomenon is related to an unfavorable PWI region adjacent to a diagnostic port outside the divertor, indicating the need for divertor geometry improvement, while in the low-ι cases, it is related to the strong influence of E×B drift in the private-flux region. With increasing the frad level, the radiation peak in the high-ι plasma shifts to other X-points, while in the low-ι plasma it remains at the same location and intensifies, leading to unstable X-point radiation, which is usually accompanied by the radiation zone penetrating the confinement region. This observation provides deeper insight into radiation instabilities in low-ι plasmas in the high-density, high-radiation regime (Winters et al 2024 Nucl. Fusion). This study also addresses the question of whether operating conditions—such as gas-fueling rate or impurity species—can influence the occurrence or severity of radiation instabilities. Additional analyses of experiments with extrinsic nitrogen and neon seeding are included.

Speaker: Dr Daihong Zhang (Max Planck Institute of Plasma Physics)

108 1.103 On the influence of impurity species and X-point geometry on accessibility and dissipation capacity of X-point radiators in fusion reactors

The X-point radiator (XPR) plasma regime displays favorable properties with regard to power exhaust in tokamaks: An H-mode-like confinement quality, a detached divertor, and the suppression of type-I ELMs are achieved simultaneously [1]. XPR scenarios may also pave the way for more compact and cheaper divertor solutions,as demonstrated on ASDEX Upgrade [2]. The parameter to control XPR stability and the fraction of power dissipated by radiation is the XPR height can be actively manipulated through impurity seeding and neutral gas fuelling.

This paper focuses on the perspectives for XPR scenarios in tokamak fusion
reactors. The XPR height depends on flux-surface geometry and seeding impurity. Among the impurities studied, argon shows the highest efficiency in radiating a large fraction of the heating power. In EU-DEMO, argon seeding resulted in an XPR height of 50 cm, corresponding to a dissipation of 90 % of the 150 MW heating power entering the pedestal region. With neon, only lower dissipation rates are achieved, and the XPR height depends only weakly on the neon concentration.

Furthermore, it is shown that even without impurity seeding, charge exchange power losses can be considerable at high neutral densities, increasing the heating power required to access the H-mode. The observed effect reproduces experimental trends in the power threshold, such as its dependence on density and on divertor closeness.

While a large flux expansion and the associated long connection length favor the occurrence of XPRs, less power is conducted to the XPR as the connection length increases, reducing the amount of dissipated power. These two opposing effects are investigated within a family of configurations related to the compact radiative divertor, where both parameters can be driven to extreme values.

These results were obtained from a reduced power and particle balance model [3] and an extension of it [4] which estimates the XPR height, the dissipated power, and the coupling to the upstream profiles. The calculated reduction in the pedestal gradient is consistent with the experiments and could explain the process of ELM suppression.

[1] M. Bernert et al., Nucl. Fusion 61, 24001 (2020).
[2] T. Lunt et al., Phys. Rev. Lett. 130, 145102 (2023).
[3] U. Stroth et al., Nucl. Fusion 62, 076008 (2022).
[4] U. Stroth et al., Plasma Phys. Contr. Fusion 67, 025001 (2025).

Speaker: Ulrich Stroth (MPPL)

109 1.104 Diffusion Pump Divertor for Fusion Devices

A new divertor concept, referred to as the Diffusion Pump Divertor [McComas et al. US Patent Application #63/919,661], is presented. This concept adapts vacuum-pump technology to enable controlled delivery and removal of vapor within fusion devices. This approach addresses longstanding limitations in present lithium vapor delivery schemes, which rely on evaporation and provide only limited control over both the quantity and spatial distribution of vapor. Poor control can adversely affect plasma performance, leading to undesirable core impurity accumulation or insufficient radiation for heat-flux mitigation. By contrast, the Diffusion Pump Divertor provides a means to inject low-Z metal vapor, such as lithium, into the divertor plasma with high precision while simultaneously removing excess vapor and contaminants during operation. In addition to improving impurity management, the introduced vapor absorbs significant plasma thermal power, offering an additional pathway for divertor heat exhaust. In this paper we present a preliminary design of the Diffusion Pump Divertor. We have evaluated performance through analytical modeling, computational fluid dynamics (CFD), and advanced SOLPS-ITER simulations. The assessment aims to determine the feasibility, operational benefits, and potential performance improvements associated with this concept in fusion-relevant conditions.
The SOLPS-ITER simulations of the boundary plasma-surface interactions employed a National Spherical Torus Experiment Upgrade (NSTX-U) magnetic equilibrium created specifically to examine high-heat-flux scenarios [Emdee and Goldston 2023 Nuclear Materials and Energy 34 101335]. Plasma-facing component envelopes as used in previous lithium vapor box modeling were used for a direct comparison with the vapor box calculations [Emdee and Goldston 2023 Nuclear Fusion 63 096003]. In these simulations, supersonic jets integral to the divertor design were represented by “transparent” surfaces emitting atomic lithium 1.62 eV kinetic energy, corresponding to a Mach 5 jet for a lithium vapor temperature of 900 K. The nozzle fluence was scanned at constant particle energy to study the impact of jet strength on plasma and neutral transport. Activation of the jet produced a measurable increase in neutral deuterium flux toward the divertor target, demonstrating diffusion-pump-like action; by one measure, the increase in pumping speed was about 20x. The ionized deuterium flux, however, decreased due to divertor detachment rather than the pumping mechanism itself. Such reduction is expected in detached divertors, where strong upstream recombination naturally limits ionized flux reaching the target. Despite the reduction in deuterium flux, the pumped flux of deuterium was still calculated to far exceed standard pumping techniques, indicating unique benefits to this divertor design.

Speaker: Andrei Khodak (Princeton Plasma Physics Laboratory)

110 1.105 SNOWFLAKE DIVERTOR STUDIES IN MAST-U TOKAMAK AND OUTLOOK FOR NSTX-U

The snowflake (SF) divertor is studied in the MAST-U tokamak as an alternative concept for next-step compact fusion devices and is planned for high-power NSTX-U tokamak experiments. The studies focus on the SF plasma transport mechanisms and their scaling with plasma current in the range 0.4-1.0 MA. The SF divertor features divertor geometry, radiation and transport enhancements (cf. standard X-point divertor) that may improve divertor power handling.
Recent MAST-U experiments have utilized 1.5-3.2 MW NBI-heated, 0.6-0.75 MA H-mode discharges to study heat and particle exhaust over eight strike points, a unique aspect of MAST-U up-down divertor symmetry. Feed-forward plasma control algorithms based on TokSys and TED simulations were used to manage the second poloidal field (PF) null location resulting in inter-null distances 0.01-0.20 m. The H-mode confinement was maintained albeit with a plasma stored energy reduction of 10-30%. The ELM regime changed from small or no ELMs to medium-sized ELMs during the SF phase. The SF variants, namely the SF-exact, SF-minus, and SF-plus configurations, were obtained with the expected geometry enhancements, e.g., the connection lengths longer by a factor of 1.5-3 (cf. standard and Super-X configurations). Divertor results included: 1) Evidence of particle and heat flux sharing over the SF strike points, from Langmuir probes, infrared and filtered divertor cameras; 2) Observations of turbulence correlation between inner and outer SOL regions in the forming SF configuration, suggestive of a shared plasma SF region, based on fast divertor imaging; 3) Plasma radiation peaking in the PF null region suggesting a potential for X-point radiator regime.
Experimental results are interpreted using an improved reduced-magnetohydrodynamic model of the churning mode and the multi-fluid code UEDGE with drifts and poloidally non-uniform transport coefficients. Two likely transport candidates for the SF zone plasma mixing, classical electromagnetic drifts and the churning mode, provide fluxes whose relative magnitudes and contributions scale differently over the range of experimental plasma currents (and PF) and SOL power widths. The UEDGE model is also used to project safe divertor operations in future 1-2 MA NSTX-U SF experiments with input power up to 10 MW. In SF configurations, the magnetic field in the additional strike points is in the opposite direction and small sections of the “fish-scaled” graphite tiles are exposed to nearly normal heat fluxes. The modeling shows that even with high degree of power exhaust redistribution into the additional strike points, heat fluxes remain below 5-7 MW/m$^2$.

Speaker: Vsevolod Soukhanovskii (LLNL)

111 1.106 Automated divertor target shape optimization using SOLPS-ITER and GOAT

It is well known that the design of the divertor component, in particular its shape, is crucial to deal with the impinging heat and particle loads. Currently, the divertor is typically designed in a trial-and-error fashion, using high-fidelity codes such as SOLPS-ITER [1]. This is often a time and resource consuming task due to the complex interaction between the divertor shape and the plasma edge transport. To speed up this process and improve the design even further, optimization-based methodologies were already presented over a decade ago, yielding promising results [2]. However, they considered a highly simplified plasma edge model. In this contribution, we extend this approach by using the state-of-the-art plasma-edge code SOLPS-ITER, in its most recent wide-grid version.

To achieve this, computational grids need to be generated in an automated and differentiable way. The grid deformation module of the Grid Optimization and Adaptation Toolbox (GOAT) was specifically designed for this purpose: it deforms an initial grid to match a newly proposed target shape while maximizing the grid quality [3]. Additionally, sensitivities are required by the optimization algorithm to automatically propose improved designs. Contrary to the continuous adjoint approach used in [2], we present a combined adjoint Algorithmic Differentiation (AD) and discrete adjoint approach to compute the sensitivities at a cost independent of the number of design variables. The latter has already been applied to compute sensitivities with respect to plasma parameters [4] and is extended here to compute sensitivities with respect to grid coordinates.
As a proof of concept, we apply the optimization framework to a COMPASS case with approximately 100 design variables for the (outer) target shape. We consider the L2-norm of the perpendicular heat load impinging on the target as cost function, and constrain the vessel shape to prevent hard to manufacture designs. The resulting design promotes detachment by increased neutral trapping near the strike points, yielding a peak heat load reduction of a factor 4. This new automated divertor shape optimization framework can now support the design of future devices such as the EU-DEMO.

[1] S. Wiesen et al., Journal of Nuclear Materials 463 (2015) 480–484; X. Bonnin et al., Plasma and Fusion Research, 11 (2016) 1403102.
[2] W. Dekeyser et al., Nucl. Fusion 54 (2014) 073022.
[3] S. Van den Kerkhof et al., Contrib. Plasma Phys. 64 (2024) e202300134
[4] S. Carli, Journal of Computational Physics 491 (2023) 112403.

Speaker: Sander Van den Kerkhof (KU Leuven)

112 1.107 Design and modeling of a closed divertor with mid-leg pumping for core-edge integration in DIII-D

A novel divertor design is presented that employs mid-leg pumping in a tightly baffled outer divertor with the goal of improving both access to detachment and stability of the detachment front, such that core plasma confinement degradation is minimized in a detached scenario. Engineering designs, high fidelity simulations and design considerations are presented for this divertor that is planned for installation in DIII-D for experiments in 2027.

The “Chimney” divertor is designed with an extended outer leg and a pump duct positioned poloidally upstream from the target in the baseline plasma shape. This volume acts as a neutrals reservoir surrounding the outer divertor target, increasing the rates of neutral friction, charge exchange and recombination that dissipate energy at low Te near the target. By then removing neutral particles before they are allowed to propagate further upstream using the ducted cryopump, the detachment front is passively stabilized with the goal of maintaining high core confinement with a detached divertor. This design also minimizes main chamber neutral density during detachment, which would reduce charge exchange neutral particle fluxes and associated sputtering of the main chamber wall in a reactor.

High fidelity 2D boundary plasma and neutral particle simulations have been performed to predict the behavior of this divertor design using both the UEDGE [1] and SOLPS-ITER [2] code suites. These simulations predict that detachment is accessible at lower upstream density with the pump duct positioned upstream from the target, and also that the detachment front (defined here as the location where Te=10 eV along the divertor leg) is passively stabilized poloidally near the position of the pump duct opening for a broad range of experimentally relevant power and particle injection.

Flexible magnetic equilibria are shown to be realizable with the existing DIII-D coil set with up to I$_\textrm{P}$=1.8 MA at B$_\textrm{T}$=2.1 T, resulting in q$_\textrm{}95$=3.3 and an estimated inter-ELM heat flux width of $\lambda_\textrm{q,Eich}$=1.6 mm. A comprehensive set of diagnostics is planned to document the behavior of this divertor as power and particle inputs and magnetic geometry are scanned. This work enables effective model validation to help ensure that design concepts can be confidently extrapolated to future devices.

[1] A. Holm et al., Nuclear Materials and Energy 41 (2024), 101782.
[2] J.H. Yu et al., Nuclear Materials and Energy 41 (2024) 101826.

This work was supported in part by the US Department of Energy under DE-FC02-04ER54698, DE-AC05-00OR22725, DE-AC52-07NA27344, DE-NA0003525, and DE-SC0023378.

Speaker: Robert Wilcox (Oak Ridge National Laboratory)

113 1.108 Empirical scaling of divertor heat loads in Wendelstein 7-X

Power crossing the last closed flux surface is guided toward the divertor plates, where it deposits on a small area (wetted area), which in a reactor-scale device would result in an enormous heat flux. For given technical constraints, the maximum wetted area achievable in axisymmetric tokamaks is determined by the power decay length ($\lambda_{q}$) [1]. For Wendelstein 7-X (W7-X), the divertor concept relies on large magnetic islands formed at the plasma boundary to govern exhaust. The field-line pitch angle ($\theta_i$) within the boundary island is approximately two orders of magnitude smaller than that at the tokamak boundary, resulting in enhanced perpendicular transport comparable in magnitude to parallel transport [2]. Due to the non-axisymmetric distribution of heat loads and the complex situation of plasma surface interaction, the power decay length $\lambda_q$ as used in tokamaks, seems inappropriate to characterize the heat load in W7-X.

In W7-X, all ten divertor units are monitored by infrared thermography systems [3]. The 3D implicit anisotropic heat-diffusion solver DELVER (Divertor Energy Load Versatile EstimatoR) [4] has been developed to calculate the heat flux on the water-cooled divertor surfaces [5].

In this study, we introduce a novel approach to quantify the heat load in W7-X by defining an effective area parameter, $A_i$. This parameter is obtained by tracing magnetic field lines to map the heat flux distribution from the divertor targets to a selected poloidal cross-section [6, figure 12 therein]. Unlike previous estimations of wetted areas directly on the divertor plates [7]—which are strongly affected by the specific geometry of W7-X—our method decouples plasma transport effects from geometric influences such as divertor shape and field-line incidence. This separation is essential for gaining deeper physics insights and for enabling reliable extrapolation to reactor-scale stellarators. The analysis is performed for a range of plasma parameters with an aim to establish a scaling law for $A_i$ in terms of relevant parameters, such as $\theta_i$, as well as boundary plasma conditions.

References:

[1] T. Eich et al., Nucl. Fusion 60 (2020) 056016.
[2] Y. Feng et al., Nucl. Fusion 46 (2006) 807.
[3] M. Jakubowski et al. Rev. Sci. Instrum. 89(2018), 10E116.
[4] S. Thiede et al., submitted to Rev. Sci. Instrum.
[5] Y. Gao et al., Nucl. Fusion 59 (2019) 066007.
[6] Y. Gao et al., Nucl. Fusion 60 (2020) 096012 (14pp).
[7] H. Niemann et al., Nucl. Fusion 60 (2020) 084003.

Speaker: Dr Yu Gao (MPPL)

114 1.109 Dynamic SOLPS-ITER modelling of the X-point radiator regime in ITER

The X-point radiator (XPR) regime (see the Ref. [1] and the references therein) is now commonly explored in most modern tokamaks using a variety of impurity seeding gases. Given its usual association with almost complete divertor detachment and benign or no ELM activity, it is an extremely attractive potential option for future reactor-scale machines.
On the SOLPS-ITER numerical simulation side, steady-state modelling has successfully reproduced the experimental XPR regime in ASDEX-Upgrade [2] and has been used to achieve the first predictive modelling of this regime for ITER [3]. One caveat is that maintaining a stable XPR regime in these simulations requires boundary conditions with fixed densities and temperatures at the inner core boundary and a high concentration of radiating impurity.
An alternative modelling strategy to achieve and stabilize the XPR regime has recently been demonstrated in [4] for ASDEX Upgrade based on SOLPS-ITER dynamic modelling. This approach uses feedback control of the XPR regime by varying the impurity seeding in a similar fashion to the experimental methods of XPR control and stabilization. The chosen control parameter for feedback is the minimum electron temperature at a selected core flux surface where the radiating region is located. In contrast to the previous stationary simulations, in this new dynamic modelling, more physically justified boundary conditions of fixed heat and particle flows from the core are applied at the inner core boundary.
We present here the first dynamic SOLPS-ITER simulations of the XPR regime in ITER following the strategy developed in [4]. Calculations are performed for neon and argon as radiating impurities. The dynamics of the transition to the XPR regime and the possibility of achieving a steady-state XPR solution in ITER is investigated. The resulting stable X-point radiator regimes are achieved with lower impurity content and power crossing the inner core boundary than in [3]. As a result, the XPR is obtained with relatively low Zeff in the core region of the simulation domain which is favourable for integration of the regime with burning plasma operation.

[1] Bernert M. et al 2025 Nucl. Mater. Energy 43, 101916
[2] I.Senichenkov et al. 2021 Plasma Phys. Control. Fusion 63 055011
[3] A. Poletaeva et al 2024 Nucl. Fusion 64 126038
[4] A. Poletaeva et al 2025 CPP to be published

Speaker: Anastasia Poletaeva (Peter the Great St. Petersburg Polytechnic University)

115 1.110 The exhaust capability of the quasi-continuous exhaust regime

The quasi-continuous exhaust (QCE) regime is naturally type-I ELM free. It combines the high density at the plasma edge needed for power exhaust with the high normalised energy confinement typical for H-mode operation. In the QCE regime large-scale type-I ELMs are replaced by high-frequency, low-amplitude filaments leading to the quasi-continuous edge transport of particles and energy [1].
Despite the high scrape-off layer density and broad power fall-off length, the regime does not routinely show a detached divertor state in ASDEX Upgrade either with or without small amounts of nitrogen seeding; larger amounts of nitrogen seeding have so far led to a back-transition to an ELMy H-mode. Similarly, small amounts of neon seeding reduced the divertor power significantly in JET without clear signs of detachment.
Although at least partial detachment in between the filaments is observed, the variation in time due to the high-frequency filaments poses an additional challenge to consider. The detachment process might be hampered by the modulation induced by the filaments which likely will need to - at least partially - be buffered for reactors. Furthermore, it is not clear if the QCE regime can be maintained with high seeding levels. As the underlying physics is believed to be ballooning modes, which - if ideal or kinetic ballooning modes - need a high enough pressure gradient to be destabilised [2]. High levels of radiated power are expected to lead to lower the kinetic gradients and, hence, the ballooning drive.
In this contribution, initial analysis of the scrape-off layer and divertor conditions will be presented in the presence and absence of impurity seeding in both ASDEX Upgrade and JET, including upstream separatrix density and temperature values and downstream target ion saturation and temperature values. These will be compared to simple detachment models and the open research points formulated.

References
[1] FAITSCH, M. et al., Nuclear Materials and Energy 26 (2021) 100890.
[2] DUNNE, M. et al., Nuclear Fusion 64 (2024) 124003.

Speaker: Michael Faitsch (Max Planck Institute for Plasma Physics)

116 1.111 Investigating the role of baffling, fuelling location, and total flux expansion on detachment front location sensitivity in MAST-U discharges

Reactors will operate in the detached divertor regime, with the hot ionizing plasma away from the target, to avoid excessive heat fluxes to the walls. A reduced sensitivity of this detachment front location is advantageous as a passive stabilization measure against variations in the upstream parameters. During transients, it can avoid the hot plasma reaching the target, or the cold neutrals reaching the X-point and possibly degrading the core performance.
Alternative divertor configurations and divertor baffling can affect the detachment location sensitivity profile. They can be investigated on the MAST-U tokamak, either by varying the strike point location in proximity of the baffle or by comparing different magnetic geometries in strongly baffled conditions. While initial investigations focused on conditions with the detachment front on the outer legs within the baffled divertor chambers, recent extensions to the diagnostic coverage allow tracking the front location also outside them. This enables experimental studies on divertor geometries with varying levels of baffling, as well as variations in front sensitivity as the front exits the baffle.
Baffling is shown to have a strong effect on the detachment front sensitivity in L-mode discharges. Shifting the outer strike point away from the baffle entrance to a more open configuration results in a strong increase in the upstream density required for detachment and a much higher sensitivity of the front location, with the front immediately jumping to the X-point and then creating a MARFE. Comparable sensitivity is instead found in conventional and Super-X divertor configurations once the detachment front is outside of the baffle, indicative of the front location being insensitive to the divertor geometry downstream of it for comparable baffling profiles.
The role of fueling location on the front sensitivity is also investigated in fueling-driven detachment conditions, comparing valves in the main chamber, PFR, and in the divertor chamber. The detachment front location is found to agree better when compared at the same divertor neutral pressure instead of the upstream density, highlighting the role of the local neutral pressure as parameter driving the detachment evolution. Fuelling in the PFR is found to be efficient in pushing the detachment front towards the X-point, while fueling in the divertor results in a strongly reduced sensitivity once the front reaches the baffle entrance, in agreement with previous modeling [1].
[1] Roberto Maurizio et al 2025 Plasma Phys. Control. Fusion in press
https://doi.org/10.1088/13616587/ae24ab

Speaker: Nicola Lonigro (UKAEA)

117 1.112 Heat flux expansion in high heat flux experiments on DIII-D

Recent experiments on DIII-D measure broadened heat flux profiles in the divertor; broadening increases with auxiliary heating at both low and high plasma current, reaching up to 3 times broadening relative to the ITPA multi-machine heat flux scaling regression[1]. These discharges push to midplane parallel heat flux $q_{||}\sim1$ GW/m$^{-2}$ to study profile broadening at previously unexplored power density levels for the
DIII-D facility.

Discharges are high triangularity H-modes with a range of plasma current ($I_p=1.0-1.9$ MA) and a combination of neutral beam and electron cyclotron heating power ($P_{aux}=6-17.7$ MW) at $B_T=2.1$ T with ion $B\times\nabla B$ drift down. A double-null shape is biased down with $dR_{sep}\approx -1.5$ cm to ensure heat flux is directed primarily to the lower divertor, which remains attached over the explored density range ($n/n_G=0.4-0.75$) to allow heat flux analysis. Outer strike point position and control are optimized to ensure $3-4$ heat flux widths ($\lambda_q$, calculated at midplane) enter the divertor over the $I_p$ range, with sweeps for diagnostic profiles. In-tile fixed Langmuir probes and a divertor viewing IR camera provide heat flux profiles in the divertor, and an edge Thomson scattering array gives high resolution pedestal profiles at the outer midplane.

Upstream scrape-off layer (SOL) heat flux width, estimated as $\lambda_q \approx 7/2 \lambda_{T_e}$ from upstream $T_e$ pedestal profiles, aligns well with the ITPA scaling. Divertor heat flux profiles are measured by Langmuir probes and IR camera, with both measurements yielding similar broadening beyond that suggested by the Eich function. Target heat flux widths range from $\lambda_{q,div}\approx 1.5 \lambda_{q,ITPA}$ for both low and high current discharges to $\lambda_{q,div}=2.7 \lambda_{q,ITPA}$ at low $I_p$ and high $P_{aux}$ and $\lambda_{q,div} \approx 3.5 \lambda_{q,ITPA}$ at high $I_p$ and high $P_{aux}$. ELITE pedestal stability analysis indicates that ELMing discharges with $I_p=1.9$ MA and $P_{aux}=17.7$ MW are near the ballooning boundary.

Future experiments will push to higher density to characterize detachment dynamics at high heat flux, and planned upgrades to ECH will push performance further towards FPP-relevant regimes.

[1] Eich, T., et al., Nucl. Fusion 53 (2013) 093031
Work supported by US DOE under default DE-FC02-04ER54698, DE-NA0003525, DE-AC52-07NA27344, DE-AC05-00OR22725, DE-SC0014264.

Speaker: Auna Moser

118 1.113 Detachment in JET-ILW L-mode deuterium and helium plasmas: Experimental and SOLPS-ITER comparison

A density scan of two heating powers, $P_{\rm{NBI}}=1$ MW and 5 MW, was performed in JET ITER-like wall (JET-ILW) experiments and SOLPS-ITER simulations of NBI-heated low-confinement mode (L-mode) helium (He) plasmas. In high-recycling conditions, ion flux to the low-field side (LFS) divertor, $I_{\rm{div,LFS}}$, is 70% lower in He then deuterium (D). SOLPS-ITER underpredicts $I_{\rm{div,LFS}}$ by ~30% for both D and He. A lower $I_{\rm{div,LFS}}$ between species of up to 50% is expected, as $\rm{He^{+2}}$ has double the charge. The larger difference is due to the higher effective ionisation cost for He, primarily the second ionisation, which reduces the recycling loop that can be supported. With the extreme divertor heat flux expected in future tokamaks, it is important that we understand the processes and uncertainties of our plasma boundary models during detachment when planning how to control the power exhaust during non-nuclear commissioning (He, H) and operational (D, DT) phases. The total radiated power, $P_{\rm{rad}}$, is equal between D and He in experiment and simulation in low and high-recycling conditions. However, the SOLPS-ITER $P_{\rm{rad}}$ is half that of bolometry. For $P_{\rm{NBI}}=1$ MW, simulated power to the LFS target is within 0.5 MW of IR camera measurements (IRTV), for both species. For $P_{\rm{NBI}}=5$ MW, SOLPS simulates double the IRTV power to the target. In JET-ILW, input (Ohmic and NBI heating) and output (Bolometry and divertor IRTV) power can be balanced for $P_{\rm{NBI}}=1$ MW, but there is a density independent 2 MW deficit for $P_{\rm{NBI}}=5$ MW. The strongest source of radiated power in the He simulations are the UV lines of the Lyman series of $\rm{He^{+1}}$, at over 80% of $P_{\rm{rad}}$, independent of density or heating power. A synthetic diagnostic of the JET-ILW VUV spectrometer, performed by the 3D ray-tracing code Cherab, underpredicts measured Lyman lines by a half. SOLPS-ITER predicts similar neutral flux into the confined region for D and He, but the simulations and bolometry indicate greater core radiation in He, as $\rm{He^{+1}}$ is a more effective radiator than $\rm{D^{0}}$. In D at $P_{\rm{NBI}}=1$ MW and He at $P_{\rm{NBI}}=5$ MW, a strong drop in the LFS target / midplane pressure ratio with increasing upstream density indicates detachment through momentum and power loss along the SOL. Strong pressure loss is not predicted in the He $P_{\rm{NBI}}=1$ MW case, which reaches radiative core collapse in the simulations due to the neutral flux into the confined region at densities lower than detachment.

Speaker: David Rees (Aalto University, Espoo, Finland)

119 1.114 Effect of divertor geometry on neutral compression, SOL width and broadening in MAST Upgrade

MAST Upgrade is designed and built to have a tightly fitting baffle to allow the use of either a conventional or long legged closed divertor configuration. This design allows the main chamber plasma to be decoupled from the divertor plasma, giving greater power dissipation before the divertor target in the closed divertor chamber and better access to detachment, whilst maintaining high performance core plasmas. This contribution will compare the effect of optimally closed conventional divertor geometries on MAST-U on the SOL widths and target profile broadening, with MAST-like open divertor geometries where the strike point is at a radius smaller than that of the baffle nose. A sudden rise in midplane Dα is observed as the strike-point decreases below Rsp < 0.75 m, which indicates a transition from baffled (closed) to unbaffled (open) divertor geometry. The amount of flux going through the gap between the strike line and the baffle nose also plays an important role in trapping the neutrals in the divertor – so called plasma plugging. It has been seen that in an optimally closed double null divertor in beam heated H-mode the λq is smaller than a comparable plasma with a more open divertor, i.e. Rsp < 0.75m. Furthermore, in an LSN geometry with a closed divertor, the λq is further reduced. Previous studies (Thornton et al [1]) of the SOL width in an open divertor configuration noted there was no correlation of the SOL width over a range of drsep between 2 and 7 mm. The S factor in the Eich fit [2], representing the SOL broadening due to cross-field diffusion, is larger in closed conventional divertor configurations compared to (MAST-like) open conventional divertor. This is likely due to the closed nature of the divertor increasing recycling and neutral interaction which can spread the power over a larger space. This is notable, given that on MAST it was postulated that the broadening in the divertor did not show any notable variance with various core and divertor parameters and was concluded to be due to the open divertor geometry.

This work has been funded by the EPSRC Energy Programme [grant number EP/W006839/1].

[1] A J Thornton et al 2014 Plasma Phys. Control. Fusion 56 055008
[2] T Eich et al 2011Phys. Rev. Lett. 107, 215001

Speaker: Sarah Elmore (UKAEA)

120 1.115 Effect of Drift Driven Fluxes, Magnetic Geometry Bias, and Operational Actuators on Upper/Lower Asymmetries in MAST-U

MAST Upgrade experimental infrared thermography and Langmuir probe data have been used to study the upper to lower outer targets asymmetries, especially the ratio of the peak of the heat flux densities, for a variety of experimental discharges. SOLPS-ITER simulations with a wide range of actuator conditions and magnetic configurations (Conventional, Elongated, and Super-X; combined with Connected and Disconnected Double Null) and with all drifts and currents activated have been used to separate the relative importance on up/down asymmetries caused by the magnetic bias, drift driven fluxes, and actuator asymmetries. In fully symmetrical connected double null simulations in the Super-X divertor configuration, the activation of the drifts in SOLPS-ITER drives an upper biased asymmetry of the peak of the heat flux density to the outer targets (up to a factor of 15) that is well ordered by the difference of the upper to lower electron target temperatures [1]. The predicted upper biased asymmetries are observed experimentally, with the cause from the simulations attributed to thermoelectric currents. However, operational correlation introduced by the standard MAST-U discharge program between the inter separatrices distance and the global plasma density/temperature complicates the separate interpretation of the role of the magnetic geometry bias and the drift driven fluxes, as both are expected to constructively cause an upper bias asymmetry of the ratio of the heat flux densities. To address this issue, new experiments will be carried out in MU05 with the aim of breaking these operational correlations.

For simulations with Conventional divertor geometry the asymmetry is weaker (up to a factor of 8) and depends on the gas puff strength. The less efficiently dissipative nature of the open CD allows upper and lower divertors to be in the same collisionality regime, whereas the lower outer divertor of the closed SXD simulations is always detached. The new MU05 data will help to validate the relative importance of drift-driven fluxes and magnetic geometry bias predicted by the different sets of SOLPS-ITER simulations. Understanding these asymmetries is important to develop control strategies for spreading power evenly among divertors, as all four divertors need to simultaneously have acceptable conditions.

This work is supported by the US DOE under contract DE-AC05-00OR22725.
This work was partly funded by the EPSRC Energy Programme (Grant No. EP/W006839/1)
[1] I. Paradela Perez et al, Nucl. Fusion, 2025

Speaker: Ivan Paradela Perez (ORNL)

121 1.116 Training of 1D and 2D surrogated model of UEDGE transport code for divertor detachment control on EAST

For future burning plasmas, mitigating divertor heat loads is critical to achieving fusion in high-performance operations. Impurity seeding to regulate heat flux to divertor targets has emerged as a promising strategy, yet injected impurities can degrade core plasma performance. This necessitates precise divertor state measurements to enable robust feedback control. Recent advancements in artificial intelligence (AI) now allow the training of surrogate models based on transport codes to enhance control accuracy. This novel approach has recently been tested on KSTAR and is demonstrated to be effective [B. Zhu et al 2025 Phys. Plasmas 32, 062508]. At this conference, we present 1D and 2D surrogate models trained via neural networks on UEDGE transport simulation data of EAST discharges. A typical EAST H-mode discharge with low single null and favorable toroidal magnetic field configuration was selected as the baseline simulation (with ExB and diamagnetic drifts turned on). The upstream plasma profiles and the outer divertor particle flux width were tuned to agree with the experiments. A comprehensive dataset comprising over 400,000 simulations (unphysical simulations like thermal quench in the core plasmas were removed systematically) was generated through parameter scans of plasma current (using FreeGS equilibrium code to change the plasma current while keeping the last closed flux surface, X points and plasma core pressure unchanged), plasma line-averaged density, impurity concentration of Neon, particle diffusion coefficient (the ratio between particle and heat diffusion coefficients is constant), and total heating power. Inputs (the scanned parameters) were normalized to their maximum values, while outputs (outer target plasma parameters for 1D models; full UEDGE fields for 2D models) underwent Z-score normalization (logarithmic transformations applied to positive fields like plasma density and temperature to increase training accuracy). The 1D model employed a fully connected architecture, while the 2D model utilized 2D-convolutional layers. Hyperparameters (network depth, layer width, epochs, learning rate) were optimized using the Optuna library. The trained models achieved accuracy in terms of determination of coefficient (i.e. R2 score) exceeding 0.97 (1D) and 0.95 (2D). These surrogate models will be integrated into EAST’s feedback control system in upcoming experiments.

Speaker: Xiang LIU (Institute of plasma physics, Chinese Academy of Sciences)

122 1.117 Effect of strike point position on the divertor heat load for Japanese DEMO

One of the critical challenges facing the design and operation of next-step high-power steady-state fusion devices is to develop a divertor solution for handling power exhaust, which necessitates access to divertor detachment at relatively low main plasma density. Now Japan focuses on the design of Japanese DEMO (JA DEMO) with ITER size. The V-shaped divertor geometry formed by the vertical and horizontal targets, as a potential candidate for JA DEMO, has a good divertor closure and then facilitates the control of divertor heat load and achievement of divertor detachment. However, for the V-shaped geometry, the interaction of the recycled neutrals with the SOL plasma is strongly dependent on the position of strike point, which has a significant effect on the divertor detachment onset or degree. So, it is high priority to explore the underlying neutral distribution and transport for the control of divertor heat load.

In our simulations, two possible ITER-like divertor geometries corresponding to the outer strike point (OSP) on the vertical and the horizontal targets respectively are used to assess the impact of OSP displacement on the divertor plasma behavior by SOLPS-ITER. The seeding rate of radiation impurity neon (Ne) is scanned with the energy flux entering the scrape-off layer (SOL) P_SOL=100MW. The modeling results show clearly that the corresponding heat flux density and electron temperature at the outer divertor target are reduced with OSP switching from vertical to horizontal target, with the lower upstream electron density for divertor detachment onset.

The different divertor plasma behavior for JA DEMO caused by the change of OSP from the vertical to horizontal plate is mainly due to the different recycling behaviors of neutrals in the outer divertor region. When OSP is on the vertical divertor plate, the recycling neutrals are towards private flux region (PFR). However, when OSP is on the horizontal divertor plate near the corner, the recycling neutrals are directed to SOL region by the reflection on the vertical plate, promoting the recycling behaviors of neutrals in the outer divertor region. Therefore, the divertor configuration with OSP on the horizontal divertor plate near the corner can help concentrate more neutrals near the outer divertor region, inducing the larger momentum loss and stronger detachment. Moreover, the V-shaped divertor geometry with OSP on the horizontal divertor plate benefits the control of heat load and electron temperature at the far SOL region.

Speaker: Dr Hang Si (National Institutes for Quantum Science and Technology(QST))

123 1.118 The impact of boron on tungsten plasma facing walls: sputter yields, near-surface morphology, and fuel retention

Tungsten (W) is the main candidate for plasma-facing materials in tokamaks, as it features, among other properties, low sputter yield and low retention of hydrogen isotopes. However, W lacks intrinsic gettering properties for mid-Z impurities, which are necessary to reduce the presence of impurities that are otherwise capable of degrading the plasma. ITER plans to use boronization as a wall conditioning process that coats the walls with a thin layer of boron (B) to reduce the partial pressure of oxygen and water. This procedure may lead to formation of W–B compounds through redeposition. Such modifications can significantly affect sputtering, fuel retention, and interaction with seeding gases.

In this contribution, we will summarize recent comprehensive experimental and computational efforts to understand the behavior of tungsten and boron containing mixed materials and their consequences in fusion devices. Layers with integral composition of W1B(1-x) with x in the range from 0 to 1 were prepared, mimicking layers expected to form in ITER. Experiments, often performed in-situ, investigated sputter yields, surface morphology evolution (surface enrichment, crack formation) after ion irradiation, hydrogen isotope retention and neon incorporation across a range of irradiation temperatures. Finally, detailed atomic-scale modeling, including sputtering simulations based on molecular dynamics as a function of surface composition, were conducted and show good agreement with the observed experimental trends.

Our results highlight the critical impact of boron incorporation on key surface properties of tungsten, with direct relevance for plasma-facing component performance and lifetime in future fusion reactors.

Speaker: Daniel Primetzhofer (Uppsala University)

124 1.119 Development and validation of an integrated SOLPS-EPED model in MAST-U

We present a novel edge-SOL coupling scheme linking SOLPS-ITER SOL transport simulations with EPED pedestal predictions and validation efforts through MAST-U dedicated experiments. This scheme employs a flux-gradient-driven density pedestal model [1], taking as inputs the separatrix conditions from a converged SOLPS solution and predicting the density pedestal top through empirically informed transport settings. This top density is used as an input to the EPED model as in [2], along with self-consistent shaping from the SOLPS equilibrium, to predict pedestal dimensions limited by ideal P-B and KBM stability constraints.

SOLPS-EPED validation efforts were conducted on MAST-U through experiments manipulating the role of neutral sourcing vs transport on the pedestal parameters through 1) modifying divertor closure and 2) adding divertor gas puffing. Reducing the divertor closure leads to considerably higher pedestal density at roughly constant pedestal pressure, which is reproduced by the SOLPS-EPED modeling. However, the strong localized HFS gas fueling required for reliable type I ELMy H-mode operation poses a challenge in characterizing the poloidal source distribution with divertor closure changes due to small variation in the inner gap in close proximity to the HFS fueling valve.

Conversely, additional divertor puffing at fixed divertor geometry and HFS fueling is shown to have a negligible effect on the density pedestal due to inefficient fueling in the conventional divertor configuration, but a significant effect on degrading the temperature pedestal. The density profile does not incur any obvious outward profile shift that could explain a reduction in P-B stability as in similar JET experiments [3]. Based on the EPED ideal-MHD and KBM constraints, the SOLPS-EPED coupled model is unable to capture this pedestal degradation, which may arise due to additional transport mechanisms, or the local HFS fueling interaction. Further experiments with varied plasma current will facilitate additional validation opportunities by moderating the natural edge density as well as isolating the effect of the HFS localized gas puff.

[1] S. Saarelma et al 2024 Nucl. Fusion 64 076025
[2] R.S. Wilcox et al Nucl. Fusion [In Review]
[3] L. Frassinetti et al 2021 Nucl. Fusion 61 126054

*Work supported by US DOE Contracts DE-SC0014664, DE-AC05-00OR2272, DE-SC0023289 and EPSRC, UK Energy Programme grant EP/W006839/1. For the MAST-U Team, see the author list of J. Harrison et. al. 2024 Nucl. Fusion 64 112017.

Speaker: Dr Davis Easley (ORNL)

19:00 Reception

Tuesday 19 May

Review Talk: Morning session

R1

125 R2 Cracking and local ageing of tungsten plasma facing components – does it matters?

Tungsten (W) is widely used as a plasma-facing material in fusion devices due to its high melting point, low sputtering yield, and excellent thermal conductivity [1]. However, due its mechanical behaviour it is prone to crack under harsh experimental fusion environments: high heat fluxes (cycling steady state loading, ELMs & disruption) [2], light impurities bombardment (H, D, He) [3]. Experiments run in high heat flux test facilities for the qualification of plasma facing components (PFCs) using tungsten as armor material highlighted macro-crack propagation through the tungsten monoblocks during steady state cyclic solicitations [4]. The presence of cracks not only degrades the structural integrity of tungsten components but may also limit the power plant availability and the plasma performance.
Since the first experiments in ASDEX-U (end of 90’s) [5], the use of tungsten as plasma facing material has become a standard practice for divertor PFC application (JET, ASDEX-U, EAST, KSTAR, WEST…). In some devices, tungsten is also used as plasma facing material for first wall limiter (EAST, WEST,…). These fusion devices accumulated hours of plasma in this configuration and report systematic cracking of tungsten elements [6-7-8]. This overview aims at presenting the cracking phenomenology (pattern, types, quantity, localisation...) observed after plasma solicitation in several fusion device environments at the divertor & the first wall regions. Rationale (thermal cycling, energetic events, …) regarding their occurrence is also discussed based on cracking patterns observation, metallographic analysis and crack initiation / propagation modelling. The impact of such cracking patterns on the dust collection process, the high impurities sources, the plasma operation and maintenance scheme are discussed. Off-normal events such as disruptions or runaway electrons may also induced a larger variety of damage patterns than cracks. Recrystallization, delamination and local melting are finally also discussed based on divertor and first wall post-mortem analysis available.

*Corresponding author e-mail: alan.durif@cea.fr

[1] Abernethy, R. G. 2016. MST, 33(4), 388-399
[2] M. Wirtz et al, NME, 12 (2017) 148-155
[3] Y. Li et al 2021 Nucl. Fusion 61 046018
[4] S. Nogami et al, FED 120 (2017) 49-60
[5] K. Krieger et al, JNM 266 -269 (1999) 207-216
[6] Chuannan Xuan et al 2025 Nucl. Fusion 65 046027
[7] Dahuan Zhu et al 2022 Nucl. Fusion 62 056004
[8] M. Diez et al, NME, Vol 41, 2024, 101746

Speaker: ALAN DURIF (CEA, IRFM, CEA, F-13108 Saint Paul lez Durance, France)

Invited Talk: Morning session

126 I6 Overview of plasma wall interactions in the new high particle fluence campaign of WEST operated with cold divertor plasma conditions

The plasma facing components of next step fusion devices will handle unprecedented heat flux and particle fluence. The WEST tokamak, equipped with an actively cooled tungsten ITER grade divertor, aims to assess the divertor performance under tokamak conditions. A first high fluence campaign was performed in WEST, in 2023, on the new actively cooled tungsten divertor based on the ITER-grade monoblock concept. The campaign consisted of the repetition of a 60 s long Deuterium L-mode pulse in attached divertor conditions, cumulating over 3 hours of plasma exposure. A deuterium fluence of approximately 5×1026 m−2 was reached in the outer strike point region, representative of a few high performance ITER pulses. This attached condition with electron temperature in the range of 20-40 eV leads to intensive tungsten erosion and to the formation of tungsten deposits prone to the release of flakes that penetrate into the plasma and can trigger disruptions.

After laser cleaning of the whole lower divertor, a second high fluence campaign was performed in WEST, in 2025, with cold edge plasma (Te<5eV) using the X-point radiator regime. The campaign consisted of 4 weeks of repetition of 75s long L-mode pulses (Ip = 370 kA, PLHCD = 3.8 MW, nle= 4×1019 m-2) with a controlled XPR phase of ~70s obtained with nitrogen seeding. Each XPR pulse cumulated a particle fluence of about 3.5×1024 m−2, 25% lower than the pure deuterium attached plasma pulses, while the heat load was divided by factor of 4. Core confinement was improved through a combination of ion dilution effects and reduced tungsten contamination with the reduction of the tungsten sources in the divertor (factor 10 reduction). This scenario was alternated with ohmic discharges of 10s in limiter or divertor configuration to reduce nitrogen legacy and ensure good repeatability of the pulses over the day. After 3 weeks among 4, already ~110 repetitive XPR pulses were performed cumulating ~3 hours of plasma operation, ~25 GJ injected and ~3.8×1026 m−2 of particle fluence, on the right track to cumulate a deuterium fluence equivalent to the previous high fluence campaign. The infrared measurements during the campaign show some divertor surface evolution but at this stage not a single UFO has been observed during the entire campaign. A comparative analysis of the two high fluence campaigns will be discussed as well as the first observations from the visual inspection and post-mortem measurements after the campaign.

Speaker: Jonathan Gaspar (Aix Marseille Univ., CNRS, IUSTI)

Oral: Morning session

127 O8 Thermal-gradient effects on recrystallization and grain coarsening in fusion-relevant tungsten monoblocks: a phase-field study

Abnormal grain growth (AGG) has been reported in tungsten monoblock divertor targets after repetitive high-heat-flux (HHF) exposure, raising questions about the specific role of steep divertor thermal gradients in triggering this behavior. In this contribution, we use an experiment-informed coupled modeling framework to ask: are fusion-relevant transient thermal gradients, by themselves, sufficient to produce AGG-like microstructures in tungsten?

A multi-order-parameter phase-field grain-growth model is coupled to a heat-conduction model of a tungsten/Cu/CuCrZr monoblock geometry. Thermal boundary conditions are calibrated to HHF loading representative of the Max Planck Institute’s GLADIS campaigns (up to ~20 MW m⁻²) and used to extract realistic near-surface temperature fields. Grain-boundary mobility is temperature dependent via an Arrhenius relation (activation energy 4.146 eV), with the prefactor fitted to published tungsten mobility data to match grain-growth kinetics. To enable tractable simulation of large domains, we evaluate cyclic HHF, single-pulse, and constant-elevated-temperature representations and show that the simplified constant-temperature approach reproduces the final grain structure for equivalent thermal exposure durations within this modeling scope.

Simulations reveal strong gradient-controlled coarsening: within the phase-field window, the experimental-informed temperature variation corresponds to a ~438% mobility increase from the colder to hotter region, producing pronounced depth-dependent grain growth. However, the grain-size distribution remains approximately Gaussian (rather than bimodal), and AGG does not emerge under the DEMO-relevant gradient fields used here. When an artificially sharpened near-surface thermal boundary layer is imposed, AGG-like oversized grains do appear, demonstrating that the framework can reproduce AGG morphologies when sufficiently steep mobility gradients exist. Collectively, these results suggest that realistic thermal gradients are a powerful accelerator of coarsening but that additional coupled drivers (e.g., stored strain/thermomechanical cycling, recrystallization-related size advantages, and/or impurity/solute drag) are likely needed to explain AGG observed in HHF-tested monoblocks.

Speaker: Prof. Narguess Nemati (Department of Mechanical and Production Engineering, Aarhus University, Katrinebjergvej, Aarhus N, 8200, Denmark)

128 O9 Qualification of plasma-facing materials via public and private sector experiments in DIII-D

A coordinated campaign of experiments in the DIII-D tokamak has advanced the qualification of plasma-facing materials (PFMs) through national laboratory and university efforts together with public–private partnerships. Candidate materials included W alloys, additively manufactured (AM) W, K-doped W, neutron-irradiated W and ceramics, high-entropy and multi-principal-element alloys, ultra-high-temperature ceramics, SiC-based composites, and boron-based concepts. These were exposed to reactor-relevant heat and particle fluxes using the Divertor Materials Evaluation System (DiMES), providing a bridge between bench-scale testing and integrated plasma–material interaction conditions expected in ITER and future pilot plants (FPPs). The campaign establishes a rapid, repeatable pathway for qualifying industry-developed materials directly under reactor-relevant tokamak conditions.
Bulk W, cold-sprayed Ta and Ta–W alloys, AM W with varied grain structures, and K-doped W were tested under H-mode plasmas. Inter-ELM heat fluxes of 2.2–2.4 MW/m² and transient ELM loads up to ~6 MW/m² were achieved, with angled samples intercepting >10 MW/m². Surface temperatures near 800 °C were recorded. Large-grain AM W exhibited reduced cracking vs ITER-grade references, while K-doped W maintained structural integrity at 750–780 °C. Alloying constituents showed selective erosion, providing new benchmarks for erosion modeling. W samples with controlled orientations (001 vs 111) reached >1500 °C in angled exposures, crossing the recrystallization threshold and enabling comparison of texture-dependent cracking and erosion. Orientation strongly influenced recrystallization onset and erosion rates, offering guidance for optimizing AM processing routes. Neutron-irradiated W and ceramic samples were also tested, probing coupled effects of irradiation damage and plasma loading.
Advanced W alloys, refractory multi-principal-element alloys, and renewable boron pebble–rod components were also qualified under integrated plasma conditions. Exposures reached inter-ELM heat fluxes of 2.2–2.5 MW/m² with surface temperatures of 600–700 °C and extended up to 15 MW/m² for angled geometries. In-situ spectroscopy identified selective erosion of alloying constituents, while boron pebble–rod prototypes demonstrated controlled boron release without plasma termination. SiC fiber composites withstood ~2.0 MW/m² in ELMy H-mode with moderate microstructural modification, while monolithic ceramics such as Si₃N₄ and B₄C proved more fragile, showing partial ejection during ELMs without major plasma impact.
These experiments, together with those from the first DIII-D materials campaign, constitute one of the most comprehensive qualifications of advanced PFMs in a tokamak environment to date. They provide mechanistic insights into cracking, recrystallization, selective erosion, and controlled material release under transient H-mode conditions, and they establish new benchmarks for modeling, accelerate materials down-selection, and inform the design of plasma-facing components for ITER and FPPs.

Speaker: Florian Effenberg (PPPL)

09:50 Coffee Break

Invited Talk: Morning session

129 I7 Energy Balance and Surface Diagnostics of Runaway Electron Beam Termination in the TCV Tokamak

We present a systematic quantification of the energy balance during controlled runaway electron (RE) beam terminations in the TCV tokamak.
Disruptions in tokamaks can generate relativistic REs capable of carrying a large fraction of the plasma current at multi-MeV energies. A sudden loss of confinement of the REs can deposit extreme energy densities onto plasma-facing components (PFCs), leading to surface melting or vaporization. Understanding how RE energy is redistributed during beam termination is therefore central to disruption mitigation and protection of PFCs. In TCV, after RE beam formation, compressing the plasma against the central column triggers a collapse in a reproducible way, both in time and in space, enabling unprecedented diagnostic access to the termination phase. The reproducibility of the collapse also permits systematic variation of beam and termination parameters, enabling controlled, repeatable tests of dissipation mechanisms. The graphite inner wall is used as a calorimeter via embedded thermocouples, measuring the energy deposition, while infrared thermography captures the spatial heating distribution. Bolometry, magnetic reconstructions, and induced-current simulations quantify radiative emission and power coupled to machine structures. Monte Carlo simulations with GEANT4 complement the experimental analysis by linking the surface heating profiles and the RE energy spectrum.
The combination of these diagnostics and analysis demonstrate that the PFCs become a powerful diagnostic for beam energy and deposition physics. Calorimetry enables direct measurements of the total RE beam energy and the combination with surface temperature measurements allow to estimate RE penetration depth of multi-MeV electrons into graphite. The absence of melting, surface ablation and damage result in more accurate measurement and observation of discharge dependent heat patterns. Results show consistent energy accounting across all channels, with representative discharges yielding approximately 10–20 kJ radiated, 14–15 kJ conducted to the inner wall, and about 4 kJ inductively coupled to surrounding structures. Combining these with reconstructed magnetic energy (≈10–20 kJ) indicates total RE beam energies of 10–30 kJ. A key finding is the clear toroidal symmetry of heat deposition, demonstrated by simultaneous infrared and calorimetric measurements at multiple sectors. Deposition depth in the order of few mm are demonstrated, consistent with simulations. This approach provides a new, experimentally grounded framework for studying discharge termination in the presence of REs by directly connecting the plasma dynamics to the PFC response. The methodology and results are broadly applicable to assessing disruption scenarios in all modern tokamaks, where quantifying, and mitigating, RE–surface interaction is critical for safe operation.

Speaker: Marta Pedrini (GNOI)

Oral: Morning session

130 O10 Coupling Microstructural Evolution Simulations to Material Property Degradation Predictions for Plasma-Facing Materials

Reliable material performance is required for plasma-facing material (PFM) candidates. Previous research has shown that plasma and neutron radiation exposure induces microstructural changes in PFMs; changes in thermal and electrical conductivities and in material hardening and embrittlement were also observed after neutron irradiation. These material property changes will negatively impact the performance of the PFMs in a fusion reactor. Despite the well-known connection between material microstructure, properties, and performance, there is a need for validated modeling capabilities connecting PFM property degradation with microstructural evolution under fusion-relevant conditions. We are developing a simulation capability to couple plasma-induced microstructural evolution to material property degradation. Our approach relies on deliberate mapping between individual simulation models and experimental characterization for validation. The open-source Multiphysics Object-Oriented Simulation Environment (MOOSE) software was used for this simulation capability development. A MOOSE phase-field model was coupled with the cluster dynamics code, Xolotl, to predict microstructural evolution. Microstructure characterization techniques, including scanning electron microscopy (SEM), transmission electron microscopy (TEM), and laser scanning confocal microscopy (LSCM) are used to validate these microstructural evolution simulations. Calculation of thermal and electrical conductivities with first principles simulations was performed for bulk material and for grain boundaries; these results are used within MOOSE models to calculate effective thermal and electrical conductivities as a function of grain characteristics. Thermoreflectance and four-probe techniques were employed to measure the thermal and electrical conductivities, respectively. A MOOSE crystal plasticity model was adapted to predict microstructure-sensitive deformation behavior, and X-ray diffraction (XRD) was used to collect bulk dislocation density data for validation. After individual simulation validation, these models are coupled to predict material property changes resulting from plasma exposure. We focused here on an experimental design to emphasize the separate effects of moderate thermal loads and plasma exposure using tungsten. Annealing of tungsten was performed under a protective environment for temperatures ranging from 500$^o$C to 1500$^o$C. The plasma exposure was completed in the Tritium Plasma Experiment at Idaho National Laboratory under a deuterium flux of 1e22 $\frac{D}{m^2s}$. This incremental approach is employed to build confidence in the modeling capability: separate-effects tests ensure that the models capture key mechanisms from single environmental conditions before predicting PFM property degradation under combined loads. We will show our early results from coupling these simulation models to predict PFM property changes from microstructural evolution. Comparisons of the simulation results with preliminary validation data will be discussed.

Speaker: Dr Stephanie Pitts (Idaho National Laboratory)

131 O11 Tungsten heavy alloys for application as plasma facing material in fusion devices

Tungsten heavy alloys (WHA) consisting of tungsten with small amounts (few weight %) of nickel (Ni) and iron (Fe) or copper (Cu) are ductile and tough materials, which could in some cases replace brittle bulk tungsten as a plasma-facing material. Therefore, such materials are considered for moderately loaded areas in current and future fusion reactors, e.g. the stellarator W7-X, the high-field tokamak SPARC and the temporary first wall of ITER.
The high heat flux test facility GLADIS was used to study the behavior of these alloys under high heat flux loading. Mock-ups designed for SPARC made of Ni (2%) – Fe (1%) tungsten heavy alloy were quasi adiabatically loaded under various loading conditions to assess their behavior and the operational limits.
Using a continuous beam with a heat flux of 30 MW/m² the effect of extreme overload was evaluated. Since the melting points and vapor pressures of W and the alloying elements are strongly different, the purpose was to investigate the sample behavior in the temperature range, where the Ni-Fe matrix exceeds its melting point. Due to the high vapor pressure noticeable evaporation of Ni and Fe occurred. This was investigated in a time- and temperature-resolved manner by optical spectroscopy. Despite considerable material loss and strong morphological changes, the thermomechanical integrity of the component was maintained. This was confirmed by post-exposure metallographic and microscopic analysis.
Using a much lower power density of 5 MW/m² the surface temperature increases much more gradually allowing for a higher time and temperature resolution up to the melting point of the Ni-Fe matrix. In this setup, pulses with a length on the order of 15 s can be employed and detailed spectroscopic data correlated with the tile temperature measured by pyrometry and thermocouples can be extracted. At the same time witness samples catching part of the evaporated elements were installed and investigated by surface analysis techniques for a semi-quantitative evaluation of the material loss. The results from this work provide a parameterization of ejected material as a function of temperature, and are used to define operational limits for SPARC.
Finally, the surface morphology of the WHA tiles is compared to the results observed after earlier exposure of similar material in AUG and DIII-D to link the results to reactor conditions.

Speaker: Johann Riesch (MPPL)

132 O12 An overview of Lithium and PSI/PMI results from the HIDRA Team

The Hybrid Illinois Device for Research and Application (HIDRA) is located at the Center for Plasma Material Interactions (CPMI) at the University of Illinois Urbana-Champaign. It is a 5-period classical stellarator with an on-axis magnetic field B0 < 0.5 T. It is capable of true steady state operation with plasma discharge length up to, t < 10,000 s in duration with typical experimental discharges on the order of t = 600 s - 1000 s. HIDRA in conjunction with a mid-plane Material Analysis Teststand (HIDRA-MAT) allows for materials to be exposed to the plasma for great lengths of time. A liquid metal injector allows lithium to be placed on the surface and be exposed to the plasma. Results have shown that lithium operation enhances plasma performance by achieving a low-recycling regime, crucial for future reactor operation. Plasma temperatures increased from Te = 15 eV base pressures to over Te = 100 eV without significant density drop. More over this performance has been shown in helium plasmas and HIDRA has demonstrated the ability to also pump and retain helium, not just hydrogenic fuel species. Langmuir probe measurements show that the plasma profile changes with lithium operation leading to the difference in performance. During operation with hydrogen, evidence of lithium vapor shielding has also been observed where the locking temperature of the surface is similar to that seen on MAGNUM at T = 750 oC. These results have further been verified in other devices such as MAGNUM-PSI where helium experiments with a lithium CPS have been performed and on EAST with a helium plasma and flowing LiMIT plate that was placed different distances from the plasma edge. Long pulse (10,000 s) operation in HIDRA shown evidence that plasma chemistry with the wall and features inside the vessel will play an extremely important role in how future machines will need to deal with impurity formation. RGA studies have shown interesting cyclical effects on the order to 30+ minutes occurring. Thus, flowing liquid lithium systems that can provide a fresh surface will be crucial in handling these impurities and achieving low recycling. Thus, in the future there will be a need for flowing systems to be installed and operated and tested in actual toroidal and 3D magnetic field geometries. This presentation will provide an overview from the last several years with lithium experiments and technology development on HIDRA and at CPMI.

Speaker: Prof. Daniel Andruczyk (University of Illinois Urbana-Champaign)

Invited Talk: Morning session

133 I8 Heat and impurity ion fluxes on the ITER first wall at high far SOL transport in ITER burning plasmas

The new ITER baseline in which tungsten (W) replaces beryllium as the chosen first wall (FW) material reinforces the need for plasma backgrounds for the analysis of W sources and transport. Crucial in this context are the FW particle/heat fluxes, which depend on the far scrape-off layer (SOL) radial transport profile. This is still very much uncertain, but there is strong evidence from current devices that broad SOL density profiles might be expected under ITER burning plasma conditions with a detached divertor. To explore this far-SOL plasma self-consistently requires numerical simulations on numerical grids extending to the FW. This is performed here with the SOLEDGE3X plasma boundary code, with standard burning plasma assumptions: 100 MW of power injected into the inner grid boundary, using neon (Ne) impurity seeding for divertor power dissipation and including helium ash. To account for uncertainties in the anomalous radial transport, two parameters are scanned: the amplitude of the far SOL transport $D_⊥^{FarSOL}$ and the distance Δr from the separatrix at which this transport enhancement begins.

We first compare the SOLEDGE3X solutions with those obtained for the same input parameters on the same magnetic equilibrium with the newly released wide-grid version of the SOLPS-ITER code. The two codes are found to yield comparable results on key metrics. The presence of shoulders in the density profile increases the FW $n_e$ and $T_e$ by up to 100x, 2x respectively compared with the standard cross-field transport used in the main ITER divertor simulation database for cases considered. This reduces the divertor target heat flux which is then partly redistributed along the FW.

In all the simulations, highly charged Ne ions (especially $Ne^{8+}$) remain the main contributors to W sputtering at the wall, consistent with earlier main chamber W source estimates supporting the ITER re-baseline [1]. The relative contribution of lower charge states tends to increase in the presence of density shoulders, due to the combined effect of the increase in T_e, changes parallel transport, and higher recombination rates from increased far-SOL density. These effects are more influenced by the reduction in Δr from 3 to 1 cm than by increasing $D_⊥^{FarSOL}$ from 1.5 to 3 $m^2s^{-1}$. The divertor is the main gross sputtering source of W throughout the far SOL transport scan. Including prompt redeposition, the outboard FW becomes the largest source of net sputtering in the presence of density shoulders.

[1] K. Schmid et al. 41 (2024) 101789.

Speaker: Srikanth Sureshkumar (CEA)

12:20 Lunch

Invited Talk: Afternoon session

134 I9 Validation of high-fidelity physics improvement in SOLPS-ITER for detached JET-ILW L-mode hydrogenic plasmas

The inclusion of hydrogenic Lyman line radiation absorption in drift-enabled SOLPS-ITER [1] modeling of JET-ILW L-mode hydrogen and deuterium discharges [2,3] resulted in improved predictions of the divertor target particle fluxes by 30% and higher heat fluxes at the onset of detachment. Lyman line absorption changes the plasma density distribution in the divertor volume in detachment and is expected to significantly impact the prediction and interpretation of spectroscopy signals critical for detachment control in ITER and DEMO. The divertor volume of JET-ILW is found to be highly opaque to Lyman-α and Lyman-β photons, up to ∼ 90% and ∼ 79%, respectively, in deep detachment. Photo-induced ionization occurs when Lyman line absorption populates higher excitation levels of hydrogen and deuterium atoms, thereby reducing the electron energy threshold for ionization. The proportion of photo-induced ionization to the total ionization rate of the scrape-off layer (SOL) increases with opacity, thereby reducing the effective ionization cost as the divertor plasma is increasingly detached. The lowering of the ionization cost prevents early onset of radiation condensation instability and allows stable SOLPS-ITER simulations at higher plasma edge densities. The decreasing ionization costs allow for further increase of ionization rate and therefore electron density, thus shifting the particle balance from power-limited to a highly recombinating divertor.
Synthetic Balmer-α diagnostic signals are produced from the highly recombinating divertor solution using the PESDT-Cherab [4] raytracing code to include surface reflections and Lymanβ absorption, improving agreement by a factor of 3 with the JET-ILW Balmer-α measurements compared to Lyman transparent solutions. The agreement in divertor target particle flux between SOLPS-ITER and Langmuir probe measurements in JET-ILW L-mode hydrogenic plasmas is further improved by reducing the uncertainties in two areas: the SOL radial transport parameters by using the SOLPS-ITER wide-grids version [5], and a more updated data on the hydrogenic molecular breakup rates for deuterium. PESDT-Cherab is extended to include Lyman line opacities, allowing the production of consistent synthetic Lyman line diagnostic signals useful for comparison with VUV measurements in existing and future devices.
References
[1] X. Bonnin et al., Plasma and Fusion Research 11:1403102–1403102 (2016)
[2] D. Reiter et al., Plasma Physics and Controlled Fusion 44, 8:1723–37 (2002)
[3] R. Chandra et al., Nuclear Materials and Energy 41, 101794 (2024)
[4] B. Lomanowski et al., Nuclear Materials and Energy 20, 100676 (2019)
[5] W. Dekeyser et al., Nuclear Materials and Energy 27, 100999 (2021)

Speaker: Ray Chandra (Aalto University, Espoo, Finland)

135 I10 Overview of Divertor and SOL Physics Results Following the Installation of the Tungsten Divertor in KSTAR

Since the installation of the actively cooled lower tungsten (W) divertor in 2023, KSTAR has conducted four experimental campaigns (>7,200 discharges; avg. 6.5–7 MW, ~10 s) addressing critical plasma-surface interactions and optimizing divertor performance. Previously, KSTAR operated with carbon wall for 15 years (~32,800 discharges). This overview synthesizes key physics and engineering results from the campaigns with the W divertor
Initial experiments confirmed that W impurities transported into the core plasma significantly degrade performance, with 2D radiation profiles measured by Infra-Red Video Bolometer (IRVB) revealing a distinct outboard-localized radiation power loss. To rigorously investigate these phenomena, diagnostic capabilities were significantly enhanced, including the development of W-flux monitors and real-time IRVB systems. Additionally, impurity transport codes were utilized to derive W concentration profiles by analyzing 2D radiation, Vacuum Ultraviolet Spectroscopy, and Compact Advanced EUV Spectrometer (CAES) measurements.
A novel “β-kicking” NBI heating scenario was developed. By optimizing NBI power injection immediately after the H-mode transition, this scenario mitigated impurity accumulation and restored plasma performance to levels comparable to the carbon-divertor phase.
Divertor detachment control was extensively explored using neon (Ne) and nitrogen (N) injection. While N injection effectively controlled the ion saturation current measured by Langmuir Probes (LPs) at the striking point on the outer divertor target, high injection rates led to increased W concentration in the core plasma. This trade-off highlights the necessity of optimizing N gas injection rates. Advanced real-time control systems were also deployed, including a feedback algorithm utilizing real-time 2D IRVB data and a UEDGE-based surrogate model. Furthermore, a novel Machine Learning-based virtual diagnostic system, trained on Absolute Extreme Ultraviolet, LPs, and historical IRVB data, demonstrated the capability to predict 2D radiation profiles and control the radiation front without direct real-time IRVB input.
Wall conditioning strategies have evolved to support high-performance operations. Impurity Powder Dropper (IPD) boronization provided a quantitative assessment of W-source reduction at the divertor target. Notably, the IPD system is now integrated into the Plasma Control System, enabling feedback-controlled injection to optimize boronization specifically for long-pulse discharges. In preparation for the full-W wall transition after the 2026 campaign, these advanced conditioning techniques, along with improved ECWC scenarios, are being established as essential tools for ensuring successful plasma start-up and effective wall conditioning.
Finally, material challenges were identified. Localized melting of W monoblocks and structural issues, such as dislodgement of the W plate, occurred. These incidents and implications for future operations are discussed.

Speaker: Hyungho Lee (KFE)

Oral: Afternoon session

136 O13 Ion temperature measurements on the lower divertor of the WEST tokamak from attached plasmas to detached and X-point radiator regimes

Ion temperature measurements on the lower divertor of the WEST tokamak from attached plasmas to detached and X-point radiator regimes

J. P. Gunn,1 P. Ivanova,2 M. Dimitrova,2,3 J. Kovačic4

1 CEA, IRFM, F-13108 Saint-Paul-Lez-Durance, France.
2 Emil Djakov Institute of Electronics, Bulgarian Academy of Sciences, 72, Tsarigradsko Chaussee Blvd., 1784 Sofia, Bulgaria.
3 Institute of Plasma Physics of the Czech Academy of Sciences, U Slovanky 2525/1a, 182 00 Prague 8, Czech Republic.
4 University of Ljubljana, Faculty of Mechanical Engineering, Slovenia.

Ion temperature plays an important role in heat loading and sputtering processes at the surface of a tokamak divertor. A reciprocating retarding field analyzer was installed on the lower tungsten divertor of the WEST tokamak and made its first measurements of divertor ion temperature in 2025. Complementing the ion temperature measurements are arrays of flush-mounted Langmuir probes, and pop-up probes that provide local measurements of density and electron temperature, and which provide a basis for understanding the ion temperature measurements in terms of collision and ion-electron equipartition times. About 200 measurements were obtained for a range of core plasma density and lower hybrid heating power. In two dedicated experiments, strike point sweeping was employed to obtain the radial profile of ion temperature at the divertor target. When the plasma is attached, with electron temperature above 10 eV, the ion temperature is 3 to 6 times higher. Ion temperatures up to 80 eV have been observed. In detached conditions both at high core plasma density without additional heating power, and in the X-point radiator regime triggered by nitrogen seeding with additional lower hybrid heating, the ion temperature collapses to become about equal to the electron temperature, around 5 eV or lower on average (see Figure). SOLEDGE3X mean-field simulations show that observed Ti/Te ratios are challenging to recover and raise questions concerning the standard model and parametrization used in edge plasma codes. Simultaneous ion temperature measurements upstream as well as on the divertor are needed to address this issue. At the time of writing of this abstract, a reciprocating retarding field analyzer is being built that will provide simultaneous measurements of the ion temperature in the scrape-off layer at the top of the machine. We expect to have these fresh measurements before the conference.

Speaker: James Gunn (CEA)

137 O14 SOLPS-ITER modelling of JET-ITER baseline discharges

Future fusion reactors require integrated operating scenarios that simultaneously address divertor power exhaust, tungsten impurity control, and good core confinement. In recent JET campaigns, dedicated “JET-ITER baseline” experiments with neon impurity seeding were conducted in D and D-T plasmas [1], developed at high input power (30-35 MW), high plasma current (2.5-3.2 MA), high density ($f_{GW}$ ≈ 0.7-0.8), and within an ITER-like configuration characterized by high triangularity and vertical divertor targets. In these discharges, simultaneous achievement of partial divertor detachment, small or no ELMs, low tungsten concentration ($C_W$ < 7×10⁻⁵), and reasonable confinement ($H_{98}$ ≈ 0.85-1.0) was demonstrated, highlighting attractive features for future device operation. Neon impurity seeding reduced divertor heat flux and tungsten sources while enhancing core confinement through improved pedestal pressure relative to unseeded plasmas.
Comprehensive modeling was carried out to interpret these experiments and to prepare for extrapolation to future devices. SOLPS-ITER simulations with drifts, using input power profiles from TRANSP, successfully reproduced key experimental measurements, including mid-plane electron and neon density and temperature profiles, divertor target conditions, and line-of-sight–resolved radiated power using exactly experimental gas fueling and seeding rates. The effects of drifts, D-Ne charge exchange, Lyman-opacity, neutral–neutral collisions, and sub-divertor structures, as well as transport assumptions—such as poloidal inhomogeneity associated with ballooning and divertor broadening, and differences between main ions and impurity species—are studied in detail to assess how they influence and contribute to reproducing experimental measurements. Comparisons between seeded and unseeded simulations, without altering transport assumptions, quantitatively reproduced the measured separatrix density drop, attributed to reduced ionization sources, while the underestimation of the pedestal density drop implies additional transport changes in that region. JINTRAC simulations [2] showed consistency with SOLPS-ITER in the separatrix region (e.g. neutral particle fluxes) and confirmed that the density reduction arises from decreased ionization sources or increased pedestal diffusivity.
Reference:
[1] C. Giroud et al., 30th IAEA FEC, Chengdu, China, Oct. 13-18, 2025.
[2] V.K. Zotta, et al., 51st EPS, Vilnius, Lithuania, Jul. 7-11, 2025.

Speaker: Ou Pan (Max Planck Institute for Plasma Physics)

Invited Talk: Afternoon session

138 I11 Comparative study of divertor detachment and tungsten screening in deuterium and helium plasmas on EAST

Dedicated EAST experiments and modeling have been performed to investigate the influence of deuterium (D) and helium (He) plasmas on divertor detachment and tungsten (W) impurity transport. As fusion reactor will be operated in the D-T plasma with He naturally existing as the product of D-T reaction, the influence of different ion species on divertor detachment remains unclear. Density ramp-up discharges on EAST reveal that the divertor detaches at a relatively higher line-averaged electron density in He plasma than D plasma. With the same line-averaged electron density, He plasma exhibits higher electron density in the core and lower electron density in the scrape-off layer (SOL) than D plasma, which indicates different characteristics for achieving detachment. Meanwhile, divertor erosion and W impurity transport have been compared between D and He discharges. Under high-recycling divertor conditions, He induces W erosion rates are nearly six times higher than by D because of the higher sputtering yields, whereas the core W concentration in He plasma is only twice higher, indicating a better divertor W screening. Under detached divertor conditions, W erosion rate (dominated by impurities) is comparable in He and D plasmas. However, the resulting core W concentration in He plasma is merely half of that observed in D plasma.
To interpret the experimental observations, SOLPS and DIVIMP simulations [1] have been performed. Simulation results demonstrate that due to the higher ionization potential of He and He+ than D, He can penetrate deeper into the core plasma, resulting in a higher density in the core and a lower density in the SOL. E×B drifts show greater impact on plasma distribution of He than D, and therefore the relative difference in outer divertor detachment onset between He and D plasmas is more significant with favorable Bt direction compared to the unfavorable Bt direction.By leveraging the newly developed kinetic impurity transport model in DIVIMP, W transport and screening have been analyzed. The results show that W stagnation point in He plasma is located farther from the divertor targets than in D plasma, which is beneficial for a better divertor W screening. Moreover, the kinetic thermal force on W (pointing upstream) is reduced under conditions of lower ion temperature and higher effective charge (Zeff). This reduction enhances tungsten screening under detached divertor conditions, particularly in He plasmas, where the relatively higher Zeff further diminishes the thermal force.
[1] Q. Li, et al., Nuclear Fusion. 2025, 65, 026069

Speaker: Rui Ding (Institute of Plasma Physics, Chinese Academy of Sciences)

139 I12 Dynamics of the X-point radiator formation in the WEST tokamak

Recent experiments on the WEST tokamak show that, with sufficient nitrogen seeding, a stable X-point radiator (XPR) can be sustained above the X-point for about 70s. This regime helps to reduce divertor heat loads by about 90% and tungsten divertor sources by up to 98% while enhancing core energy confinement by about 25%, highlighting the XPR’s potential for future fusion power plants.
For the first time, time-dependent simulations in SOLEDGE3X-EIRENE, including drifts, have reproduced experimentally observed millisecond-scale sequence of rapid divertor transitions from attached to XPR, with microsecond resolution, and achieved a stable XPR after the transition. This represents a major step beyond previous steady-state or drift-free models and demonstrates that self-consistent boundary plasma modeling can be a powerful tool for predictive control of XPR regimes in current and future fusion devices.
The simulations reveal that just before the onset of the XPR regime, a coherent X-point-centered vortex structure forms, connecting the common SOL, PFR, and closed-flux regions. This vortex is triggered by a transient electric potential well arising from steep temperature gradients in the LFS common SOL and is shaped by anomalous diffusion and drifts. It drives cross-field transport that rapidly increases the nitrogen concentration above the X-point, mixes cold and hot particles, steepens temperature gradients, and deepens the potential well (to –31 V), sustaining a self-amplifying cycle for several milliseconds until saturation and gradual decay.
The vortex plays a critical role in accessing a stable XPR: by advecting cold particles and providing ~91% of the total nitrogen source (~$1.1\times 10^{20}\rm{s}^{-1}$) into XPR region before XPR onset, it raises the neutral density and impurity concentration without causing edge over-cooling, thereby preventing disruptive MARFEs and bridging the hot and cold solutions. The vortex also balances steep impurity gradients (2–3% in the HFS common SOL/PFR vs. 0.2–0.5% in the LFS common SOL/closed-flux region), allowing the closed-flux nitrogen concentration to exceed the ~0.66% threshold required for XPR sustainment.
This vortex mechanism explains the clear hysteresis observed during XPR access. By effectively “locking in” a high impurity concentration, the resulting plasma — which subsequently evolves into a deeper, colder, and denser MARFE — requires the nitrogen seeding rate to be reduced by approximately 50% below the access rate before the plasma reverts to the attached state. Simulated radiation maps show good qualitative agreement with experimental fast-camera 2D reconstructions, indicating vortex-driven impurity transport as the underlying physical mechanism enabling stable XPR access.

Speaker: Dr Hao YANG (M2P2 Aix-Marseille Univ, CNRS, Centrale Méditerranée, 13013 Marseille, France)

16:00 Coffee Break

Postersession 2: Tracks A, B, D, F, H, I and J

140 2.058 Influence of boron-containing films on optical and emissive properties of tungsten

Accurate wall temperature monitoring is essential for the safe operation of magnetic confinement fusion devices like ITER, and is routinely performed using infrared (IR) thermography. However, the fully metallic nature of the ITER First Wall (FW) introduces strong reflective effects, where hot optics and thermal scene parasitic light and variable emissivity of surfaces can lead to significant errors in the temperature estimation [1]. Consequently, reliable modelling and interpretation of IR measurements requires detailed knowledge of the optical and emissive response of surfaces exposed to representative heat-flux conditions, obtained through systematic experimental characterization [2]. Moreover, previous studies have also demonstrated that the tungsten surface’s optical response, quantified through Bidirectional Reflectance Distribution Function (BRDF) and reflectivity measurements, are correlated with its topography [3].
In ITER, the FW will undergo Glow Discharge Boronization (GDB) using diborane (B2H6 or B2D6) gas mixed with He as a carrier gas. This process is intended to getter impurities and stabilize plasma operation by forming boron-based films on plasma-facing components [4]. However, boron film will alter the optical and emissive properties of the FW, affecting the accuracy of the IR-based temperature diagnostics.
This study aims to investigate the influence of boron film composition and thickness on the optical, emissive and morphological properties of ITER-relevant FW surfaces. Boron-containing coatings with a thickness of several tens to hundreds of nanometres with various compositions are deposited via magnetron sputtering. The resulting surfaces are characterized using a comprehensive set of characterization techniques, including emissivity measurements, scanning electron microscopy (SEM), confocal microscopy, X-ray photoelectron spectroscopy (XPS), BRDF measurement, and spectrophotometry.
By establishing quantitative correlations between emissivity, optical properties, and surface morphology, this work aims to provide a critical input for the calibration and accuracy of improvement of ITER’s IR thermographic diagnostic systems.
1. Le Bohec, M., Steiner, R., Natsume, H., Kajita, S., Yaala, M. B., Marot, L., Aumeunier, M. H. (2024). Relationship between topography and BRDF for tungsten surfaces in the visible-spectrum. Optik, 303, 171750.
2. Aumeunier, M. H., Gerardin, J., Talatizi, C., Le Bohec, M., Yaala, M. B., Marot, L., ... ASDEX-U team. (2021). IR-thermography in metallic-environments of WEST and ASDEX-U. Nuclear Materials and Energy, 26, 100879.
3. Retailleau, F., Aumenier, M. H., Marot, L. (2025). BRDF determination based on topography measurement and Monte Carlo ray-tracing. Optik, 172476.
4. Wauters, T., Hagelaar, G. J. M., Pitts, R. A. (2025). Modeling input to the ITER GDB system. Nuclear Materials and Energy, 42, 101891.

Speaker: Sanah Hussain (GNOI)

141 2.059 Research and Development of Boronization in HL-3

Boronization has been proven to be an effective wall-conditioning technique in tokamak devices, significantly reducing impurities such as oxygen and carbon, and enhancing plasma confinement by minimizing radiative losses. Following ITER’s new baseline, the application of boron coatings on plasma-facing components, especially tungsten-based first walls, has been prioritized to mitigate tungsten erosion and its subsequent contamination of the plasma. Experimental campaigns on devices such as EAST, DIII-D, and ASDEX Upgrade have confirmed that boron films suppress impurity influx by acting as a gettering layer. However, critical challenges remain in optimizing boronization processes, including real-time boron powder injection techniques and initial film deposition via glow discharge using diborane. Key issues such as film longevity, fuel retention, and the formation of tungsten-boron co-deposited layers require further investigation to address non-uniformity, erosion, and other concerns.
To address the challenges of boron film uniformity and fuel retention in large-scale tokamaks, a dedicated boronization test platform and boronization system for HL-3 have been developed. This platform enables systematic validation of boronization processes through independent control of critical parameters, such as glow discharge current, gas pressure, and discharge duration, allowing iterative optimization of coating efficiency. Building on operational experience from existing devices like HL-2A, multi-point gas injection and distributed pumping configurations have been integrated into the HL-3 boronization system to mitigate non-uniform boron deposition on plasma-facing components. Furthermore, the platform supports the synthesis of diverse film types, including pure boron and tungsten-boron co-deposited layers, for specialized test samples on HL-3. In-situ diagnostics, such as surface analysis, are employed to evaluate film longevity and fuel retention dynamics under deuterium or helium discharge conditions. These efforts are essential to meet ITER’s engineering targets and accumulate foundational experience for future fusion reactors.

Speaker: Prof. Chengzhi Cao (Southwestern Institute of Physics)

142 2.105 Characterisation of steady state plasma at the linear plasma device – JULE PSI

Linear plasma devices can generate steady state long duration plasma and are ideal for lifetime plasma exposure studies on plasma-facing materials (PFM) in a fusion reactor. The plasma exposures are followed by in-situ and ex-situ measurements on surface erosion and fuel retention respectively. Additional possibilities such as sample biasing and sample temperature variation enable a wide scenario of experimental conditions on PFM. JULE-PSI is a hot cathode-based linear plasma device, specifically designed to operate in a radiation compatible hot cell environment. Its main objective is the study of plasma-wall interactions on irradiated PFM with in-situ measurements of fuel retention.
The integration and operation of a linear plasma device in a radiation environment is challenging and requires extensive modification of the machine as compared presently operated linear plasma devices. Hence, presently the machine alongside modifications is being characterised on a test-stand outside the radiation area for machine stability and plasma performance. The plasma is characterized through a single tip Langmuir probe and a segmented dump with current measurement capabilities. A series of experiments using argon, neon, helium and hydrogen gas have been run under steady state conditions with the aim to achieve stable plasma with a top hat profile, electron density of ~1017-1019 m-3, an electron temperature of 3-10 eV and an ion flux to the target of ~1021-1023 m-2s-1. Optical spectroscopy using an overview spectrometer (OES, wavelength range 300-890 nm) was used in conjunction to the Langmuir probe for verification of the plasma profile.
In this contribution, results from different magnetic configurations measured using the Langmuir probe and the segmented dump will be presented. A stable plasma configuration is seen in the cross-field discharge configuration as compared to a direct arc magnetic configuration. The optical spectroscopy results show a good agreement with the Langmuir probe results and will be detailed alongside hyperspectral imaging results. Additionally, the results from JULE-PSI test-stand will be compared to the existing linear plasma device PSI-2, which has a cylindrical hollow LaB6 cathode. The differences in the Langmuir probe results, highlight the influence of cathode geometry in plasma profile and transport. Lastly, an outlook to the project alongside integration into the hot cell environment is discussed.

Speaker: Rahul Rayaprolu (Forschungszentrum Jülich)

143 2.107 Improved modelling and sampling of particles reflected from non-planar surfaces

The interaction of neutral and charged particles with the plasma-facing wall sets the boundary conditions for the confined plasma and is thus essential for important aspects like L-H power transition thresholds or impurity radiation. For example the transition from carbon to tungsten as wall material in ASDEX Upgrade reduced the L-H power threshold by about 25% which has been traced to the altered reflection properties of the surface [1]. Thus a proper description of the sputtering and reflection properties of plasma-facing walls is crucial. However, in most of the modelling of plasma-wall interaction in codes like EIRENE data tables computed for the case of atomistically flat surfaces are being used to generate the Monte Carlo (MC) samples – although in practice the plasma exposed surfaces exhibit a non-smooth morphology and roughness. While the simplification of a flat surface has been shown to be acceptable in many circumstances for the qualitative or even semi-quantitative assessment of sputtering yields (see e.g. [3]) the situation is different for the reflection properties. While for atomistically smooth samples and non-perpendicular particle impact forward reflection is always dominating sample roughness can cause pronounced backward reflection together with a sign-reversal of the integrated tangential momentum [4] – thus drastically altering some assumptions presently used in codes like SOLEDGE3X-EIRENE or SOLPS-ITER. Unfortunately due to their size a straightforward inclusion of the species, energy and angle dependent reflection distributions in EIRENE is challenging and a more efficient approach is desirable. In the contribution we present a practical method to generate simultaneously an arbitrary number of Monte Carlo samples following the reflection distribution. The key idea is the use of an adaptive recursive quadtree to assign samples according to the underlying probability density and this can be made efficient by exploiting recurrence relations of hemispherical harmonics. The performance is being demonstrated by applying the code to hydrogen, helium and neon-reflection distributions computed by SDTrimSP-7 for amorphous and crystalline ((100) and (111)) tungsten samples. More than one million MC-samples per second can now routinely be obtained – thus paving the way to include realistic reflection distributions in PWI-codes like EIRENE.

[1] L. M. Shao et al, Plasma Physics and Controlled Fusion 58 p. 025004, 2016
[2] R. Arredondo et al, Nuclear Materials and Energy 18 p. 72, 2019
[3] U. von Toussaint, R. Preuss, Nucl. Mat. Energy 41, p. 101817, 2024

Speaker: Dr Udo von Toussaint (Max-Planck-Institute for Plasmaphysics)

144 2.115 Modelling the Roles of Dislocations and Coalescence on the Nucleation and Growth of Helium Bubbles in Tungsten

The trapping of helium in crystal defects in tungsten is one of the mechanisms of bubble nucleation. Due to its low solubility and high mobility, helium tends to accumulate by self-trapping or trapping in defect sites such as dislocations, grain boundaries, and vacancies, leading to bubble formation. When these bubbles come close enough, they can coalesce and form large bubbles. Large bubbles near the surface can burst due to their internal pressure and cause changes in the surface of the material [1]. These phenomena have been extensively studied at the atomic scale by molecular dynamics, both for bubble formation along dislocations [2] and for bubble coalescence [3].

In this work, we propose a model of cluster dynamics in a one-dimensional continuous medium describing the formation of helium bubbles by trap mutation as well as by nucleation at dislocations and vacancies. This model, extended from [4], is implemented in FEniCSx [5] and it includes bubble depletion model depending on the bubble distance from the surface and bubble coalescence model. The analysis revealed the influence of helium trapping and bubble nucleation at dislocations and preexisting vacancies on bubble concentration and size. Several hypotheses are considered regarding the evolution of dislocation densities, such as their annihilation during bubble nucleation and their creation during growth. The results obtained were compared with experimental observations (radius size, porosity and growth kinetic) from literature [1,6,7].

The project leading to this publication has received funding from the ANR under Grant No ANR-23-CE08-0026-04 (HEBUTERNE). This work has also been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No 101052200—EUROfusion).
[1] Ialovega et al. Nuclear Fusion, 62(2022)126022
[2] Wang et al. Comput. Mater. Sci. 230 (2023) 112457
[3] Zhan et al. Comput. Mater. Sci. 187 (2021) 110076
[4] Delaporte-Mathurin et al. Sci. Rep. 11 (2021) 14681
[5] I. A. Baratta et al. preprint (2023) see https://fenicsproject.org
[6] Corso et al. Nucl. Mater. Energy (2025)101894
[7] Pappalardo et al. submitted to J. Phys. D: Appl. Phys. (2025)

Speaker: Ms Emna frikha (Sorbonne Paris Nord University, Laboratory of Process and Materials Sciences, LSPM, CNRS, UPR 3407, F-93430, Villetaneuse, France)

145 2.106 Centralized database on integrated testing of advanced materials in the DIII-D tokamak

A repository storing key performance data collected from integrated material testing on the DIII-D tokamak is being actively developed in collaboration with the Clean Air Task Force Material Database for Fusion (MatDB4Fusion) initiative to aid in material down-selection for commercial fusion reactor deployment. Advancing the technological readiness level of plasma-facing materials (PFMs) requires combined effect testing to assess synergies that may impact survivability. The Divertor Material Evaluation System (DiMES) [1,2] is a material testing platform installed in the lower divertor of the DIII-D tokamak with world-leading diagnostic coverage. Coupon-sized (⌀ 0.6-2.5 cm) samples of varying compositions and geometries are loaded into DiMES, exposed to reactor-relevant heat and particle fluxes, and characterized using a comprehensive suite of spectroscopy and plasma diagnostics. Three decades investigating different materials across a wide range of plasma scenarios has produced one of the largest sources of combined effect data that is now being synthesized and made accessible to the PMI community via MatDB4Fusion.

The DiMES material database provides key performance metrics for several reactor-relevant PFMs exposed to combined effect loading across a range of plasma conditions. Gross and net erosion rates are based on in situ emission spectroscopy and post-mortem Rutherford Backscattering Spectroscopy measurements. Thermal handling is quantified via embedded thermocouples and infrared (IR) camera temperature measurements and is linked qualitatively to changes in surface morphology (e.g., recrystallization, melting) quantified ex situ before and after exposure via microscopy. Fuel retention is quantified by post-mortem nuclear reaction analysis and thermal desorption spectroscopy. Edge-localized mode (ELM)-resolved measurements of local plasma conditions (e.g., electron temperature, particle flux, and heat flux) using Langmuir probes, Thomson scattering, spectroscopy, and IR imaging link the material response to loading conditions, offering insight on the underlying physics processes at the plasma-material interface.

Constructing a comprehensive material property database will reduce risk and stimulate growth in fusion reactor design. Characterization of changes during combined effect loading provides actionable insight on performance once in operation. Valuable data from integrated plasma exposure testing performed on the DiMES testing platform has now been compiled and coupled to MatDB4Fusion to support the growing commercial fusion industry. Curating high-quality datasets with a standardized protocol will enable machine learning tools to detect performance trends, predict material behavior, and accelerate design cycles for next-generation PFMs.

[1] Wong C.P.C. et al 1998 J. Nucl. Mater. 258–263 433-439
[2] Rudakov D.L. et al 2017 Fusion Eng. Des. 124 196-201

Speaker: Gregory Sinclair (General Atomics)

146 2.064 Measurements and OEDGE predictions of the scrape-off layer plasma conditions and Balmer emission in JET-ILW horizontal and vertical divertor plasma configurations

In JET ITER-like-wall low-confinement mode plasmas the radial scrape-off layer (SOL) measured profiles of electron temperature and density at the low-field side (LFS) midplane and the ion fluxes to the LFS target plates were the same, within the uncertainties of the measurements, for the same volume-averaged core density profiles, independently of the LFS strikepoint connected to a horizontally or vertically inclined target plate. The similarity of the profiles is in contrast with simulations performed using the coupled edge fluid, kinetic neutral Monte-Carlo code EDGE2D-EIRENE [1,2], predicting electron temperatures at the LFS strikepoint ($T_{e,LFS-sp}$) significantly lower in the vertical than in the horizontal divertor plasma configuration for the same high-density conditions at the LFS midplane. Understanding the impact of the divertor geometry on the neutral particle confinement and the reduction of power loads to the plasma facing components are critical for validating divertor design tools for next-generation fusion-power plants.

For low-recycling conditions, i.e., $T_{e,LFS-sp}$ above 20 eV, reconstructions of the LFS plasma using the OEDGE code constrained by the measured ion saturation current, $j_{sat}$, and $T_e$ profiles at the LFS target predict the radial profiles of $T_e$ and electron density ($n_e$) in the SOL at the LFS midplane within 10% of the measured profiles for both divertor plasma configurations. For high-recycling conditions, $T_{e,LFS-sp}$ of 5-10 eV, OEDGE predicts $T_e$ at the LFS midplane for both divertor plasma configurations, but overpredicts $n_e$ at the LFS midplane by factor of 2. For the vertical divertor configuration, lowering the assumed $j_{sat}$ by 30% improves the agreement of the simulated $n_e$ with the LFS midplane measurements. Furthermore, the agreement between the simulated and the measured line-integrated Balmer-𝛼 emission across the LFS divertor plasma is improved with lowering the assumed $j_{sat}$.

Scans of the OEDGE input and model parameters for both divertor plasma configurations are carried out to characterize their impact on the 2D poloidal $T_e$ and $n_e$ profiles in the divertor and to assess their validity against 2D poloidal $T_e$ and $n_e$ profiles inferred from tangentially viewing cameras [3]. The role of recombination and molecular processes on the predicted Balmer-𝛼 and Balmer-𝛾 emission is assessed.

[1] Groth et al., J. Nucl. Mater., 463 (2015), 471.
[2] Moulton et al., Nucl. Fusion 58 (2018) 096029.
[3] Karhunen et al., Nucl Mater. Energy., 25 (2020) 100831.

Speaker: Henri Kuivasniemi (GNOI)

147 2.108 Development of combined joint method for W/RAFM steel with copper intermediate layer

Combined joint method of brazing and diffusion bonding, for between tungsten (W) and Reduced Activated Ferritic/Martensitic (RAFM) steel (W/RAFM steel) with a pure-Cu intermediate layer, was developed. Flat plate type joint sample of W/pure-Cu/RAFM steel with the joint area of 20 × 20 mm² was selected in this experiment, in which W and the 1 mm thick pure-Cu was jointed by brazing with Ni-11%P filler material under the 960°C, 10 min, as a primary process. Then, pure-Cu and RAFM steel was jointed by simple diffusion bonding with 40 MPa uniaxial compressive load under the 720°C, 1 hour, as a secondary process. All joint procedures were conducted in the vacuum furnace.
Thickness of W, pure-Cu, RAFM steel and Ni-11%P filler were 3, 1, 5 mm and 38 micrometer, respectively. The most challenging point in this study was whether sufficient joint quality was able to obtain between Pure-Cu and RAFM steel joint under the relatively low bonding temperature of 720°C, that is below the tempering temperature (750°C) for RAFM steel and is the specified temperature for post-welded heat treatment (PWHT). Therefore, obtaining a suitable joint quality at this temperature would lead to advantage from the viewpoint of utilizing the RAFM steel to design the DEMO reactors. To confirm a joint quality for W/pure-Cu/RAFM steel sample, inspections for the joints were performed by using digital microscope, scanning electron microscope (SEM) and transmission electron microscope (TEM). To understand the thermal transfer properties of the joints, heat loading experiments were also performed by using an electron-beam device (ACT2) with actively cooled condition of the backside of the RAFM steel.
Although some micro-level voids were observed in the W/pure-Cu joint interface, they did not show any negative effects for thermal transfer properties during electron-beam heating. The pure-Cu/RAFM steel interface showed almost no joint defects such as voids and cavities in nanometer level, and inter diffusion of Fe and Cu elements were confirmed with the range of 1.0~2.0 micrometer. Thermal transfer property of W/pure-Cu/RAFM steel sample did not change from the initial condition up to the heat loading value of ~2.4 MW/m² with cyclic heat loading. On the other hand, thermal transfer properties degraded in the range of 2.4~2.8 MW/m² of repeated heat load due to the crack formation in the W bulk.
The combined joint method demonstrated one of the superior advantages to manufacturing the diverter heat removal component below PWHT temperature of the RAFM steel.

Speaker: Masayuki Tokitani (National Institute for Fusion Science)

148 2.016 Fatigue and crack growth life assessment of hypervapotron divertor under thermo-electromagnetic coupling condition

The divertor is a key component of the nuclear fusion reactor tokamak that being responsible for power exhaust and impurity removal via guided plasma. During the operation of tokamak, the divertor is subjected to multi-physical field loads such as thermal and electromagnetic, meanwhile, the micro-cracks so-called “self-castellation” in the tungsten have been observed during experiments of Experimenta Advanced Superconducting Tokamak (EAST) and International Thermonuclear Experimental Reactor (ITER), so it is essential to carry out fatigue and crack growth life for the divertor target. The hypervapotron divertor concept was developed on ITER, which offers advantages such as simple manufacturing process and high heat removal efficiency. This paper aims to optimize the hypervapotron divertor concept and carry out fatigue and crack growth life assessment to make it suitable for the China Fusion Engineering Test Reactor (CFETR) divertor. Firstly, the Euler-Euler model coupled with RPI boiling model is used to carry out the thermo-mechanical optimization analysis of hypervapotron target under 10 MW/m2 steady-state thermal load. Secondly, the thermal-electromagnetic coupled analysis of the hypervapotron divertor target under VDE (vertical displacement event) and MD (major disruption) conditions was conducted through elastic and elastic-plastic analysis of RCC-MR code. Furthermore, a crack growth life model accounting for cooling water velocity, initial crack length, and heat flux density was established based on linear elastic fracture mechanics. Finally, a fatigue and crack growth life prediction model was constructed using multiple linear regression under thermo-electromagnetic coupling condition, that was verified theoretically based on the parameter characteristics of the model, a fatigue and crack growth life experimental scheme was also designed to modify the divertor life prediction model. This study demonstrates that the hypervapotron divertor target with downstream triangular fin channel can significantly reduce the temperature of the CuCrZr tube, thereby lowering the risk of catastrophic failure, which can meet the CFEDR design requirements withstand the highest surface heat fluxes, up to 10 MW/m2 during steady state operation and 20 MW/m2 during slow transient operation. Meanwhile, the systematic analysis method and the fatigue and crack growth life prediction model of hypervapotron divertor provided in this paper have enhanced the reliability of assessments and production efficiency in order to suggest timely divertor design modifications of the tokamak.

Speaker: Shijun QIN (Institute of Plasma Physics, Chinese Academy of Sciences (ASIPP))

149 2.017 EMC3-Eirene simulations of first wall heat and particle loads in ITER Start of Research Operation scenarios

EMC3-EIRENE scrape-off layer (SOL) simulations, with a domain extended to the first wall, are completed for the two scenarios which constitute the main targets of the Start of Research Operations (SRO) campaign in the ITER 2024 re-baseline: $B_T=5.3$ T, $I_P=15$ MA hydrogen L-modes and $B_T=2.65$ T, $I_P=7.5$ MA deuterium H-modes. These simulations also form part of a coordinated activity with the ITER Scientist Fellow Network boundary plasma modelling activity to cross-validate the EMC3-EIRENE, SOLPS-ITER and SOLEDGE3X codes on the SRO mission objective scenarios. The 3D EMC3-EIRENE boundary plasma calculations, including heat and particle loading onto the shaped wall geometry of the inertially cooled tungsten (W) Temporary First Wall (TFW) [1], will support studies of W sourcing, and erosion-migration of boron (B) layers deposited during boronization wall conditioning. An initial comparison of the predicted plasma, divertor, and wall loading profiles between EMC3-EIRENE and SOLPS-ITER are in reasonable agreement.

Additionally, EMC3-EIRENE modeling of non-axisymmetric SOL plasmas caused by the application of Resonant Magnetic Perturbations (RMP) for Edge Localized Mode (ELM) suppression is performed for the $B_T=2.65$ T deuterium H-mode SRO scenario. The perturbed magnetic field geometry is calculated by the MARS-F plasma response code for a scan of plasma current from $I_P=6.5→8$ MA ($q_{95}=3.4→2.7$). RMP coil phasing is optimized to maximize the displacement of the X-point $\xi_X$, a metric correlated with ELM suppression. For fixed $30$ kAt RMP current the magnetic footprint width on the outer divertor target varies between $17$ and $38$ cm when RMP current is scaled to maintain constant $\xi_X$=2 mm, the footprint width varies between $1.5$ and $25$ cm. Initial modeling results show that TFW particle load and temperature are modified by these RMP scenarios, particularly near the secondary null at the top of the vessel, and are expected to have a non-negligible impact on W sources and B layer lifetimes.

[1] R. A. Pitts et al., “Physics basis and status of the ITER tungsten First Wall”, Int. Conf. on Plasma Surface Interaction, 2026.

Speaker: Jonathan Van Blarcum (ITER Organization (IO))

150 2.018 Modelling of limiter load and its dependence on the wall clearance in Quasi-Continuous Exhaust regime in the ASDEX Upgrade tokamak

In future fusion reactors, the transient heat loads associated with large edge-localised modes (ELMs) are not tolerable, motivating strong interest in regimes with small or suppressed ELM activity. Among these, the Quasi-Continuous Exhaust (QCE) regime has emerged as particularly promising. It is characterised by the absence of large ELMs with a small confinement degradation. The QCE regime is accompanied by a broadening of the scrape-off layer (SOL), which can reduce heat fluxes to the divertor but may increase particle and heat loads on poloidal limiters [1,2].
Most research has focused on the compatibility of QCE with divertor detachment and divertor integrity, while its impact on other plasma-facing components (PFCs), such as limiters, has received less attention. Although several studies have addressed fluxes in the far SOL [5,6], a quantitative estimation of peak limiter heat fluxes in the QCE regime is still lacking. This work aims to address this gap.

Three ASDEX Upgrade (AUG) discharges, ranging from an ELMy H-mode to a QCE discharge achieved through increased gas puffing, have been simulated using the three-dimensional code EMC3-EIRENE [4]. The simulations are validated against experimental data, yielding two-dimensional maps of heat and particle fluxes on limiter surfaces. A peak parallel heat flux of approximately 5.5 MW m⁻² is found on one limiter, corresponding to a peak perpendicular flux of about 1.8 MW m⁻² when accounting for the limiter geometry. This value corresponds to roughly 40% of the peak perpendicular heat load on the divertor target, indicating that first-wall heat loads in the QCE regime are significant.
An estimate of the ITER first-wall heat load has also been performed using a simple analytic model, suggesting that the limiter–separatrix distance may influence the heat and particle fluxes impacting ITER PFCs. To experimentally test this model, scans of the separatrix–wall clearance have been carried out in QCE discharges at AUG, and corresponding modelling activities are ongoing. The goal is to assess the influence of the wall gap on limiter loads and to investigate the role of recycled particles in the formation of the density shoulder and the toroidal asymmetries observed in recent AUG experiments [3].

[1] M.Faitsch et al., Nucl. Mater. Energy (2021)
[2] A.Redl et al., Nucl. Fusion (2024)
[3] B.Tal et al., Nucl. Fusion (2024)
[4] Y.Feng et al., J. Nucl. Mater. (1999)
[5] T.Lunt et al., Plasma Phys. Control. Fusion (2020)
[6] T.Lunt et al., J. Nucl. Mater. (2015)

Speaker: Luca Scotti (University of Milano-Bicocca)

151 2.019 Quantifying nanobubbles in helium-plasma irradiated tungsten: Density and pressure estimation using TEM-EELS measurements

In magnetic confinement nuclear fusion reactors, the tungsten (W) divertor (the main plasma-facing component) is exposed to extreme fluxes of helium (He) and hydrogen isotopes as well as high thermal loads. Despite tungsten’s strong thermo-mechanical properties (e.g. high melting point and high erosion resistance), the low solubility of He in W leads to the formation of near-surface He nano-bubbles [1]. These bubbles severely damage the microstructure, particularly when they burst [2], potentially creating what it is referred to as “fuzz” [3] and consequently affecting the properties and lifetime of the material [4]. The aim of this work is to study the formation and growth of He bubbles on W samples, through measurements of the size and spatial distribution of the bubbles formed near the surface, as well the estimation of the density of He trapped in each bubble, thus allowing us to calculate the pressure inside these bubbles.
In this study, two sets of samples were analyzed: Post-mortem W monoblocks extracted from the WEST divertor after the C4 He plasma campaign performed in 2019 [5], as well as W polycrystalline samples that were irradiated in the laboratory using a He plasma. The laboratory irradiations were performed at a constant fluence of 4.5.1023 m-², flux around 1.1019 m-²s-1, incident ions energies of 79 eV and at temperatures of 823 K or 973 K.
Bright-Field Transmission Electron Microscopy images proved the existence of sub-nanometer sized bubbles in the first 20 nm bellow the surface in the WEST samples, as well as showing an increase in overall bubble size as a function of the temperature in the laboratory samples. On the other hand, the He K-edge was acquired using STEM-EELS for various bubbles, resulting in the estimation of the He density in bubbles, using the framework developed by Walsh et al. [6]. The evolution of the He density as a function of the bubble size and the He K-edge energy shift, as well as the He pressure calculations will then be discussed and compared to previously published work.

[1] H. Iwakiri et al. JNM 283 – 287 (2000), 1134 – 1138
[2] M. Alfazzaa et al. NME 42 (2025), 101883
[3] K. Saito et al. NME 42 (2025), 101859
[4] S. Kajita et al. Nuclear Fusion 49 (2009), 095005
[5] E. Tsitrone et al. Nuclear Fusion 62 (2022), 076028
[6] C. A. Walsh et al. Philosophical Magazine A 80 (2000), 1507-1543

Speaker: Dr Ayoub BENMOUMEN (Physics of the Interactions of Ions and Molecules (PIIM))

152 2.020 EMC3-Eirene wide-grid modelling of ITER $Q_{DT}$ = 10 scenarios with realistic tungsten first wall geometry

Three-dimensional simulations providing boundary plasma backgrounds and taking into account the realistic wall geometry are a key input for the study of plasma-wall interaction issues raised by the revised ITER Research Plan which accompanies the new 2024 Baseline [1]. Particularly important areas are erosion and material migration of tungsten (W) and boron, first wall heat loads, edge fueling strategies, ICRH coupling, etc. Under the auspices of the ITER Scientist Fellow Network, a coordinated effort is underway to provide a database of such backgrounds and to validate plasma boundary codes against each other. Here we report on new EMC3-Eirene code simulations of the $Q_{DT}$ = 10 ITER scenario with a 3D numerical wide-grid covering the full realistic W first wall geometry in one 20$^\circ$-sector.

The simulations are performed for the standard ITER burning plasma magnetic equilibrium ($I_p$ = 15 MA and $B_t$ = 5.3 T) with scrape-off layer (SOL) power, $P_{SOL}$ = 100 MW and follow specific input parameter guidelines set for the inter-code validation against the 2D codes SOLPS-ITER, SOLEDGE3X operating with wide-grids on the same equilibrium [2,3]. This specification includes the target $Q_{DT}$ = 10 H-mode pedestal width and pedestal top values of Te, ne and fixed upstream separatrix density of $n_{e,sep}=4\times10^{19}~\mathrm{m^{-3}}$. Importantly, prescriptions are also given for profiles of SOL cross-field transport coefficients which lead to broad density shoulders designed to investigate worst case wall interactions. Neon (Ne) seeding is used for divertor heat load control with two target values of separatrix averaged Ne concentration, $c_{Ne}=\left\langle n_{Ne,sep}/n_{e,sep} \right\rangle$ = 0.6% and 1.2%. The simulations also include the improved treatment of Ne impurity recycling and pumping recently developed for EMC3-EIRENE [4]. Overall global trends are reasonably matched between the three codes, with each finding enhanced interactions at the lower-outer and upper main wall regions corresponding to intersections with the secondary separatrix. The paper will discuss in detail the results of the 2D to 3D code comparison.

References:
[1] A. Loarte et al., Plasma Phys. Control. Fusion 67 (2025) 065023
[2] J. Van Blarcum et al., “EMC3-Eirene simulations of first wall heat and particle loads in ITER Start of Research Operation scenarios”, this conference
[3] S. Sureshkumar et al., “Heat and impurity ion fluxes on the ITER first wall at high far SOL transport in ITER burning plasmas”, this conference
[4] H. Frerichs et al., “Recent improvements in EMC3-EIRENE modeling for divertor detachment in burning plasmas”, this conference

Speaker: Dr Manni Jia (Institute of Plasma Physics, Chinese Academy of Sciences)

153 2.021 Design Considerations for Main Wall Plasma-facing Components in Magnetic-confinement Fusion Devices

Magnetic-confinement fusion energy (MFE) devices, including Pilot Plants and DEMOs, need to breed tritium if the D-T fuel cycle is used. Breeding blankets pose a new challenge for main wall plasma-facing component (PFC) design not faced until now since they must have thin (~mm’s) armoring to achieve a Tritium Breeding Ratio (TBR) > 1 [1]. By contrast, the ITER first wall thickness is ~10 cm and can handle up to ~ 5 MW/m2 of deposited power flux density [2] – an order of magnitude higher than expected by a breeder blanket wall [3]. In this work, this challenge is addressed by separating the function of the wall into 3 distinct regions; 1) thin blanket-wall interface; 2) divertor target region; and 3) main wall, which could include the divertor entrance as well as protection limiters around the blanket-wall. Here, the focus is on tokamaks because the scrape-off layer (SOL) in this MFE system is the most understood. Nonetheless, it is foreseeable that a similar separation of function applies to any MFE reactor concept.
The main wall PFCs purpose is the removal of SOL plasma fluxes (heat and ions) not captured by divertor targets before striking the blanket-wall interface. Controlling SOL plasma contact with this interface requires a balance between plasma transport parallel and perpendicular to the magnetic field. This work uses a new code, DIV3D, for SOL power handling that simulates this parallel/perpendicular balance as well as allows for non-axisymmetric PFCs [4]. The code features strict global power conservation and estimates the cross-field plasma transport into magnetically-shadowed regions, which is significant when dealing with protection limiters that can produce large shadowed regions near the blanket-wall interface. This workflow can analyze limiter surfaces to protect against transient operation, e.g. off-normal events and the startup and ramp-down phases, as well as allows for a more compact vacuum volume by optimizing the wall to confined-plasma gap. This new tool and proposed PFC workflow is aiming for a more general and systematic description of plasma loading to MFE PFC surfaces.
Work supported in part by US DOE under DE-AC05-00OR22725 and DE-FC02-04ER54698.
[1] P.C. Stangeby, E.A. Unterberg et al., Plasma Phys. Control. Fusion 64 (2022) 055018.
[2] R. Mitteau et al., J. Nucl.Mater. 463(2015) 411.
[3] Y. Miyoshi et al., Fusion Eng. Des. 151(2020) 111394.
[4] J.H. Nichols et al., Fusion Eng. Des. 219 (2025) 115278.

Speaker: Zeke Unterberg (Oak Ridge National Lab)

154 2.022 Two-point model with filamentary transport

The well-known two-point model [1, 2, 3] is a simple way to connect midplane and target conditions in the SOL. To further connect parallel and perpendicular SOL transport, average radial diffusive and advective transport of particles and heat is often assumed. However, the radial transport in the SOL is turbulent, consisting of plasma filaments of excess particles and heat driven by interchange motion [4]. This turbulence gives rise to correlations between fluctuations not accounted for when considering average plasma parameters [3].

Recent progress in stochastic modeling of filaments has resulted in self-consistent expressions for both single-point fluctuations and radial profiles due to filamentary transport [4, 5], giving expressions consistent with experimental measurements [6, 7]. In this contribution, we extend the stochastic model to simultaneous models for particle density and temperature as well as perpendicular particle and heat fluxes. We compare the results to expectations from diffusion-advection models for the simple SOL [1], and consider the effect of including the filamentary transport on target conditions.

References
[1] P. C. Stangeby. “The Plasma Boundary of Magnetic Fusion Devices.” CRC Press, 2000.
[2] F. Militello. “Boundary Plasma Physics.” Springer, 2023.
[3] W. Fundamenski. “Power exhaust in Fusion Plasmas.” CUP, 2009.
[4] O. E. Garcia et al. Phys. Plasmas 23.5 (2016). doi: 10.1063/1.4951016.
[5] J. M. Losada, A. Theodorsen, and O. E. Garcia. Phys. Plasmas 30 (2023).
doi: 10.1063/5.0144885.
[6] S. Ahmed et al. Plasma Phys. Controlled Fusion 65.10 (2023). doi:
10.1088/1361-6587/acf2c1.
[7] O. E. Garcia et al. “Blob structures and density shoulder formation in Alcator C-Mod”. Oral contribution at TTF 2025.

Speaker: Audun Theodorsen (UiT - the Arctic University of Norway)

155 2.023 Temperature threshold reduction of W fuzz formation affected by surface pre-treatment

In future tokamak reactors, tungsten (W) is expected to serve as the main plasma-facing material. When subjected to helium (He) plasma—a byproduct of fusion reactions—and heated above 900 K, W surfaces can develop nanofiber-like structures known as "fuzz," which significantly degrade thermal conductivity [1]. This phenomenon has been linked to the dynamics of helium bubble formation and rupture, but the detailed mechanisms, particularly those governing growth from tens of nanometers to micrometers, are still under active study.
Previous studies have consistently shown that the lowest temperature required for tungsten fuzz formation is approximately 900 K [2]. However, in our work, we found that when the initial surface already contains nanoscale features, fuzz can develop at temperatures significantly below 900 K. This observation first emerged from He plasma and W co-deposition experiments performed on Si nanocone substrates. At a substrate temperature of 790 K, W fuzz was observed to grow on top of the Si nanocones [3]. Motivated by this result, we directly irradiated nanostructured W substrates with He plasma. Fuzz was first produced at 1000 K under He plasma, which can be considered as an initial, after which the temperature was lowered while irradiation continued. Remarkably, fuzz growth persisted even at ~750 K [4]. To exclude the possibility that high-temperature He-plasma pretreatment was responsible for the reduced fuzz-growth temperature, we pre-textured W surfaces using femtosecond-pulse laser irradiation to generate periodic ripple structures (LIPSS). At ~750 K, the typical fuzz morphology did not form; instead, the resulting surface resembled that of untreated W under the same conditions. However, at the edges of the LIPSS region, where the laser energy was excessive, nodule-like features transformed fully into fuzz at 750 K. This indicates the importance of the sharp structure to reduce the temperature threshold.
The growth rate of fuzz with an initially nanostructured surface at low temperature shows different comparing to the conventional high temperature irradiation. Thermal desorption spectroscopy (TDS) was performed to investigate the role of He on the above phenomenon. This study provides a new perspective on understanding the further growth of fuzz and lowers the practical barrier for applying metal fuzz structures.
[1] S. Kajita et al., Nucl. Fusion 49 (2009) 095005.
[2] G. De Temmerman et al., Plasma Phys. Control. Fusion 60 (2018) 044018.
[3] Q. Shi et al., Nucl. Mate. Energy 39 (2024) 101668.
[4] Q. Shi et al., J. Nucl. Mate. 617 (2025) 156134.

Speaker: Quan Shi (the University of Tokyo)

156 2.024 Characteristics of boron layers deposited during boronization and boron powder injection in KSTAR

One of the most deleterious impurities is oxygen, which originates from the inevitable oxidation process in the material inside the tokamak. Boron (B) is an effective oxygen getter by easily forming chemical compounds such as boron oxide. Boronization is a commonly used method of wall conditioning that has been employed in various devices[1-3].
In KSTAR, boronization was performed using carborane (C2B10H12)[4]. B layers are replenished during operations using an impurity powder dropper (IPD), which can inject precise, reproducible amounts of powdered materials into plasma discharges[5].
In KSTAR, wall conditioning effects, such as reduced impurity emission and radiated power, following B injection have been reported[6]. However, the duration of these beneficial effects is limited by the erosion of the boron layer during plasma operation. For this reason, two samples were collected: a dust sample from the lower tungsten cassette divertor (TCD), and a coupon from the outer mid-plane region, collected using a specially designed structure with a slit opening.
Post-mortem analysis using ellipsomtery revealed that thickness of the TCD dust and the coupon sample were 3 um and 400 nm, respectively. The XPS spectra showed that all samples exhibited boron (B1s), carbon (C1s), tungsten (W4f), and oxygen (O1s) peaks. However, their chemical compositions differed. For the B1s spectra, the TCD dust exhibited two peaks at 187.5 eV and 190.6 eV, corresponding to elemental boron and boron nitride (BN) [7], respectively. In contrast, the coupon sample showed only a single peak at 191.2 eV, which is attributed to a B–O bond. The O1s peak for both samples appeared at approximately 532 eV, which can be assigned to either C–O or B–O bonding[8]. In case of C1s spectra, both samples displayed a large peak for C-C (284 eV). The coupons sample had an additional peak at ~287 eV, corresponding to a C-O bond. The W4f spectrum also differed between the samples. The dust samples showed peaks at 34.8 and 36.8 eV corresponding to the species of W2O5 and WO3[7], whereas the main W4f peaks for coupon sample is at ~ 30 eV. The observed differences in chemical composition indicate that the boron layers experienced different plasma exposure conditions depending on their retrieval locations. In this work, we report a comprehensive characterization of the boron-containing coated layers, including their elemental compositions and chemical states.

Speaker: SOOHYUN SON (KFE)

157 2.025 SURFACE DAMAGE OF ITER-GRADE TUNGSTEN UNDER SEQUENTIAL IRRADIATION WITH HELIUM ION BEAM AND HIGH HEAT FLUXES OF QSPA HYDROGEN PLASMA

The synergistic effects of tungsten exposure to combined hydrogen and helium particle fluxes need to be extensively studied for the realization of a fusion reactor project. IGP tungsten (PLANSEE) with transversal grain 12125 mm3 was sequentially irradiated with a helium (He) ion beam and hydrogen plasma generated by QSPA at a surface temperature close to room temperature (RT). The ion energy of He was 4 MeV, and the total fluence up to 2∙1024 ion/m2. Sputtering-like relief pattern of the exposed surface. Residual compressive stresses of -430 MPa were detected on the sample surface after He ion beam irradiation. An increase in the number of dislocations was observed, while the change in the lattice parameter remained negligible. High heat flux QSPA plasma streams (the duration of each pulse is 0.25 ms) with surface energy load of 0.9 MJ/m2 relevant to ITER ELM caused pronounced melting of the exposed surface. QSPA plasma irradiation resulted in partial annealing of the residual stresses, reducing them by a factor of about four to -120 MPa. The dislocation density decreased to a level close to the initial state, while the change in the lattice parameter remained negligible. Surface melting, a large crack network (up to 0.6 mm), and an intergranular crack network (up to 40 µm), as well as pores, were detected on the exposed surface after the combined irradiation. Surface changes and cross-sectional studies are discussed as well.

Speaker: Dr Vadym Makhlai (National Science Center Kharkiv Institute of Physics and Technology, Kharkiv, Ukraine)

158 2.026 Growth and coalescence of He bubbles in W

In the context of nuclear fusion, the inner walls of the reactor are exposed to extremely high ion fluxes, particularly at the divertor, which collects most of the particles escaping from the plasma. Tungsten (W) is used for this component because of its high melting point, erosion resistance, low hydrogen retention and good thermal conductivity[1]. However, helium implantation induces the formation of bubbles, which can alter the material properties. Moreover, tungsten oxidizes easily, even at room temperature, and this oxidation is enhanced in fusion environments due to oxygen impurities and extreme temperature[2]. The formation of a surface oxide layer can reduce the thermal conductivity and change the He retention [3]. It is therefore essential to understand how the presence and thickness of a thin oxide layer influence helium bubble formation and evolution.

In this work, we investigate helium implantation in W(110) single crystals covered with high-purity tungsten oxide thin films of controlled thickness. Oxide layers ranging from 7 to 50 nm were produced by thermal oxidation under low oxygen pressure (10⁻⁴ Torr) and characterized prior to implantation using grazing-incidence X-ray techniques. Helium implantation was performed at room temperature at 400 eV, below the displacement threshold of tungsten atoms, enabling us to isolate the effects induced solely by helium. The structural and morphological evolution of the samples was monitored in situ at the ESRF BM32 beamline.

During implantation, grazing incidence small angle X-rays scattering GISAXS measurements show a progressive broadening of the specular rod. Implantation was stopped at a fluence of 3 × 10²⁰ m⁻². After implantation, GISAXS patterns exhibit a diffuse low-q halo, indicating the formation of non-faceted helium bubbles. Post-mortem transmission electron microscopy reveals helium bubbles in the sample coated with the thinnest oxide layer (7 nm). In this case, the bubbles are homogeneously distributed throughout the WO₃ (~3 nm) layer and are larger than those observed in the tungsten substrate (~1 nm). In W, bubbles are mainly located at the W/WO₃ interface and remain detectable up to ~10 nm in depth. These results demonstrate that oxide thickness plays a decisive role in helium bubble formation, and is therefore a key parameter for understanding the behaviour of oxidized tungsten under fusion-relevant conditions.

Bibliography
[1] G. Laval, Nuclear Fusion: EDP Sciences, 2007.
[2] C. V. Ramana and al., J. Phys. Chem. B, vol. 110, pp. 10430–10435, 2006.
[3] A. W. Kleyn and al., Vacuum, vol. 80, pp. 1098–1106, 2006.

Speaker: Hozane Blanche NGONGANG ELOKO (Aix Marseille Univ, CNRS, CINAM, Marseille, France)

159 2.027 The effect of the radio frequency sheath on the sputtering of plasma facing materials

The presence of the radio frequency (RF) sheath can cause enhanced sputtering of plasma facing materials. Compared to a DC sheath with the same average potential, an RF sheath causes a broadening of the ion energy distribution function (IEDF) and results in enhanced light ion sputtering, especially for average RF sheath potentials close to the material sputtering threshold. The effect of RF sheaths on the erosion of fusion research relevant materials is being studied on the Radio Frequency Plasma Interaction Experiment (RF PIE). The RF PIE consists of an electron cyclotron resonance plasma source (2.45 GHz, 5 kW) with a biased and heated RF electrode that is used to simulate antenna surfaces in contact with the edge plasma for RF biases up to 500 V. The erosion of the material surface is being studied spectroscopically as a function of ion energy using a mirror-linked UV imaging spectrometer for measuring plasma line emission in and near the sheath. Materials of interest include plasma-sprayed tantalum pentoxide on silicon nitride, which is relevant for understanding the erosion of the plasma-exposed window material in the helicon plasma source being developed for the Material Plasma Exposure eXperiment (MPEX). Other plasma facing antenna materials being explored include tungsten, carbide dispersoid-strengthened tungsten , and titanium diboride. Calculations of the IEDF using hPIC2 predict a high energy tail of the distribution that can be close to 2X higher than the average sheath energy, which results in finite sputtering when the average sheath energy is below the sputter threshold energy. Predictions of the expected sputtering yield for DC and RF sheath conditions using semi-empirical sputtering formulas for helium and deuterium sputtering of these materials are consistent with experimental observations of changes in the line emission intensity as a function of ion energy. The SDTrimSP code is being used to predict sputtering of compounds (e.g., tantalum pentoxide) by calculating partial sputter yields in steady-state conditions, and results compare well with experimental measurements.

*This work sponsored by US DOE under DE-AC05-00OR22725

Speaker: John Caughman (Oak Ridge National Laboratory)

160 2.028 3D modeling of target-plasma interactions at oblique angles of incidence in MPEX-like linear plasma devices

We present three-dimensional EMC3-EIRENE modeling of an MPEX-like linear plasma configuration, showing that access to a high-recycling regime can be maintained even at significantly oblique target angles via utilization of near-target neutral baffles. MPEX (Materials Plasma Exposure eXperiment) plans to expose candidate material targets to extreme plasma fluences in tokamak divertor-relevant unbiased sheath regimes. Tilting the targets relative to the plasma column leads to the formation of a more reactor-relevant Chodura sheath, but also reduces the ion flux to the target due to both geometric expansion of the plasma footprint and a reduction in ionization of recycled neutrals.

Simulations in the geometry of Proto-Lite (a helicon-only prototype of MPEX) show that in a high-density upstream plasma regime, tilting the target to 85 degrees relative to surface normal decreases the recycling-driven target ion flux by over a factor of 2 relative to a standard 0 degree to normal target. However, simulations show that the inclusion of neutral baffling around the target sample holder reflects recycled neutrals into the plasma column, bringing the target back into the high-recycling regime and recovering up to 80% of the lost ion flux.

Comparisons are shown to early experiments in Proto-Lite, which has been upgraded to allow both tilted and standard target holder configurations. D-alpha emission trends at the target confirm the escape of recycled neutrals from the plasma column at significant target tilt. Experiments with baffle plates opposite the tilted target (in a geometry reminiscent of a tokamak slot divertor) exhibit enhanced D-alpha emission, indicating improved ion flux to the target. Implications for initial experiments in MPEX – which only includes a standard target configuration in its baseline project scope – are discussed.

This work is supported by the DOE Office of Science, Office of Fusion Energy Science, under contract number DE-AC05-00OR22725.

Speaker: Jake Nichols (ORNL)

161 2.029 Non-Equilibrium Ion Dynamics in Tungsten: A Machine Learning Approach to Radiation Damage

The resilience of plasma-facing materials (PFMs) under intense radiation is a significant challenge for feasible fusion reactors. Computational predictions of cumulative radiation damage are needed to explain experimental observations but are obscured due to the complex ultrafast dynamics that occur after energetic collisions with the lattice. As a result, scientific understanding of PFM degradation is inadequate, creating the need for model advancement. This project seeks to overcome aforementioned challenges through advanced computational techniques to study radiation damage accumulation in tungsten.

We introduce a novel class of highly precise machine-learned interatomic potentials (ML-IAP) for Molecular Dynamics (MD) simulations that directly account for electronic structure changes when shifting from ground state to warm-dense matter. ML-IAP are trained on a diverse set of density functional theory (DFT) defect calculations to directly capture the altered electronic states that occur during initial stages of neutron collisions. Employment of these excited state ML-IAP can significantly improve the accuracy of mesoscale damage accumulation models (NRT-, CRC-DPA) as MD is the determining factor in their parameterization. In particular, the consequences of accounting for excited state dynamics in MD are amplified at higher energies (>10keV) of the recoil spectra. This research demonstrates advancements in machine learning for materials development in extreme conditions and illustrates the importance of semiclassical dynamics in understanding radiation damage accumulation in PFM. Extensions of the work to other excited state processes in material is also discussed.

Speaker: Gabrielle Koknat (Sandia National Laboratories, USA)

162 2.030 Fundamental quantities in ion-solid interaction at fusion relevant energies - interatomic potentials of Cr assessed using low-energy ions

In a future fusion reactor, energetic ions and neutrals escaping from the plasma will interact with plasma facing materials (PFM) on the first wall of the reactor. The impact of these particles will lead to modifications of the PFM, which in turn affects the performance, durability and safety of the reactor. Reliable modelling is therefore necessary to accurately predict these interaction and material modifications. One fundamental quantity of ion-solid interaction is the interatomic potential, which describes the force field between atoms. It plays a crucial role in predicting collision cascades and thus sputter yields, implantation profiles and near-surface modification under ion bombardment. However, the commonly employed screened Coulomb potentials are known to be inaccurate for fusion relevant energies, and experimental attempts of a quantitative assessment are scarce [1]. Time-of-Flight Low Energy Ion Scattering (ToF-LEIS) allows the assessment of interatomic potentials at energies or interaction distances relevant for the plasma surface interactions in fusion reactors.

In this work, we experimentally investigate the interatomic potential of He and Cr, with complementary measurements using deuterium (D) planned to assess isotope effects for fusion plasmas. Cr is of particular interest as it is a component of steel (i.e. EUROFER97 [2]), considered as structural material for future fusion reactors, while also being explored as a constituent of SMART (self-passivating metal alloys with reduced thermo-oxidation) materials for use as PFM [3]. Specifically, we performed angular scans of the scattering yield on a Cr(100) single crystal, using a beam of 3keV He+ ions. By adjusting the screening length with a correction factor in Molecular Dynamics (MD) simulations, we were able to reproduce the width and intensity of the minima and maxima in the experimental angular scans. The resulting improved potentials were subsequently compared to commonly used interatomic potential models as well as earlier studies on Fe and Cu, elements with similar atomic number in order to identify potential trends on the interatomic potential strength. Furthermore, the effect of the ion energy on the observed interatomic potential was investigated to assess how the interaction strength evolves with ion velocity and thereby improve the predictability across an extended range.

Speaker: Athanasios Bamidis (Uppsala University)

163 2.031 Analysis of Combined Radiation Damage and Plasma Exposure on Tungsten Components for Fusion Reactor Applications

This study addresses critical knowledge gaps in tungsten-chromium (W-Cr) alloy behaviour under fusion-relevant conditions. As fusion energy progresses,developing materials for extreme reactor environments is crucial. This research conducts a comprehensive comparison of W-Cr alloys and ITER-grade tungsten, focusing on their response to simultaneous radiation damage and plasma-material interactions. Using advanced experimental techniques and computational modelling, the study quantifies radiation effects,characterises plasma exposure effects, and evaluates their combined impacts. A key innovation is the development of machine learning models to predict long-term material behaviour and usage of neural networks in image analysis. By benchmarking W-Cr alloys against pure tungsten and creating predictive models this project provides crucial insights for future fusion reactor design. Aligning with key research missions on neutron tolerance and safety, this project integrates cutting-edge materials science with fusion technology development, aiming to accelerate progress towards sustainable, carbon-free fusion energy.
The project's multidisciplinary approach combines radiation physics, plasma science, and advanced materials characterization techniques. The findings will not only advance our understanding of W-Cr alloys but also contribute to the broader field of extreme environment materials, potentially impacting other areas of energy research and industrial applications. By addressing key challenges in fusion reactor materials, this project plays a vital role in the global effort to develop clean, sustainable energy solutions for the future.

Speaker: Dr Deepankara Vrushabhadas Shastri (School of Metallurgy and Materials, University of Birmingham, Edgbaston B15 2TT, United Kingdom)

164 2.032 Laser-Induced Breakdown Spectroscopy for Characterization of Plasma-Facing Components in the ADITYA-U Tokamak

Plasma–wall interactions play a critical role in determining plasma performance, impurity behaviour, and component lifetime in present-day tokamaks and future fusion reactors. Processes such as impurity deposition, re-deposition, erosion, and fuel retention in plasma-facing components (PFCs) are strongly influenced by wall conditioning techniques and local plasma conditions. In the ADITYA-U tokamak, wall conditioning methods including lithiumization, boronization, and glow discharge cleaning (GDC) are routinely employed to control impurities and improve discharge performance. While residual gas analysers (RGA) provide valuable information on global gas composition, they do not offer direct, spatially resolved measurements of surface impurity content on PFCs.

To overcome this limitation, an in-situ Laser-Induced Breakdown Spectroscopy (LIBS) diagnostic has been developed for elemental characterization of PFC surfaces in ADITYA-U. LIBS enables rapid, remote, and multi-element analysis with minimal sample preparation by focusing a pulsed laser onto the surface and spectrally analysing the resulting plasma emission.

The LIBS system is integrated with the tokamak through a dedicated diagnostic port and isolated from the main vacuum vessel using a gate valve, allowing measurements to be performed between plasma discharges and under controlled ambient gas conditions. The diagnostic allows us to introduce various materials, such as Graphite, Tungsten, stainless steel, etc., into the plasma to check their viability as PFCs. This sample can then be retracted on a shot-to-shot basis from the vessel, and LIBS analysis can be performed.

This contribution presents the design, calibration, and first results of the LIBS system for PFC characterization in ADITYA-U tokamak. Furthermore, an ex-situ LIBS setup has been developed. Experiments were performed to analyse LIBS spectra of various exposed materials retrieved from the vacuum vessel at the end of a campaign and to compare them with unexposed reference materials.

Speaker: Bharat Hegde (Institute for Plasma Research)

165 2.033 Survivability of boron dust and powder in edge fusion plasmas

The recent ITER re-baseline to an all-tungsten (W) reactor necessitates frequent boronizations, that will also lead to the production of intrinsic boron (B) dust whose transport and survivability should be well understood. Furthermore, it has been experimentally demonstrated that B powder injection can be utilized for real-time wall conditioning, pedestal control and turbulence suppression [1,2]. Thus, the intricate coupling between impurities from vaporization of injected B dust and edge plasmas needs to be elucidated. Surprisingly, in works that report simulations of B dust in fusion devices [3], there has been no discussion of the surface & material properties of B (light semi-conducting) that are fundamentally different from those of W (heavy refractory metal) or C/Be (light conducting).

The MIGRAINe dust dynamics code boasts a complete description of dust microphysical processes and mechanical impacts as well as a reactor relevant plasma collection model (that considers thin sheath and electron magnetization effects) [4]. We report the extension of MIGRAINe’s capabilities to B dust. MIGRAINe has been upgraded with state-of-the-art B models that describe all aspects of: (i) secondary electron emission and electron backscattering up to 5keV based on reliable experimental results and Monte Carlo simulations with Geant4, (ii) physical sputtering and ion reflection of H, D, T, He, B, Ne, W up to 5keV based on ERO SDTrimSP6 Monte Carlo simulations, (iii) thermal properties of thermodynamic nature (heat capacity, latent heat) and evaporative nature (vaporization, thermionic emission) based on reliable experimental results, (iv) optical properties (refractive index) based on measurements in an extended frequency range that are used to calculate the adhesive force via Lifshitz theory of van der Waals forces and the hemispherical emissivity via Mie scattering theory.

We report on MIGRAINe simulations of B dust survivability in plasma profiles relevant for Scrape-off-Layer and divertor plasmas. We focus on the cooling and heating contributions in different plasma scenarios and scan different sizes. The simulations of B dust are compared with those for W dust under identical conditions. Emphasis is put on the role of thermionic emission (dominant for W but insignificant for B dust) and secondary electron emission (important for B but insignificant for W dust).

[1] S. Ratynskaia et al., Rev. Mod. Plasma Phys. 6, 20 (2022).
[2] F. Nespoli et al., Nat. Phys. 18, 350 (2022).
[3] F. Nespoli et al., Nucl. Mater. Energy 42, 101837 (2025).
[4] S. Ratynskaia et al., Plasma Phys. Control. Fusion 64, 044004 (2022).

Speaker: Lorenzo Boccaccia (KTH Royal Institute of Technology)

166 2.034 Machine Learning for Hydrogen Recycling Modeling

Understanding the mechanisms of hydrogen recycling is a key factor in accurately predicting and managing plasma behavior in nuclear fusion reactors. In the present study, we develop a machine learning (ML) model capable of predicting the translational energy distributions and rovibrational states of hydrogen atoms and molecules emitted from tungsten plasma-facing components. These predictions play a critical role in evaluating the influence of recycled hydrogen on edge plasmas through neutral transport simulations.

The training dataset for the ML model[1] is generated using molecular dynamics (MD) simulations[2-4], which replicate hydrogen atom injection into hydrogen-saturated tungsten under various conditions, including different incident energies, material temperatures, and hydrogen-to-tungsten (H/W) ratios. These simulations provide detailed information on the resulting energy and rovibrational distributions.

To strike a balance between computational efficiency and predictive accuracy across a broad parameter space, we employ a fully connected neural network trained on 120 datasets derived from 24 distinct simulation scenarios, enhanced via random sampling for data augmentation. This ML model demonstrates high fidelity in reproducing the emission behavior of hydrogen species under monochromatic injection conditions.

The model is further generalized to more realistic plasma conditions by incorporating a shifted-Maxwellian distribution for the incident energy, accounting for the energy gain of ions in the sheath region. Two methodologies are proposed: (1) numerical integration of the monochromatic ML model over the shifted-Maxwellian distribution, and (2) development of a new ML model trained directly on pre-integrated data with four input parameters—H/W ratio, material temperature, and ion and electron temperatures. The analysis reveals that elevated electron temperatures enhance atomic hydrogen emission, while lower electron temperatures favor the release of molecular hydrogen.

References
[1] S. Saito, et al., “Machine Learning-Based Hydrogen Recycling Model for Predicting Rovibrational Distributions of Released Molecular Hydrogen on Tungsten Materials via Molecular Dynamics Simulations”, accepted to Nucl. Mater. Energy, (2025).
[2] S. Saito, et al., Contrib. Plasma Phys. e201900152 (2020).
[3] S. Saito, et al., Jpn. J. Appl. Phys. 60, SAAB08 (2021).
[4] S. Saito, et al., Nucl. Fusion 64 126067 (2024).

Speaker: Seiki Saito (Yamagata University)

167 2.035 Comprehensive Simulation of Plasma Transient Events—Showstopper for the magnetic fusion energy production!

A serious obstacle to a successful magnetic fusion energy production is reactor performance during abnormal events. Abnormal events include plasma disruptions, edge-localized modes (ELMs), vertical displacement events (VDEs), and runaway electrons. While tremendous efforts are being pursued to find ways to mitigate such events, a credible reactor design must be able to tolerate few of these transient events without significant mitigation techniques. Accurate and detail prediction of plasma-facing and nearby components response to instabilities are essential for safe and reliable operation of future fusion devices. We have recently upgraded our comprehensive HEIGHTS (High Energy Interaction with General Heterogeneous Target Systems) simulation package to enable detail 3-D investigation of the overall aspects of plasma-material interaction (PMI) phenomena during the transient events. Advanced numerical tools and solution methods in an integrated parallel environment were used to efficiently couple major key processes during the transient events, and in particular disruptions and giant ELMs. These include dynamic interaction of the escaping core plasma particles with the evolving and propagating secondary divertor vapor/plasma. The collisions, heating, and scattering of the disrupting energetic plasma particles with the propagating secondary “mini” dense plasma in tokamak magnetic field structure in realistic rector configuration are critical in assessing the damage to all interior components including hidden structure and the first wall which were not directly exposed to these transient events. Despite developing numerous efficient numerical techniques and solution methods such calculations take several weeks on supercomputers to complete due to implementing very detail physics and exact 3D geometry. Our results show, for the first time, that unmitigated and mitigated transient events could cause significant damage to most interior and hidden components including the first walls that were not directly exposed to these events.
The simulation showed that disruptions and high-power ELMs cause excessive target melting and erosion of candidate materials of W and Li and potential plasma contamination. Recent simulation predicted serious damage from plasma disruptions to most divertor nearby and hidden locations including baffles, dome, and even parts of the first wall due to the secondary radiation and ions scattering resulting from the dynamically evolving vapor cloud on the original disruption area of the W divertor plate. Various and new mitigation methods were also suggested analyzed with serious disadvantages were identified for each of these methods. The current ITER divertor design may need modifications to tolerate these events, even with mitigations.

Speaker: Ahmed Hassanein (GNOI)

168 2.036 β-decay induced self-charging of tritiated tungsten dust in fusion reactors

Dust in D-T fusion reactors will be radioactive due to interactions with 14MeV neutrons and retention of tritium, which undergoes β-decay into helium with a 12.32yrs half-life. Tritium radioactive decay releases 18.6keV partitioned between an electron and anti-neutrino. The mean electron kinetic energy of 5.7keV suffices to traverse 200nm of W, implying that tritiated W dust constantly emits electrons.

It has been documented that tritiated carbon dust collected from TFTR post D-T operation exhibited unusual mobility, levitated in presence of electrostatic fields and even spontaneously boiled [1]. Many consequences of tritiated dust self-charging have been conjectured [2]. Nevertheless, even very recent Monte Carlo simulations of β-decay induced self-charging have been oversimplified [3]. They were based on the EmOpt4 Geant4 package that is incapable of treating secondary electron emission due to beta electrons and, thus, significantly underestimate self-charging.

We performed Monte Carlo simulations of the self-charging rate of tritiated W dust with the MicroElec Geant4 extension that is tailor made for the relevant electron energy range [4], since it utilizes single scattering algorithms, employs the dielectric formalism that treats secondary electron generation due to electron excitation & plasmon decay and considers reflection of internal electrons from the surface potential barrier. Moreover, we empirically modified the surface barrier so that it reproduces reliable electron emission data at normal incidence and validated the updated model against reliable electron emission data at oblique incidence. Finally, we introduced a hybrid physics list implementation that smoothly combines MicroElec for <7keV with EmOpt4 for higher energies circumventing practical limitations in MicroElec’s applicability.

Self-charging rates are evaluated for different W dust radii after averaging over the uniform angular distribution & known energy spectra of the beta electrons and over experimental tritium depth profiles. Regardless of dust size, an empirical expression is obtained for the depth resolved electron escape yield that stems from the elementary theory of secondary electron emission. Moreover, up to one-order-of magnitude differences are revealed between the MicroElec and EmOpt4 predictions. Finally, the effect of the magnetic field due to prompt electron re-deposition and the effect of positive dust charge due to Coulomb attraction on the self-charging rates is evaluated.

[1] C. Skinner et al., Fusion Sci. Technol. 45, 11 (2004).
[2] J. Winter, Phys. Plasmas 7, 3862 (2000).
[3] C. Grisolia et al., Nucl. Fusion 59, 086061 (2019).
[4] Q. Gibaru et al., Nucl. Instrum. Meth. Phys. Res. B 487, 66 (2021).

Speaker: Tommaso Rizzi (KTH - Royal Institute of Technology)

169 2.037 Self-consistent integrated modeling of processes taking place during plasma material interactions in Tokamaks

Detailed 3D models of processes in materials subsurface and in plasma above the surface were recently developed and integrated into a single package, for the first time. This avoids uncertainties in predicting materials performance in fusion devices with complex geometries, magnetic, and electric fields. The upgraded ITMC-DYN+ package includes comprehensive sets of Monte Carlo and deterministic models to simulate a) collisional interactions in material and in plasma, b) particle motion in electromagnetic fields, c) thermal processes of diffusion and chemical reactions; d) D/T trapping in defects and intrinsic defects growth and evolution to other types. These models enable self-consistent analysis of time- and space- dependent changes in plasma facing materials (PFMs) and their effect on erosion/redeposition, potential plasma contamination, D/T retention and recycling, material properties degradation. The integrated models were benchmarked against experiments at DIII-D regarding material erosion/redeposition and D retention as well as against laboratory experiments on D retention and defects growth in tungsten-based materials.
We simulated the response of several PFMs during DIII-D discharges and calculated fluxes of generated impurities depending on plasma and material properties. Detailed simulations of tungsten-based alloys, proposed to strengthen W microstructure, and tungsten with deposited coating layers were performed to predict high Z material erosion, T co-deposition, and potential plasma contamination at various material surface conditions. Time-dependent, multi-species modeling showed the effects of interplay between preferential sputtering, local and extrinsic impurities deposition, and ExB drift on divertor material erosion, enrichment, and growth of mixed amorphous layers. Changes in material surface and in high-Z impurity source for current graphite walls at DIII-D and future metallic walls were critically assessed in this study. Our calculations for the conditions of graphite-free walls showed, e.g., that fast surface enrichment in W in compounds with low Z content results in near full suppression of W erosion and orders of magnitude reduction of low Z impurity from the divertor in comparison with a graphite wall environment where W erosion does not decrease with time. Our successful simulations of various experiments showed 1) the effects of surface contamination or enrichment in W on D diffusivity and retention, 2) changes in the surface microstructure induced by collisional interactions and by D supersaturation, 3) correlation between irradiation conditions and D retention in blisters and in dislocation networks. Ignoring such processes integration to predict performance of materials in complex PMI environment during steady state and transient operation may not yield correct results.

Speaker: Tatyana Sizyuk (ANL, USA)

170 2.038 Composed domain solutions of plasma transport model for 2- dimensional Scrape-off-Layer setup in statistical limit of stochasticity

A Composed Domain Matrix Based (CDMB) approach to the edge-SOL plasma transport is developed and demonstrated to provide access to statisti- cal limit n → ∞ of the solutions converged from a conventional stochastic im- plementation of the SOL transport simulation model, where n is multiplicity of random variable realizations. The developed framework involves unwind- ing a multiple dimensional orthonormal grid domain to a one-dimensional state-space, rendering the CDMB simulation approach independent of num- ber of spatial dimensions. The developed steady-state simulation modules implement a coupled particle-momentum transport in a 2-dimensional Scrape- off Layer (SOL) having a convective parallel and diffusive cross-field transport coefficients with prescribed diffusivity across the field aligned degree of free- dom. The benchmark of composed domain solutions is done against both analytical solutions as well as numerically converged 2-dimensional solutions from a Markov-Chain Monte- Carlo simulation code MCMC. For mapping the solutions to the edge and SOL domains of a diverted tokamak plasma, the rectangular radial-poloidal sub-domains are subjected to periodicity along the poloidal dimension of the core while target-like recycling boundaries along the radial dimension. The plasma density and velocity distributions are obtained for plasma sources purely at the core boundary and are found to produce good agreement with the MCMC simulation procedure. The ap- plications of the CDMB approach to potential speed-up of the stochastic simulation procedures, usually desired during the fast scenario development and control for fusion grade device, are explored.

Speaker: NIDHI PANDAY (Institute For Plasma Research, India)

171 2.039 Net erosion of molybdenum coated ICRH limiter tiles in ASDEX-Upgrade

Ion cyclotron resonance heating (ICRH) is one of the essential heating systems in fusion devices, and the only mean of depositing energy directly on ions to boost fusion reactions. To maximize the power coupled to the plasma, ICRH antennas, surrounded by protection limiters, must be placed as close as possible to the plasma. If the antenna is too far, little power can be injected, but if it is too close, plasma surface interactions (PSI) kick in and deteriorate the performance. Optimal operation therefore often boils down to a trade-off between good wave coupling efficiency and well-mitigated interactions (either heat loads or erosion).

Three tiles of an ASDEX Upgrade (AUG) 3-strap ICRH antenna tungsten limiter were coated with molybdenum, to use it as trace material [1] allowing dedicated investigations by means of spectroscopy. Due to pre and post-exposure analysis of these tiles, the erosion during a whole AUG campaign with a mixture of discharge scenarios with and without ICRH can be determined.

This contribution focuses on the net erosion measurement of the 5 µm thick Mo-coated tiles after exposing them during the campaign 2024/25 in AUG. The two Mo-coated tiles from the left limiter, one in the upper half and the other one in the lower half, were analyzed by ion beam techniques (IBA) and scanning electron microscopy (SEM) before and after exposure in AUG. The erosion was determined by layer thickness change obtained with IBA and direct observation of FIB-prepared µm-rulers with SEM. The maximal Mo net erosion is below 1 um. Along the toroidal direction, the tile of the lower half (L3) shows two erosion maxima, one as expected at the point with the smallest distance to the plasma. A second maximum is about 1 cm farer away from the plasma and about 2 cm closer to the antenna strips. The Mo-coated tile from the upper half (L10) shows only marginal erosion. On both tiles deposition composed of boron and tungsten was observed in grooves even in the erosion areas. These results allow to strengthen the assessment of the life time of the ICRH limiter tiles.

[1] Urbanczyk G et al., 2020 Nucl. Fusion 60, 126003

Speaker: Martin Balden (GNOI)

172 2.040 Research on Plasma-Surface Interactions at the Budker Institute of Nuclear Physics

The Budker Institute of Nuclear Physics SB RAS (BINP SB RAS) conducts research on the interaction of plasma, electron beams, and laser radiation with materials under conditions simulating the loads on plasma-facing components (PFCs) in fusion devices. The current research is primarily focused on investigating the durability of first wall and divertor materials for projects like ITER, TRT, DEMO, as well as limiter and expander section for the advanced open trap GDMT. These facilities are characterized by quasi-stationary heat loads on PFCs with heat flux of $\sim 5 - 30$ MW$/$m$^2$, and transient loads with durations of $\sim 0.1 - 100$ ms and heat flux of $\sim 0.1 - 10$ GW$/$m$^2$.

The development of the BINP experimental setups led to the creation of the specialized BETA complex to study material erosion caused by heating expected during magnetic plasma confinement. The implementation of \textit{in situ} diagnostics for observing surface modification processes (including tungsten and promising ceramic compositions) during heating is particularly important. This approach, utilizing electron and laser beams for heat load simulation, is fundamentally unavailable for setups based on plasma sources.

The current BINP plasma-surface interaction research program focuses on:
- Upgrading the BETA complex facilities for material thermal loading (based on electron and laser beam sources);
- Developing new experimental stand based on a LaB$_6$ cathode electron beam source, designed to reproduce up to $\sim 10^8$ ELM-like pulses;
- Developing diagnostic systems for the quasi-stationary helicon plasma source facility to study not only material erosion but also hydrogen isotope retention and co-deposition;
- Design and construction of a station for studying fast transient processes at the SKIF synchrotron radiation (SI) source;
- Investigating samples using SI at the "Plasma" station (VEPP-4, BINP SB RAS);
- Experimental study of tungsten and promising ceramic materials behavior under exposure to plasma, electron beams, and laser radiation.

Key obtained results include:
- Determination of critical thresholds for various material damage mechanisms (cracking, melting and intense tungsten dust ejection, explosive graphite erosion, brittle fracture of monolithic ceramics and ceramic coatings);
- Experimental observation of the dynamics of residual plastic deformation in tungsten under repeated pulsed heating;
- Development of a theoretical model of tungsten cracking under pulsed heating;
- Experimental investigation of erosion of a wide range of ceramic materials (boron and silicon carbides, titanium and zirconium diborides) under transient thermal load exposure.

Speaker: Dr Dmitrii Cherepanov (BINP SB RAS)

173 2.041 Limiter phase nickel migration in the JET ITER-Like wall

ERO2.0 simulations of the limiter phase in the JET ITER-like wall (JET-ILW) predict nickel to erode primarily from the low-field side (LFS) midplane of the inconel 625 vacuum vessel wall due to charge-exchange neutrals. Inconel is a nickel alloy, mixed with iron and chromium. The sputtered nickel ionizes mostly in the LFS scrape-off layer and is transported by the plasma flows towards the contact point between plasma and beryllium limiters, depositing onto the limiter surfaces. Nickel is observed in post-mortem analysis of the divertor tiles [1]. Previous studies on nickel migration in the JET-ILW focus on the diverted phase of the plasma, with typical distances of 6cm from the limiter apex and 45cm from the inconel wall, neglecting the limiter phase despite the considerable time accumulated during plasma startups and shutdowns [2].

Nickel deposition into the divertor is negligible in comparison with the diverted phase, consistent with previous Be migration studies in limiter plasmas [3], during the 2011-2016 JET-ILW campaigns, with the high-field side (HFS) limited plasma adding nickel to the primarily Be deposit layer on the order of 1e17/cm² on tile 1. In the LFS limited plasma the nickel content is predicted to be 4e17/cm² on tile 1. Deposition into the divertor peaks on tile 10, where in the HFS limited plasma the deposition of nickel is on the order of 3e17/cm² and 1-2e18/cm² in the LFS limited plasma. The erosion rates for nickel in HFS and LFS limited scenarios are 8.4e19/s and 1.9e20/s respectively. The limiter configuration erosion rates are comparable to the lower end of the predicted rates in diverted plasmas [2].

Two limiter plasmas, representative of the 2011-2016 campaigns, are investigated using SOLPS-ITER, one limited by the HFS limiters and the other by the LFS limiters. SOLPS-ITER replicates the radial electron density profile as observed in experiments in both HFS and LFS limited plasmas. For the HFS limited plasma, SOLPS-ITER simulations match Te measurements from the core up to the last 3-4 cm from the radially outer edge of the grid, where Te is underpredicted by an order of magnitude. Te measurements for the LFS limited plasma are only available near the grid edge. Te at the closest measurement to the separatrix is 1.5x the predicted value.

[1] A. Widdowson et al., Nucl. Mater. Energy, 19:218, 2019
[2] P. Virtanen et al., Nucl. Mater. Energy, 42:101864, 2025
[3] S. Brezinsek et al., Nucl. Fusion, 55:063021, 2015

Speaker: Pyry Virtanen (Aalto University)

174 2.042 Thermal transition of W-on-W adhesion from van der Waals interactions to metallic bonding

The adhesion of dust on plasma-facing components (PFCs) plays a pivotal role in remobilization, wall mechanical impacts, resuspension during loss-of-vacuum accidents, collection activities and removal techniques [1]. Theoretical evaluations are possible with: (i) the van der Waals approach with adhesion assumed to emerge from the cumulative interaction between instantaneously induced multipoles [2]; (ii) elasticity theory with adhesion incorporated via the surface energy due to metallic bonding [3]. For metallic contacts, as in W-on-W, metallic bonding adhesion is nearly 100 times stronger than van der Waals adhesion. However, the experimental adhesion force agrees with the van der Waals result, since nm-scale roughness suffices to exceed the 0.3nm distance required to establish the metallic bond [4].

Experiments have quantified the effect of surface roughness, fusion-relevant coatings, atmospheric impurities and heat treatment [4,5,6]. In particular, electrostatic detachment measurements combined with vacuum furnace exposures have revealed that moderate but prolonged temperature excursions below the W recrystallization range can increase adhesion by orders of magnitude [6]. Unfortunately, dielectric breakdown at high field strengths prevented accurate adhesion measurements and only permitted lower bound estimates.

We surpass the limitations of earlier measurements employing an optimized electrostatic detachment set-up that features: (i) a HV power supply with a 60kV maximum (compared to 30kV) so that higher electrostatic fields are produced, (ii) 50um W dust (compared to 14um) so that higher electrostatic forces can be exerted, (iii) re-design of the insulation and cabling so that higher electrostatic fields can be tolerated prior to arcing. Thus, adhesion increase due to prolonged thermal treatment could be measured for the first time.

The results reveal that there is a sharp transition from the weak van der Waals value to the nearly 100 times stronger metallic bonding value with the transition point depending on the duration and strength of the thermal treatment. There is a switching of the dominant W-on-W interaction from long-range weak van der Waals to short-range strong metallic bonding which is enabled by atomic diffusion that slowly eliminates the nm-scale roughness.

[1] S. Ratynskaia et al., Rev. Mod. Plasma Phys. 6, 20 (2022).
[2] J. Israelachvili, Intermolecular and Surface Forces, Academic Press (2011).
[3] K. Johnson , Contact Mechanics, Cambridge University Press, (1985).
[4] G. Riva et al., Nucl. Mater. Energy 12, 593 (2017).
[5] P. Tolias et al., Nucl. Mater. Energy 15, 55 (2018).
[6] P. Tolias et al., Nucl. Mater. Energy 24, 100765 (2020).

Speaker: Marco De Angeli (Institute for Plasma Science and Technology - CNR Italy)

175 2.043 In-situ measurements of boron erosion and redeposition in deuterium plasma in the linear plasma device PSI-2

The new ITER baseline foresees boronizations for reliable plasma operations in the initial stages of experimental campaigns. However, boron can enhance the fuel retention in the wall and decrease the availability of ITER. Therefore, it is essential to understand the erosion and redeposition mechanisms of boron and its codeposition with hydrogen isotopes.
We performed studies of boron erosion and redeposition in deuterium plasma in the linear plasma device PSI-2. PSI-2 produces a steady-state plasma column with a diameter of ~70 mm. PSI-2 is equipped with a quartz microbalance (QMB) system, an in-situ deposition monitor, placed 175 mm away from the plasma axis. A disc of pure boron with an exposed diameter of 55 mm, fixed by a Mo mask, was used as the target. The axial distance between target and QMB was varied 5-35 cm by changing the target position. According to its solid angle, the QMB detector collects a fraction of ~10^-4 - 10^-5 of particles eroded from the target. The temperature of QMB was kept at RT by water cooling. Simultaneously the eroded boron was monitored by optical emission spectroscopy (OES) of the B I 2p-3s transition at 249.7 nm. The incident ion flux was 5x10²² m^-2s^-1 and the ion energy was varied 30 - 100 eV by biasing the target. After the exposure, the quartz detector was analysed by scanning electron microscopy (SEM) and nuclear reaction analysis (NRA).
The maximum deposition rate at QMB of 10 nm/hour was obtained for 100 eV and a target position 15-20 cm away from QMB. For 60 eV and 30 eV the deposition rate dropped by factors of 2.2 and 6, respectively. The drop of the OES signal intensity was less steep, 1.8 for 60 eV and 4.5 for 30 eV. In general, this behaviour reproduces the data available for boron sputtering from previous ion beam and OES studies.
SEM observed a layer of boron deposited on the detector plate with a thickness of 10-20 nm, in agreement with QMB. The estimated layer density is 1.1 g/cm³. NRA measured 10x10¹⁵ B/cm² and 5x10¹⁵ D/cm² in the boron layer. A significant D amount of 20x10¹⁵ D/cm² was detected in the substrate. The SEM and NRA results showed that the quartz plate is not the optimal substrate for the post-mortem analyses. Additional experiments with a polished tungsten substrate are planned.

Speaker: Arkadi Kreter (Forschungszentrum Jülich)

176 2.044 In-situ investigation of oxygen trapping in tungsten, boron and mixed thin-films during thermal annealing using ion beams

Oxygen (O), as a common medium-Z impurity in vacuum vessels, is one of the main concerns for magnetic confinement fusion devices, in part due to its high chemical reactivity and its potential to act as seed for sputtering of high-Z materials. For devices with full tungsten (W) walls for which the adsorption of O is low relative to other plasma-facing materials (PFM), wall conditioning techniques such as boronization become necessary for successful plasma startup. Continuous re-deposition steps during operation will lead to the formation of mixed W and boron (B) layers on plasma-facing surfaces, in which the presence of O can be expected [1].

The aim of this study is to provide a deeper understanding of the trapping of O in plasma-deposited thin films of W, B as well as in W:B mixtures under the effects of thermal annealing. Magnetron sputter deposition was employed to grow thin-films of B, W and W:B mixed layers. The films were irradiated by 1 keV O$_2$, equivalent to 500 eV per O, to a nominal fluence of 4×10$^{16}$ O/cm$^2$ followed by in-situ stepwise thermal annealing. Ion-beam analysis (IBA) was used to monitor the presence of O before, during and after irradiation and annealing of the samples.

It was found that the O-content following implantation is significantly higher in pure boron when compared with W, and mixed W:B layers. Annealing reduced the O-content in all compositions but had the strongest effect for pure B, which was heated to a maximum temperature of 500°C and for which the retained amount of O was reduced by close to 50%, corresponding to 4×10$^{15}$ O/cm$^2$. Results highlight that the oxygen affinity to pure boron layers can be potentially reduced for mixed layers containing tungsten. Finally, the effect of pre-loaded deuterium in boron layers on the subsequent trapping of oxygen (as expected for the use of diborane gas used in boronization) will also be presented and discussed.

1. A. Marin et al., J. Nucl. Mater. 604 (2025) 155525, doi: 10.1016/j.jnucmat.2024.155525.

Speaker: Daniel Gautam

177 2.045 Neon retention in tungsten, boron and mixed thin-films under the effects of thermal annealing studied by isotopic tracing

Neon (Ne) is considered to be used in magnetic confinement fusion devices for radiative plasma edge cooling and reduction of divertor heat loads. However, uncontrolled release of medium-Z impurities such as Ne will degrade plasma performance and lead to erosion of plasma-facing materials (PFM) by sputtering of high-Z elements. It is thus important to understand the Ne-retention behaviour in wall materials. Furthermore, for devices with tungsten (W) as PFM, wall conditioning techniques such as boronization will be unavoidable for startup. As a consequence, the presence of boron (B) and W:B mixtures can be expected: their formation has been previously observed in WEST [1].

The retention of neon was investigated by an isotopic tracer technique using ion beam analysis (IBA) to atomic concentration depth profiles of thin-films of mixed W and B as well as for pure W and B layers, grown both on silicon- (Si) and W-substrates by means of magnetron sputter deposition. Two Ne-isotopes, $^{20}$Ne and $^{22}$Ne, were implanted to a fluence of 3×10$^{16}$ at./cm$^2$ at different energies (35-190 keV) in order to achieve well-separated implantation profiles at specific depths in the thin-films (close to the surface and a deeper region). Thermal annealing in combination with time-of-flight elastic recoil detection analysis (ToF-ERDA) was employed to investigate the retention and depth distribution of the Ne-isotopes in each composition for a range of temperatures between room temperature and 1000°C.

This first-of-its-kind study has shown that both isotopes remain at their original implantation depth for the full range of temperatures and for all studied compositions, thus suggesting that no diffusion, intermixing nor desorption occurred. Substrate type did not play any significant role in the retention properties of the grown films. The research program is being extended to examine the impact of plasma impurities on the stability of the Ne-containing layers.

1. A. Marin et al., J. Nucl. Mater. 604 (2025) 155525, doi: 10.1016/j.jnucmat.2024.155525.

Speaker: Daniel Gautam

178 2.046 Reference boron layers for investigating the formation and characteristics of boron-based co-deposits in tokamaks

The new baseline of ITER foresees tungsten (W) as the sole plasma-facing material in the reactor vessel. To ensure proper conditioning of the all-W device, in particular for controlling oxygen levels in the plasma and enabling reliable start-up of the experimental operations, regular boronizations have been proposed. An unavoidable consequence of boronizations, however, will be the formation of (thick) boron (B)-containing co-deposited layers on the plasma-facing components, possibly as an outcome of several material migration and erosion steps in the reactor vessel. To obtain insights on the properties and formation mechanisms of such layers, an extensive programme has been set up by the EUROfusion Consortium to produce and characterize various B-containing reference layers with pre-defined specifications. In addition, the physics behind their erosion and deposition behaviour will be investigated via dedicated experiments in laboratory conditions and in linear plasma devices. In this contribution, we will report on advances made for the topic during the last two years.

Several reference layers with varying thicknesses (from ~50 nm to several micrometers) and compositions, ranging from pure B to mixed W-B deposits and with inclusions of plasma fuel (here, deuterium (D)), oxygen (O), and gaseous impurity elements have been prepared, exposed to plasmas or controlled particle beams, and analysed in the participating laboratories. Pure B layers can easily oxidize at ambient conditions, however typically no significant compositional changes are observed after extended exposures to air. The layers are also relatively brittle whereas inclusion of D seems to strengthen and stabilize them but only at low enough temperatures; once the deposits are treated at temperatures >200-250°C, cracks start to occur and the fuel content to decrease. Comparison against films originating after boronizations on different tokamaks is ongoing. Several similarities have already been observed such as a common tendency for layer oxidization and the fact that the composition does not notably changed with ageing. Ongoing studies include assessing the role of porosity on oxygen gettering, permeation of fuel into the substrate, and the dependence of the layer properties on their thickness and substrate. Finally, exposures to plasmas in linear machines have highlighted the role of chemical erosion in sputtering of the films at low energies as well as a consistent picture of erosion in varying flux regimes.

Speaker: Antti Hakola (VTT)

179 2.047 Erosion behaviour of compact and porous boron layers exposed to deuterium plasma in the GyM linear device

As part of ITER’s 2023 re-baselining, the first-wall material was changed from beryllium to tungsten. The absence of the beneficial ability of Be to getter low-Z impurities like oxygen will be compensated by covering the armour tiles with a boron layer of 10-100 nm (boronisation). Broadening the understanding of physics behind B erosion and re/co-deposition is a top priority for ITER to effectively control plasma-wall interactions.

This work investigates erosion of pulsed-laser deposited B coatings: 100 nm compact films (35% porosity), representing pristine B layers, on mirror-polished W substrates, and 1-8 $\mu$m porous films (50-95% porosity), mimicking re-deposits, on silicon wafers. B model systems were exposed to deuterium plasma in the GyM linear device under ITER first-wall-relevant conditions (3-4e20 m$^{-2}$s$^{-1}$ ion flux).

Within the EUROfusion work-programme, thin B films were exposed to E$_{ion}$=76 eV for comparison with experiments in PSI-2, on RF-magnetron-sputtered coatings [1], and in a D$^+$ source, on plasma-sprayed coatings [2]. Here, a low ion fluence $\Phi$=4e23 m$^{-2}$ was selected to avoid complete layer erosion. After exposure, samples were characterised using mass loss, AFM, SEM, ToF-ERDA, ToF-MEIS, SIMS, Raman spectroscopy, and XPS.

Regarding the pristine B layer, SIMS data interpretation was hindered by partial oxidation and is being disentangled through comparison with pure B samples and related compounds. ToF-ERDA shows $\sim$13.3 at.% of oxygen. Combining areal density ($\sigma$) and film thickness, a calculated mass density of 2.39 g/cm$^3$ is obtained. Post-exposure, $\sigma$ corresponds to 6.8 nm. High-resolution ToF-MEIS analysis indicates W at the surface, suggesting its incorporation into the B-O layer. The effective sputtering yield (Y$_{eff}$) agrees within a few percent with [1–2], despite differences in layer structure and a two-orders-of-magnitude variation in ion flux.

Porous B films were exposed to E$_{ion}$=43-223 eV and $\Phi$=7.0e23-2.8e24 m$^{-2}$. Y$_{eff}$ of all layers, notwithstanding differing morphologies, is generally within 20% of SDTrimSP results for E$_{ion}$>43 eV, while at 43 eV it reaches up to four times higher, possibly due to ion-assisted chemical erosion at 550 K. The limited sensitivity of Y$_{eff}$ to film morphology may, for instance, have important implications for plasma-cleaning of B-based re/co-deposits on ITER first-wall optical diagnostic mirrors. Although the structure and porosity of re/co-deposits may vary depending on mirror location in the vessel, their erosion rates are expected to be similar, potentially enabling a single plasma-cleaning strategy for all optical diagnostics.

[1] M.Sackers, et al., Nucl. Mater. Energy 45(2025)102003
[2] E.Hechtl, et al., J. Nucl. Mater. 196-198(1992)713-6

Speaker: Andrea Uccello (Consiglio Nazionale delle Ricerche, Istituto per la Scienza e Tecnologia dei Plasmi (CNR-ISTP))

180 2.048 Experimental evidence of the loss of boronization effectiveness by plasma redeposition in WEST

Boron (B) layers deposited on the PFCs erode quickly off of surfaces that receive high incident ion fluxes such as divertor targets, RF antenna protection limiters, and startup limiters. However, main wall surfaces away from the ion flux areas or in areas shadowed by other components may sustain their layers for significantly longer. To study the short time scale erosion and migration of B around WEST, samples were installed on the reciprocating collector probe (RCP) and exposed in the far-SOL during two dedicated experiments in the spring of 2024 and spring of 2025. The samples collected from the RCP after exposures with and without the use of isotopically enriched B-10 powder in the impurity powder dropper (IPD) both have substantial B layers deposited on them suggesting that transport through the SOL is a major contributor to material buildup in addition to local erosion and redeposition.
To determine campaign integrated deposit thickness and infer layer growth rates in net deposition zones, two sets of samples were installed around the main chamber of WEST on the central solenoid and the outboard side main wall during the same campaigns as the RCP experiments. One set was exposed to only glow discharge boronization (GDB) while the other set was exposed to GDBs plus IPD injections to compare the migration of B from the two techniques. In the samples collected from the GDB-only campaign a factor of two asymmetry in the total B density, a proxy for layer thickness, was observed on samples installed roughly 150° toroidally apart. Analysis of the deposited layer’s structure shows discrete layers where undisturbed B from GDB and mixed plasma modified layers from deposition in the main chamber can be differentiated. Samples collected from the IPD experiment show a similar long-range asymmetry in total B and an enhanced isotope ratio near the lower inner target, as expected based on prior modelling of B transport from the IPD [1]. It is proposed that far from the divertor the modification of B layers by ion bombardment is low and the primary evolution occurs from charge exchange neutrals that either erode or deposit onto and cover the B layers depending on their energy. Interpretive modelling supporting the hypothesis that undisturbed boronization layers are being capped with redeposits during operation is also presented.

Work supported by U.S.D.O.E. under grant number DE-SC0020414

[1] K. Afonin et al., Nuclear Fusion, vol. 63, no. 12, Dec 2023.

Speaker: Sean Kosslow (University of Tennessee - Knoxville)

181 2.049 Study on divertor tungsten erosion during inter-ELM and intra-ELM phases of EAST H-mode plasmas with resonant magnetic perturbations

Divertor tungsten (W) erosion during inter-ELM and intra-ELM phases in EAST H-mode plasmas is investigated experimentally and well reproduced via simulations, with a focus on the impact of resonant magnetic perturbations (RMPs). Experimental results indicate that the ELM frequency (f_ELM) increases with the application of RMPs, and the W erosion source per ELM cycle decreases during both inter-ELM and intra-ELM phases. The time-averaged W erosion flux decreases in the inter-ELM phase but increases in the intra-ELM phase. As a result, with the application of RMPs, the total time-averaged W erosion flux shows slight increases when f_ELM<150 Hz while slightly decreases when f_ELM>150 Hz. The ERO code and Free-Streaming Model (FSM) [1] are used to reproduce the experimental data and elucidate the underlying mechanisms. In the inter-ELM phase, the low divertor plasma temperature makes the average incident energy of deuterium (D) below the W sputtering threshold, leaving carbon (C) impurities as the primary cause of erosion. Conversely, during the intra-ELM phase, the energetic D+ originated from the pedestal region dominates the W erosion. ERO results further reveal that the inter-ELM W gross erosion rate decreases with applying RMPs, mainly due to the reduced particle flux near the strike point and the enhanced C deposition on the W surface. Regarding the intra-ELM phase, FSM simulations indicate that the W gross erosion rate also decreases under RMP applications, which is caused by the reduction of C fraction in the plasma. Therefore, although the dominant sputtering species vary in different phases, the RMP-induced erosion mitigation is dominated by carbon impurities throughout the entire ELM cycle. These findings advance the understanding of RMP effects on impurity behavior and tungsten erosion.
References
[1] D. Moulton, et al., “Quasineutral plasma expansion into infinite vacuum as a model for parallel ELM transport.” Plasma Physics & Controlled Fusion, 2013, 55: 275-284.

Speaker: Yiqin Zhu (SWJTU)

182 2.051 Helium Topology Controlled Erosion of Tungsten Under High velocity Dust Impacts: Molecular Dynamics Insights

Tungsten (W) plasma-facing components in magnetic-confinement fusion are simultaneously exposed to intense helium (He) implantation and to high-velocity dust remobilization events. While each driver is known to degrade W surfaces, their combined impact remains insufficiently constrained because He implantation generates a spectrum of subsurface defect morphologies (e.g., bubbles, bubble arrays, and platelets) that can fundamentally alter how impact energy is dissipated and how material is removed. Here, we use large-scale molecular dynamics simulations (up 300 million particles) to isolate how He defect topology governs impact induced cratering and ejecta production in fusion relevant tungsten.

We simulate normal incidence high velocity impacts of a nanoscale W projectile onto single crystal W targets containing representative near-surface He morphologies: (i) pristine W, (ii) sparse bubble arrays, (iii) dense bubble arrays, (iv) a population of isolated bubbles with varied sizes and depths, and (v) a He platelet. By tracking crater evolution, cumulative ejected mass and fragment statistics, helium redistribution, and depth resolved dislocation activity, we identify distinct deformation and failure regimes that are controlled primarily by topology rather than He inventory alone.

In all cases, an early shock dominated stage is followed by a slower relaxation stage. However, the late time response diverges strongly with He morphology, yielding a robust hierarchy in retained crater volume and ejecta: pristine < bubble arrays < isolated bubbles ≪ platelet. Bubble arrays act as a mechanically compliant, gas-venting layer that promotes localized plastic accommodation at the implanted depth, sustaining steady production of fine debris without catastrophic opening. Isolated bubbles produce intermittent, spatially localized micro-collapse events that lead to step-like crater and ejecta evolution. In contrast, the platelet morphology triggers a qualitatively different failure mode: progressive decohesion culminating in delayed delamination and a sudden, large burst of ejecta, consistent with opening-dominated damage.

Ejecta kinetics for all non platelet cases are well described by a stretched exponential law, whereas the platelet requires an additional discrete “burst” term to capture delamination driven mass loss. Dislocation density signatures (a narrow spike for arrays versus a broader near surface elevation for platelets) provide a mechanistic diagnostic linking subsurface He topology to macroscopic erosion outcomes. These results support helium aware erosion and lifetime models for W plasma facing materials and identify He platelet as a critical topology to mitigate.

Keywords: tungsten; plasma-facing materials; helium bubbles; helium platelets; dust impact; molecular dynamics.

Speaker: Dr Prashant Dwivedi (Czech Technical University in Prague)

183 2.052 ERO2.0 simulations of physical and chemical erosion of boron nitride-boron pebble targets in PISCES-A

We present ERO2.0 simulations that reproduce the experimentally observed similarity in erosion rates between boron nitride–boron (B-BN) pebbles and solid boron (B). Low-Z plasma-facing materials like B are attractive for fusion reactors due to their low core plasma radiation compared to high-Z materials such as tungsten. However, pure B erodes and evaporates rapidly, limiting component lifetime and causing significant tritium co-deposition. A recent solution embeds B pebbles in a boron nitride matrix [1], forming sacrificial B-BN rods extruded into the divertor. The pebbles detach under plasma loading to be transported away for reconditioning and re-extruded in a closed loop. This renewable divertor concept is being considered for Thea Energy's planar-coil stellarator [2]. Laser-heating experiments simulating high heat fluxes showed acceptable B-BN recession rates. In PISCES-A linear plasma tests, B-BN and solid B samples exhibited similar erosion rates, but B-BN demonstrated a two-order-of-magnitude reduction in deuterium retention.
We apply the ERO2.0 impurity transport and erosion code [3] to these PISCES-A experiments to validate the code and prepare for extrapolation to reactor conditions. Additional comparisons are made with B exposure experiments at DIII-D [4], PSI-2 linear plasmas [5], and ion-beam data. The simulations qualitatively confirm the experimental observation that B-BN and solid B experience similar net erosion. The more oblique incidence of deuterium projectiles on the curved pebble surfaces increases gross erosion, which is compensated by enhanced redeposition of sputtered B within the pebble structure. Simulations also predict a modest shift in the angular distribution of sputtered B toward more oblique angles for the pebble geometry. For quantitative comparison with experimental erosion rates, we examine sputtering yields from the SDTrimSP code [6]. Initial simulations underestimated erosion by about a factor of four, but agreement improves significantly by adopting a lower surface binding energy for B. We also discuss experimental uncertainties, including the deuterium ion energy distribution and factors affecting sputtering yields, such as nitrogen from the BN binder. Finally, we assess the contribution of chemical erosion and compare it with BD A–X molecular band emission measured during the PISCES-A experiments.
[1] E. Martinez-Loran et al., Nucl. Mater. Energy 2025
[2] D.A. Gates et al., Nuclear Fusion 2025
[3] J. Romazanov et al., Phys. Scripta 2017
[4] A. Ottaviano et al., presented at APS-DPP 2025
[5] M. Sackers et al., Nucl. Mater. Energy 2025
[6] A. Mutzke et al., IPP Report 2024-06

Speaker: Juri Romazanov (FZJ)

184 2.053 MD simulation of the sputtering effect and surface damage of low-energy He irradiation on WTaCrV refractory high entropy alloy

Tungsten (W) has long been regarded as the primary candidate for plasma-facing wall materials in fusion reactors. Nevertheless, W suffers from several drawbacks, including fuzz formation and the generation of high‑Z impurities that contaminate the plasma core. In recent years, refractory high‑entropy alloys (RHEAs) have attracted increasing attention due to their superior mechanical properties and enhanced irradiation resistance. However, the effects of helium on RHEAs remain insufficiently understood.
In this study, we employed molecular dynamics (MD) simulations to investigate the sputtering behavior of low‑energy He ions on WTaCrV. An empirical interatomic potential developed by Xiong et al. was adopted to describe metal–helium interactions. The influence of incident energy, incident angle, temperature, crystalline orientation, and elemental composition was systematically examined, followed by analyses of sputtering yield, stopping range, and helium clustering.
The results reveal that the atomic yield of WTaCrV exceeds that of pure W, with preferential sputtering observed for low‑Z elements such as Cr and V. The angular distribution of sputtered atom velocities was also characterized. Surface binding energy (SBE) calculations indicate that, within WTaCrV, the SBEs of V and Cr are significantly elevated compared to their pure elemental states due to chemical complexity, though they remain lower than those of W and Ta. Orientation dependence was evident, with the {111} surface exhibiting the lowest yield and the {110} surface the highest. Regarding stopping range, the {111} surface showed the deepest He penetration, attributed to channeling effects. Overall, the average penetration depth of He atoms in WTaCrV is shorter than in W, a consequence of pronounced lattice disorder that suppresses channeling, thereby concentrating helium effects in the near‑surface region. Furthermore, helium clusters in WTaCrV were found to be more reduced and discretely distributed, which can be explained by distinct diffusion processes of He atoms. The transport pathways of He were strongly influenced by the surrounding atomic environment. Migration trajectories, examined using the nudged elastic band (NEB) method, suggest a broadened distribution of migration barrier energies.
Finally, more literature modeling work related to the W and alloys sputtering by hydrogen, He, and impurities gas were compared and discussed.

Speaker: Ze Chen (City University of Hong Kong)

185 2.054 Chemical Erosion and Deuterium Retention of Boron Thin Films using in-situ Ion Beam Analysis in Upgraded Pilot-PSI

The recent decision from ITER to replace the beryllium first wall with tungsten requires boronization on the first wall to capture sufficient oxygen and minimise plasma cooling caused by high-Z plasma material interactions. The deposited layer thickness and frequency of boronizations are critical, since redeposition in remote locations leads to thick deposits, potentially causing flaking and enhanced tritium retention. These lead to concerns for safe machine operation via dust production from the deposits and tritium accumulation within them.

Upgraded Pilot-PSI (UPP) experiments were carried out on thin (0.8 $\mu$m) boron films pre-deposited on Si substrates by Pulsed Laser Deposition (PLD). The experiments aimed to distinguish chemical and physical erosion by measuring erosion rates at ion impact energies between 2.26 and 72.26 eV, both below and above the physical sputtering threshold. ITER-relevant high-flux deuterium plasmas were used to load the targets at surface temperatures of 400 K, while in-situ ion-beam analysis was used to determine the erosion via Elastic Backscattering Spectrometry (EBS) and deuterium retention via Nuclear Reaction Analysis (NRA), allowing contamination-free, time-resolved measurements in between plasma exposure. In-situ ion-beam analysis was performed using a 1.5 MeV H beam for Elastic Backscattering Spectrometry (EBS) and a 2.6 MeV $^3$He beam optimised for boron and deuterium thickness determination. During plasma exposure, Thomson Scattering (TS) measured deuterium plasma conditions of $T_e = 0.93$ eV and $n_e = 0.16\times 10^{20}$ m$^{-3}$, corresponding to a deuterium flux of $8.61\times 10^{22}$ D/m$^2$. These parameters provide ITER-relevant high-flux conditions for quantifying chemical versus physical erosion mechanisms and investigating deuterium retention.

The erosion experiments revealed an erosion yield of $(9.22 \pm 1.04)\times 10^{-4}\,\text{atoms/ion}$ at $E_{\text{ion}} = 2.26\pm0.06$ eV. Deuterium retention ($\phi$) in the boron layer was measured after 600 seconds of low impact energy ($E_{\text{ion}} = 2.26\pm0.06$ eV) deuterium plasma exposure, remained below $\phi=15\times10^{15} \text{ atoms/m}^2$ and exhibited a power-law time (t) dependence described by $\phi = 2.04\, \times \text{t}^{0.28}$. Additional outgassing measurements showed a significant (approximately 40$\%$) long-term outgassing after 60 hours in vacuum. These results indicate that substantial chemical erosion occurs below the physical sputtering threshold energy and needs to be considered in order to accurately extrapolate the expected boron layer lifetime and boronization frequency for future fusion reactors. Conversely, the deuterium retention in the boron is low, likely limited by the chemical erosion. A broader discussion of the energy and temperature dependence of the erosion and retention processes will be presented.

Speaker: Jort van Kesteren (DIFFER)

186 2.055 First-in-kind plasma transport modeling of EC wall-conditioning of JT-60SA with EMC3-EIRENE code

Wall conditioning is an essential technique to control particle recycling on the wall surfaces and to realize stable discharges of fusion plasma. In recent large devices, superconducting coils are installed for high performance and long discharges, and Wall Conditioning with Electron Cyclotron resonance (ECWC) is planned as an inter-shot wall conditioning to avoid frequent de-energization of magnet system. ECWC plasma is produced by EC resonance and tends to localize, and therefore control of poloidal field is needed to extend the plasma toward the wall surface for conditioning.
An ECWC plasma usually has no conferment region and has strong plasma-neutral interactions. These physical conditions are similar to divertor plasma, and therefore a divertor plasma transport code EMC3-EIRENE was employed. This code can use a general mesh structure apart from magnetic flux surfaces. A simple cartesian-type mesh at a reference toroidal position was used, and the meshes at different toroidal positions were made by field-tracing in the range from -10 to 10 degrees, i.e., 1/18 of a torus with periodic condition. The first wall shape of JT-60SA of Operation Phase 1 was used, and two discharge conditions were chosen: discharge numbers E101164 with O1-mode EC and E101166 with X2-mode. The heating distributions were assumed to be vertically elongated profile at the resonance position near the inboard first wall for O1-mode and localized two-point regions at the resonance along the trace of the EC wave including a reflection for X2 mode. The heating power were chosen as 800 kW and 180 kW estimated from the experiment conditions, respectively.
Our first-in-kind modeling with the three-dimensional code has successfully generated EC plasma with the magnetic field and the first wall of the real device. The preliminary results showed a qualitative agreement with 2D image observations in experiments: 1) the radiation region with O1-mode tends to elongate vertically near the first wall corresponding to the heating region, 2) the radiation region with X2-mode tends to localize in a certain magnetic field configuration, and 3) a magnetic field configuration with a simple horizontal poloidal field component gives wide plasma covering almost all the device for O1-mode but no wide plasma for X2-mode. However, our results suggested that the radiation distribution significantly depends on electron temperature and density affected by a discharge condition, and therefore direct comparison with experiment results are needed to validate the results and will be addressed as future issues.

Speaker: Gakushi Kawamura (National Institutes for Quantum Science and Technology)

187 2.056 Chemical and physical properties of boron layers produced during non-uniform glow discharge boronization in the WEST tokamak

Following ITER re-baseline and the replacement of beryllium by tungsten (W), there is a need for assessing the efficacy of diborane glow discharge boronization systems [1,2]. Such systems are regularly used in full-W tokamaks such as ASDEX upgrade and WEST as standard conditioning method.
To support ITER research into fuel retention and lifetime of boron layers, a non-uniform boronization was performed in WEST in 2025, using all six anodes and only three of the six B₂D₆ gas inlets, leading to a toroidally non-uniform distribution of the ionization rate of the glow discharge. Two mobile probes on the upper part of the main chamber were used to expose samples during the boronization. The probes were located at two toroidal positions: on sector Q3A near an active gas inlet and on sector Q4B near an inactive gas inlet. Each probe was carrying 21 samples distributed over 5 vertical positions (level#1 to #5, level#1 being the upper part of the probe and level#5 the bottom) on its four faces.
Here, we present the physical and chemical characterization of boron layers deposited on W samples located on level#2 and level#5 of the Q3A and Q4B probes. First, RBS/NRA measurements were performed to quantify the boron (B), deuterium (D) and oxygen (O) contents. Results show that B amount is higher at level#5 than at level#2 and that the D/B ratio is larger at the Q3A location, and that the O/B ratio is higher at Q4B.
Confocal microscopy and electron microscopy techniques on FIB-cut cross sections and lamellae were then performed to analyse the roughness, thickness and nanostructure of the boron layers. The latter are found to be amorphous and dense with no crack or large porosities. The boron layers at the bottom of the probes are thicker than the ones on the upper part (110-170 nm at level#5 vs 50nm at level#2 for Q3A). Moreover, the layers at Q4B#5 are thicker (200-250 nm) than those on Q3B#5, which is located near an active gas inlet. A 10 nm native oxide (WO₃) layer was detected on the W surface, affecting O/B quantification in thin boron layers themselves. By combining RBS/NRA data with the structural characterisation, the density and chemical content of the boron layers will be discussed in relation with their distance and orientation on the probe faces with respect to the anode and the active/inactive gas inlets.

[1] https://doi.org/10.1016/j.nme.2024.101854
[2] https://doi.org/10.1016/j.nme.2025.101891

Speaker: Dr Céline MARTIN (Aix Marseille Univ, CNRS, PIIM, UMR 7345)

188 2.057 Analysis of Solid Boron Injection for ITER

Following the decision to begin ITER operation with a full tungsten first wall and divertor, it became necessary for ITER to also include boron wall conditioning techniques to avoid excessive influx of tungsten into the plasma that would radiate away the plasma stored energy and make the path to long pulse discharges up to Q=10 quite challenging. As a result, Glow Discharge Boronization (GDB) with deuterated diborane (B2D6) gas has been adopted on ITER, but it can only be performed while the magnetic fields are off, which limits the frequency of boronization and may complicate plasma operation. Another method that is being considered by the ITER Organization that would allow the introduction of pure boron and could be performed on demand as needed during plasma operation is solid boron injection (SBI) either through pellet injection or with an Impurity Powder Dropper (IPD). This additional boron wall conditioning technique could become important as risk mitigation for obtaining and sustaining high performance plasma operation on ITER.
To help ITER assess this technique, modeling of the ablation, transport, erosion and redeposition of solid boron has been performed using the Dust Injection Simulator (DIS), EMC3-EIRENE, Walldyn, and Neutral Gas Shielding models for a range of boron particle sizes and injection velocities. Some of the main questions this modeling addresses are 1) how deep should boron penetrate into the plasma to condition the plasma facing components (PFCs) sufficiently, 2) where does the boron end up on the first wall and divertor, 3) is a single injection location sufficient to condition the PFCs, and 4) which plasma conditions are suitable for SBI? The DIS modeling scans boron particle sizes from 1 µm – 1.5 mm and injection velocities from 5 – 50 m/s and NGS modeling scans sizes from 500 µm – 5 mm and injection velocities from 20 – 200 m/s, assuming either dropping powder or injecting pellets from a single upper port or injecting pellets along the existing lower outboard, lower inboard, or inboard midplane pellet guide tubes. Initial results with DIS and EMC3-EIRENE show that particles > 100 µm can cross the separatrix and penetration increases with size and velocity in a low power L-mode scenario. For a Q=10 scenario, particles > 500 µm can cross the separatrix and the NGS model allows penetration of 4 – 8.5 cm for the larger and higher velocity pellets.

Speaker: Joseph Snipes (Princeton Plasma Physics Laboratory)

189 2.060 Characterization of boron layers on tungsten substrates by picosecond laser-induced breakdown spectroscopy

Boronization is used in present-devices as wall conditioning technique to getter oxygen, reduce other intrinsic impurities in the vessel, and control the hydrogen recycling as well as improving plasma performance. The technique is also foreseen as reference wall conditioning method for the new ITER baseline with full-tungsten (W) wall. However, the effectiveness of the deposited Boron (B) layer thickness and its homogeneity after the boronization process is uncertain as well as the exact knowledge about the interlink of boron layer lifetime with improved wall conditions.
In this study, an approach of the picosecond-laser-induced breakdown spectroscopy (ps-LIBS) is investigated to analyze the depth distribution of B-layers on W-substrates in a vacuum environment mimicking in-vessel deposited B-layers in devices like ITER. At first, based on the dynamic behaviour of the LIBS-plasma in a vacuum, the appropriate spectral lines and acquisition settings for B and W were determined. Sequentially, the depth distribution of two types of B-films on W-substrates with the thicknesses of 130 nm and 260 nm were measured under different laser spot sizes (diameter:142-1518 μm)[1]. The measured average ablation rate of ps-LIBS shows a notable decrease with increasing laser spot size. The spectral lines of B II and W I exhibit distinct intensity distributions under different spot sizes. The interface between B-films and W-substrates, as well as the thickness of the B-films, were determined using the normalized intensity and intensity ratio method, respectively. The results from ps-LIBS measurements regarding the depth are in good agreement with those obtained through the FIB-SEM and EDS. Finally, the B-layers with a range of different thickness (10-150 nm) on the W-substrates were used to establish quantitative curve and obtain the limit of detection [2]. This calibration or quantitation is consistent with results of LIBS detection on ultrathin B-layers of two W samples exposed in W7-X during boronization.
These initial findings demonstrate the feasibility of using ps-LIBS to characterize, in situ within fusion devices, the thickness and uniformity of thin boron films on W substrates down to the 100-nm scale and below [3].
[1] H. Wu, R. Yi, A. Houben, et al., Nucl. Mater. Energy 101812, 41 (2024)
[2] H. Wu, R. Yi, S. Brezinsek, et al., Nucl. Mater. Energy 102018, 45 (2025)
[3] S. Mittelmann, J. Oelmann, S. Brezinsek, et al., Appl. Phys. A-Mater. Sci. Process 672, 126 (2020)

Speaker: Huace Wu (Forschungszentrum Jülich)

190 2.061 Molecular Dynamics Investigation of Carbon-Based Hydrogen Recycling Control in Plasma-Facing Walls

Carbon-based plasma-facing materials (PFMs) have been regarded as attractive candidates for fusion reactors owing to their high compatibility with plasma performance and favorable thermal properties. However, concerns over tritium retention and material activation have led to the adoption of metallic PFMs in recent devices. In contrast, experiments on the QUEST device identified a hydrogen transport barrier at the interface between carbon-containing redeposition layers and the metallic substrate. This finding suggests that active surface control of metallic PFMs could enable partial utilization of carbon’s advantageous properties while maintaining acceptable activation levels. Building on this insight, Hanada et al. proposed a new operational concept in which a small amount of carbon is intentionally introduced into the plasma (“carbon doping”) to form a carbon-containing redeposited layer over the entire plasma-facing surface, while a low-temperature region (<150 °C) serves as a “carbon pump” that preferentially collects carbon. This system aims to regulate the in-vessel carbon inventory and suppress hydrogen isotope retention below critical limits. To develop such a system, it is essential to understand the microscopic behavior of hydrogen within carbon-containing redeposited layers, including its bonding states, diffusion pathways, and molecular hydrogen formation. Recent first-principles molecular dynamics (FPMD) studies by Kusaba et al. have shown that H₂ formation in hydrogenated amorphous carbon (a-C:H) is strongly governed by the saturation of C–H bonds, with molecular hydrogen emerging primarily in voids. Motivated by these findings, the present study investigates the structural evolution and hydrogen retention characteristics of carbon-doped redeposited layers through atomistic simulations, systematically varying carbon density, hydrogen concentration, and temperature. Special attention is given to C–H bond saturation behavior, the kinetics of H₂ formation, the development of nanoscale voids on metallic substrates, and carbon re-condensation behavior on low-temperature surfaces. These analyses clarify the fundamental physical mechanisms by which the carbon-pump concept suppresses hydrogen isotope retention.
The outcomes of this study establish a scientifically grounded hydrogen-recycling model tailored to metallic-wall fusion systems and provide a quantitative basis for optimizing integrated operation scenarios that combine carbon doping with active carbon pumping. The results contribute to the design of next-generation plasma-facing components that leverage controlled carbon behavior while maintaining low activation and safe tritium handling.

Speaker: Prof. Hiroaki Nakamura (National Institute for Fusion Science)

191 2.062 Enhancement of Surface Finish and Porosity Reduction in Graphite Plasma-Facing Components of the ADITYA-U Tokamak through Wall Conditioning

The main drawback of graphite as a plasma-facing component (PFC) material is its high retention of fuel gases, such as hydrogen and its isotopes deuterium and tritium, during prolonged wall conditioning, tokamak plasma operation, and fusion experiments. Consequently, graphite is not universally acceptable for fusion-grade plasmas due to its tritium retention characteristics. Nevertheless, graphite PFCs are widely used in many tokamaks for the generation of low-Z, well-confined high-temperature plasmas. Therefore, studies on graphite PFCs and their behavior under tokamak plasma operation and wall conditioning are crucial for fusion research.
In the ADITYA-U tokamak, the plasma-facing components (PFCs), consisting of graphite tiles, cover more than 15% of the in-vessel surface area along with the stainless steel vessel wall and other in-vessel components. These PFCs have been installed and operated as toroidal inner belt limiters, outer poloidal limiters, and safety limiters since the commencement of ADITYA-U tokamak operations in 2017 [1]. The PFC samples were collected from the inner belt limiters and had been exposed to various plasma conditions, including tokamak plasma discharges, as well as wall conditioning methods such as glow discharge cleaning (GDC). Among the different wall conditioning techniques, H₂-GDC and Ar–H₂ mixed-fuel GDC were most frequently employed during this period for cleaning the vessel wall and the PFC samples [2]. Lithium coating under the influence of H₂-GDC was also carried out to control carbon and oxygen impurities [3].
Hydrogen fuel retention and its effects have also been studied during ADITYA-U plasma discharges, showing high retention in freshly installed PFCs and periodic reduction through wall conditioning methods developed for ADITYA-U. The collected PFC samples, along with a reference graphite sample of the same grade, were characterized using X-ray diffraction (XRD), scanning electron microscopy (SEM), and Raman spectroscopy. The properties of the PFC samples were compared with those of the reference graphite. These characterization studies provide valuable insights into changes in graphite porosity reduction, hydrogen retention reduction, and surface modifications leading to improved surface finishing. The results demonstrating surface enhancement of graphite plasma-facing components in the ADITYA-U tokamak are of significant importance for the development of treated graphite PFCs suitable for fusion-grade plasma applications.
References:
[1] “Design..limiter..Tokamak” K.M. Patel, et al., Fusion Engineering and Design Volume 222, January 2026, 115520
[2] “Plasma performance ... argon–hydrogen..ADITYA-U tokamak”, K.A. Jadeja, et al. Nuclear Fusion 64 (2024)106048
[3] “Lithium wall ... fuel control”, K. A. Jadeja et al. Nucl. Fusion 62 (2022),016003

Speaker: Dr Kumarpalsinh Jadeja (Institute for Plasma Research, Gandhinagar-382428, India)

192 2.063 Deuterium Implantation and Thermal Extraction in the Actively Pumped Open-Surface Lithium LOop (APOLLO)

Liquid lithium plasma-facing components (PFCs) present several advantages for fusion applications, including enhanced plasma performance, protection of underlying structural materials, and mitigation of transient melting phenomena associated with solid PFCs. As a low atomic number (low-Z) material, lithium exhibits strong gettering capabilities and enables operation in a low-recycling plasma regime. However, the absorption of hydrogenic species by lithium poses a significant challenge for future fusion reactors, where stringent limits on tritium inventory must be maintained.

To address this issue, the University of Illinois Urbana-Champaign (UIUC), in collaboration with Tokamak Energy Ltd., has developed the Actively Pumped Open-Surface Lithium Loop (APOLLO). This experimental platform comprises a circulating liquid lithium loop, a free-surface lithium PFC operating within a magnetic field, a deuterium plasma source or electron beam heating system, and a distillation column designed for the extraction of hydrogenic species. The PFC incorporates a computationally optimized distributor that uniformly delivers lithium from an inlet pipe across a 7.5 cm-wide, additively manufactured refractory metal ordered mesh situated within a free-surface flow channel. Lithium flows across the mesh with average surface velocities of up to 10 cm/s and mass flow rates reaching 12 g/s, while being simultaneously exposed to an electron cyclotron resonance (ECR) hydrogen/deuterium plasma source.

Plasma characteristics are diagnosed using an array of 16 Langmuir probes, a retarding field energy analyzer (RFEA), and actinometric spectroscopy. After exiting the PFC via a collector, the lithium is transported to the inductively heated Hydrogen Distillation Experiment (HyDE), where it undergoes thermal treatment at temperatures up to 700 °C to remove hydrogenic species and other impurities. Deuterium uptake in flowing liquid lithium exposed to an ECR plasma is investigated as a function of lithium flow rate and plasma operating conditions. The efficiency of deuterium thermal extraction is quantified using a resistive impurity probe and thermal desorption analysis. A zero-dimensional (0D) model of the APOLLO system is employed to interpret and contextualize the experimental results.

Speaker: Braden Moore (University of Illinois Urbana-Champaign)

193 2.065 EMC3-Eirene simulation of neutral source effects on density build-up in the W7-X island divertor

Achieving high plasma density near the divertor target in magnetically confined fusion devices is beneficial for two main operational objectives: firstly, the high densities lead to high neutral densities near the pumping duct, allowing for efficient particle exhaust. Secondly, it allows for significant radiation (heat exhaust) in the SOL with minimal impurity concentration ($P_{rad}∝n_e^2 c_{imp}$). In W7-X, the island divertor is being tested as a possible exhaust solution for a future stellarator reactor. The divertor in its current form is an open divertor. It is optimized for high configuration flexibility rather than for any specific divertor configuration. Therefore, its performance may not be optimal in every configuration.
According to the stellarator two-point model [2], larger field-line pitch increases the weight of parallel transport, which is hypothesized to have a beneficial effect on the density build-up. To verify this, three different island divertor configurations (low-iota, standard and high-iota) in W7-X, with different field-line pitches ($Θ_{i,avg}=.002,.0038,.0073$ respectively) were simulated with EMC3-Eirene. While the low-iota configuration exhibited the expected behavior with field-line pitch, showing the poorest density build-up, no changes in achievable downstream densities were observed when the field-line pitch was further increased by a factor of about two from the standard to the high-iota configuration.
The reason for this deviation from the expected scaling was determined to be the close proximity of the X-point to a region of a vertical divertor plate, which introduces a limiter-like component into the high-iota divertor configuration in that the neutral particles recycled there can easily penetrate the confined region. Although the vertical target only receives approximately 10% of the total recycling flux, the deeper location of the ionization source leads to convective transport, reducing the maximum achievable density in the divertor to a degree that compensates for the field-line pitch effect. When the vertical target was removed from the simulation, thus removing this core source of particles (and therefore more closely matching the assumptions made in two-point models), the expected scaling with the field line pitch was recovered. This work highlights the importance of the island’s neutral screening efficiency on the density build-up capability of the island divertor and possible consequences of external gas puffing in the private flux region in the divertor region. For example, purposefully injecting gas near the X-point, as an attempt to induce X-point radiators, may be detrimental in the island divertor in terms of pumping.

Speaker: Victoria Winters (University of Greifswald/Max Planck Institute for Plasma Physics)

194 2.066 Impact of Lyman opacity on divertor plasma and neutral conditions in DIII-D

Standalone EIRENE simulations of neutral deuterium emission across the DIII-D low-field side (LFS) divertor target accurately predict the measured Lyman and Balmer series radiative emission in both attached and detached divertor conditions when constrained by the measured 2D Divertor Thomson Scattering (DTS) electron temperature (Te) and (ne), and Langmuir probe (LP) ion current density. A rigorous description of the spatial distribution and absolute densities of hydrogenic neutrals, and validation of plasma radiation predicted by scrape-off layer (SOL) codes is critical for the design of future fusion devices. Hence, for the measured ion flux to the divertor target and the plasma conditions, and disregarding edge-localized modes and other transients, the combination of the plasma recycling and neutral transport models in the neutral Monte-Carlo code EIRENE, and published atomic and molecular rates tabulated in the AMJUEL data repository, reproduces the measured neutral deuterium emission within the uncertainty of the measurements and simulation setup. These studies separate the validation of neutral particle transport, and atomic and molecular physics, from significantly more uncertain plasma transport in plasma-neutral coupled SOL codes, such as EDGE2D-EIRENE and SOLPS-ITER.

In detached conditions, inclusion of photon opacity of Lyman-alpha and Lyman-beta line emission in the EIRENE simulation further improved the agreement with the measured Lyman-alpha and Balmer-alpha emission well within the uncertainties of the measurements. In contrast, EIRENE executed on plasmas predicted by EDGE2D-EIRENE overpredict Lyman-alpha emission by up to a factor of 8, and underpredict the Balmer-alpha and the Balmer-gamma emission by factors of 2 and 10, respectively. Furthermore, EDGE2D-EIRENE predicts an order of magnitude lower molecular and atomic densities compared to EIRENE when constrained by DTS and LP measurements, implying that the momentum losses due to plasma-neutral collisions are significantly underpredicted in EDGE2D-EIRENE.

SOLPS-ITER simulations of DIII-D low and high-confinement plasmas, including cross-field drifts and Lyman line emission photon opacity are presented to assess the impact of photon opacity on the predicted 2D plasma distribution, and Lyman and Balmer emission intensity in the LFS divertor. To assess the fidelity of the updated neutral gas and radiation models in SOLPS-ITER, the predicted 2D poloidal profiles of ne and Te are compared to the measured profiles.

*See the author list of C.T. Holcomb et al., Nuclear Fusion 64 (2024) 112003.

Speaker: Mathias Groth (Aalto University, Espoo, Finland)

195 2.067 Development of multi-diagnostic Bayesian analysis tool for inference of plasma parameters in the W7-X island divertor

The W7-X stellarator is an experimental device to study the reactor relevance of this concept. Its complex three-dimensional geometry makes full plasma characterization challenging, particularly in the scrape-off layer (SOL) where plasma behaviour can be highly localized. Understanding transport and physics in this region is essential for reactor design. Therefore, a Bayesian inference framework is implemented for W7-X which enables the integrated reconstruction of key plasma parameters such as electron density and temperature from a limited diagnostic coverage.

This contribution presents the first steps in the implementation of the framework and its adaptation to W7-X geometry. Currently the work is focused on specific poloidal planes and so only 2D results will be shown. The framework has been benchmarked by performing inversions of bolometry data with the full Bayesian approach, including Hamiltonian Monte Carlo sampling for uncertainty quantification. The obtained tomograms are compared against previous results from Gaussian Process tomography, with the methods agreeing well with each other. Furthermore, results from adding divertor spectroscopy data as prior constraints to the tomographic inversions are showcased. Also next steps which include preparing for inferring plasma parameters at the divertor site using divertor spectroscopy and the MANTIS [1] imagining system will be highlighted.

[1] A. Perek et al, "MANTIS: A real-time quantitative multispectral imaging system for fusion plasmas.", Rev. Sci. Instrum. 90, https://doi.org/10.1063/1.5115569 (2019)

Speaker: Linnéa Björk (MPPL)

196 2.068 Analysis of the secondary strike line in Wendelstein 7-X’s higher plasma beta scenarios

The divertor plays a crucial role in plasma and particle exhaust.
After heating the surrounding plasma, He ions need to be diverted through the scrape-off layer into the divertor chamber.
Here plasma neutralizes on divertor targets in high-flux regions, known as strike lines.
In Wendelstein 7-X (W7-X) this exhaust is accomplished via an island divertor, where resonant magnetic islands intersected by the divertor target plates form the scrape-off layer and private flux region.
In W7-X’s standard configuration one dominate strike line per divertor module is typically observable.
However, with increased plasma beta $\beta$, defined as the ratio between the plasma pressure and magnetic pressure $(\beta = p_{plasma}⁄(B^2⁄(2\mu_0)))$, a secondary strike line appears $\left[1\right]$.
This additional strike line may be linked to changes in divertor shadowing or to variations in perpendicular and localized transport effects.
The strike lines of experiment programs with the available range of $\beta$ were evaluated.
Their poloidal and toroidal dimensions and magnitudes were characterized.
To investigate the underlying mechanisms, a simulation of the strike lines, using EMC3-Lite $\left[2\right]$, is being performed.
The simulation is compared to experimental results, measured via the infrared and H$_\alpha$ imaging diagnostic systems $\left[3, 4\right]$.
The study examines different $\beta$ effects on the magnetic topology and key parameters, including the perpendicular diffusion coefficient, to assess their influence on secondary strike line formation and to identify the driving physical processes.

$\left[1\right]$ Y. Gao et al 2024 Nucl. Fusion 64 076060
$\left[2\right]$ Y. Feng et al 2022 Plasma Phys. Control. Fusion 64 125012
$\left[3\right]$ J. Marcin et al 2018 Review of Scientific Instruments 89 10E116
$\left[4\right]$ T. Kremeyer et al 2025 JINST 20 T04003\

Speaker: Sebastian Draeger (MPPL)

197 2.069 Investigating the main dependencies in heat-flux decay length scalings using first-principles turbulence codes

Handling and mitigating heat fluxes to plasma-facing components remains one of the primary challenges for magnetic fusion devices. In future high-field compact machines relying on high-temperature superconducting coils (e.g., SPARC [J. D. Lore 2024]), the heat-flux decay length $\lambda_q$, which characterizes the width of the Scrape-Off Layer (SOL), is predicted to be critically small according to experimental scaling laws derived from medium-sized tokamaks (Eich Law [T. Eich 2013], Scarabosio Law [A. Scarabosio 2013]). Nevertheless, extrapolation to currently unexplored conditions present large uncertainties and some studies point out a possible change of turbulence regime breaking the validity of the scaling laws in reactor-like conditions (e.g., XGC1 gyrokinetic code [C. S. Chang 2017], BOUT++ two-fluid code [Z. Y. Li 2019]). In this context, our objective is to provide a physically motivated, turbulence-simulation-based estimate of $\lambda_q$ for a high-field compact tokamak. To this end, as a first step, parameter scans based on 3D turbulence simulations in slab geometry — selected for their reduced computational cost — were carried out using the SOLEDGE3X code [H. Bufferand 2024] [R. Düll 2024] [V. Quadri 2024]. In particular, the main $B_{\mathrm{pol}}^{-1}$ dependence in the scaling laws was recovered. At fixed $B_{\mathrm{tor}}$, this behavior can be attributed to the linear increase of $\lambda_q$ with the parallel connection length $L_{\parallel}$ (proportional to the safety factor), as structures can propagate farther in the SOL due to reduced parallel damping. Additionally, the effects of the major radius $R$, toroidal field $B_{\mathrm{tor}}$, magnetic shear $s = r/q \cdot dq/dr$, and collisionality $\nu_{ei}$ were investigated. Finally, preliminary results of similar scans performed in toroidal geometry are presented, highlighting the effect of geometry on $\lambda_q$.

Speaker: Hugo Corvoysier (CEA, IRFM, F-13108 Saint Paul-lez-Durance, France)

198 2.070 Predicting MANTIS performance in W7-X: Synthetic Modeling of the Expected Operational Regime Using EMC3-EIRENE

Modern large-scale magnetic confinement fusion devices use divertors to optimize particle and heat exhaust. At the Wendelstein 7-X stellarator (W7-X), this is achieved with an Island Divertor concept, where magnetic islands guide particles and heat along field lines to dedicated target plates. Given the 3D-nature of the island divertor geometry, a large part of the island needs to be diagnosed to assess and quantify transport, which is challenging with the current set of W7-X diagnostics. To address this, the newly installed MANTIS imaging system will be combined with helium gas puffs in upcoming experimental campaigns to provide localized, targeted measurements of island plasma parameters in order to infer transport dynamics.

To support the implementation, simulation studies were conducted using the EMC3-EIRENE plasma fluid transport code. A new synthetic forward modeling tool was developed to predict the performance and operational viability of MANTIS. In addition, we incorporate data from existing W7-X diagnostics, including helium puff spread, electron temperature, and density profiles, into the analysis to provide as much data and realistic plasma parameters for the evaluation. The resulting predictions enable optimized design and deployment of the MANTIS system. This contribution presents the simulation results, performance predictions for MANTIS, and the current status of its integration into W7-X.

Speaker: Joey Louwe (MPPL)

199 2.071 Recent improvements in EMC3-EIRENE modeling for divertor detachment in burning plasmas

Particle and power exhaust remains one of the major challenges for burning plasma operation in next generation tokamaks and stellarators. Plasma detachment via impurity seeding is a possible solution which has been widely explored on present devices and which is the default approach adopted on ITER. For future devices, predictive modeling is the only tool available both for divertor design and scenario development. While 3D plasma boundary models are essential for stellarators, 2D (axisymmetric) models are typically applied for tokamaks. Nevertheless, 3D models can also be necessary for tokamaks in order to account for non-axisymmetric magnetic perturbations (for edge localized mode control), toroidally localized gas puffing or non-axisymmetric plasma-facing components (PFCs).

Recent improvements of the 3D plasma boundary model EMC3-EIRENE will be presented. In particular, an extension of the coupling between the plasma and neutral gas codes allows for a more realistic treatment of impurity recycling and pumping. Specifically, this includes reflection of impurity neutrals on baffles which has not been included up to now. Furthermore, the new version can account for additional neutral-neutral collisions (such as D-Ne, D2-Ne and Ne-Ne), and it can evaluate charge state resolved impurity fluxes to PFCs (e.g. for subsequent erosion studies). The gas puff can be feedback controlled in order to match a given impurity concentration at the separatrix or total radiated power. Furthermore, a core source for He ash can be included. Simulations for ITER under burning plasma conditions with Ne seeding show a strong asymmetry between the inner and outer divertor targets with the old code version where a sudden onset of deep detachment is found at the inner divertor target. Conversely, a more gradual onset of detachment is found with the improved code which includes transport of - and collisions between - neutral impurities below the dome, and this is found to be more in line with corresponding SOLPS-ITER simulations.

Speaker: Heinke Frerichs (University of Wisconsin - Madison)

200 2.072 Collisional-radiative analysis of atomic and molecular hydrogen emissions for electron density and electron temperature diagnostics

0ne of the major challenges on the pathway to fusion power plants is safe control of large plasma heat load to the strike points. To keep the plasma heat load below acceptable limits, fusion reactors are expected to operate in detached divertor regime. The detached divertor is facilitated by electron-ion recombination and molecular activated recombination (MAR). Although experimental studies on the MAR processes under realistic divertor plasma conditions have been limited, they have been attracting considerable attention in recent years, owing to improvements in plasma diagnostic systems and analysis techniques.

To understand contribution of MAR reactions to the plasma detachment, reaction rates of the entire reaction chains must be evaluated. Such analyses require various plasma parameters, including ro-vibrational temperatures of hydrogen molecules in ground electronic state, densities and temperatures of electrons and ions, fractions of atomic and molecular ions. Electron temperature ($T_\mathrm{e}$) and electron density ($n_\mathrm{e}$) are easily obtained than other parameters by using a Langmuir probe (LP). However, presence of molecular ions ($\mathrm{H}_{2}^{+} $ and $\mathrm{H}_{3}^{+}$) and negative ions ($\mathrm{H}^{-}$) is usually ignored in LP analyses, introducing uncertainties in the evaluated values, particularly when MARs are strongly facilitated.

In this work, we propose a diagnostic method for determining $n_\mathrm{e}$ and $T_\mathrm{e}$ based on a collisional-radiative analysis of atomic and molecular hydrogen emissions. $n_\mathrm{e}$ can be inferred from line intensity ratio of two Balmer lines, without assuming fractions of molecular and negative ions. To incorporate molecular reactions into the analysis, we employ a collisional-radiative code developed for hydrogen molecules. In addition, the population escape factor is introduced to account for the radiation trapping effect.
As a preliminary experiment, we have tested the applicability of this technique for $n_\mathrm{e}$ determination using hydrogen ionizing plasma, and good agreements have been confirmed with measurements from a Langmuir probe. Currently we are extending the technique to $T_\mathrm{e}$ determination using intensity ratio of the Balmer series and the hydrogen Fulcher-$\alpha$ band. We plan to test the applicability of the $T_\mathrm{e}$ determination for a wide range of plasma conditions.

In the presentation, we will introduce details of our method, present experimental results, and discuss applicability of the method. Impacts of the radiation trapping effect and selection of the Balmer lines on electron density determination will also be addressed.

This work is supported by JSPS KAKENHI grant number JP24K00607.

Speaker: Dr Hiroyuki Takahashi (Tohoku University)

201 2.073 Kinetic modelling of W impurity transport in SOL/divertor plasmas including W sources from first wall by IMPGYRO

Recently, W impurity transport in SOL/divertor plasmas is attracting more interest in the community as the new baseline of ITER replaces the first wall material from Be to W. In this study, we perform W transport simulations that include W erosion of the first wall in ITER using the kinetic impurity transport code IMPGYRO [1], aiming to clarify W transport processes in the SOL/divertor plasma.
Background plasma parameters for the burning plasma operation of ITER are computed by the boundary plasma transport code SOLPS-ITER [2]. To take the sputtering of W from the first wall into account, we adapt IMPGYRO to the SOLPS-ITER result obtained by the wide-grid version of SOLPS-ITER [3]. The main simulation conditions of SOLPS-ITER are: total power across the core side boundary of 100 MW, Ne seeding, and no drifts. IMPGYRO estimates the sputtered flux of W based on the local Te, Ti, and the particle flux of bombarding Ne ions towards the first wall and divertor plates. Two cases of IMPGYRO simulations, Case A: W sputtering only from divertor plates, and Case B: W sputtering only from the first wall, are carried out.
In Case A, W impurities are mainly sputtered from the outer divertor plate, while in Case B they originate from the low-field side of the first wall. The ratio of W flux crossing the core side boundary to the sputtered W flux is 0.03% for Case A and 0.09% for Case B. These results indicate that the W impurities sputtered from the first wall are more likely to accumulate into the core. In Case A, most W impurities sputtered from divertor plates stay in the vicinity of the divertor plates and are eventually re-deposited due to the friction force. To the contrary, W impurities sputtered from the first wall near the outer midplane tend to be trapped by the thermal force and transported to the separatrix by diffusion. Detailed contributions of W sputtered from each wall segment to the W flux across the separatrix and associated transport processes, as well as the impact of drifts will be presented.

[1] S. Yamoto, et al., Comput. Phys. Commun. 248 (2020) 106979.
[2] X. Bonnin, et al., Plasma Fusion Res. 11 (2016) 1403102.
[3] N. Shtyrkhunov, et al., PET2025, P2-32, Sep. 2025, Leuven, Belgium

Speaker: Dr Shohei YAMOTO (QST)

202 2.074 Divertor physics basis for the Gauss Fusion power plant conceptual design

We present a physics-grounded workflow that identifies and iteratively refines island-divertor configurations for GIGA, a first-of-a-kind stellarator fusion power plant. The approach couples magnetic field line tracing and diffusion analysis with edge modeling to inform divertor placement, sizing and geometry. This workflow adaptively maximizes divertor closure while guaranteeing a suitable heat load pattern on the dedicated surfaces.

Accordingly, we successfully generated a conceptual divertor design aimed at maximizing particle exhaust via geometrical means. Starting from vacuum field configurations, we used field-line diffusion analysis to assess strike-point localization, connection lengths, island lobe coverage and geometrical effects for multiple divertor configurations. These metrics inform divertor placement, aperture shape and sizing, as well as baffle topology to achieve high closure without compromising pumping access.

We are now advancing toward edge simulations to validate detachment performance, heat-flux distribution, and neutral exhaust behavior under reactor-relevant conditions. In parallel, our work on the development of engineering scaling laws for separatrix quantities (contribution by Matteo Moscheni) allows for identifying the suitable parameter sub-space to preferentially investigate. As additional physics is brought into the modeling, the divertor concept will continue to evolve within this adaptive, physics-driven workflow.

The methodology also provides a framework for systematic sensitivity studies, enabling the assessment of design robustness under varying operational scenarios and physics assumptions. The evolving divertor configuration is presented alongside key results from the optimization workflow, highlighting how additional physics insights influence design choices.

Speaker: Dana Douqa (Gauss Fusion)

203 2.075 Comparison of experimental impurity radiation dynamics in the scrape-off layer of W7-X with EMC3-EIRENE simulations.

The leading concept for power exhaust in stellarators is the island-divertor configuration. At Wendelstein 7-X (W7-X) the scrape-off layer (SOL) of the standard magnetic configuration is characterized by a 5 island-chain. The islands topology and their interaction with the divertor targets lead to a complex behavior of the impurity radiation structure. Moreover, the more dominant bi-normal transport compared to the SOL of tokamaks further complicates predictions. In this context, the 3D transport code EMC3-Eirene is often used to assess the SOL of W7-X. This contribution will focus on the validation of such simulations with experimental total radiated power data.
In [S. Togo et al (submitted) Plasma Phys. Control. Fusion], density scans have been simulated with EMC3-Eirene for 5 different magnetic configurations with planar-coil currents $I_{PC}=-1.5,-1,0,1,1.5$ kA, corresponding to an increasing outward radial shift of the islands. W7-X density scan-experiments with corresponding magnetic configurations are chosen and data from the 3 available total radiated power measurement diagnostics analyzed (imaging bolometer, triangular cross-section bolometry system and bean-shaped bolometry system). However, while the imaging bolometer offers poloidal and toroidal coverage of the upper divertor region, the latter two are characterized by a mainly poloidal coverage of the triangular and tear-shaped cross-sections respectively. In order to compare the simulated and experimental radiation data, synthetic diagnostic data generation tools and tomographic inversions are employed. This allows to compare the dynamics of the poloidal localization of the radiation features, for the different geometries. The simulations show a shift of the target-localized radiation towards the island O-points, when increasing the density at the separatrix. The features are gradually spread throughout the SOL and, in the $+1.5$ kA case, eventually start penetrating the separatrix for $f_{rad}\simeq1$. A similar effect has been observed in [V.R Winters et al (2024) Nucl. Fusion 64 126047] for the low-$\iota$ magnetic configuration of W7-X. Moreover, the observed broadening of the radiation features seen in the simulations is associated with a change in the density scaling of $f_{rad}$. Previous observations have already seen a discrepancy between simulations and experiments in the dependence of these trends on the radial position of the islands. For better characterization, experimental density profiles will be employed in addition to the bolometry data. Since the mentioned studies suggest a strong correlation between these dynamics and the ability to reach high $f_{rad}$, this systematic experimental validation of the trends predicted by EMC3-Eirene is expected to be of relevance to the topic.

Speaker: Kevin Andrea Siever (MPPL)

204 2.076 Surrogate boundary model development for KSTAR tungsten divertor operation based on SOLPS-ITER simulations

Plasma detachment via impurity seeding is the most common and widely explored solution for the heat exhaust problem to date. The physics of this process has been usually studied with dedicated 2D edge transport codes such as SOLPS-ITER. However, scenario development and testing of control schemes with rapid pulse design simulators on future devices requires reduced models that run on much shorter timescales, yet provide reliable predictions for key quantities such as heat flux and radiation profile. This can be achieved, for example, by extracting a scaling law from a database of higher fidelity simulations [1] or training a neural network [2][3].
Here, we report on an ongoing effort to constitute a database of SOLPS-ITER simulations for KSTAR operating with a lower tungsten divertor with the main wall armoured with carbon PFCs. The database, comprising more than 700 simulations (all from coupled fluid/kinetic neutral code runs), a representative vertical-target magnetic equilibrium of discharge #34560 ($I_p$ = 500 kA, $B_T$ = 1.8 T, and $q_{95}$ = 4.5). The power crossing the core boundary ($P_{SOL}$) in the database varies from 4 to 7 MW. Three different impurity species, nitrogen, neon, and argon, are considered and the ratio of radiative power dissipated in the SOL to PSOL ranges from 10% to 40%.
We first use the database to demonstrate that the existing semi-empirical scaling law [4], linking effective charge ($Z_{eff}$) to the line averaged electron density and total radiation power ($P_{rad,tot}$), is in promising agreement with both simulation and experimental data. This agreement validates the database and gives physical insights for the impurity leakage of different impurities. In a second step, a deep neural network has been trained to predict the 2D radiation distribution and heat flux profiles at both divertor targets from 0D input parameters, including $P_{SOL}$, $P_{rad,tot}$, impurity species, electron temperature and density at the outer midplane. The resulting surrogate model reproduces the target heat flux profiles an order of magnitude faster than a full SOLPS-ITER simulation, with a relative error below 10% on the test set. These results indicate that the proposed approach is a promising predictor of the divertor–SOL plasma environment.
References
[1] H.D. Pacher et al., J. Nucl. Mater. 463 591–595 (2015)
[2] S. Dasbach et. al., Nucl. Mater. and Energy 34 101396 (2023)
[3] Ben Zhu et. al., Phys. Plasmas 32, 062508 (2025)
[4] G. F. Matthews et al., J. Nucl. Mater. 241 450-455 (1997)

Speaker: Chanyeong Lee (Korea Advanced Institute of Science and Technology)

205 2.077 Intensity-calibrated VUV spectroscopy at UPP in attached and detached hydrogen and deuterium plasmas

Spectroscopy in the vacuum-ultraviolet (VUV) range, i.e. below 200$\,$nm, gives access to the resonant transitions of atomic (Lyman series) and molecular hydrogen (Lyman and Werner bands). In order to get insight into the relevance of molecules and their recombination in edge plasmas, intensity calibrated emission spectroscopy is a valuable diagnostic. In the VUV spectral range, however, obtaining such spectra is a complex task, since radiation standards for intensity calibration are not readily available and the spectroscopic system needs to be directly attached to the vacuum system at hand. This contribution shows the application of two flexible and “easy-to-apply” VUV spectroscopic systems to the Upgraded Pilot-PSI (UPP) device at DIFFER.

A VUV mini-spectrometer (Resonance© VS7550) gives access to the spectral range between 117 and 420$\,$nm with a resolution (FWHM) of about 0.2$\,$nm. In addition, a self-developed VUV diode system, comprising a VUV-sensitive photo diode and interference and edge filters for spectral resolution [1], is applied to obtain access to the spectral range below 113$\,$nm. Both systems are intensity-calibrated against a 1m-VUV-spectrometer at a laboratory inductively-coupled plasma (ICP) experiment, that is in-turn calibrated within a spectral range of 46$-$300$\,$nm [2]. In combination, absolute emissivities can be obtained for the H$_2$/D$_2$ Werner and Lyman bands, the H$_2$/D$_2$ dissociation continuum as well as the H/D Lyman and Balmer series (from H$_\delta$/D$_\delta$).

The systems have been applied simultaneously to the UPP device, observing the plasma close to a tungsten dummy surface. The VUV detection cones covered the entire plasma column diameter, while Thomson scattering provides $n_\mathrm{e}$ and $T_\mathrm{e}$ profiles. Varying the source power, the gas flow rate and the confining magnetic field strength, the transition from attached to detached plasmas was observed under several conditions. The Lyman series dominates the spectrum, while molecular transitions are detected with an emissivity of about one order of magnitude less. The transition from attached to detached regime (increasing $n_\mathrm{e}$, decreasing $T_\mathrm{e}$) is clearly visible in the spectra: molecular radiation is strongly reduced, while atomic radiation via the Balmer and Lyman series increases, including emergence of the Balmer continuum arising from high quantum states of H and D close to the ionization edge. The spectra hence indicate a strong influence of recombinative processes, i.e. radiative & 3-body recombination from H$^+$/D$^+$ and dissociative recombination from molecular ions (H$_2^+$/D$_2^+$).
$\\$
[1] R.Friedl, C.Fröhler-Bachus, U.Fantz, Meas.Sci.Technol. 34 (2023) 055501.
[2] C.Fröhler-Bachus, R.Friedl, S.Briefi, U.Fantz, JQSRT 259 (2021) 107427.

Speaker: Roland Friedl (University of Augsburg, AG Experimentelle Plasmaphysik)

206 2.078 Quantification of SOL filamentary transport and parametric dependencies across different regimes in TCV.

Understanding and evaluating particle and energy exhaust in the divertor and scrape-off layer (SOL) regions of a tokamak are essential for the sustainable operation of future fusion devices such as ITER. Cross-field transport in the SOL is known to be dominated by intermittent, elongated convective structures called filaments or blobs. In this study, filament-induced cross-field heat and particle fluxes in the SOL region of the TCV tokamak are quantified using Gas Puff Imaging (GPI) and Thermal Helium Beam (THB) measurements of radial velocity, filament packing fraction, density, and temperature. Fluxes calculated from the combined GPI–THB analysis are compared with Langmuir probe measurements and benchmarked against GBS and SOLPS-ITER simulations, to assess the GPI–THB approach for estimating SOL fluxes. A parametric study is further conducted to examine the dependence of filamentary fluxes on plasma density, plasma current, and toroidal magnetic field orientation. No significant differences are observed between forward and reversed field configurations under comparable plasma conditions. The parametric dependence of fluxes across different TCV regimes, including QCE, LSN, and DN, will also be reported.

Speaker: Kaushlender Singh (École Polytechnique Fédérale de Lausanne (EPFL), Swiss Plasma Center (SPC), CH-1015 Lausanne, Switzerland)

207 2.079 Reconstructing the separatrix density in SPARC for pulse planning and control

The SPARC tokamak is projected to have high unmitigated heat fluxes to the divertor plasma-facing components, with q|| up to 10 GW/m2 in H-mode, due to its high magnetic field and current (8.7 MA, 12.2 T). Type-I ELMs will cause unacceptable energy fluence on the target for routine operation due to the high pedestal pressures in H-Mode conditions. The separatrix density is an important parameter for access to detachment and scenarios with reduced ELM activity such as the quasi-continuous exhaust regime. SPARC is predicted to be highly opaque to edge fueling [1].

At CFS, we are preparing a hierarchy of models to estimate the separatrix density as a function of engineering parameters, with the goal of predicting it as a function of injected neutral particle flux. The simplest is a 0D model which calculates the divertor neutral pressure from the injected neutral particle flux balanced with pumping throughput and calculates the separatrix density from a multi-machine attached H-mode scaling [2]. The separatrix density is predicted within 50% accuracy for the AUG database, and we additionally test the model on a database from Alcator C-Mod and extrapolate to SPARC. This model is used to find a lower limit for the separatrix density with no injected particle flux using the available pumping throughput and assumed background particle source. The model is extended to solve for divertor neutral pressure and separatrix density consistent with recycling fluxes. We compare the predictions of the 0D models to the Extended Lengyel model [3], SOLPS simulations, and SOLPS-NN predictions [4]. We use these tools to evaluate SPARC scenarios planned for the first campaign.

[1] S. Mordijck et al. 2024 Nucl. Fusion 64 126034
[2] D. Silvagni et al. IAEA FEC 2025
[3] T. Body et al. 2025 Nucl. Fusion 65 086002
[4] S. Dasbach et al. 2023 Nucl. Mater. Energy 34 101396

Speaker: Sean Ballinger (Commonwealth Fusion Systems)

208 2.080 Coupled turbulence and transport simulations of multiple ELM cycles within Edge Plasma Coupling Simulation framework

The tokamak divertor is subjected to huge heat load, including both the transient heat load due to the edge localized modes (ELMs) and steady-state heat load in between ELMs. Exploring the edge plasma solution compatible with the high-performance plasma is critical for the fusion reactors. Numerical simulations are indispensable for both understanding the edge plasma physics and predicting the edge plasma behavior. However, the cross-field transport coefficients adopted in the transport codes such as SOLPS-ITER, UEDGE, EDGE2D-EIRENE and SOLEDGE2D-EIRENE are usually given empirically or by fitting experiments. On the other hand, although the turbulent cross-field transport of the edge plasma can be simulated by lots of turbulence codes such as BOUT++ and JOREK, it is hard to achieve a self-consistent simulation of the edge plasma transport due to the large gap between the turbulence and transport time scales. One effective way to realize a self-consistent edge plasma simulation is the coupling simulation by the transport and turbulence codes [1, 2].

For the purpose to implement the self-consistent coupling simulation of the edge plasma automatically and efficiently, a simulation framework called EPCS (Edge Plasma Coupling Simulation) is developed recently [3]. At the present stage, BOUT++ [4] is chosen as the turbulence code and SOLPS-ITER [5] as the transport code. To simulate multiple full ELM cycles, a time-dependent coupling simulation workflow is developed [6]. Based on the time-dependent workflow, the simulations are conducted for the grassy ELM experiment [7] and the neon-seeding experiment (where both detachment and grass ELM regime are achieved) [8] in EAST. For both cases, multiple ELM cycles are successfully simulated, which have the ELM frequencies consistent with the experiments. Especially, the influence due to the injection of neon impurities on the divertor detachment and the ELM size and frequencies is qualitatively reproduced. Further analysis of the detailed mechanisms will be reported in the conference.

References
[1] T.D. Rognlien, et al., Contrib. Plasma Phys. 44 (2004) 188–193.
[2] D.R. Zhang et al., Nucl. Fusion 60 (2020) 106015.
[3] T.Y. Liu et al., Plasma Phys. Control. Fusion 67 (2025) 055004.
[4] X.Q. Xu et al., Phys. Plasmas 7 (2000) 1951–1958.
[5] S. Wiesen et al., J. Nucl. Mater. 463 (2015) 480–484.
[6] T.Y. Liu et al., submitted to Plasma Phys. Control. Fusion.
[7] G.S. Xu et al., Phys. Rev. Lett. 122 (2019) 255001.
[8] Q.Q. Yang et al., Nucl. Fusion 60 (2020) 076012.

Speaker: Shifeng Mao (University of Science and Technology of China (CN))

209 2.081 EMC3-EIRENE predictions of radiative detachment scenarios in W7-X equipped with a tungsten based divertor

Wendelstein 7-X (W7-X) operates with carbon-fiber composite (CFC) plasma-facing components (PFC) forming the island divertor used for particle and power exhaust. Erosion of PFCs leads to an influx of carbon (C) into the edge plasma where C radiation significantly cools down the divertor in the applied hydrogen plasmas. Rising the radiation in the divertor by hydrogen fueling and/or impurity seeding of low-Z species like N2 or Ne allows access to the “radiative detachment" regime with reduced target heat and particle loads [1,2] at high electron densities and moderate electron temperatures in the divertor of about 10 eV. However, carbon-based PFCs are not tolerable for a future fusion reactor due to unacceptable fuel retention. A promising alternative PFC material is tungsten (W), which shows in general low sputtering, low fuel retention, and good thermomechanical properties.
As a first systematic approach to study the impact of potential W PFCs in W7-X, we use EMC3-EIRENE as plasma boundary simulation code for 3D magnetic configurations. When C as instrinsic radiator is absent, impurity seeding is unavoidable for plasma operation at elevated power and particle fluxes [3]. We investigate detachment scenarios with the standard magnetic configuration of W7-X equipped with a W divertor in the current divertor design and analyze the dependence of detachment relevant parameters (e.g. target heat loads) on Ne impurity concentration and operation conditions (e.g. heating power). First results indicate a deeper penetration of recycled H neutrals into the confined region as a consequence of the increased reflection probability of H ions at the strike lines.
It is anticipated that low impact energies and cold divertor operation with electron temperatures below 5 eV are required to restrict W erosion caused by the seed impurities to avoid significant W accumulation in the core. At these temperatures, volume recombination processes become increasingly relevant and hence are also investiged since up to now, the research on volume recombination in the island divertor has been limited to the C divertor [4].
Both the hydrogen recycling and seed impurity concentration need to be adapted to get access to this reactor-relevant regime and eventually, a change of the island divertor geometry with higher baffeling might be unavoidable.

[1] Y. Feng et al 2021 Nucl. Fusion 61 086012, [2] F. Effenberg et al 2019 Nucl. Fusion 59 106020,
[3] S. Brezinsek et al 2019 Nucl. Fusion 59 096035, [4] Y. Feng et al 2025 Nucl. Fusion 65 066008

Speaker: Daniil Ryndyk (FZJ)

210 2.082 Impact of radiation trapping effect with spatial distribution on He I line intensity ratio method in the study on detached divertor plasma

One of the most important subjects in detached divertor plasma studies is reliable diagnostics of the electron density ($n_e$) and temperature ($T_e$). He I line-intensity-ratio-method (LIR-method), based on a visible spectroscopy and a collisional-radiative (CR) analysis, offers a non-invasive measurement of these parameters. When this method is applied to plasmas with high neutral densities, reabsorption of the emitted photons called the radiation trapping becomes problematic. A practical way to account radiation trapping is to introduce the Optical Escape Factor (OEF) in CR analysis. Typically, OEFs are introduced to the resonant transitions with ground state, i.e., $1^{1}\mathrm{S}$–$n^{1}\mathrm{P}$ series. $n$ represents the principal quantum number. Calculations of the OEFs are based on spatial profiles of $1^{1}\mathrm{S}$ atoms and $n^{1}\mathrm{P}$ atoms. So far, several models have been proposed for calculation of the OEFs [1], and these models assume spatially uniform ground state atoms. However, sometimes ground state atoms have non-uniform distribution due to neutral depletion [2] or particle transport, making above assumption no longer valid.
In this situation, we propose a new model for the OEF calculation to account for the spatial distribution of the ground state atoms. Based on the model described in Ref. [1], we have extended the formulation so that it can consider any arbitrary spatial distribution of ground-state atoms. To investigate impact of the distribution assumed in the CR analysis, recently we have conducted a preliminary experiment using RF plasma. Visible lines of wavelength at 667.8 nm ($2^{1}\mathrm{P}$-$3^{1}\mathrm{D}$), 706.5 nm ($2^{3}\mathrm{P}$-$3^{3}\mathrm{S}$), and 728.1 nm ($2^{1}\mathrm{P}$-$3^{1}\mathrm{S}$) were collected for the LIR-method. In addition, the line emission from $3^{1}\mathrm{P}$ level was also collected to calculate OEFs. A Langmuir probe was also used to obtain $n_e$ and $T_e$. Results of the LIR-method using several different distributions for ground-state atoms showed that determination of $n_e$ and $T_e$ was significantly dependent on the distribution selected. However, agreement in $n_e$ and $T_e$ was relatively poor at peripheral region of the target plasma, regardless of the distribution selected. One probable reason for this disagreement is uncertainties in the OEF for $1^{1}\mathrm{S}$-$2^{1}\mathrm{P}$ transition. Currently we are trying to evaluate more reliable OEFs by adding emission from $2^{1}\mathrm{P}$ level to the CR analysis. In the presentation, we will report impact of the spatial distribution of the ground state atoms on LIR-method, based on the improved OEF calculation.

[1] Y. Iida et al, Phys. Plasma 17, 123301 (2010).
[2] R. M. Magee et al, Phys. Plasma 20, 123511 (2013).

Speaker: Shogo Otsuka (Tohoku Univ.)

211 2.083 Edge modelling of low and high radiative partially detached JET high power plasmas in the ITER-like configuration

To ensure an operating regime compatible with its tungsten divertor, ITER must operate at least in a partially detached state. Considering the required level of power crossing the separatrix ($P_{SOL}$) to sustain H-mode, such a partially detached state will be achieved through strong radiative power losses obtained via extrinsic (seeded) impurities. However, impurities can also influence plasma behaviour through additional mechanisms, such as modifying transport properties or affecting neutral penetration.
To quantify these effects, edge modelling has been performed for three JET pulses with and without neon seeding of the JET ITER-baseline scenario, characterized by high plasma current ($Ip=2.5\ MA$), input power ($P_{in}=27-33\ MW$) and triangularity ($\delta=0.38$) and a closed vertical target divertor configuration (VV). Using the Soledge2D and EDGE2D codes, transport profiles have been estimated in a comparative way, and by carefully matching outer-midplane kinetic profiles (electron density and electron/ion temperatures), Langmuir probe target profiles and total radiated power. Particle and heat fluxes from TRANSP interpretative simulations have been used to provide core boundary conditions for the edge model at top-of-pedestal region.
The results show that whilst increasing the seeding level an increase in particle diffusivity is required throughout the entire region between the core boundary and the far SOL to match radial (density) profiles. Conversely, the electron heat diffusivity decreases across the same region, except for a localized increase near the top of the pedestal. In all cases, reproducing the measured ion temperature at the separatrix requires the ion heat diffusivity to remain very low, likely close to its neoclassical value, in the near- and far-SOL.
To assess the impact of seeding on neutral dynamics, the estimated radial transport coefficient profiles were applied for cases with varying impurity concentrations corresponding to different levels of radiative dissipation. The results indicate that, with all other parameters held constant, the electron density at the separatrix decreases as the impurity concentration and total radiation increase as previously observed in (partially) detached conditions. This behavior results from the combined effect of reduced free energy available for ionization (the so-called power-starvation effect) and the different distribution of electrons originating from deuterium and impurity atoms. At constant pressure, as the available power decreases, the temperature drops while the electron density in the divertor region increases, reducing deuterium ionization and neutral penetration. Electrons originating from impurities only partially compensate for the loss of deuterium-derived electrons at the separatrix.

Speaker: Paolo Innocente (Institute for Plasma Science and Technology, CNR, 35127 Padova, Italy)

212 2.084 Progress towards Integrated Scenarios for Exhaust: experiments and new self-consistent core-edge modeling framework

Recent dedicated experiments combined with a new modeling suite have advanced the crucial topic of core edge integration and power exhaust for fusion plasmas substantially. We report on experimental findings with reactor relevant seeding gases in various configurations and in multi-machines that establish a core edge integrated boundary solution. Those include highly radiative plasmas in high performance plasmas, in negative triangularity [1] plasmas as well and closed detached divertors. A new core-edge integrated modeling framework has been validated on these experiments and is available for extrapolation to future power plants.
Core-edge integrated simulations with the new developed SICAS framework (SOLPS-ITER Coupled to ASTRA-STRAHL) [2] provides self-consistent background plasma and impurity transport from the divertor to the core with good agreements with experimental data. A key capability of this new integrated core-edge tool is the possibility to model self-consistent scenarios and phenomena such as the X-point radiator (XPR) physics for which a tool like SICAS is required. This framework opens new possibilities in integrated modeling of fusion devices for the interpretation of current experiments, prediction for ITER as well as reactor design. By capturing the complex interplay between impurity transport, core confinement, and edge dissipation, SICAS provides a physics-based foundation for designing new integrated scenarios for exhaust.
[1] L Casali et al 2025 Plasma Phys. Control. Fusion 67 025007, [2] A. Welsh et al 2025 Nucl. Fusion 65 044002
Work supported by the U.S. Department of Energy, under Award(s) DE-SC0023100, NRC 31310022M0014.

Speaker: Livia Casali (GNOI)

213 2.085 Multi-spectral coherence imaging to characterize the scrape-off layer flows in TCV and Magnum-PSI

Plasma flows are fundamental in understanding various aspects of the power exhaust in fusion devices, including helium exhaust, impurity transport, detachment processes, and distribution of the diverted heat flux [1-3]. A comprehensive and validated understanding of the plasma flows is required for evaluating and optimizing power exhaust strategies [4,5].
A multi-spectral coherence imaging spectroscopy (CIS) setup has been developed to simultaneously acquire the plasma flows of several impurity charge states in TCV (C$^+$, He$^+$,C$^{2+}$), together with an achromatic CIS system on Magnum-PSI (H, Ar$^+$). CIS is a camera-based interferometric technique to obtain 2D integrated plasma velocity profiles from the spectral emissions’ Doppler shift. The spatial resolution (1232x1028 lines-of-sight) and velocity precision (±200 m/s) of legacy CIS systems were improved in a new diagnostic configuration, consisting of a temperature-stabilized field-widened polarization interferometer with a full quadrant polarization sensitive camera, allowing for the observation of steep velocity gradients in 2-dimensions with a single ex-situ calibration. The diagnostic was validated against high-resolution spectroscopy in the Upgraded Pilot-PSI with excellent agreement.
Measurements on Magnum-PSI and TCV characterize the flows from the near-surface plasma region to the full scrape-off layer, providing new insights into the performance of alternative divertor configurations and detachment strategies for future fusion reactors. In Magnum-PSI, hydrogen and impurity argon flow measurements indicate collisional coupling between the main plasma and impurities that extends into the pre-sheath. Initial TCV observations indicate strong gradients in the velocity profile (shear) across the last-closed flux surface, with strong collisional coupling between He$^+$ and C$^{2+}$, whereas C$^+$ exhibits a lower velocity, indicating imperfect thermalization, in quantitative agreement with previous spectroscopic studies [6]. Investigations of the impact of plasma flows on detachment will be presented.

[1] P.C. Stangeby and D. Moulton, Nuclear Fusion (2020), 60(10):106005
[2] A. Zito et al., Nuclear Fusion (2025), 65(4):046022
[3] A.V. Chankin et al., Journal of Nuclear Materials (2001), 290-293:518-524
[4] Y. Wang et al., Nuclear Fusion (2024), 64(5):056040
[5] A.V. Chankin et al., Nuclear Fusion (2007), 47(8):762
[6] Lorenzo Martinelli, EPFL PhD thesis (2023), 10158

Speaker: Mark Cornelissen (Eindhoven University of Technology)

214 2.086 Influence of 3 T magnetic field on RF plasma self-bias and erosion pattern for the ITER Wide Angle Viewing System

The ITER Wide Angle Viewing System (WAVS) relies on in-vessel metallic mirrors to transport optical signals from the plasma to out-of-vessel diagnostics [1]. Under the conditions of ITER, energetic plasma particles induce sputtering of plasma-facing components, and the resulting sputtered material can re-deposit on these mirrors, resulting in reflectivity degradation [2, 3].

In-situ plasma removal using radio-frequency (RF) capacitively coupled plasma (CCP) is one of the proposed methods to mitigate this issue. Experiments were conducted in a 3 T homogeneous magnetic field using a medical Magnetic Resonance Imaging (MRI) system at the University Hospital of Basel. The WAVS First Mirror Unit (FMU) consists of two mirror electrodes (M1 and M2), one or both of which can be powered with RF, while the remaining surfaces act as grounded areas. Systematic measurements of the self-bias voltage were performed for varying discharge conditions, including gas species (argon and helium), pressure (0.1–10 Pa) and mirror capacitance, as well as the angle θ between the mirror surface and magnetic field.

The results revealed a strong dependence of the self-bias on the grounded-to-powered area ratio Ag/Ap, which is affected by the magnetic confinement. Erosion and deposition patterns were examined in cases where one or both mirrors were powered. Erosion occurred primarily where a powered mirror faces a grounded surface, whereas overlapping discharges in the dual-powered case resulted in low ion kinetic energy and preferential net deposition at the overlapping area.

These observations were supported by one-dimensional simulations of plasma potential evolution under representative conditions. These results provide a critical insight into RF plasma behaviour in fully magnetised, geometrically asymmetric discharges and support the development of the robust in-situ cleaning strategies for ITER optical diagnostics.

[1] S. Vives et al 2024 Rev Sci Instrum. 95, 113508
[2] S. Rode et al 2024 Nucl. Fusion 64 086032
[3] A. Litnovsky et al 2019 Nucl. Fusion 59 066029

Speaker: Tomás Sousa (University of Basel)

215 2.087 Spectral ellipsometry for in-vessel deposition measurements on plasma-facing components in Wendelstein 7-X

Erosion, re- and co-deposition processes in Wendelstein 7-X (W7-X) lead to the formation of deposits on plasma-facing components (PFCs). Owing to the carbon-based wall-tiles, these deposits are carbon-dominated including hydrocarbons, oxides,and boron co-deposits. Characterizing these deposits is essential for understanding material migration and fuel retention.

A handheld spectroscopic ellipsometer has been developed for the first combined in-vessel thickness and refractive index measurements of the deposits in W7-X. Previous material migration investigations at W7-X relied on spectroscopic observations, on ex-situ analyses [1], including marker-layer experiments [2]. These methods provide valuable detail but have drawbacks, e.g., they require the removal of PFCs. In-vessel colorimetry [3] enables wider-range studies; however, it is limited by the use of three measurement wavelengths (red, green, blue), the lack of relative phase information from both polarization components, and the assumption of uniform optical properties of the deposits across the vessel.

The newly developed diagnostic is a compact white-light ellipsometer (frame 8.7 × 44 cm) with a stepwise rotational analyzer, operating at an incidence angle of 70°. Light from a tungsten-halogen source is coupled via optical fibers, and reflected spectra (350 - 950 nm) are recorded for analysis. Deposited layer properties are derived by fitting the measured spectra; Bayesian data evaluation provides full error propagation, increasing robustness to geometric and instrumental uncertainties [4].

Laboratory tests on SiO₂ standards (nominal thickness 492 nm) demonstrated an overall measurement accuracy of around 1 nm [5]. The system was validated on amorphous C:H layers (7 nm and 30 nm) and ex-situ W7-X wall components. Subsequently, the in-vessel measurements on W7-X PFCs are conducted on stainless steel wall panels, poloidal closure and pumping gap panels, covering an area of ~71 m². The approach enables wide-range, resource-efficient mapping of deposits around the torus. Future developments include optimization for rough layers and substrates, and multi-layer modeling.

References:
[1] C. P. Dhard et al., Phys. Scr. 96 (2021) 124059.
[2] M. Mayer et al., Nucl. Fusion 62 (2022) 126049.
[3] G. Motojima et al., Nucl. Mater. Energy 43 (2025) 101934.
[4] U. v. Toussaint et al., AIP Conf. Proc. 872 (2006) 272-9.
[5] M. Krychowiak, Rev. Sci. Instrum. 95 (2024) 113511.

Speaker: Laura Dittrich (IPP)

216 2.088 BH molecular emission in the divertor region of LHD: dependence on magnetic axis, plasma parameters, and multi-view spectroscopy

Boron remains a key material for real-time wall conditioning in fusion devices due to its strong gettering properties and its impact on impurity control and hydrogen recycling. Previous LHD experiments have identified boron monohydride (BH) emission as a sensitive indicator of surface processes during boron powder injection, with localized signatures near the divertor and rotational temperatures around 3600 K.

In this study, we extend earlier work by analyzing a new dataset obtained on the Large Helical Device with two magnetic-axis configurations and systematic scans of heating power and plasma density. Spectrally resolved BH emission was recorded with two complementary instruments:
(1) a multi–line-of-sight visible spectrometer used in previous campaigns, enabling spatially resolved characterization of the divertor region; and
(2) a high-resolution echelle spectrometer providing access to BH together with additional species such as H₂ (Fulcher-α), although the latter is still under evaluation.

Preliminary analysis of the visible-range data shows that the BH Q-branch remains strongly localized near the divertor for all operational conditions studied, consistent with surface-associated production mechanisms. Rotational temperatures are obtained by generating a synthetic BH spectrum using tabulated transition coefficients and fitting the calculated Q-branch envelope to the measured spectra. The resulting temperatures are similar to earlier results (≈3.3–3.7 kK), with only weak dependence on magnetic-axis position, density, or heating power. The persistence of these high rotational temperatures suggests that BH formation involves energetic surface-driven chemistry rather than purely sputtering-based processes.

These new results indicate that BH emission retains clear diagnostic sensitivity to wall interaction processes across different magnetic configurations and plasma conditions in LHD. Completion of the echelle-based analysis—particularly including H₂ Fulcher-α rotational temperatures—will further clarify the chemical and thermal environment of boron-containing species in the divertor region.

Speaker: Arseniy Kuzmin (Kyoto University, Kyoto, Japan)

217 2.089 Poloidal flow measurement of impurity ions in DIII-D divertor

A novel spectroscopic technique has been used to perform the direct measurement of impurity poloidal flow in the DIII-D divertor. Particle dynamics in the open-field-line scrape-off-layer (SOL) and divertor, to which both parallel and perpendicular transport contribute, are critical for predicting heat and particle loads on plasma-facing components. While the parallel component of impurity flow is routinely determined from Doppler shifts, quantifying the much smaller cross-field velocity (< 100 m/s) has remained elusive, despite its expected essential role in the dynamics.
Here, we report the first direct poloidal-flow measurement in the DIII-D upper-outer divertor using Flowmetry via Absorption-Cell-Enhanced Spectroscopy (FACES). To achieve 100 m/s resolution, systematic uncertainties, such as the temperature and humidity drifts, must be compensated. In the FACES technique, a Te2 absorption cell inserted in front of the entrance slit of high-resolution spectrometer embeds molecular absorption lines into the Doppler-broadened impurity emission line from the divertor. These molecular lines serve as an in-situ wavelength ruler. This in-situ calibration enables us to achieve 50 m/s resolution, which is an order-of-magnitude improvement over conventional spectroscopy.
We have measured the flow speed of C²⁺ ions from the C III (465 nm) line with FACES. The data reveal C²⁺ ions streaming from the target toward the X-point at 0.3 - 2.0 km s⁻¹, opposite to the direction expected from main-ion parallel flow, ∇B drift, and E × B drift due to the sheath electric field.
This flow direction is consistent with the impurity particle balance: because the C2+ ion source should be located near the target plate and the sink should be upstream, the C2+ ions must flow from the target toward the main plasma. On the other hand, friction with the downward main-ion flow should be large enough to push the impurity ions back to the target plate. A simple one-dimensional model does not account for the large (> 1 km/s) poloidal flow of the impurity ions, indicating the two-dimensional structure must be considered.
In the presentation, we will discuss the relation to the parallel flow and its possible drivers, including frictional coupling to neutral recycling flows, poloidal electric-field, and flow reversal, with the aid of numerical modelling.

*Work supported by the US DOE under contracts DE-FC02-04ER54698 and DE-AC05-00OR22725.

Speaker: Keisuke Fujii (Oak Ridge National Laboratory)

218 2.090 CF-LIBS Depth Profile Analysis of Inner Divertor from In-situ JET LIBS Robotic Arm measurements: Impact of Spectral Averaging on Depth Resolution and Quantification

Laser-Induced Breakdown Spectroscopy (LIBS) is a key diagnostic for analysing plasma-facing components in fusion devices [1], where erosion, migration, redeposition and fuel retention produce complex multi-element layers containing Be, W, Mo, H/D/T isotopes and various impurities [2,3]. Accurate characterisation of these layers is essential for understanding material migration, surface composition and fuel retention in ITER and future pilot fusion plants.

In this work, we analyse in-situ LIBS data from the JET robotic wall-inspection system acquired during recent D/T campaigns to obtain quantitative depth profiles of major and minor elements and retained hydrogen isotopes. Sub-nanosecond LIBS spectra were recorded at 840 wall positions, with several hundred laser shots per location, using a broadband echelle spectrometer (255–760 nm) and a narrow-band Littrow spectrometer centred on the Hα line (656 nm). The experimental setup is described in detail in [4–6].

Quantification uses the Calibration-Free LIBS (CF-LIBS) method applied to interference-free lines of W, Mo, Be, Ti, Ni and Cr. Electron temperature (approximately 0.7 eV) is obtained from multi-element Boltzmann plots using several hundred lines, while electron density is evaluated with the Saha equation. This ensures the precision required for reliable CF-LIBS quantification. Spectral averaging is necessary to reduce noise, but it directly affects depth resolution: minimal averaging (2–5 shots) preserves high resolution for major elements, while minor-element detection requires stronger averaging, reducing depth precision. Case studies on inner divertor tiles demonstrate CF-LIBS depth profiles over ~500 laser shots with depth resolution down to only a few shots. Quantified species include W, Mo, Be, H/D/T, Ti, Ni, Cr and impurities such as Cu and Ca, while C and O could not be detected due to the limited spectral range.

References:
[1] G. S. Maurya et al., Journal of Nuclear Materials, 541 (2020) 152417.
[2] J. P. Coad et al., Journal of Nuclear Materials, 313–316 (2003) 419.
[3] P. Veis et al., Nuclear Materials and Energy, 25 (2020) 100809.
[4] J. Likonen et al., Nuclear Materials and Energy, 45 (2025) 102021.
[5] J. Ristkok et al., Nuclear Materials and Energy, 44 (2025) 101968.
[6] R. Yi et al., Nuclear Materials and Energy, 45 (2025) 102016.
[7] A. Ciucci et al., Appl Spectrosc, 53 (1999) 960.
[8] P. Veis et al., Phys. Scr., T171 (2020) 014073.
[9] M. Hornackova et al., Eur. Phys. J. Appl. Phys., 66 (2014) 10702.

Speaker: Pavel Veis (FMPI, Comenius University, Bratislava, Slovakia)

219 2.091 Fine-Tuning of Machine Learning Diagnostics of Electron Density and Temperature from Helium Line Emissions

While the helium line intensity ratio method has been used to measure electron, ne, and temperature, $T_e$, by combining measured line intensities with a collisional radiative model (CRM) [1], one of its difficulties is to include the photon transport and metastable atom transport. A machine learning (ML) approach has been considered as an alternative method to measure $n_e/T_e$ from the line ratios. If training data is sufficiently available, it can be a useful diagnostic tool. The challenging issue in this approach is developing a global model that can be applied to other devices. In this study, we collected an OES dataset and $n_e/T_e$ data from four linear divertor simulators, and we investigated the fine-tuning method to develop a global cross-machine model.
Data from the following four linear devices are used: Magnum-PSI, NAGDIS-II, and PISCES-A, and Lotus-I. Line emissions at 447.1, 492.2, 501.6 + 504.8, 667.8, 706.5, and 728.1 nm are used. The dataset includes 24, 64, 6, and 3 discharges (radial profiles) and 960, 417, 342, and 70 data points from Magnum-PSI, NAGDIS-II, PISCES-A, and Lotus-I, respectively. Laser Thomson scattering was used in Magnum-PSI and a Langmuir probe was used in the other devices to obtain $n_e/T_e$. In addition to a deep neural network (DNN) model, physics-informed ML approach [2] was also tested, where a pre-trained NN with a CRM tuned with experimental data [3].
It was shown that a DNN model trained with the dataset from three devices leads to an error of ~100%, when applying it to a remaining unseen fourth device for both ne and $T_e$, which is significantly higher than the model applied to the seen devices. This is primarily due to device-specific parameters such as plasma radius and different ranges of ne and $T_e$, which hinder the model’s generalizability across devices. To address this, we additionally performed fine-tuning using data from the target device itself. It was found that errors significantly decreased by the fine-tuning process with a small amount of data. Furthermore, the reduction in the errors was more significant for physics-informed models. The results suggested that the physics-informed model has an advantage when using fine-tuning with a limited dataset.

[1] S. Kajita, D. Nishijima, Journal of Physics D: Applied Physics 57 (2024), 423003.
[2] S. Kajita, https://arxiv.org/abs/2506.20117.
[3] M. Goto, J. Quantitative Spectroscopy and Radiative Transfer 76 (2003) 331.

Speaker: Shin Kajita

220 2.092 Measurements of heat flux distribution profile on the main limiter based on calorimetry methods in EAST

High heat flux on the first wall is one of the critical issues in tokamak experiments. An integrated diagnostic method was developed to investigate heat flux in the SOL in EAST, with combination of fine calorimetry measurements and three-dimensional magnetic field-line tracing using the PFC Flux code. The measurements help to identify heat loads on the wall material and understand SOL transport.
The main limiter was upgraded to full tungsten plasma-facing components for higher power handling ability. It consists of seven independently water-cooled tungsten-copper monoblock strings. High-precision thermal sensors (with 0.05 K resolution) and flowmeters (with 5% accuracy) were incorporated into each string, which have been used to precisely measure the heat power variations. And the heat power was then used to validate the surface heat load distribution simulated by PFC Flux code[1]. The code identifies a regime of short magnetic connection length, characterized by a precipitous drop in the connection length beyond a critical value of R or Z. In this regime, the heat flux with different exponential decay lengths is determined by diffusive or convective cross-field transport in the SOL. The methodology enables to obtain heat flux evolution with high resolution of kW/m². The heat flux profile in the SOL can be thereby reconstructed.
A ~30 s long-pulse L-mode discharge was used to validate the method's reliability which allows the water-cooling system to reach thermal equilibrium. This discharge featured low density and low recycling, and thus the simple SOL theory can be used. As a result, an exponentional decay length of 12 mm with a peak parallel heat flux of 0.8 MW/m² at the limiter tangency point was reconstructed, which is in line with the reciprocating probe data. Furthermore, asymmetric heat load deposition obtained by PFC-Flux code matches the measurement via calorimetry. This phenomenon verifies the dependence of heat flux characteristics on local connection lengths. There is a localized discrepancy observed on the central limiter string, which indicates additional non-thermal heating sources. Such measurement capability is help for investigating power deposition patterns on the first wall and the power balance in tokamak experiments.
[1] B.F. Gao, et al., Nucl. Fusion 65 (8), (2025) 086005.

Speaker: Cong Cao (Institute of Plasma Physics, Hefei Institutes of Physical Science Chinese Academy of Sciences, Hefei 230031, China)

221 2.093 Investigation of Deuterium Retention in Fe–Zr/Y Multilayers Using Femtosecond Laser-Induced Mass Spectrometry

Monitoring and detection of deuterium (D) retention in plasma-facing components in a tokamak is essential for safety and fuel management. In this work, we present femtosecond laser-induced ablation coupled to quadrupole mass spectrometry (fs-LIA-QMS) to investigate D depth profiles in Fe–Zr/Y+D–Fe multilayer samples (each layer ~1 µm) prepared on Si substrates. Fs-LIA-QMS combines ultrafast laser ablation, which removes material layer by layer with minimal thermal damage, with real-time mass spectrometry detection of the ablated species. This approach allows quantitative analysis of gaseous species like hydrogen isotopes with high depth resolution (~10–20 nm per pulse) and minimal sample modification [1].
A comparative study of top-hat and Gaussian beam profiles and their impact on depth resolution was performed. The top-hat beam was produced using an Airy beam shaper (320–450 nm) combined with a theta lens (f = 163 mm) positioned at 162.6 mm along the focal axis, producing a uniform ablation profile with uniformity ~0.71. Analysis using a confocal microscope yields crater radii of approximately 14 µm (top-hat) and 6.5 µm (Gaussian) at 5.6 µJ/pulse. The top-hat beam was selected for its uniform intensity distribution, providing better layer-by-layer depth information of D retention. Each layer was irradiated with 256 laser pulses at 10 kHz, with two QMS measurement cycles synchronized to each laser cycle. Depth profiling was performed over 500 laser cycles, enabling high-resolution ablation in depth. Real-time QMS quantification of D₂ released during ablation was performed, and the D concentration as a function of depth was determined using a calibrated deconvolution procedure.
The obtained fs-LIA-QMS depth profiles are compared with Nuclear Reaction Analysis (NRA) based depth profiles and both resolve the D retained in the Zr layer. A detailed discussion of the obtainable depth-resolved information from fs-LIA-QMS in comparison to the NRA measurement will be presented. The results demonstrate that fs-LIA-QMS is a sensitive, high-resolution technique for characterizing hydrogen isotope retention in fusion-relevant multilayer structures but also highlight the importance of beam-shape selection for accurate depth profiling.
Reference:
[1] S. Mittelmann, M. Mayer, U. von Toussaint, B. Buchner, A. Theodorou, T. Dürbeck, W. Jacob, and T. Schwarz-Selinger, “Femtosecond Laser-Induced Ablation – Quadrupole Mass Spectrometry (fs-LIA-QMS) experiment for the detection of trapped hydrogen isotopes and helium in nuclear fusion relevant materials,” Spectrochim. Acta Part B: At. Spectrosc., vol. 233, p. 107283, 2025. https://doi.org/10.1016/j.sab.2025.107283

Speaker: Shweta Soni (IPP)

222 2.094 Wide-area spectroscopic observation and machine learning analysis of attached and detached helium plasmas

Reducing the enormous heat flux on the divertor target is key to realizing future fusion reactors. One promising approach is to increase the neutral gas pressure near the divertor to promote strong plasma recombination. This leads to a detached plasma state, in which heat and particle fluxes to the target are significantly reduced. However, current understanding is insufficient to reliably predict detachment behavior in future devices, so it is essential to conduct complementary researches with experiments and simulations.
In the linear plasma device NAGDIS-II, which can produce a steady-state cylindrical plasma approximately 20 mm in diameter and ~2 m in length using DC discharge, an integrated transport code named “DISCOVER” is being developed to achieve high-accuracy simulations of detached plasma (H. Natsume et al., Physics of Plasmas 32, 022508 (2025)). As complementary experimental work, we have developed a two-dimensionally movable spectroscopic system, enabling measurement of plasma emission across a wide region by scanning both the axial and azimuthal directions. To eliminate the reflection effect from the wall, a reflection prevent plate was also installed.
This study measured wide-area helium emission distributions together with radial profiles of electron density and electron temperature at a fixed axial position by using a Thomson scattering system. By changing the discharge current and the gas flow rate, we created a large dataset. The emission distributions show a clear transition from an ionization-dominated attached plasma to a recombination-dominated detached plasma as the gas pressure increases. Using emission spectra and plasma parameters obtained at the same axial position, a machine-learning (neural network) model was trained. This was then applied to estimate wide-area plasma parameter distributions. Details of the full results will be presented at the presentation.

Speaker: Hirohiko Tanaka (Nagoya University)

223 2.095 Progress in high time and mass resolution mass spectrometry in Wendelstein 7-X

The diagnostic residual gas analyzer (DRGA) at Wendelstein 7-X (W7-X) provides remote measurements of the exhaust gas composition sampled from the sub-divertor region. A new compact time-of-flight mass spectrometer has been installed and operated routinely during the latest operation campaign OP2.3. It records full mass spectra, enabling comprehensive monitoring of exhaust-gas dynamics. The system achieves a temporal resolution of 0.1 s and a mass resolution of approximately m/Δm = 500 at mass 4. Both raw spectral data and directly processed data for a predefined set of masses are available, facilitating rapid trend analysis as well as detailed post-processing.
The measurement fidelity allows for the analysis of trace gases. Two applications are presented.
First, an enrichment study of Neon between the core plasma, divertor, and exhaust was carried out during experiments with auxiliary Neon seeding for edge-radiation control. No significant impurity enrichment was observed.
Second, helium transport in the sub-divertor region is examined by comparing measured signal ratios. The measured concentrations match well with the model results.
At present, these analyses are limited to relative signal intensities, as no absolute calibration has yet been performed. Calibration of the instrument response and determination of the transfer function between the sub-divertor and the analysis chamber are planned.
Initial measurements using a He–D₂ gas mixture have been successfully performed. The resulting spectra show a clear separation of the expected mass peaks, confirming the system’s capability to distinguish species of similar mass. This represents an important preparatory step for the W7-X deuterium campaign currently in preparation.
These results demonstrate the utility and versatility of the instrument.

Speaker: Georg Schlisio (IPP HGW)

224 2.096 Results from the Hydrogen Absorption and Desorption Experiment in a Stellarator (HADES) Campaign in HIDRA

Lithium (Li) as a plasma facing material is of great interest due to its ability to retain and pump recycled reactive atoms, decrease instabilities and increase plasma performance. The most recent Li experimental campaign in the HIDRA device at the University of Illinois, Urbana-Champaign, named the Hydrogen Absorption and Desorption Experiment in a Stellarator (HADES) campaign, has shown increased hydrogen retention during in-operando Li evaporation into a hydrogen plasma. This results in a low recycling regime operation, supported by spectroscopy, Residual Gas Analyzer (RGA), and pressure data. Spectroscopy data has shown that Li ions, not just excited neutrals, are needed for retention effect to be observed. A wall heating element next to the HIDRA wall, placed in the path of high Li deposition on the walls, was used to test the location of the retention mechanism, serving as a controlled heating source. Wall heater desorption tests showed increased hydrogen desorption, measured by RGAs, post Li evaporation into hydrogen plasma shots. The desorption increased as the wall heater surface reached 300 C, signifying hydrogen alpha desorption from the Li. The LiH formation could not be investigated due to the heater’s maximum temperature capability, limiting the temperature to about 350 C, too low for LiH decomposition; investigation of this is part of future work. The hydrogen retention at HIDRA’s stainless steel walls by Li is hypothesized to be due to both physical trapping and chemical bond formation between the two species.

Speaker: Nina Mihajlov (University of Illinois, Urbana-Champaign)

225 2.097 Qualification of Grain Oriented, Additively Manufactured Tungsten as a Plasma Facing Material

High-density (99.8%) electron beam powder bed fusion (EB-PBF) tungsten (W) has been successfully manufactured and deployed in plasma material interaction investigations that included DIII-D tokamak and Tritium Plasma Experiment (TPE) exposures. Since the tailored microstructure of EB-PBF W is fundamentally different from conventionally manufactured W, critical knowledge gaps were addressed such as thermal cycling, anisotropic thermal management, erosion properties, and fuel permeability mechanisms. EB-PBF offers distinct advantages over conventional processing, including reduced oxygen content, designing intricate geometries and controlling crystallographic textures through tuning the electron beam parameters.
Our experiments demonstrate that EB-PBF W with controlled crystallographic structures maintains structural integrity and erosion resistance comparable to manufactured W with the notable distinction that hydrogenic species preferentially occupy higher-energy defect sites. This suggests its promise as a candidate for plasma facing components (PFCs). EB-PBF-built W samples with dominant grain orientations, flat and angled geometry, and orientations to the AM build direction were exposed to DIII-D H-Modes with average surface heat fluxes of 36.5 MW/m² and 37.8 MW/m². Scanning electron microscopy revealed increased surface roughness on the samples after plasma exposure and migrated W was identified on the graphite sample holder, which both indicate erosion occurred. However, no substantial differences in material gross erosion rates were found between different grain orientations using visible W spectroscopy.
However, fuel retention could differ from conventionally manufactured W due to the distinct microstructure. A previous study revealed distinctive deuterium release in AM-W compared to typical traps in sintered W which suggests that retention may be due to trapping at high-energy defect sites, but at higher fluence retention could be more limited. After deuterium plasma exposure in TPE, thermal desorption spectroscopy identified distinctive desorption behavior compared to the conventional counterpart, likely arising from microstructural differences induced by varied textures. Higher temperature desorption in EB-PBF-built W relative to sintered W could point to a reduction in low energy trap sites but increased trapping in higher-energy defect sites consistent with past investigations.
Our work demonstrates that EB-PBF-built W exhibits only minor effects of crystallographic orientation on erosion and material integrity which highlights the potential of tailoring microstructure to minimize fuel retention, making it an attractive manufacturing process for PFCs.
This work is supported by US DOE under DE-FC02-04ER54698, DE-NA0003525, DE-AC02-09CH11466, DE-AC52-07NA27344.

Speaker: Aaliyah Zuniga (GNOI)

226 2.098 Integrated PMI benefits of liquid lithium: steady power exhaust and ELM control in EAST

High-confinement (H-mode) operation is an attractive path toward reactor-relevant performance, but large edge-localized modes (ELMs) can produce impulsive particle and heat loads that challenge plasma-facing components (PFCs). Liquid lithium plasma-facing surfaces offer an intuitively self-healing interface under intense bombardment. This work reports experiments on EAST employing a midplane flowing liquid lithium limiter (FLiLi) as a boundary-condition actuator for ELM control[1]. During FLiLi operation, a clear real-time modification of wall recycling is observed, with rapid changes in edge recycling indicators correlated with access to ELM-suppressed phases[2]. The phenomenology is consistent with strong lithium pumping and reduced recycling, and it is qualitatively similar to the ELM suppression obtained in NSTX during lithium evaporation[3] and ELM mitigation with solid Li granule injection[4], where recycling control and modified edge conditions were linked to ELM-free or strongly ELM-mitigated operation.
In addition to the beneficial ELM suppression, spectroscopic measurements during the ELM-suppressed phase show ramp-up of high-Z impurity line emission and the onset of core radiation accumulation, indicating an operational boundary that must be managed for long-pulse application. These observations motivate integrated strategies that combine recycling control (to suppress ELMs and reduce transient loads) with impurity/radiation management (toavoid radiative collapse or confinement degradation).
Overall, the results highlight a complementary effect relevant to PSI: midplane liquid-lithium boundary control can suppress ELMs upstream—thereby reducing transient power loading on downstream PFCs—while liquid lithium surfaces remain promising candidates for resilient PMI interfaces in future steady-state scenarios.

[1] G. Z. Zuo. et al. 2020 Phys. Plasmas. 27 052506
[2] G. Z. Zuo. et al. 2020 Phys. Scr. 2020 014008
[3] R. Maingi1. et al. 2009 Phys. Rev. Lett. 103 075001
[4] Z. Sun. et al. 2021 Nucl. Fusion. 61 066022

Speaker: Y.Z. Xu. (Institute of Plasma Physics, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei, Anhui 230031, China)

227 2.099 Cyclical loading of different tungsten grades under COMPASS-U divertor conditions in Magnum-PSI

The COMPASS-Upgrade tokamak will be a high field compact device (R = 0.9 m, Bt = 5 T, Ip = 2 MA, tflattop = 2-10 s) capable of studying diverse magnetic configurations at high plasma densities [1]. The small machine size, high B and relatively high heating power means that scrape-off layer widths are predicted to be small (λq~0.6–1.2 mm), and attached power densities up to 100 MW m-2 could be achieved [2]. Due to the lack of active cooling, such loading would result in rapid heating of the W divertor tiles up to 2000 °C within a fraction of the total pulse length.
In order to evaluate the expected response to such repeated loading under relevant heat and particle loading conditions, the Magnum-PSI device [3] was used to compare three candidate tungsten grades (ALMT, AT&M, Plansee). In order to design a heating and cooling cycle which could achieve the desired conditions a finite element model was created using COMSOL Multiphysics. Preliminary testing was carried out to better evaluate the unknown input parameters to this model: through-plane conductivity of the grafoil thermal interface material (TIM) HT-1220 and the heat transfer coefficient of the Magnum-PSI jet-impingement cooling. This was carried out via heat loading measurements on a dedicated test mock-up using five type-K thermocouples to provide data while the plasma power was varied from 1.5 – 6.5 MW m-2 and TIM layer number, vacuum gap diameter or applied compression was changed.
Based on this input a cyclical loading scheme was applied to nine 10x12x3 mm tungsten samples. Each of the three grades was heated for 100 cycles up to either 2000 °C, 1800 °C or 1600 °C, with a heating time of 4.5 seconds and a cooling period of 39.5 seconds. From inspection by SEM, despite the short total duration, recrystallization was observed in all samples. Additionally, for the 2000 °C samples, the central region displayed cannibalistic grain growth and strong plastic deformation. Surrounding this region heavy roughening was also observed, but no cracking was detected. All grades behaved similarly. Overall the results show that strong microstructural and mechanical changes should be expected, but that the tiles should survive without failure.

[1] P. Vondracek et al., Fusion Engineering and Design, 169, 112490, (2021)
[2] M. Komm et al., Nucl. Fusion, 64, 076028, (2024)
[3] H.J.N. Van Eck et al., Fusion Engineering and Design, 142, 26–32, (2019)

Speaker: Thomas Morgan (DIFFER)

228 2.100 Molecular Dynamics Simulation of One-Dimensional Migration of 1/2<111> Dislocation Loops in Tungsten

In magnetic confinement nuclear fusion devices, Tungsten (W) is considered the most promising plasma-facing materials (PFMs) for divertor, which will be exposed to high-energy neutron irradiation, high flux plasma irradiation at high temperatures (up to 1200 °C during steady-state operation). High-energy neutron irradiation can directly generate dislocation loops through collision cascades, or produce self-interstitial atom (SIA) that subsequently aggregate to form loops. Moreover, during the accumulation of helium from plasma irradiation and the subsequent formation of nano-scale helium bubbles, SIAs can also be created; these can further aggregate into dislocation loops as the bubbles grow. To simulate the long-term defect evolution of tungsten at the mesoscale, it is of great significance to conduct detailed studies on processes such as cluster migration. For body-centered cubic (bcc) metals like tungsten, the 1/2<111> dislocation loop is widely recognized as the most energetically favourable configuration and can undergo rapid one-dimensional (1D) migration along the <111> direction even at relatively low temperatures. However, due to the considerable difficulty in experimentally observing their thermal migration directly, research on the migration behavior of dislocation loops remains insufficient.
In this study, molecular dynamics (MD) simulations were performed using Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) to investigate the 1D migration of 1/2<111> dislocation loops containing 13 to 99 SIAs NI over a temperature range of 400–2000 K for a duration of 5 ns. To ensure statistical reliability, each combination of size and temperature was simulated 10 times with different random seeds. The mean square displacement (MSD) of the center of mass of the loop was analyzed as a function of time. Based on the Arrhenius equation, the pre-exponential factor D0 and activation energy Ea for the migration of smaller loops were fitted as functions of NI. Furthermore, a sharp decrease in migration coefficient was observed for larger dislocation loops, similar to that reported in iron, and a further study of this phenomenon was carried out.

Speaker: Zhaofan Wang (University of Science and Technology of China)

229 2.101 Qualification of tungsten as plasma-facing components for COMPASS-U: Effects of laser and plasma damage on microstructure and deuterium retention

High-energy plasma interacting with plasma-facing components (PFCs) causes surface damage (e.g., vaporization, sputtering) via intense heat and particle fluxes, compromising reactor lifetime, performance, and safety. Effective PFCs require properties like low sputter yield, high melting point, thermal conductivity, moderate activation, and low gas permeation. Meeting diverse, location-specific requirements (e.g., first wall demands versus divertor's need for high heat load resistance) is a major material challenge.Tungsten (W) is a primary candidate for PFCs, offering high thermal conductivity, sputtering threshold, and melting temperature, but its embrittlement, particularly from overheating, is a significant drawback. As fusion technology advances, demands for power exhaust and lifetime continue to rise. In light of these challenges, this study aims to qualify W components for the COMPASS-U device, specifically focusing on its inertial W components in the divertor region. These components employ W high thermal inertia to resist and manage intense, rapid thermal loads from the plasma. Therefore, evaluating their performance under steady-state and transient heat loads, understanding material evolution, microstructure, surface condition, and D2 retention characteristics is critical. W samples (Plansee) were exposed to plasma in the PSI-2 facility and characterized post-exposure by Atomic Force Microscopy (AFM), Scanning Electron Microscopy (SEM), and fuel retention measurements via Thermal Desorption Spectroscopy (TDS) in a high vacuum system up to 1700 °C. A Pfeiffer mass spectrometer monitored gas release, with simultaneous temperature tracking (pyrometer, type-C thermocouple). Samples underwent 100,000 laser pulses (0.5 ms, 0.4 GW/m²) while exposed to D2/He (6%) plasma (flux 3.2 × 10²¹ m⁻² s⁻¹), with a total fluence of 3.4 × 10²⁵ m⁻² at 700 °C. After initial exposure, samples were heated to 1000 °C (1 hour) to outgas, then exposed to D2 plasma (flux 3.7 × 10²¹ m⁻² s⁻¹, 1 hour), accumulating a total fluence of 1.4 × 10²⁵ m⁻² at ~300 °C. AFM and SEM characterization revealed significant cracking and surface roughening on laser-damaged samples, contrasting with smoother surfaces after only D2 plasma exposure. Despite these severe microstructural changes, total desorbed deuterium for laser-damaged samples was approximately 9.8 × 10¹¹ atm/cm² (peak desorption at 896 K). This retention level showed no increase compared to plasma-only exposed samples (1.276 × 10¹² atm/cm²). This comparative analysis comprehensively evaluates combined laser and plasma damage effects on W microstructure and fuel retention, offering key insights for COMPASS-U W divertor components.

Speaker: laura Laguardia (ENEA-CNR)

230 2.102 Integrated SOLPS-ITER and ERO2.0 modelling of helium–tungsten plasma–surface interaction for an H-mode ASDEX Upgrade discharge

Understanding helium-induced plasma-surface interaction (PSI) on tungsten is essential for quantifying divertor erosion and impurity sourcing in magnetic confinement fusion devices, as He will inherently accumulate as a fusion ash. Experiments at ASDEX Upgrade have provided erosion, deposition and morphology data for He-based L- and H-mode discharges [1,2]. Here we present a modelling study to support the experimental campaign by providing physical insights into the H-mode PSI results, extending earlier L-mode analysis [3] and examining the role of edge-localised modes (ELMs).
The edge plasma was modelled with the state-of-the-art SOLPS-ITER mean-field boundary plasma code [4,5], while PSI and impurity transport were simulated with ERO2.0 [6], a 3D Monte-Carlo code. Optimised SOLPS-ITER simulations showed that both He$^{+}$ and He$^{2+}$ charge states contribute to divertor particle fluxes and that the inclusion of radiating impurities in the divertor region is required to reproduce experimental data. Additional simulations included hydrogenic species to account for H-based neutral beam injection heating. Overall, these simulations provided divertor plasma profiles, particle flux composition at the outer strike point (OSP) and optimised plasma background conditions for subsequent PSI and impurity transport modelling.
ERO2.0 simulations were performed for both inter- and intra-ELM phases. Inter-ELM modelling shows that a simplified He$^{2+}$-only plasma retains most of relevant PSI information, while the inclusion of impurity species strongly modifies the erosion pattern. O$^{6+}$ was chosen as a representative proxy for light impurities (B, C, N, and O) typically present in ASDEX Upgrade, capturing both impurity-driven sputtering and W self-sputtering amplification and allowing individual erosion contributions to be quantified. For the intra-ELM phase, a new approach was adapted from literature data [7,8], accounting for 1 keV He$^{2+}$ energy deposition at an 85° incidence angle at the OSP and an ELM frequency of 200 Hz - as experimentally recorded. This enabled ERO2.0 to reproduce the experimentally observed net erosion near the OSP – 120 nm over the 150-250 nm experimental range – and of net deposition in the near private flux region.

[1] A.Hakola et al. Nuclear Fusion, 64 (2024)
[2] M.Rasiński et al. Nuclear Materials and Energy, 37 (2023)
[3] G.Alberti, et al. arXiv preprint arXiv:2506.03883 (2025)
[4] S.Wiesen et al. Journal of Nuclear Materials, 463 (2015)
[5] X.Bonnin et al. Plasma Fusion Research, 11 (2016)
[6] J.Romazanov, et al. Physica scripta 2017.T170, (2017)
[7] A.Kirschner et al. Nuclear Materials and Energy 18, (2019)
[8] H.A.Kumpulainen et al. Nuclear Materials and Energy 33, (2022)

Speaker: Andrea Mastrogirolamo (Politecnico di Milano)

231 2.103 Evidence of Helium Retention by Lithium Interaction in EAST

Helium is a byproduct of the DT fusion reaction and if not removed quickly enough will pose a real issue as a contaminant of the fusion plasma. Helium is difficult to remove and so finding ways of removing it sufficiently quickly is of importance. Recent results in HIDRA and MAGNUM-PSI have shown that liquid lithium can certainly do the pumping and the FLIER results from the early 2000’s show that flowing lithium can remove the helium. Thus, this possibly provides a path to removing helium. Experiments were performed in EAST where the LiMIT/FLiLi limiter plates that were developed between a US-China collaboration from 2016 – 2020 were reused and deployed in helium plasmas. The limiters were moved from 12 to 3 cm from the plasma edge with flowing lithium and the plasmas were 10 s long. Evidence shows that there was a reduction in the local helium pressure around the divertor where lithium that that interacts with the helium is deposited and co-deposition of helium in lithium occurs. Correlating the particle flow through gas puff balance, plasma density, plasma temperature and stored energy is important in trying to understand the effect of lithium on helium. This paper will compare results in EAST with similar experiments that have been performed in HIDRA and MAGNUM-PSI previously.

Speaker: Mr Haoxuan Yu (University of Illinois Urbana-Champaign)

232 2.104 Dynamics of hydrogen isotope exchange in co-deposited LiD layers using operando ion beam

Liquid metals, and lithium (Li) in particular, are increasingly considered strong candidates for next-generation divertor concepts in fusion reactors due to their excellent power-handling capabilities and self-healing behavior. However, lithium’s high affinity for hydrogen isotopes and its tendency to form stable hydrides introduce important challenges for tritium (T) inventory control and fuel-cycle management in future power plants. Since some degree of tritium retention in Li-based divertors appears unavoidable, the development of efficient T-extraction methods is essential. In this work, we investigate hydrogen isotope exchange (IE) in lithium as a potential tritium mitigation technique and directly compare its efficiency with conventional thermal outgassing. Because of tritium’s radiological constraints, deuterium (D) and hydrogen (H) are used as experimental proxies.

We present the first dynamic isotope-exchange experiments performed in lithium using the newly commissioned linear plasma device Upgraded Pilot-PSI (UPP). UPP uniquely enables ITER-relevant plasma exposure combined with operando ion beam, allowing us to directly monitor Li and D contents during plasma interaction. Four identical lithium-filled capillary porous structures (CPSs) were exposed to deuterium plasma leading to Li evaporation. Co-deposited LiD layers formed on a nearby heated stainless-steel foil. Following complete lithium depletion, the CPS was exposed to H plasmas, converting the co-deposited LiD layers into LiH.

Isotope-exchange behavior was investigated from 220 °C to 400 °C. Across all conditions, deuterium concentration during hydrogen plasma exposure followed an exponentially decaying temporal behaviour, with decay constants exhibiting Arrhenius scaling. This temperature dependence demonstrates that isotope exchange in lithium is a thermally activated process. To evaluate reversibility, two samples were subsequently re-exposed to deuterium plasma, restoring LiD with an approximately 1:1 Li:D atomic ratio. This confirms that the IE mechanism is fully reversible.

For the remaining two samples, IE was directly compared to thermal outgassing. After D plasma exposure, the LiD layers were first held in vacuum at the same temperatures used during IE. During this period, Li₂O rapidly formed on the LiD surface, significantly reducing long-term D release through thermal desorption alone. Nevertheless, subsequent hydrogen plasma exposure demonstrated that isotope exchange remained highly effective, even in the presence of oxidized surfaces, and substantially faster than thermal outgassing.

Our results demonstrate that isotope exchange is a highly efficient deuterium removal method, capable of extracting essentially all retained D. IE can potentially be employed in tokamaks before in-vessel maintenance or integrated into a lithium loop system for tritium recovery, enhancing the tritium extraction rate.

Speaker: MARIA MORBEY (GNOI)

233 2.109 Recrystallization Kinetics of Tungsten Under Fusion-Related Conditions: Possible Pathways to Enhanced Microstructure Stability

Tungsten (W) is the leading candidate for plasma-facing components in future magnetic confinement fusion reactors due to its high melting point, low fuel retention, excellent thermal conductivity, etc. However, the elevated operating temperatures, high neutron flux and plasma flux expected in devices such as ITER, BEST, and DEMO can accelerate recrystallization, leading to grain boundary embrittlement, a higher ductile-to-brittle transition temperature, and mechanical softening. Understanding how irradiation influences the intrinsic recrystallization behavior of tungsten has therefore become a central challenge in predicting material lifetime and designing microstructures with improved stability.
Existing experimental and modeling studies provide essential but limited insight into the kinetics of irradiation-affected recrystallization. Irradiation introduces point defects, voids, dislocation loops, and transmutation products that can simultaneously act as recrystallization drivers and inhibitors. On the one hand, defect accumulation increases stored energy and promotes diffusivity, lowering the onset temperature for recrystallization; on the other hand, the segregation of hydrogen isotopes, He, and other transmutation products may retard dislocation and grain boundary mobility. Current efforts reported in the literature span post-irradiation microstructural analyses and annealing experiments, ion-irradiation studies, phase-field modeling, cluster dynamics simulations, and extended Johnson–Mehl–Avrami–Kolmogorov (JMAK) kinetic frameworks. Yet, systematic comparisons remain challenging due to variation in irradiation conditions, flux, thermal histories, and initial microstructure.
In this work, we analyze available experimental and simulation datasets to identify consistent trends governing W recrystallization, with and without irradiation, and to outline potential pathways to enhance microstructure stability. By evaluating defect-induced energy storage, grain boundary mobility suppression, and the influence of alloying strategies, we establish a conceptual kinetic map that highlights the competing mechanisms controlling recrystallization onset and progression.

Speaker: Chao Yin (University of Science and Technology of China)

234 2.110 Microstructural and mechanical property characterization of tungsten fiber-reinforced copper alloys

A significant challenge on the path to magnetic confinement fusion power plants is the robust exhaust of power and particles. In particular, plasma‑facing components (PFCs) in the divertor region must endure challenging particle and heat fluxes while simultaneously sustaining damage from fusion‑neutron irradiation. This leads to a progressive degradation of key material properties such as strength, toughness and thermal conductivity. For next‑step devices — including ITER and DEMO‑class reactors — divertor target concepts have therefore been developed that, in essence, join a monolithic tungsten (W) armor to a high‑conductivity copper (Cu) alloy heat sink. This architecture is intended to combine the plasma‑compatibility of W with efficient heat removal capability by a Cu‑based structure. In such a design, the mechanical stability of Cu‑based alloys is essential and must be ensured throughout the PFC's operational lifetime. However, the degradation of mechanical properties of Cu-alloys due to high-temperature operation and neutron irradiation remains a substantial challenge for divertor PFC design. One promising route to maintain the mechanical properties at elevated temperatures and even under neutron irradiation is the use of high-strength drawn W fibers in a Cu alloy matrix. The contribution will present work regarding the characterization of such advanced composite materials. In this context, material specimens based on W fiber weaves with two different fiber volume fractions have been infiltrated with oxygen-free Cu, Cu-Chromium and Cu-Chromium-Zirconium alloy, in order to study the influence of the different preforms and matrix materials on the composites properties. The specimens were analyzed by means of scanning electron microscopy (SEM) with electron backscatter diffraction (EBSD) to reveal their microstructures after infiltration. In terms of mechanical property characterization, tensile tests at room temperature and above have been carried out while the specimens showed an increased strength compared to the monolithic Cu-based alloy matrices. Additionally, fractography was carried out to understand the failure mechanism. The results of this study led to the fabrication of tensile specimens for a neutron irradiation campaign in the BR2 reactor (Mol, Belgium) which will be conducted in the near future.

Speaker: Dr Hanns Gietl (Max Planck Institute for Plasma Physics, 85748 Garching, Germany)

235 2.111 A Systematic Analysis of Runaway Electron Energy Deposition in Plasma-Facing Components

Runaway Electrons (REs) generated during disruption events pose a critical challenge for the operation of next-generation fusion devices such as DEMO. The thermal quench causes intense heat and particle loads on plasma-facing components (PFCs), while the current quench can induce a strong toroidal electric field that accelerates electrons to relativistic energies. A detailed understanding of how REs deposit their energy on PFCs is therefore crucial to proper design the reactor plasma-facing components.

This work presents a systematic predictive modelling approach based on the FLUKA Monte Carlo code [1] to quantify RE energy deposition under DEMO-relevant conditions. A comprehensive look-up table is built by performing a parametric scan over four key variables: electron kinetic energy, pitch angle, magnetic field intensity, and magnetic incidence angle on the solid surface. The explored energy range was identified through RE distribution function calculations on DEMO disruption simulations, considering intrinsic argon impurities.

The study first considers a reference configuration consisting of a flat tungsten monoblock, where both the total deposited energy and the depth at which 90% of the energy is absorbed are evaluated. RE impact on the surface is analysed systematically to identify the influence of each parameter on the deposited energy. The methodology is then extended to a shaped limiter geometry. This allows assessment of the role of geometrical effects, by performing a detailed comparison with the reference case.

The results suggest that the pitch angle may have a notable influence on the interaction between REs and PFCs. For identical electron energies and magnetic field parameters, small pitch angles lead to significantly higher loads. Under the worst-case conditions, more than 90% of the RE incident energy is deposited in the W target, and electrons can deliver a considerable amount of energy up to a depth of 25 mm. These insights provide a solid basis for predicting RE-induced loads and for optimising the PFC design and mitigation strategies.

References:
[1] A. Ferrari, P.R. Sala, A. Fassò, and J. Ranft (2005). FLUKA: A Multi-Particle Transport Code. CERN-2005-010, INFN/TC_05/11, SLAC-R-773.

Speaker: Mr Alessio Villa (Politecnico di Torino)

236 2.112 INVESTIGATING THE INFLUENCE OF DOPING AND ADDITIVE MANUFACTURING TECHNIQUES ON THE PLASMA MATERIAL INTERACTION RESPONSE OF TUNGSTEN

Tungsten (W) is the primary candidate material for plasma-facing components in the fusion divertor. However, the operational limits of conventional W necessitate the development of advanced variants to mitigate issues regarding brittleness and thermal stability. This research focuses on determining how specific material modifications, compositional doping and additive manufacturing (AM), collectively influences the plasma material interaction (PMI) behaviour under divertor-relevant conditions. Experiments were conducted using the linear plasma device PSI-2 to simulate high-flux plasma exposure. The study investigates two distinct material pathways: First, we evaluated the effect of advanced doping on PMI response. While these doped W variants are primarily engineered to deliver increased low-temperature ductility, superior resistance to recrystallisation and radiation induced grain growth, this study isolates how the modified microstructure influences surface evolution, damage mechanisms and retention of deuterium during plasma exposure, benchmarking this behaviour with ITER grade W. Secondly, we investigated the modification of PMI response driven by AM production methods. A comparative analysis is presented between Laser Powder Bed Fusion (L-PBF) and Electron Beam additive manufacturing. L-PBF W is used to establish the baseline PMI response for additively manufactured microstructures. This is contrasted with components produced via Electron Beam methods, a technique selected for its ability to maintain significantly higher base temperatures during fabrication. We assess how the distinct thermal histories and reduced residual stresses characteristic of the Electron Beam process alter the material's resilience to plasma-induced damage compared to the L-PBF baseline. By correlating these manufacturing and compositional variables directly with the surface morphological changes observed in PSI-2, this work clarifies the viability and progression of these novel W variants for future fusion energy applications.

Speaker: Simon Corah (School of Metallurgy and Materials, University of Birmingham)

237 2.113 Controlled Exposure of High-Temperature Ceramics in the Scrape-Off Layer of WEST

The WEST tokamak hosts a comprehensive set of diagnostics dedicated to monitoring heat flux, radiation, impurity transport, and local plasma parameters, providing detailed characterisation of its scrape-off layer (SOL). Among the available diagnostics, mobile Langmuir probes positioned at strategic SOL locations can be equipped with sample holders, enabling controlled exposure of material samples to extreme heat fluxes (10⁶–10⁸ W/m²) and ion fluxes (10²²–10²⁴ m⁻²s⁻¹). These capabilities allow WEST to operate as a testbed for evaluating advanced materials under plasma conditions that are otherwise challenging to reproduce. In particular, the accessible SOL heat and ion fluxes are suited for studying the thermal response and ablation dynamics of High-Temperature Ceramics (HTCs), whose excellent thermomechanical properties position them as promising heat-shield ablators for atmospheric re-entry vehicles. Two dedicated experiments have been carried out to test different HTCs thanks to the exploitation of two custom-made sample holders, fitted on two different exposure devices, targeting respectively the lower and the upper end of the accessible heat and ion flux range. The first system performs short plunges up to the separatrix on the top of the machine using a hydraulic drive, and exposed 10 BN samples to cumulated fluences up to 10²⁴ part/m². The second device, installed on the divertor, exposes a single sample to strike-point fluxes thanks to an in-situ magnetic drive, consisting of a coil that moves when a voltage is applied to it, due to the Lenz force in the tokamak magnetic field. Using the divertor device and a dedicated magnetic configuration, the ablation of a polycrystalline SiC sample has been successfully achieved, as confirmed by strong Si emission lines in VUV and visible spectroscopy, observed in both core and edge channels immediately after the exposure trigger, indicating high sublimation rates and motivating further plasma transport studies thanks to the unusual impurity release. Furthermore, divertor upper-view IR imaging shows a saturation of the sample temperature at ~1400 °C (underestimated by limited spatial resolution), behaviour consistent with a phase change. The average parallel heat flux on the sample’s leading edge, estimated with Fiber Bragg Gratings and flush-mounted divertor Langmuir probes, is about 45 MW/m². In-situ Langmuir probe collectors, together with complementary diagnostics, provide a complete record of the plasma conditions during the samples’ exposure. The experimental protocol is presented, as well as the analysis and interpretation of the data from the relevant diagnostics, and finally the post-mortem analyses on different HTCs samples.

Speaker: Bianca De Martino (CEA/IRFM)

238 2.114 The design of 3D-printing solid tungsten-liquid lithium combined divertor target plate and its interaction with steady/transient plasma

3D-printing tungsten-based porous matrix coupled with liquid lithium was designed and its intrinsic wicking capacity as well as its interaction with high-density plasma investigated through simulation and experiment. On account of present available fabrication technology, porous channels, inclination angle and arrangement were designed and optimized as finger-type unit based array with critical size of 150 μm. The simulative results showed that the porous matrix temperature, relevant thermal stress and wicking rate of lithium would be consistently within the safety limit of evaporation exhaustion and structural collapse under 10 MW/m2 heat load without the consideration of vapor shielding effect and redepositing effect. The experimental results verified its excellent wicking ability, temperature locking function and redeposition behavior. The liquid lithium wetting stability and composite structure reliability were confirmed through real-time surface monitoring during plasma irradiation and post-mortem appearance examination. Furthermore, ≥10 individual target parameters were optimized and relevant performance with plasma was investigated. It reveals an obvious enhancement in permeability and capillary performance factor, and significant suppression in matrix temperature and thermal stress even under high heat-load transient plasma (≥100 MW/m2). These results reveal the application potential and feasibility of 3D-printing porous capillary structure in plasma-facing components and provide a reference for further liquid-solid combined target design.

Speaker: Zongbiao Ye (Sichuan University)

239 2.116 DTT divertor Qualification activities: assessment of W/Cu Joining Technologies for Flat Tile end of the Central Target

The first divertor of DTT [1] will accommodate different magnetic configurations while maximizing heat exhaust capability. Therefore, most of its plasma-facing surface is based on ITER-like plasma-facing units (PFUs), consisting of CuCrZr cooling tubes protected with tungsten monoblocks [2]. To intercept the strike points of the reference magnetic configurations (Signe Null, X-Divertor, and Negative Triangularity), four targets are integrated into the divertor: Inner and Outer Targets for the SN configuration, and Central and Horizontal Targets for compatibility with X-D and NT.

The design and qualification of the Central Target (CT) are particularly challenging. Due to variations in radial position and PFU number, CT monoblocks are toroidally wider than those of the Inner and Vertical Targets, increasing their susceptibility to macro-crack formation under high thermal loads. In addition, the central location of the CT prevents shielding of PFU ends by First Wall modules, requiring a sharp 90° bend that cannot be protected by monoblocks. As a result, the current CT design includes a 25 mm long section clad with tungsten flat tiles, designed to withstand a heat load of 5 MW/m², mainly radiative.

To qualify both the design and manufacturing process, dedicated mock-ups were produced. These consist of copper blocks joined to tungsten flat tiles using different joining technologies and subsequently joined to the CuCrZr tube by diffusion bonding (HRP). Although not fully representative of the final component, the mock-ups were designed to assess the thermal fatigue performance of the W/Cu joint with tile dimensions foreseen in the design (26.5 mm toroidal and 8 mm poloidal). Two of the manufactured mock-ups—one with the W/Cu joint produced by diffusion bonding and one by molten copper casting—were tested at the GLADIS facility. The test protocol consisted of 5000 cycles at 5 MW/m² and 300 cycles at 10 MW/m². The high heat flux (HHF) tests were conducted in successive steps, and non-destructive ultrasonic inspections were performed to monitor joint quality as the number of cycles increased. In this work, the results obtained during the tests are described and the final metallographic investigation is presented.

[1] F. Romanelli et al., Nucl. Fusion 64, 111, (2024)
[2] S.Roccella et al., IEEE TRANSACTIONS ON PLASMA SCIENCE, 52, 9, 2024
[3] P. Innocente et al., Nuclear Materials and Energy, 33, 2022, 101276

Speaker: Selanna Roccella (ENEA)

240 2.117 Surface Morphology and Composition Changes of Additive Manufactured W Components Exposed to Liquid Li or Stellarator Plasma

Liquid lithium (Li) is considered an attractive candidate for Plasma Facing Components (PFC) in magnetic confinement fusion devices due to its high heat and particle exhaust capabilities, self-replenishing property, and potential for confinement improvement. Additive-manufactured structures made of refractory metal, such as tungsten (W), gained interest as solid substrates for Li PFCs to improve the uniformity and stability of Li flow.

Separate experiments are performed to expose additive manufactured W components to either liquid Li at 650 [$^\circ$C] for 100 [h] or a stellarator plasma of Hybrid Illinois Device for Research and Applications (HIDRA) up to 10000 [s] (fluence ~10$^{26}$ [m$^{-2}$] and electron temperature Te ~ 5-20 [eV]). Surface characterization with Scanning Electron Microscope (SEM) with Energy Dispersive X-ray Spectroscopy (EDS), X-ray Diffraction (XRD), X-ray Photoelectron Spectroscopy (XPS), and Laser Scanning Confocal Microscope (LSCM) is performed before and after the Li or plasma exposures to analyze surface morphology and composition changes.

Speaker: Kenta Kawashimo (University of Illinois Urbana-Champaign)

241 2.118 Recent analysis of tungsten dust in ASDEX Upgrade

Dust particles are recognized as a significant safety issue in the future fusion devices. When present in large quantities they may create an operational risk for the fusion device. Even small amounts, mobilized during a discharge, can cause additional radiation and influence the plasma performance. Dust particles can be observed by cameras, but these events are hard to interpret. In the 2024/25 campaign many dust events were observed during plasma operation in ASDEX Upgrade (AUG). These may be resulting from the in-vessel work during the installation of the new upper divertor, issues arising from the newly installed components or due to the new plasma operation scenarios. During the 2024/25 campaign, damages on the outer divertor tiles caused tungsten (W) melting, and as a result tungsten droplet production. Melting was also observed on the W coated protection tiles at the midplane and on the edges of the tiles at the new upper divertor. Besides melting, W particles also originate from arcing, sputtering and subsequent re-erosion.

Post-mortem analysis of the dust samples offers classification of dust particles. At AUG silicon (Si) wafers mounted in a protection box are used to collect dust mobilized by the plasma. Five dust boxes had been installed at the low and high field side of the midplane and in the divertor region. After exposure, the wafers were measured by an automated procedure using scanning electron microscopy, allowing identification of thousands of particles by size and composition [1].

For comparison samples installed at the same positions for the 2022 and 2025 campaign are used. The size distribution was determined for particles with different W content. Typically, the W containing particles also show oxygen and carbon content. The size distribution is limited by the magnification used for analysis and the number of pixels used to select a particle. Analysis of a surface area of 2 x 2 $\mathrm{mm}^2$ shows that, for the 2022 campaign, a total of 2700 articles were found in the 0.1 to 300 $\mu\mathrm{m}$ size range, whereas in the 2025 campaign, 7800 particles are observed. Using the size distribution similar amounts were found for particles below 0.3 $\mu\mathrm{m}$ and bigger than 3 $\mu\mathrm{m}$ during both campaigns. In the range between 0.3 and 3 $\mu\mathrm{m}$, up to 4 times more particles are produced in 2025. These results are discussed in the view of different tungsten dust production mechanisms.

[1] M.Balden et al., Nucl. Fusion, 54 (2014) 073010

Speaker: Dr Sangeetha Sasidharan (Max-Planck-Institut für Plasmaphysik, Garching, Germany)

242 2.119 Development of Retarding Field Analyzer for Spatial Distribution Measurements of Ion and Electron Temperature in Recombining Plasma

Excessive heat flux to the divertor plates is one of the most serious issues in magnetic fusion reactors. A detached divertor, in which neutral gas injection enhances radiation and volumetric recombination, is the most promising candidate for mitigating the divertor heat load. Thus integrated divertor simulation codes for future fusion reactors are actively developed. Linear plasma devices, which have high discharge reproducibility and allow detailed measurements in recombining plasmas, have been used for fundamental research on the divertor plasma physics. Additionally, they are recently playing an important role in benchmarking divertor simulation codes through detailed measurement in recombining plasmas. However, most measurements have focused only on electrons and excited atoms, and measurements of ions and comparisons with simulation results have rarely been performed. This has made it challenging to evaluate the accuracy of ion-related reaction and heat-transport models. In our previous work, we fabricated a Retarding Field Analyzer (RFA) and successfully obtained the radial profile of the ion temperature[1]. We are now developing a compact RFA that can translate along magnetic field lines. The axial RFA allows measurements of the axial distributions of both ion temperature and electron temperature in recombining plasmas.
The axial RFA is being introduced into the linear RF divertor simulator DT-ALPHA to measure detailed axial distributions of recombining plasma. DT-ALPHA can generate both helium ionization plasmas and steady-state recombining plasmas. Plasma is produced by $13.56\ \mathrm{MHz}$ radio-frequency discharge. Working gas is supplied from the upstream side of the device. In addition, helium secondary gas is supplied from the downstream side to enhance volumetric recombination. Typical parameters of the recombining plasma are approximately $T_\mathrm{e}\sim1.0\ \mathrm{eV}$, $n_\mathrm{e}\sim10^{18}\ \mathrm{m}^{-3}$.
The newly developed axial RFA consists of four grids and a collector. The distance between electrodes is $100\ \mu\mathrm{m}$, which allows analysis of the energy distributions of ions and electrons without being affected by space-charge limitations even in high density recombining plasmas. The outer and aperture diameters are $5\ \mathrm{mm}$ and $1\ \mathrm{mm}$, respectively, both of which are sufficiently smaller than the plasma diameter in DT-ALPHA. In the presentation, details of the axial RFA and experimental results for both helium ionization plasmas and recombining plasmas will be reported and discussed.
This work is supported by JSPS KAKENHI (Grant Nos. 24K00607 and 25KJ0528).
[1] S. Kagaya et al., Phys. Plasmas 32, 112103 (2025).

Speaker: Shigetaka Kagaya (Tohoku university, Sendai, Japan)

Wednesday 20 May

Review Talk: Morning session

R1

243 R3 Stellarator island divertor physics in Wendelstein 7-X: Current understanding and remaining challenges towards a reactor exhaust concept

Wendelstein 7-X (W7-X) operation successfully demonstrated power and particle exhaust with the island divertor. It is compatible with high performance operation using steady-state pellet fueling and homogenous detachment up to the current maximum heating power of 8 MW. However, the current open divertor geometry suffers from an unfavorable, approximately linear scaling of divertor density and neutral pressures with upstream density, which challenges the extrapolation of power and particle exhaust to reactors. The limitations seem to be the combined effect of the stronger prominence of cross-field transport due to lower fieldline pitch, the open divertor geometry and stellarator-specific geometry aspects, such as the target-shadow region. Luckily, W7-X provides a large flexibility in operational and magnetic configuration parameters. Modeling and experiment shows that this enables W7-X to operate in different boundary regimes that are dominated by different physics. This allows investigating the role of the complex geometry and relevant, leading-order physics.
Deducing physics purely from experimental measurements with limited spatial resolution is challenging in W7-X due to strong poloidal and toroidal asymmetries, but the fast advancement of imaging diagnostics allows to address this challenge. We observed clear signatures of the complicated 3D structure of the island divertor plasma, such as strongly toroidally localized radiation features or details of the target heat loads. In addition, first reactor-relevant scaling approaches for heat transport, the impact of drifts, and radiation capabilities are emerging.
New mean field and turbulence codes for the drift-reduced Braginskii model in 3D-geometries can now capture and allow understanding experimental observations such as the drift-driven evolution of counter-streaming flows towards uni-directional patterns, strong poloidal ExB drift flows along the last closed flux surface and particular turbulent features in the island.
Pushing reactor relevance of stellarators, the island geometry became part of the magnetic optimization. A more refined insight into relevant geometric island parameters (shear, pitch, size) to tailor their properties emerges. Divertor target optimization (closed vs open) addresses the observed limitation of the density-build up in W7-X. Both aspects show promising results for a consistent path towards a reactor-relevant exhaust performance.

The large variability of the magnetic configuration space of island divertors and other concepts, such as chaotic boundaries, make it a challenge to focus the research for fast progress. However, it also provides opportunities for innovative geometries to further improve the stellarator exhaust. Initial results show the potential for additional heat flux broadening, radiation control or more ‘tailored’ transport properties.

Speaker: Felix Reimold (MPPL)

Invited Talk: Morning session

244 I13 An Overview of Divertor Research in LHD

The Large Helical Device (LHD) experiments will conclude in December 2025, marking nearly three decades of divertor research. This milestone provides a unique opportunity to summarize the achievements and lessons learned from the development of the helical divertor concept and its role in steady-state stellarator operation.$\newline$
Early LHD experiments demonstrated that the open helical divertor produced neutral pressures below 0.1 Pa, leading to confinement degradation and limitations in long-pulse operation. To overcome these issues, the Closed Helical Divertor (CHD) was conceptually designed in the late 2000s using magnetic field-line tracing and three-dimensional neutral transport simulations with the EIRENE code. The CHD, fully implemented in the 2010s, introduced a geometrically closed divertor structure with integrated in-vessel pumping systems located in the inner toroidal sections. A combination of cryo-sorption and non-evaporable getter pumps provided a total effective pumping speed exceeding 80 m$^3$/s, enabling direct exhaust of more than 50% of the fueling particles [1]. This significantly reduced wall recycling, improved density controllability, and enabled long-pulse plasma operation.$\newline$
Systematic investigations of particle control during long-pulse discharges revealed the critical role of divertor pumping. In inward-shifted magnetic configurations ($R_{\mathrm{ax}}$ = 3.60 m), strong localization of particle flux was observed on the inboard side where the CHD modules were installed. Forty-second ECH-heated discharges demonstrated that divertor pumping suppressed wall saturation, maintained stable temperature profiles, and enabled superior density control. In contrast, discharges without pumping suffered from density rise and confinement degradation due to enhanced wall recycling [2]. Profile analyses showed higher core electron temperatures with pumping, exhibiting electron internal transport barrier (e-ITB)-like characteristics. Heat-conductivity evaluations confirmed that the confinement improvement was associated with reduced wall fueling and sustained low core thermal diffusivity.$\newline$
More recently, ultra-high neutral pressures up to 2.4 Pa—comparable to those in tokamaks—have been observed in $R_{\mathrm{ax}}$ = 3.55 m configurations. These regimes are associated with enhanced neutral compression, near-wall condensation, and radiation-driven detachment, achieved without significant degradation of core confinement [3].$\newline$
Overall, the LHD divertor program has established effective strategies for particle exhaust and recycling control in stellarators, providing a robust experimental foundation for steady-state reactor concepts. As LHD completes its final experimental campaign, its divertor studies leave a lasting legacy for next-generation devices such as W7-X and for the design of DEMO-class reactors.
[1] G. Motojima+, Nuclear Fusion, 59, (2019), pp. 086022.
[2] G. Motojima+, Phys. Scr. 97 (2022), pp.035601.
[3] U. Wenzel+, Nuclear Fusion 64 (2024), pp. 034002.

Speaker: Gen Motojima (National Institute for Fusion Science)

Oral: Morning session

245 O15 Explaining observed potential and drift phenomena in Wendelstein 7-X with the mean-field Braginskii model

Particle drifts play a crucial role in the understanding of the edge physics of current and future fusion devices. While tokamak simulations routinely include drift effects, they have been largely neglected in stellarator simulations until now. The low field line pitch in a stellarator amplifies the relative importance of drifts compared to the parallel transport. Experiments in Wendelstein 7-X show observations which we can explain by the inclusion of drifts and currents. In this contribution, we present a mean-field Braginskii model that is applicable to stellarator geometries. The model is built on the BOUT++ framework and is a derivative of the Hermes model. It includes a self-consistent description of the electrostatic potential as well as drift effects. We will present first-of-its-kind simulations of the scrape-off layer of Wendelstein 7-X and use the simulations to interpret and address multiple experimental observations.

Strong poloidal ExB velocities on the order of $km\,s^{-1}$ and multiple shear regions are measured by Langmuir probes and the gas puff imaging diagnostic [Killer 2025]. The simulations reproduce ExB velocities of similar magnitude and indicate a strong increase in flow velocities in areas of densely packed flux surfaces. A region of strong ExB shear forms inside the scrape-off layer, which is qualitatively consistent with the measurements.
Coherence imaging diagnostics measure a dependence of the stagnation point on field direction for low density plasmas [Perseo 2021,Kriete 2023]. The same phenomena occur in the simulations due to parallel momentum transport by poloidal ExB flows.
Detailed analysis of the simulations shows that the electrostatic potential in the scrape-off layer mainly arises from the sheath potential drop, which gets modulated in the parallel direction by the parallel temperature and pressure gradients. Shallow gradients between the X-points compared to stronger gradients between the X-points and the divertor targets lead to potential isocontours that are predominantly aligned with the LCFS of the confined plasma.
The inclusion of drifts in the boundary modeling tools now allows for an advanced quantitative understanding of observations and should provide a more robust foundation for extrapolations to future reactors.

Speaker: Tobias Tork (MPPL)

09:30 Coffee Break

Oral: Morning session

246 The PULSE regime: pellet-utilized long-pulse high-performance operation in W7-X

Stellarators intrinsically avoid pulse-length limitations of the magnetic confinement, since the rotational transform is generated by external coils rather than by a plasma current. We report on the development of discharge scenarios that enable long-term stabilization of high-performance plasmas in the optimized stellarator W7-X. For the first time, dimensionless parameters of reactor relevance are achieved within stellarator plasmas sustained for up to 43 s. The attained fusion triple products are comparable to those of JET at similar pulse lengths but at about one third of the plasma volume.

The reported scenarios exploit repetitive pellet fueling combined with second-harmonic electron-cyclotron O-mode heating, which together reliably raise the ion temperature beyond the ion-temperature clamping limit and sustain a quasi-stationary operational cycle at high density and plasma energy. Surprisingly, these plasmas exhibit indications of central fueling. In order to stay within technical heat-load limits, targeted impurity seeding within this PULSE regime (Pellet-Utilized Long-pulse Steady-state Enabled) mitigates heat loads without accumulation of the seeding species.

The PULSE-regime proves robust: irregularities in pellet size reduce plasma performance, but recovery is observed once nominal fueling is restored. In first experiments, central ion temperatures exceeding 2 keV at central densities above $1.5\times 10^{20}$ m$^{−3}$ were obtained, reaching central beta values up to 4% while approaching confinement as expected form the ISS04 scaling ($f_{ren} \approx 0.8 \ldots 1.1$).

While first experiments (reported here) were limited by technical availabilities of components, the PULSE-regime appears to be controllable by adapting the pellet mass, velocity and frequency as well as heating actuators. Therefore, the approach is suggested to be further assessed as a potential scenario for stellarator reactor operation.

Speaker: Andreas Dinklage (MPPL)

Invited Talk: Morning session

247 I14 Divertor dynamics of pellet-fueled discharges in DIII-D

Experiments and modeling in DIII-D highlight the key parameters determining the dynamic change in divertor conditions with pellet fueling. In future reactors, pellet injection will be necessary for effective core fueling[Kukushkin2003NF] since a high-performance scenario typically necessitates a scrape-off layer (SOL) that is opaque to neutrals. Due to increasing alpha heating with density, the detachment level in reactors will depend on the density evolution cycle imposed by pellet ablations[Wiesen2017NF]. Integrated modeling in ITER finds that the associated increase in power crossing the separatrix ($P_{SOL}$) via convection may result in divertor reattachment with each pellet.
To explore this phenonmenon, experiments have been performed in DIII-D injecting up to 4.5$\times10^{21}~$atoms/s of D pellets from the high-field side, in a range of plasma scenarios ($B_t=+/-2.2~$T, $|I_p|=0.6~-~2.0~$MA, $P_{inj}=3.0~-~9.0$~MW, $n_G=0.24~-~0.82$). These include type-I ELMy and ELM-free (via negative triangularity, RMP suppression and QH-mode) scenarios. In the majority of cases, pellet injection (with density rise up to $\psi_N\sim0.6$) results in transient detachment of the divertor if starting attached, and a transient deepening of the detachment level if starting detached, lasting $\sim25~$ms. This is consistent with results in EAST attached plasmas with shallow core fueling ($\psi_N>0.94$)[Deng2017PST]. Well-diagnosed ELMy plasmas demonstrate transient detachment with target temperature $T_{e,t}<2~$eV and up to 35% reduction in outer target power loads ($P_{targ}$). Meanwhile, ELM-free scenarios reveal that the transient detachment is preceded by a transient increase in $P_{targ}$ lasting $\sim1~$ms. However, in an ELM-free scenario with relatively steep edge density gradients, no detachment enhancement is observed with pellet injection and $P_{targ}$ increases over $\sim25~$ms.
Time-dependent SOLPS-ITER simulations have been performed to study these dynamics, including time-dependent EIRENE. The pellet is emulated with a transient plasma source inside the confined region, and the radial source profile can be scanned. Experimentally observed divertor dynamics are qualitatively reproduced in simulation, with divertor conditions impacted over two distinct timescales. On shorter timescales ($\tau_1$), transient increases in $T_{e,t}$ and $P_{targ}$ are observed, with the magnitude determined by the pellet-induced $P_{SOL}$ increase. On longer timescales ($\tau_2$), transient reductions in $T_{e,t}$ and $P_{targ}$ are observed, with the magnitude determined by the increase in particle flux crossing the separatrix. The simulated timescales are governed by the pellet deposition location and the energy and particle confinement times, spanning $\tau_1\gtrsim2~$ms and $\tau_2\gtrsim10~$ms in a H-mode plasma, increasing with pellet penetration depth. To ensure continuous core-edge integration in future devices, the pellet deposition location may need to be optimized to the plasma scenario.

Speaker: Sophie Gorno (ORNL)

Oral: Morning session

248 O17 Integration of high normalized beta core with divertor detachment in DIII-D

Recent DIII-D experiments have demonstrated the compatibility of divertor detachment with a high normalized beta core, a significant step towards solving the core-edge integration issue for steady-state fusion. The high-confinement, high-beta operation leveraging both high beta hybrid and high poloidal-beta approaches greatly improves core performance under dissipative divertor operation. Experimental analysis and simulations reveal that a closed divertor with impurity seeding and opaque edge, a widened pedestal with optimized radiation front, and improved core stability play key roles on the improvement of core-edge integration. These results pave a promising path to improve the integration of high-performance core and dissipative edge.
Using a high-density hybrid scenario, simultaneously high-confinement high-beta core (H98 ~1.1, βN ~3.1), low collisionality pedestal with v*ped~0.5 and Te,ped ~1keV, naturally small ELMs with frequency >500Hz, and partial detachment with significant pressure loss near the strike point have been obtained. With moderate gas puffing, the pedestal density gradient was reduced via edge neutral opacity, which promotes divertor dissipation and the achievement of small ELMs. Increasing heating power does not significantly increase the detachment threshold, which is attributed to the broadening of upstream profiles. Additional divertor nitrogen gas puffing further enhances the divertor dissipation to achieve partial detachment.
Separately, using high-poloidal-beta plasma operation, DIII-D experiments have demonstrated the compatibility of high-confinement core with nearly-full-discharge divertor detachment. Partial detachment with divertor temperature Te<10eV was achieved shortly after the L-H transition and sustained for the entire flattop phase, while the energy confinement is significantly improved with H98 increased to 1.7 and maintained, which is attributed to the formation of a large-radius internal-transport-barrier (ITB).
Starting from the low-density hybrid scenario, shallow XPR (peak radiation right inside the X-point) with complete divertor detachment and high beta (βN ~3.0, H98~1.25) have been simultaneously achieved using an ITER-similar shape and nitrogen puffing. With stronger N2 impurity injection, the plasmas exhibit a deep XPR regime with 2x core radiation peaking in the pedestal, while ELMs are strongly mitigated. However, the confinement is significantly reduced to H98<1.0, which is attributed to the 50% lower pressure and 30% colder pedestal temperature. By leveraging the large-radius ITB in high poloidal-beta plasmas, deep XPR plasma with ELM suppression could be integrated with the high beta core, while the confinement maintained. The ITB compensates the potential performance degradation from the weaker pedestal due to the formation of XPR, even with radiation front deeply inside the separatrix

Speaker: Huiqian Wang (General Atomics)

Invited Talk: Morning session

249 I15 Plasma-Neutral Interactions and Atomic Processes in Detached Divertors under High-Density Transient Pulses

Edge localized modes (ELMs) deliver bursts of energy and particles that can exceed the survivability limits of plasma-facing components. While plasma detachment provides a promising scenario for reducing steady-state heat fluxes, its compatibility with transient events remains poorly understood. It is unclear whether detached plasmas can sustain their protective role under ELM-like pulses or instead undergo reattachment leading to excessive heat loads. Addressing this issue is essential for developing reliable divertor scenarios.
Pulsed plasma experiments were performed in the superconducting linear device Magnum-PSI by superimposing high-density pulses onto steady-state detached helium (He) plasmas. The device produces ITER-relevant high flux plasmas and ELM-class pulses [1]. Neutral pressure near the target was independently varied, enabling systematic changes from weakly to deeply detached plasma regimes. Diagnostics included time-resolved laser Thomson scattering, optical emission spectroscopy, and ion current measurements at the target. This combination of controllable pulsed plasma generation and comprehensive diagnostics examine how transient events interact with plasma detachment and neutral dynamics.
The experiments revealed that the dynamic response of detached plasmas to the heat pulses is strongly governed by transient recycling fluxes and atomic processes. At lower neutral pressures, while the initial part of the pulsed ion flux reached the target, the tailing part was truncated. Since this attenuation was not observed in upstream diagnostics, it is attributed to a localized feedback effect caused by dynamic recycled neutrals. Modeling with a coupled plasma-neutral fluid code demonstrated that the dynamic pressure induced by the recycled neutrals provided sufficient momentum loss to stagnate the pulsed plasma before reaching the target. At higher pressures, enhanced electron-ion recombination decreased the overall pulse intensity; however, its mitigating role was limited when strong pulses depleted neutrals in the plasma column. Time-resolved spectroscopy indicated that recombination occurred sequentially: He$^{2+}$ first recombined to He$^{+}$, followed by recombination into He atoms, with a measurable time delay. This finding has direct implications for He ash behavior in ITER, where He$^{2+}$ is expected to be the dominant charge state transported by ELMs. These results highlight two key mechanisms: dynamic suppression of pulsed ion flux by recycled neutrals and the limitation of energy dissipation due to neutral depletion and delayed recombination. These effects provide new insights into the transient response of detached divertors and emphasize the necessity of considering local plasma-neutral coupling for future fusion reactors.

[1] H.J.N. van Eck, et al., Fusion Eng. Des. 142 (2019) 26.

Speaker: Yuki Hayashi (Graduate school of Frontier Sciences, The University of Tokyo)

Oral: Morning session

250 O18 Collisional sheath impact on ITER divertor plasma and power exhaust

Plasma detachment is a common divertor regime in tokamaks and stellarators. Plasma and power exhaust properties in this regime are described by classical magnetized plasma models including the plasma sheath [1]. However, the classical sheath model can fail with increasing plasma density and collisionality when plasma ions can be demagnetized and no longer follow magnetic field lines. Consequently, in very high density, cold divertor plasmas (density at the sheath entrance, nd > 1021 m-3) a new collisional sheath regime will be formed [2]. In the collisional sheath, plasma transport is dominated by elastic and inelastic plasma and neutral particle collisions. As a result, the sheath properties differ significantly from the classical situation. For example, the plasma flow is subsonic, the normalized divertor heat load and sheath potential drop are reduced and the energy and angular distribution of absorbed particles flatten. Under such conditions, models of particle and power exhaust require revision.
In the present work, we use sophisticated kinetic particle-in-cell simulations with the BIT1 code to explore new properties of the collisional sheath and describe the consequences of their formation in next generation fusion devices such as ITER. We demonstrate that the divertor target plasma profiles will broaden, reducing peak heat loads by a factor ~2, and that a significant portion of these loads will be carried by neutral particles.
In the presence of the collisional sheath the plasma density in front of the target will exceed the value estimated from the classical expressions, affecting plasma-impurity reactions and net sputtering rates of divertor material. For example, in the case of tungsten (W) target material (as in ITER) prompt re-deposition of sputtered W will be negligibly small (below 20%), and contrary to expectations, when typically, more than 90% of sputtered W is promptly redeposited, even a small fraction of energetic ions (e.g. highly ionized impurity ions) can cause non-negligible net erosion. Our model indicates that for detached, high density, cold ITER divertor plasmas, W net erosion resulted by impact of seeded impurity ions (Ne) can reach 1020 atoms/m2s. An additional negative consequence of the collisional sheath will be increased penetration of fuel ions and neutrals into the divertor tile gaps. We finally discuss the impact that the collisional sheath may have on Langmuir probe measurements in ITER.

[1] P.C. Stangeby, The Plasma Boundary of Magnetic Fusion Devices, IOP, 2000.
[2] D. Tskhakaya, invited talk at the 47th EPS conference, Barcelona, 21-25.06.2021.

Speaker: Dr David Tskhakaya (Institute of Plasma Physics of the CAS, Za Slovankou 1782/3, 182 00 Prague 8, Czech Republic)

251 O19 Unprecedented characterisation of ion kinetics in the tokamak divertor from attached to detached regimes

A new Tangential Divertor Spectrometer System (TDSS) on the TCV tokamak provides, for the first time, simultaneous measurements of ion temperature and parallel flow poloidal and radial profiles in the divertor Scrape-Off Layer (SOL). High spectral resolution enables the identification of the emission region through Zeeman splitting, allowing for the reconstruction of radial profiles using the unique plasma shaping capabilities of TCV. Understanding the main ion and impurity kinetics is essential, as they significantly affect the in-out symmetry of radiation in the divertor and can lead to increased core contamination. In addition to previously presented temperature and density measurements using divertor spectroscopy, dedicated experiments were carried out in Ohmic L-mode plasmas with divertor conditions ranging from detached to attached, combining TDSS with Thomson Scattering, multispectral imaging (MANTIS) and reciprocating probe measurements to characterise the SOL over a range of densities. The magnetic field and plasma current directions were changed to identify the impact of magnetic drifts on the radial profiles of flow and temperature. Measurements of C$^{2+}$, He$^{+}$, and N$^{+}$ line emissions revealed that these impurity species exhibited similar flow velocities and temperatures along the divertor leg. However, C$^{+}$ was systematically colder and slower, consistent with weaker entrainment and thermalisation with the main ion population. For the first time, radial profiles of ion $T_i$ and electron temperature $T_e$ revealed $T_i> T_e$ in the divertor far SOL and $T_i \leq T_e$ near the separatrix. Magnetic drifts had a profound effect on the ion flow profile, resulting in a significant change in ion energy transport along the leg. With increasing core density, the impurity flow increases for the so-called favourable (for H-mode) magnetic field direction but is unchanged for the unfavourable magnetic field direction. SOLPS-ITER simulations with magnetic drifts reproduce the upstream and divertor electron density and temperature profiles, and capture the experimentally observed dependence of the C$^{2+}$ flow profile on core density and magnetic field direction. This combined experimental and modelling study revealed that ion kinetics are somewhat decoupled from the electrons in the outer divertor leg and depend strongly on plasma conditions. These findings underscore that ion-driven energy fluxes can constitute a substantial fraction of the total SOL power balance, making their accurate characterisation essential for understanding divertor energy fluxes.

Speaker: Richard Ducker (EPFL)

12:20 Lunch

14:00 Outings

Thursday 21 May

Review Talk: Morning session

R1

252 R4 Progress in understanding and modelling of fuel retention in view of future fusion devices with metallic walls

For future burning-plasma fusion devices that will operate with a D-T fuel mix, the scarcity and radioactivity of tritium represent critical constraints on reactor safety and availability. Therefore, it has long been recognized that tritium retention and accumulation in Plasma-Facing Components (PFCs), along with its permeation into structural materials and cooling loops, require continuous monitoring and control. These topics have been addressed by extensive experimental studies, ranging from fundamental laboratory-scale investigations to machine-scale experiments in tokamaks and stellarators, with analysis methods ranging from local in-situ diagnostics to global gas balance measurements and ex-situ post-mortem analysis. In parallel, modelling efforts have provided important insights into hydrogen transport and retention, spanning first-principles calculations, atomistic and continuum approaches, up to whole-device scale analysis. The recent ITER re-baselining with tungsten as the material for both the first wall and the divertor, together with the prospect of next-step burning-plasma devices expected to reach high neutron fluence, call for a comprehensive summary and critical revision of the accumulated knowledge.
This contribution reviews recent advances in the description and understanding of fuel retention mechanisms in metallic PFCs, based on both experimental and modelling studies. This includes the implications of the transition from a beryllium to a tungsten first wall in ITER, with particular attention to boronization and associated fuel co-deposition; the role of self-ion- and neutron-induced displacement damage; isotope effects in retention and permeation; the application of in-situ fuel retention diagnostics; global-scale fuel recovery and accountancy; aspects of tritium breeding; progress in code development and validation; integrated modelling, and associated uncertainty quantification. Notable recent progress includes the development of open-source macroscopic rate-equation codes FESTIM and TMAP8 for advanced tritium transport modelling; the development of the global-scale simulation framework Hydrogen Inventory Simulations for PFCs (HISP) and its application to fuel retention and outgassing analysis in tokamaks; the recent record D–T campaigns at JET and subsequent tritium clean-up experiments, including successful demonstration of in-situ laser-based local fuel retention measurements that are also in planning for ITER; experimental and modelling studies addressing the creation and annealing of displacement damage in tungsten and fusion relevant steels. The remaining knowledge gaps and forthcoming challenges in view of reactor-scale fusion devices will be summarized.

Speaker: Dmitry Matveev (FZJ)

Oral: Morning session

253 O20 He and D retention in bulk-W lamellae from the JET divertor measured by Laser-Induced Desorption with melting

In the final divertor setup of JET, the divertor bottom consisted of 4 toroidal rings of thin, separated bulk-W lamellae called “tile 5” or load-bearing septum replacement plates (LBSRP). [1]
Three of the lamellae studied here had been exposed during JET plasma operation only during the ITER-like wall campaign 1 (ILW1) and another three lamellae in addition during the ITER-like wall campaign 3 (ILW1 + ILW3). The divertor plasma time in the ILW1 campaign was 13 h and in the ILW3 campaign 18.5 h. The base temperature of the LBSRP was 70° C, while up to 1200° C was reached during plasma operation. They were only cooled inertially.
All 6 W lamellae were then cut in small samples and analysed ex situ by Laser-Induced Desorption (LID) with melting of 2 mm diameter spots using a 3 ms Nd:YAG laser pulse of 64 J in FREDIS [2]. The laser melts about 20 µm of the surface, the retained gases desorb and are quantified by Quadrupole Mass Spectrometry (QMS). FREDIS is equipped with a high-resolution QMS, that can separate the mass peaks of He and D₂ and measures masses 1 to 6 amu/e, while an overview QMS measures masses 1-50 amu/e with low resolution. The high-resolution QMS was especially useful in this case as the He and D₂ peaks at 4 amu/e were similarly high and thus distinct from each other.

LID-QMS measurements with melting were performed in 27 positions on the top surface of the lamellae, which does not show colour differences, while the side surfaces show colourful deposits. On the side surfaces of the lamellae 18 radial scans with 3 positions each were performed in a distance of 2, 5 and 7 mm from the top surface. In some positions the retention is about one order of magnitude larger on the side surface than on the top surface but mostly has a rapid decrease in the radial direction away from the plasma.

Afterwards some of the samples were cut through the centre of the laser spots, polished and etched. These metallographic cross-sections show the melting depth indicated by vertical grain boundaries, the heat affected depth, cracks at some laser spots and bubbles at the bottom of the molten volume.

[1] Ph. Mertens et al., Fus. Eng. Des. 84 (2009) 1289-1293,
doi:10.1016/j.fusengdes.2008.11.055
[2] M. Zlobinski et al., Fus. Eng. Des. 146 (2019) 1176-1180,
doi:10.1016/j.fusengdes.2019.02.035

Speaker: Dr Miroslaw Zlobinski (Forschungszentrum Jülich GmbH)

Invited Talk: Morning session

254 I16 Hydrogen co-deposition with fusion-relevant materials

Tritium accumulation is going to be a serious safety issue in future fusion devices. Accumulation in co-deposited layers is one of the main channels of hydrogen isotope accumulation in tokamaks, at least for low Z-materials [1]. Even W layers can contain up to 5-10 at.% of hydrogen, with some works reporting up to 20 at.% D content [2]. Their properties vary greatly with deposition conditions, such as deposition temperature, deposition rate, hydrogen particle energy and flux during deposition. The range of parameters encountered in fusion devices complicate experimental study in laboratory conditions. This makes it necessary to develop predictive models and scaling equations for hydrogen content in co-deposited layers [3,4]. There are still a lot of questions remaining, such as the difference between retention of different hydrogen isotopes (H vs. D vs. T) and their mixtures, the effect of He on hydrogen accumulation, the role of hydrogen implantation energy on the structure of the co-deposited layer and hydrogen accumulation.
In this work, a review of available experimental and theoretical data is presented for hydrogen isotope co-deposition with a number of fusion-relevant materials, including W, B, and Li. The commonly used methodologies for laboratory studies of co-deposition, and their pitfalls are discussed, such as the differences between post-mortem and in-vacuo analysis, the difficulties in setting up and interpreting experiments with varied hydrogen implantation energies, etc. Applicability of both empirical scaling [3] and theoretical diffusion-advection model [4] are discussed and compared, based on available experimental data, both from literature sources and newly produced data. Experimentally observed difference between H, D and H-D mixture co-deposition with W and B is shown, with an eye towards possible implications for co-deposition in future D-T using fusion devices, including ITER.
Based on the available data, the rate of possible tritium accumulation in ITER is estimated and compared with other predictions. Possible ways of removing tritium from co-deposited layers are demonstrated, such as exposure to light hydrogen isotopes at elevated temperatures up to ~200 °C, which is shown to have > 98% effectiveness in removing deuterium from W-D and Li-D co-deposits.

Speaker: Stepan Krat (NRNU MEPhI)

Oral: Morning session

255 O21 Simultaneous self-damaging and deuterium plasma exposure leading to increased deuterium retention in tungsten

Increased hydrogen isotope retention in materials damaged by energetic neutrons has been a major concern for the operation of fusion reactors, as it leads to the loss of precious fusion fuel. Laboratory studies on the effect of such damage on deuterium (D) retention in tungsten (W) rely primarily on self-ion damaging of the material as a surrogate for neutrons. Most such studies were performed in the sequential manner—the W samples were first self-damaged and later exposed to D to populate the defects. The first studies conducted in the simultaneous manner with low-flux D atoms and ions [1,2] revealed an important role of the synergistic effects and D retention, differing largely from the sequential case.

The new POSEIDON experimental system, consisting of the PISCES-RF linear plasma device coupled with a 3 MV ion beam accelerator, allows us to conduct simultaneous damaging and plasma exposure at fusion-relevant D fluxes. Four W samples were exposed to D plasma at 333 K to a flux of $1.8\times 10^{22}\;\mathrm{D/m^2s}$ for 3 hours, resulting in a fluence of $1.9\times 10^{26}\;\mathrm{D/m^2}$. After 1 hour of plasma exposure, a defocused 20 MeV W$^{6+}$ ion beam was used to self-damage the samples during the ongoing D plasma exposure. The damage rate was around $6\times 10^{-5}\;\mathrm{dpa/s}$, and the damaging time varied from 40 s to 2 hours, resulting in 0.002, 0.02, 0.2, and 0.35 dpa. After the damaging, the D plasma exposure continued until the desired fluence was reached.

He-3 nuclear reaction analysis (NRA) was utilized to measure D depth profiles in the simultaneously damaged and plasma-exposed samples. The depth profiles indicate that D did not diffuse throughout the entire expected thickness of the damaged region. Still, D concentration shows a 3- to 4-fold increase compared to sequentially self-damaged and plasma-exposed samples [3], despite D not yet reaching the depth of the peak damage. Preliminary thermal desorption spectroscopy (TDS) spectra show a strong increase of the low-temperature peaks and a moderate decrease of the high-temperature peaks for an overall ~50% increase in total D amount. Additional D plasma exposures will be conducted to populate all the remaining defects, and additional NRA and TDS analyses will be performed.

This work is supported by US-DOE-FES under agreement DE-SC0022528.

[1] S. Markelj et al., Nucl. Mater. Energy 12 (2017) 169
[2] S. Markelj et al., Nucl. Fusion 59 (2019) 086050
[3] T. Schwarz-Selinger, Mater. Res. Express 10 (2023) 102002

Speaker: Anže Založnik (UCSD)

09:50 Coffee Break

Invited Talk: Morning session

256 I17 X-point radiation stability, operational space and performance impact in DIII-D H-mode discharges

Stable X-point radiation (XPR) was obtained via impurity seeding in DIII-D experiments ($I_p$=0.8-1.3MA, $P_{inj}$=6-12MW) for characterization of XPR access, validation of radiation stability models, and assessment of impact of XPR on H-mode pedestal. XPR regimes [1] have gathered interest for future devices thanks to the simultaneous elimination of steady-state and transient heat fluxes without external ELM-control actuators. While such regimes were achieved in many tokamaks, extensive diagnostics, high $P_{inj}$ and flexible shaping in the DIII-D open divertor enable improved understanding of operational access requirements, radiation stability and impact on confinement.
X-point radiation was accessed from detached conditions via impurity seeding ($CD_4$, $N_2$). After a rapid transition at detachment onset, the radiation front gradually moves inside the X-point, indicating controllability of XPR access. At high $I_p$, deeper XPR resulted in back-transition to L-mode while maintaining X-point radiation, without unstable MARFE evolution. A narrower operating space was observed with C as primary radiator in terms of accessible depth of XPR before back-transition. Unlike theoretical predictions, no difference in unstable MARFE evolution was observed between C or N-dominant radiators. Thomson scattering measurements inside the X-point indicate $T_e\sim$1-2eV, $n_e\sim3-6\times10^{20}m^{-3}$ with a reduction in $p_e$ with respect to upstream in the last 1$\%$ of $\psi_N$. $I_p$ scans isolated the role of connection length, $L_{\parallel}$, on stability and accessible XPR depth, with MARFE evolution observed at longer midplane-to-XPR $L_{\parallel}$, consistent with analytical models. Once the XPR is established, only a marginal role of X-point height $h_X=0-12cm$ is found, highlighting the potential for compact radiative divertor configurations.
Access to XPR regimes was accompanied by a reduction in confinement $\sim20\%$ compared to detached conditions. While the total radiated power fraction is nearly unchanged from deep detachment to XPR, the core radiated power fraction increases leading to a reduction in $T_{e-ped}$ and enabling access to ELM mitigation. The XPR phase was accompanied by a transition from Type I (100Hz, $\Delta W_{ELM}/W\sim1.2-1.5\%$) to Type III ELMs (300Hz, $\Delta W_{ELM}/W\sim0.3-0.5\%$). While the SOL power lost via ELMs remained $\sim10-20\%$ of $P_{INJ}$, the energy lost per ELM and the peak divertor heat fluxes were reduced by 4$\times$ and 10$\times$ respectively, indicating increased ELM buffering. Pedestal dilution can become large with carbon and nitrogen concentrations simultaneously up to 3$\%$. Coupling of XPR regimes to advanced tokamak scenarios and closed divertors will be necessary to offset confinement degradation and limit core dilution to enable their application in future devices.
[1]Bernert,NF(2021). Supported by U.S.DOE DEAC52-07NA27344,DE-FC02-04ER54698,DE-AC05-00OR22725,DE-NA0003525.

Speaker: Filippo Scotti (Lawrence Livermore National Laboratory)

Oral: Morning session

257 O22 Time-dependent SOLPS-ITER simulations of type-I ELMs in ASDEX Upgrade H-mode plasmas: experimental validation and impact on particle exhaust

Fuel and impurity density control is essential to ensure reliable tokamak operations. This requires a deep understanding of the physical mechanisms determining the transport of particles in the plasma edge and the pumping of these in form of neutrals.
Divertor compression is the main physics-based figure of merit characterizing particle exhaust efficiency. High-fidelity transport models succeed in reproducing the experimental compression of the main plasma species, but underestimate impurity compression, at least in H-modes (A.Zito,NF2025). Reduced models suggest that a possible reason for this mismatch is the absence, in high-fidelity models, of MHD-related phenomena, such as edge localized modes (ELMs). ELMs are indeed regarded as an efficient mechanism for flushing plasma particles from the core, especially impurities, thereby increasing their transport towards the divertor (T.Pütterich,JNM2011).
The impact of ELMs on particle exhaust has now been quantitatively assessed by performing accurate, time-dependent SOLPS-ITER simulations of a type-I ELM cycle on an ASDEX Upgrade discharge, contrasted with high-time-resolution measurements across the plasma edge/SOL and divertor (M.Cavedon,PPCF2017), including He, Ne and Ar as trace impurities. To this aim, new code capabilities have been implemented to allow more sophisticated time-varying transport assumptions, to better resolve the fast ELM-related transients.
The periodic collapse of the pedestal has been mimicked through a modulated increase of edge radial transport coefficients, to emulate the sudden loss of confinement and consequent explosive ejection of particles and energy from the core. Additionally, a periodically increased anomalous parallel flow velocity in the SOL is used to emulate the free-streaming motion of the hot core-ejected filaments (D.Moulton,PPCF2013).
As a result, it is found that the ELM-averaged divertor compression of the main species is weakly affected, but that of impurity species is significantly increased w.r.t. comparable simulations but without ELMs, with effect scaling with the impurity charge. The impact of ELM bursts on the subdivertor gas density depends on a complicated interplay between pedestal collapse, setting the relative particle loss per species from the core, and increase in temperature of the divertor, affecting the leakage of recycled particles. For impurities, whose H-mode inter-ELM edge transport follows the neoclassical predictions, a higher pedestal, also scaling with the impurity charge, results in a higher relative loss of particles from the core once this collapses, explaining the numerical observations. The achieved results highlight the importance of properly including transient events when performing predictive simulations in view of future devices, especially when concerning impurity transport.

Speaker: Dr Antonello Zito (Max-Planck-Institut für Plasmaphysik, Garching, Germany)

258 O23 Validating edge plasma simulations in TCV: a multi-diagnostic, experiment-driven approach

This work presents an extensive SOLPS-ITER validation effort on the TCV tokamak, built around a global, multi-diagnostic comparison between simulations and systematic experimental databases. Accurate modelling of the scrape-off layer (SOL) and divertor plasma is essential for predicting wall erosion, impurity transport, and heat dissipation in present and future tokamaks. Quantitatively validating these models against experiments is therefore a prerequisite for reliable extrapolation to ITER-scale devices and for informing power-exhaust solutions in next-generation reactors.

In typical L-mode plasmas, SOLPS-ITER reproduces TCV conditions with consistently good accuracy across different divertor closures, matching upstream, divertor-leg and target profiles, the degree of detachment, and divertor neutral pressure. A key advancement is the first direct comparison between simulated and experimentally measured parallel flows in the divertor, made possible by the new tangential divertor spectroscopy system. The agreement significantly strengthens confidence in the modelled SOL flows and in the implementation of the drift physics.
Building on this validated L-mode baseline, the analysis is extended to the inter-ELM phase of type-I ELMy H-mode plasmas. A dedicated H-mode experimental database was developed with the same magnetic equilibrium as the L-mode set and a low ELM frequency (~60-120 Hz), enabling detailed diagnostic characterisation of the inter-ELM phase. By comparing L- and H-mode discharges at matched upstream density, the study isolates the impact of the edge transport barrier on power exhaust. In the modelling, a pedestal-like temperature and density structure is reproduced by locally reducing the anomalous cross-field diffusivities, χₑ and Dₙ, near the separatrix. This approach allows the resulting differences in target profiles to be quantitatively assessed and directly compared to experiments. The role of neutral compression in the divertor is quantified by comparing both L-mode and H-mode scenarios in baffled and unbaffled TCV configurations.

Enabled by controlled experiments in a mid-size tokamak, this work establishes a robust validation methodology applicable to other devices. It represents a significant step toward predictive, validated edge-plasma modelling for the design and optimisation of future fusion reactors.

Speaker: Elena Tonello (SPC - EPFL)

259 O24 Evaluating trade-offs in integrated power exhaust solutions for next-step devices through multi-machine projections of the separatrix operating space

Progress in mapping the tokamak operational space to the separatrix plasma parameters via the separatrix operating space framework (SepOS [1]) has enabled cross-machine comparisons of small/no-ELM regime access (e.g., QCE [2] or EDA H-mode [3]). In this work we extrapolate the SepOS projections to SPARC and ITER, and, crucially, introduce detachment access criteria, thus formulating the combined power exhaust constrained SepOS (i.e., PE-SepOS) to evaluate integrated power exhaust solutions in next-step devices.

Through the interpretation of multi-device (ITER, JET and SPARC) SOLPS-ITER datasets and foundational work demonstrated on JET [4,5], we develop an as simple as possible description of the SOL net power and momentum losses in dissipative regimes to link the main power exhaust quantities (q_dep,t, T_e,t, c_z) with the SepOS parameters (n_e,sep, T_e,sep, Z_eff, α_t). Leveraging normalizations for machine parameters (R, a, B_t, I_p) through the SepOS LH transition boundary minimum, while taking advantage of self-similarities in SOL transport physics, we demonstrate the utility of a normalized PE-SepOS framework in identifying accessible operational points for given exhaustible P_sep requirements. In applying the PE-SepOS framework to project the SPARC and ITER operational space, we find inherent trade-offs, namely: i) accessing high impurity radiation scenarios leads to pronounced reductions in n_e,sep (e.g., 50% n_e,sep reductions at c_Ne=2%) as a consequence of power starvation, and ii) a compromise between high radiative fraction and high density/neutral pressure is required for access to small/no-ELM regimes at sufficiently high density, high α_t conditions, with the divertor dissipative regime transitioning to pronounced detachment.

We further show that by establishing robust correlations between actuators Γ_D2, Γ_Ne, P_sep and physics parameters T_e,t and c_z, the PE-SepOS framework can be used to inform experimental strategies in scaling up from low heating power to reactor-scale values of P_sep/R. Taking advantage of universal trends enabling projections of density and impurity seeding scans, the PE-SepOS thus provides a framework for mapping out the edge plasma operational space in a scalable manner.

[1] Eich T. et al 2025, NME 42 101896
[2] Faitsch M. et al 2023, NF 63 076013
[3] Miller M.A. et al 2014, NF 65 052002
[4] Lomanowski, B. et al 2022, NF 62 066030
[5] Lomanowski, B. et al 2023, NME 35 101425

The work was supported in part by the US DOE Office of Fusion Energy Sciences under contract DE-AC05-00OR22725 by the Innovation Network for Fusion Energy (INFUSE), and by Commonwealth Fusion Systems.

Speaker: Bart Lomanowski (Oak Ridge National Laboratory)

260 O25 Unified Scaling Laws Linking Gas Puffing Rates and Separatrix Quantities: A Multi-Machine Empirical Study

The transition from a hot, clean core to a dissipative, engineering-compatible edge renders separatrix density $n_{\text{sep}}$ and impurity concentration $c_{\text{sep}}$ central to confinement, pedestal stability and detachment control. Predicting them reliably—and quantifying their link to the true experimental actuators of fueling and seeding puffing rates—remains a key objective for reactor-relevant operation and tritium fuel cycle design.

However, papers that first proposed historical scalings with divertor neutral pressure $p_{0}$ reported qualitative violations of the classical power-law dependence
\begin{equation}
n_{\text{sep}} \sim p_0^{\alpha}
\end{equation} in impurity-seeded plasmas [Kallenbach_2018, Schweinzer_1999]. Although the physical cause of these violations is known [Lore_2022], no mathematical model has yet captured the observed range of exponents—spanning from negative to positive $\alpha$ values. Similarly, proxies for impurity concentration, such as
\begin{equation}
c_{\text{sep}} \sim n_{\text{sep}}^{\beta}
\end{equation} produced conflicting results. Reported $\beta$ values range from the theoretical −2 [Lengyel_1981] to about −5 [Järvinen_2023], at times even featuring non-log-linearities in $c_{\text{sep}}(n_{\text{sep}})$ [Body_2025].

In this work, we apply an empirical macro-to-micro approach. We build on our previous macroscopic multi-machine study [Moscheni_2025], combining experimental and numerical data from 32 magnetic-confinement devices. That analysis identified a non-linear function linking $n_{\text{sep}}$ with fueling $\Gamma_{\text{D}}$ and seeding $\Gamma_{\text{Z}}$ at detachment onset, reproducing high-level trends across machines with factor-2 accuracy.

Here, we apply a mathematically equivalent formulation to a new purpose-built database of 50+ SOLPS-ITER edge-plasma simulations [Bonnin_2016], covering devices from SPARC to EU-DEMO and spanning one order of magnitude in $n_{\text{sep}}$ and two in $c_{\text{sep}}$.

Once disentangled from $p_0$, trends for $n_{\text{sep}}$ and $c_{\text{sep}}$ become consistent with the unified functional form
\begin{equation}
\begin{bmatrix}n_{\text{sep}} \ c_{\text{sep}}\end{bmatrix} = \begin{bmatrix}N_0 \ C_0\end{bmatrix} \times (\Gamma_{\text{D}} + \Gamma_{\text{Z}})^{\begin{bmatrix}N_1 \ C_1\end{bmatrix}} \times (\Gamma_{\text{Z}})^{\begin{bmatrix}N_2 \ C_2\end{bmatrix}}
\end{equation}which links separatrix quantities to the puffing rates. This formulation reproduces all qualitative behaviours reported in the literature while explaining apparent contradictions in $\alpha$ and $\beta$ as different manifestations of the same underlying relationship. By comparing the fitted $N_i$ and $C_i$ across machines, we assess their invariance and their dependence on the trajectory followed in the two-dimensional space of puffing rates.

Future work will extend the database to additional divertor configurations, explore the underlying physics [Lomanowski_2025], and pursue experimental validation alongside ongoing studies—predominantly involving un-seeded scenarios [Silvagni_2025].

Meanwhile, the derived scalings offer practical guidance for edge-plasma modelling, experiment planning, and real-time operation in current experiments and next-generation power plants.

Speaker: Matteo Moscheni (Gauss Fusion GmbH, Parkring 29, 85748 Garching bei München, Germany)

12:10 Lunch

Invited Talk: Afternoon session

261 I18 Performance enhancement through real time wall conditioning and profile modification : A cross configurational analysis of Impurity Powder Dropper Experiments

The PPPL Impurity Powder Dropper (IPD) represents a novel capability for simultaneously controlling main ion recycling, modifying edge plasma conditions, and managing material surfaces in reactor-relevant environments. Coordinated experimental campaigns across multiple plasma confinement devices over a range of magnetic configurations have demonstrated that controlled low-Z particulate injection enables real-time wall conditioning while also providing substantial performance enhancements through modification of edge and core plasma profiles.
In tokamaks, powder injection acts as an in-situ boronization tool capable of reducing impurity sources in real-time during plasma operation. For example, ASDEX-Upgrade results show cumulative injection of only a few hundred milligrams of B produces substantial reductions in O and W emission, suppresses inter-ELM W influx, and supports recovery of low-ν* RMP ELM suppression, with post-mortem material analysis confirming micron-scale boron deposition across the divertor and limiter surfaces. Similarly, KSTAR results demonstrate parallel PSI benefits as BN injection reduces Dα emission by 30–40%, mitigates or suppresses ELMs, and lowers global recycling.

Stellarators exhibit complementary performance pathways. Short-pulse B₄C injections on W7-X produce rapid edge ablation leading to field aligned transport providing toroidal equilibration of ablated material and divertor deposition consistent with EMC3-EIRENE/DIS modeling. Material assimilation undergoes a characteristic two-stage confinement response whereby a transient stored-energy loss due to the particulate ablation is followed by a profile shift leading to a recovery of 20–40% higher ion temperatures and total stored energy. This response correlates with steepened edge ion-temperature gradients while spectroscopic and profile measurements indicate transitions toward ion-root confinement driven by edge density modification and Er restructuring. Comparably, on the Large Helical Device (LHD), B and BN powder injections have produced measurable main ion density control through generation of a low recycling wall and reduced intrinsic impurity sources. These injections have also steepened edge Te/Ti gradients with a broadband reduction in turbulence thus enabling the transition to a higher confinement ITG-suppressed state.

Cross-configuration analysis indicates that, despite differences in magnetic geometry and parallel transport paths, the performance gains observed on both tokamaks and stellarators share common underlying mechanisms through dynamic modification of edge impurity sources, redistribution of radiative power, and subsequent modification of temperature and density profiles. These results underscore the versatility of impurity powder injection across confinement concepts and highlights the emerging role of IPD systems as multifunctional tools for real-time wall conditioning, and as dynamic actuators for turbulence regulation, impurity control, and profile shaping.

Speaker: Robert Lunsford (Princeton Plasma Physics Laboratory)

262 I19 Effects of low-Z powder injection on wall recycling, impurity content, and ELM behavior in KSTAR with a W divertor

Boron powder injected into KSTAR lower single null, H-mode discharges with a W monoblock divertor reduced impurity influx from the plasma-facing components (PFCs). The wall conditioning effect provided by the B powder injection resulted in a ~40% reduction in radiated power, ~30% reduction in core electron density, and ~20% reduction in $Z_{eff}$. W-I and O-II line emission brightness was also reduced by as much as 50%, both during and after B injection. A dynamic particle balance found that conditioning via B powder injection resulted in lower wall fueling at start-up, by as much as 99%. While these results qualitatively agree with previous results from full-C KSTAR, WEST, DIII-D, AUG, and EAST, there is a key difference: these KSTAR results show a reduction in W-I line emission during and after B injection, while AUG, EAST, and WEST showed a constant or increased emission during injection, associated with B induced sputtering. Ongoing analysis is underway to understand the origin of this discrepancy. The reduced W influx from the wall supports the use of solid boron injection as a risk-mitigation strategy for controlling core impurity content and radiated power to facilitate the achievement of Q = 10 in ITER.
B powder was also injected into Ne-seeded plasmas to evaluate the effects of combined impurity injection. With simultaneous B injection, the Ne-seeded plasmas had a lower divertor W flux compared to Ne-seeding alone. BN powder injection resulted in periods of reduced ELM activity, which correlated with the excitation of edge harmonic modes. New results from the 2025-2026 KSTAR experimental campaign will also be shown, where we will report on the first Li injection experiments in KSTAR.
Reducing the impurity content and radiative losses is crucial for reactor operation, as well as for KSTAR as it prepares for a full-W upgrade of the first wall in 2026. B powder injection offers a promising wall conditioning technique that can be deployed in real-time. Further work will focus on understanding the long-term dynamics of these B layers, including their lifetime and hydrogen retention properties.

This work is supported by the U.S. Dept. of Energy under Contract DE-AC0209CH11466, and by the Korean Ministry of Science and ICT under the KFE R&D Program ‘KSTAR Experimental Collaboration and Fusion Plasma Research (KFE-EN2401-15)’.

Speaker: Hanna Schamis (Princeton Plasma Physics Laboratory)

Oral: Afternoon session

263 O26 Recycling Control for Long-Pulse Plasma Confinement: Statistical Evidence and a 1000 s H-Mode in EAST

Particle recycling and wall conditions play a crucial role in plasma confinement and long-pulse operation in tokamaks, yet their quantitative impact has not been clearly established. In EAST superconducting tokamak with full metal wall, a systematic statistical analysis was performed to assess the dependence of plasma stored energy on density, heating power, wall conditions, and recycling.

The analysis, with all double-null divertor configuration, shows that the plasma stored energy follows a multi-parameter scaling of the form

$$ W_{\mathrm{MHD}} \;\propto\; n_e^{0.67}\, \mathrm{Li}^{0.11}\, P_{\mathrm{heat}}^{0.28}\, p_{div}^{-0.11}, $$ where $n_e$ is the line-averaged electron density, $P_{\mathrm{heat}}$ is the total auxiliary heating power, $\mathrm{Li}$ represents lithium emission intensity, and $p_{div}$ is the lower divertor neutral pressure characterizing recycling. This scaling demonstrates a clear and independent negative dependence of confinement on recycling, in contrast to the positive dependences on density, heating power, and lithium wall conditioning. The relatively weak power exponent indicates that, once particle-related effects are explicitly included, confinement degradation previously attributed to heating power is largely driven by recycling-related particle processes. Wall conditioning is found to be an effective means to regulate recycling and improve confinement. Compared with boron-coated conditions, lithium conditioning systematically reduces the recycling level and enhances effective wall pumping, leading to improved stored energy under otherwise similar operational parameters. These statistically identified confinement trends are directly demonstrated in the recent achievement of a **1066 s H-mode discharge** in EAST. Particle balance analysis during the stationary phase shows stable density control at approximately $3.0\times10^{19}\,\mathrm{m^{-3}}$, a nearly constant global recycling coefficient of $R_{\mathrm{global}}\approx0.95$, and a dominant contribution from wall pumping, accounting for about 70% of the total particle exhaust. The results confirm that controlled recycling and enhanced wall pumping are essential for maintaining both improved confinement and long-pulse H-mode operation.

Speaker: Yaowei Yu (Institute of Plasma Physics, Chinese Academy of Sciences)

Invited Talk: Afternoon session

264 I20 Comparing the effectiveness of non-uniform and uniform glow-discharge boronisation in full-tungsten devices in support of ITER

Glow-discharge boronisation (GDB) is a standard wall-conditioning technique in fusion devices to ensure reliable plasma operation. With ITER’s re-baseline decision to adopt tungsten (W) as the main-chamber material, GDB is foreseen as a key method to achieve favourable plasma–wall conditions in its full-W environment. To enable reliable extrapolation to ITER, dedicated studies on existing full-W devices are required. Since ITER’ first operations campaign will employ a partial-anode configuration, investigations at ASDEX Upgrade (AUG) and WEST have examined the uniformity and properties of boron (B) layers from partial versus standard GDBs.
Plasma restart on a pristine, unboronised W wall proved difficult in both AUG and WEST. In contrast, sustained plasma discharges were achieved immediately after non-uniform GDBs employing only part of the available hardware—two of four anodes in AUG and three of six gas inlets in WEST—showing that even partial B coverage ensures reliable start-up.
To validate predictive simulations of GDB performance for ITER, quartz microbalances (QMBs) were installed at seven AUG locations to monitor in-situ layer growth, while witness samples on manipulators in AUG and WEST enabled post-mortem ion-beam analysis of layer thickness, composition, and density.
The B layer thickness, oxygen (O) content, and retained deuterium (D) fraction—typically 5–50% relative to B—depended strongly on substrate material and on activation of nearby glow anodes (AUG) or gas inlets (WEST). In WEST, samples near active inlets showed higher B thickness and D fraction but lower O content. In contrast, in AUG, higher B deposition occurred when adjacent anodes were inactive, indicating local erosion during GDB may play a role.
Both the in-situ QMB measurements in AUG and the sample analyses from the two manipulators in WEST revealed a considerably smaller toroidal variation of the B deposition than predicted by modelling, likely explaining the successful restart even after a non-uniform GDB. This discrepancy may be attributed to the sticking probability of the B-containing precursor molecules, measured to be approximately 0.3 at AUG, whereas the initial modelling made no assumption on sticking.
Apart from the witness samples exposed during GDBs, AUG wall samples installed for the entire campaign were analysed to assess long-term B migration during plasma operation and the additional tritium inventory that could be retained in ITER as a result of successive GDBs.
The results indicate that ITER’s planned GDB system should ensure reliable plasma start-up. Future work will address the persistence of GDB effects in more detail.

Speaker: Sehoon An (Max-Planck-Institut für Plasmaphysik, Garching, Germany)

265 I21 Efforts for H-mode plasma achievement without low-Z material coating in EAST with full metal walls

Full metal walls are a high priority choice for future fusion reactors like ITER and CFETR, but they present significant challenges in plasma-wall interaction and high-Z impurity control. Low-Z material coatings (e.g., lithium, boron) are commonly employed to improve plasma confinement and achieve H-mode, but they add considerable complexity to reactor operation and maintenance due to tritium retention in co-deposits.
To investigate the feasibility of achieving H-mode plasmas without relying on any low-Z material coating on full metal walls, a trial experiment was carried out in EAST superconducting tokamak with molybdenum first walls and tungsten divertors. Long-term baking and GDC/ICRF associated plasma cleanings were used for light impurities removal, and ECRH was used for pre-ionization and central heating to prevent tungsten impurity penetration into the plasma core.
Following one week air exposure for maintenance in the 2025 spring campaign, solely through a few days baking and GDC/ICRF seven consecutive stable H-mode plasmas (Ip ~400kA, ne~4.5×1019/m³, duration >14s, 1.95 MW ECRH+ 1 MW LHCD+0.2 MW ICRH auxiliary heating) were successfully realized after 461 plasma discharges exercises. In these H-mode plasmas, the H₉₈(y,2) factor ranged from 0.95 to 1.0, the plasma stored energy was approximately 130 kJ.
This result validates that even without a low-Z coating, EAST can still achieve high-performance plasma operation, but it requires intensive impurities removal and an effective heating scheme. This research provides a good reference for H-mode plasma operation in future fusion reactors with full metal walls with potentially simplifying wall conditioning strategies and reducing operational complexity. This research was funded by National Key Research and Development Program of China (2024YEF03000200).

Speaker: Dr Jiansheng hu (Institute of plasma physics, HIPS, Chinese academy of sciences, 230031, Hefei, China)

15:30 Coffee Break

Postersession 3: Tracks B, C, E, G and J

266 3.091 Development and High Heat Flux Testing of Cold Sprayed Tungsten-based Coatings for the ITER Temporary First Wall

For the ITER plasma-facing first wall the original plan of using beryllium as a wall material has been changed to using tungsten. In this configuration the first operational phase, will be performed with an inertially cooled “Temporary First Wall” [1]. The elements of this wall will experience different loading conditions depending on their respective locations. Correspondingly, the design comprises a mixture of bulk tungsten blocks, tungsten heavy alloy tiles, and tungsten-coated steel components. For this application a tungsten-based solution of cold sprayed (CS) coatings on steel is being developed.
In CS powder particles are accelerated to velocities on the order of 1000 m/s by an inert gas through a Laval nozzle and the bonding to a substrate occurs by plastic deformation upon impact. Therefore, the deposition of pure tungsten coatings by CS is currently not possible and W coating trials with admixtures of chromium, iron and steel have been performed as a follow up of the successful production of W/Ta coatings by CS [2]. This was done together with our Industry Partner Impact Innovations GmbH, Germany. With the finally chosen powder composition, Tungsten powder with 3 wt% admixture of AISI 316L steel, we obtained a coating with a tungsten content of 70 vol% and test samples on 80 * 80 mm² sized substrates with coating thicknesses of 100 and 200 µm, respectively, were produced.
The test samples were high heat flux tested in our facility GLADIS in accordance with the requirements given by ITER: In a first step we applied increasingly high power densities from 0.5 to 2.0 MW/m² with a fixed pulse duration of 4 s. At 2 MW/m² we then increased the pulse duration until a peak surface temperature of 800°C was reached. Finally, we performed a cyclic loading to a peak surface temperature of 800°C. To decrease the cool-down time in between pulses, this was done at a power density of 4.0 MW/m², which corresponded to a pulse duration of 2.8 s. After 100 pulses no deterioration of the coating due to the loading was observed by visual examination. A thorough comparison of pre- and post-characterization will be presented.
In a next step, larger test samples (~ 200 * 300 mm²) will be coated in order to demonstrate the scalability of the process.
[1] R. Pitts et al., Nucl. Mater. Energy 42 (2025) 101854
[2] K. Hunger et al., Fus. Eng. Des. 221 (2025) 115400

Speaker: Hans Maier (MPI for Plasma Physics)

267 3.071 EMC3-EIRENE simulations of boron transport in wall conditioning experiments on EAST upgraded divertor

Three-dimensional (3D) coupled plasma fluid and kinetic neutral edge transport Monte Carlo code EMC3-EIRENE has been used to study boron (B) transport in EAST H-mode experiments with B powder injection for the real-time wall conditioning. A good agreement between experimental measurements and simulation results has been obtained for the electron density and temperature profile with the help of the parameters scanning of particle and energy cross-field transport coefficients. The transport characteristic of B1+~B3+ ions shows an asymmetric distribution along toroidal direction while a symmetric distribution is gained for B4+ and B5+ ions. The calculated profile of B1+~B3+ ions mostly locate at the upper private region, which is attributed to the dominated friction force. While the B4+ and B5+ ions penetrate into the closed magnetic surface due to the leading thermal force, which exhibits a uniform density distribution at closed magnetic surface. The comparative study indicates that B1+~B3+ ions have a more obvious impurity screening effect, in comparison with B4+ and B5+ ions. The impact of B powder injection on the deposited features of boron ions has also been investigated by the EMC3-EIRENE code. The B powder injection gives rise to a strong toroidal asymmetry of B ion deposition on the upper outer divertor target, whereas a faint asymmetric distribution of B ion is found on the upper inner divertor target.

Speaker: Chao Wang (Northeast Agricultural University)

268 3.069 Integrated modelling of boron injection and its impacts on carbon erosion in HL-3

Abstract:
In current magnetic confinement fusion devices, boron (B) powder injection has emerged as an efficient wall conditioning technology for improving plasma boundary characteristics and maintaining plasma stability. Experimental evidence shows that the B injection scheme exerts a significant influence on the distribution of B impurities in the scrape-off layer (SOL). Furthermore, the optimization of injection scheme varies considerably across different experimental devices [1-2]. Meanwhile, injected B may not only dilute the main plasma but also induce erosion of divertor target. Consequently, investigating the transport behavior of injected B is essential, and it can also provide guidance for the design of B injection experiments.
HL-3, a newly renamed tokamak derived from HL-2M, is situated at the Southwestern Institute of Physics (SWIP) in Chengdu [3]. To investigate the transport behaviors of B impurity in HL-3, the study conducts integrated simulations of B transport in both the edge and core regions. The impurity source and edge transport processes of B are modeled using the EMC3-EIRENE code [4-5]. Meanwhile, the accumulation of B in the core plasma is calculated via the STRAHL code, with transport coefficients derived from the ONETWO and TGYRO codes. The simulated B distribution in both core and edge plasma show good agreement with experimental measurements obtained from the HL-3 device. Further, the effects of B injection on the C erosion on divertor targets are analyzed in detail.

References:
[1] K. Afonin, A. Gallo, R. Lunsford et al Nucl. Mater. Energy 40 101724 (2024)
[2] A. Bortolon, R. Maingi, A. Nagy et al Nucl. Fusion 60 126010 (2020)
[3] X.R. Duan, M. Xu, W.L. Zhong et al Nucl. Fusion 64 112021 (2024)
[4] Z.X. Wen, Z.H. Gao, B. Liu et al Nucl. Mater. Energy 42 101881 (2025)
[5] Z.H. Gao, S.Y. Dai, P. Qin et al Nucl. Fusion 65 106021 (2025)

Speaker: Mr Zixuan Wen (Dalian university of technology, China)

269 3.070 Influence of nitrogen recycling on toroidally asymmetric heat load on CFETR X-divertor with EMC3-EIRENE modelling

Nitrogen with the high chemical activity can generate the complex chemical compounds with other elements (i.e. hydrogen, deuterium) in the plasma volume or in the divertor target material. These chemical processes of nitrogen impurity cannot be simulated in the modelling, which is attributed to the limited version of the current EMC3-EIRENE code. Nevertheless, these chemical effects can be summarized and modelled in an effective impurity recycling coefficient of the EMC3-EIRENE code. Accordingly, influence of nitrogen recycling on toroidally asymmetric heat load distribution with nitrogen seeding on the Chinese Fusion Engineering Testing Reactor (CFETR) X-divertor configuration has been performed by the three-dimensional (3D) edge fluid transport code package EMC3-EIRENE. The sensitivity studies of the recycling coefficient of nitrogen ions scanned from 0% to 100% show that the larger recycling coefficients of nitrogen ions lead to a significant increment of nitrogen particle content with different charge-states in the simulation volume. The increased contributions to total power loss induced by nitrogen impurities are found both for the in- and out-board seeding positions. While the peak values of electron temperature and heat load rapidly decrease with the larger recycling coefficients of nitrogen ions, especially for the high recycling coefficient cases. A more obvious toroidal asymmetry of heat load distribution is gained for the high nitrogen recycling coefficient compared to the low recycling coefficient scenarios. This is attributed to the toroidally enlarged deposition area of nitrogen ions on divertor targets and the higher power loss by nitrogen impurities as the recycling coefficient of nitrogen ions increases.

Speaker: Tian Xie (Northeast Agricultural University)

270 3.050 Mesoscale structures and anomalous transport driven by the RDW turbulence

Even though the importance of the mesoscale structures (e.g. “avalanches”, “blobs”, vortices) in
the magnetized plasma transport was recognized a long time ago (e.g. see [1-4]), still the physics
of such objects has many open issues and the extensive studies of these phenomena, both
theoretical and experimental, are continue (e.g. see [5] and the references herein).
The avalanche is usually portrait as a “front” of enhanced plasma (particle/energy) flux
propagating “ballistically” in radial direction. As a “proof” of such phenomenon following from
the simulations, the researchers present the poloidally averaged flux, j(t,x), as the function of
time t and the radial coordinate x. For the case of the radial ballistic propagation of the flux
enhancement, δj, the function δj(t,x) produces a straight “stripe” on (t, x) plane (e.g. see Fig. 2
from [6]).
Our simulations of the resistive-drift-wave (RDW) turbulence with the modified
Hasegawa-Wakatani (mHW) equations [7] for a small electron adiabaticity parameter α also
show similar “stripes” of δj(t,x) on the (t, x) plane (see Fig. 4 from [6]). However, it is well
known that for α <1 the RDW turbulence becomes rather similar to the 2D fluid turbulence [8-
10]. It does not exhibit any front-like phenomena and is dominated by a long leaving mesoscale
vortices (e.g. see Fig. 1 from 2D fluid turbulence modeling [8] and Fig. 3 from the plasma RDW
turbulence modeling with the mHW equation [10]).
Further examination of the results of our RDW turbulence modeling reveal that the origin
of the “stripes” on the function δj(t,x), is a “pairing” of the different sign but similar magnitude
vortices forming poloidally directed “dipoles”, which can propagate ballistically in radial
direction on a large distance before the pair is disintegrated due to the interactions with other
vortices. In addition to that, such a “dipole” results in “dragging” radially the whole plasma and
enhances the perturbation of plasma density/energy in between the vortices, which increases the
flux even more. We notice that whereas the positive perturbation is advected by the dipole
“down the hill” of plasma density (they can be the “seeds” of the blobs in the SOL!), the
negative one moves “up the hill”. Finally, no proximity to “marginal stability” is needed for the
advection mechanism described here. Whether such a turbulent mesoscale advection mechanism
works for the turbulence driven by other plasma instabilities relevant to fusion devices is a
matter for our further studies.
[1] T. Hwa and M. Kardar, Phys. Rev. A 45 (1992) 7002-7023.
[2] P. H. Diamond and T. S. Hahm, Phys. Plasmas 2 (1995) 3640-3649.
[3] X. Garbet and R. E. Waltz, Phys. Plasmas 5 (1998) 2836-2845.
[4] S. I. Krasheninnikov, Czech. J. Phys. 48 (1998), 97-112; Phys. Lett. A 283 (2001) 368-370.
[5] K. Ida, Rev. Mod. Plasma Phys. 6 (2022) 2.
[6] Y. Zhang and S. I. Krasheninnikov, Plasma Phys. Contr. Fusion 62 (2020) 115018.
[7] R. Numata, R. Ball, and R. L. Dewar, Phys. Plasmas 14 (2007) 102312.
[8] J. C. McWilliams, J. Fluid Mech 219 (1990) 361-385.
[9] J. Laurie, et al., Phys. Rev. Lett. 113 (2014) 254503.
[10] A. R. Knyazev and S. I. Krasheninnikov, Phys. Plasmas 31 (2024) 012502.

Speaker: Prof. Sergei Krasheninnikov (UCSD)

271 3.092 Plasma Flow Measurements for Investigating Impurity Transport in the Helicon-Based PMI Device PISCES-RF

Linear plasma devices have widely been utilized to examine plasma-facing materials under fusion reactor relevant conditions. In addition to DC arc sources, helicon plasma-based devices have also been developed. However, impurity generation from dielectric vacuum windows surrounded by an RF antenna remains a critical issue, as observed in Proto-MPEX [1], since unwanted deposited impurities can have an adverse impact on plasma material interaction (PMI) studies.

To address this issue, we have explored plasma flow in PISCES-RF, where no source impurities deposition on a PMI target has been detected [2]. Radial profiles of the parallel plasma flow were measured using two radially reciprocating Mach probes at two axial locations z = 421 mm ($\alpha$) and 675 mm ($\beta$) near the RF source exit, defined as z = 0 mm. Doppler shift spectroscopy of a He II line (468.6 nm) was also carried out at two axial locations z ~ 573 mm ($\delta$) and ~ 1047 mm ($\epsilon$) near the PMI target at 1202 mm.

In this experiment, measurements were first performed in helium plasmas, and then we confirmed that Mach probe-measured Mach numbers agree reasonably with those from Doppler shift measurements. In helicon discharges, it is found that there is a stagnation point around the $\delta$ and $\beta$ locations; the plasma flow is directed toward the RF source and PMI target in the upstream and downstream of the stagnation point, respectively. With increasing gas pressure, the stagnation point tends to move closer to the RF source. At the $\epsilon$ location, the Mach number reaches ~ 0.23, insensitive to the RF input power, and decreases to ~ 0.1 with increasing gas pressure. In deuterium plasmas, the existence of a stagnation point around the $\beta$ location was also confirmed. These experimental findings can explain impurity migration observed in [2,3], and are consistent with SOLPS-ITER simulations [4]. Thus, the plasma flow toward the RF source is thought to play an important role to prevent the transport of impurities from the RF source to the PMI target.

References:
[1] C.J. Beers et al., Phys. Plasmas 28 (2021) 103508.
[2] M.J. Baldwin et al., Nucl. Mater. Energy 36 (2023) 101477.
[3] G. Dhammale et al., Plasma Phys. Control. Fusion 66 (2024) 095015.
[4] M.S. Islam et al., Plasma Phys. Control. Fusion 67 (2025) 025002.

Acknowledgement:
This work was supported by the Japan-US Cooperation in Fusion Research and Development and the US DOE Cooperative Agreement No. DE-SC0022528.

Speaker: Yuta KINASHI (Plasma Reseach Center, University of Tsukuba)

272 3.025 High heat flux exposure of advanced tungsten composites and chromium in the DIII-D divertor under Fusion Pilot Plant-relevant conditions

A number of advanced tungsten (W) based and chromium (Cr) plasma-facing materials were exposed to reactor-relevant divertor plasmas in the DIII-D tokamak using the Divertor Materials Evaluation System (DiMES). The objective was to evaluate the surface response and thermo-mechanical performance of next-generation divertor materials under steady-state and transient heat loads typical of future power-plant operation.

The 5 cm diameter DiMES holder contained seven 6 mm diameter samples, some of them angled at 10º towards the incident heat and particle fluxes to reach more reactor-relevant conditions. The materials tested included micro-structured W (2 angled samples), long- and short- W fiber strengthened W (Wf/W, angled and flat, respectively), a flat W sample strengthened with W and silicon carbide (SiC) fibers (WfSiCf/W), and two flat Cr samples. The experiment was performed in H-mode deuterium discharges with Bₜ ≈ 2 T, Iₚ ≈ 1.1 MA, and up to 4.5 MW of neutral-beam heating, producing average incident heat fluxes of ~4.5 MW m⁻² on flat surfaces and ~15 MW m⁻² on 10° angled surfaces. A 5 cm outer strike point sweep over DiMES at ~2 Hz minimized spatial variation in the incident heat flux and provided near-uniform (within 15-20%) exposure conditions for all samples. Diagnostics included IR thermography, filtered visible imaging and spectroscopy to capture sample surface temperature and erosion during the plasma exposures.

Preliminary analysis of spectroscopic data and post-exposure microscopy indicates that both angled micro-structured tungsten and tungsten fiber-based composites maintained structural integrity, with minor edge melting and cracking, but no observable major crack propagation. The flat WfSiCf/W sample exhibited localized surface smoothing and minor fiber-interface restructuring but preserved bonding and toughness, demonstrating efficient energy dissipation through SiC interlayers. The chromium sample showed the least surface changes; only light surface erosion was observed and emission near Cr-I lines, enabling initial benchmarking of gross erosion and redeposition models.

These results confirm that micro-structuring and composite approaches to improving W properties exhibit superior resilience to transient and cyclic heat loads, while chromium’s radiative properties and controlled erosion behavior make it a viable candidate for passively mitigating divertor heat flux. The experiment provides crucial input toward down-selection and modeling validation for FPP plasma-facing material strategies.

Work supported by the Department of Energy under Award Numbers DE-FC02-04ER54698, Sandia National Laboratories: DE-NA0003525, Princeton Plasma Physics Laboratory: DE-AC02-09CH11466. This work/co-authors received funding from the EUROfusion Consortium via the Euratom Research and Training Programme (Grant No. 101052200)

Speaker: Žana Popović (General Atomics)

273 3.026 Evaluation of erosion and redeposition of irradiated plasma-facing materials by high-energy particles in fusion plasma

In a fusion device, the high-energy ions and neutrals of impurities cause excessive erosion and re-deposition of plasma-facing materials (PFMs) that lead to a reduction in fusion power output and strongly affect the divertor’s lifespan. This study examined impurity types, concentration of chemical species, impurity distribution, and co-deposits on the plasma-facing components (PFCs). Collisional drag from plasma flow accelerates impurities to elevate the impact energies of impurity deposits. The test tiles of molybdenum (Mo), tungsten (W), and carbon (C) were exposed to fusion plasma in the Experimental Advanced Superconducting Tokamak (EAST) to investigate co-deposition and sputtering yield by high-energy particle bombardment and high heat flux. The re-deposition patterns of local and global impurities on each test tile were analyzed using laser-induced breakdown spectroscopy (LIBS) and scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM-EDX). The characterisation of the test tiles reveals that enhanced gross erosion of the PFCs is caused by impurity entrainment. An uneven, thin layer co-deposited with W, Mo, Cu, Cr, Fe, Li, and Ti was observed on the test tiles. The observed high re-deposition mitigates the migration of eroded materials and suppresses net erosion. Consequently, this phenomenon reduces core dilution and enhances the operational lifetime of plasma-facing components (PFCs).

Speaker: Muhammad Imran (ASIPP)

274 3.027 Microscopy study of plasma- and laser-induced effects in tungsten components exposed to helium and hydrogenic plasmas in JET

Following the third deuterium-tritium campaign, DTE3, and the end of JET operations in 2023, a selection of plasma-facing components was removed from JET vessel which included the modules of bulk tungsten divertor tile, so called “tile 5” or LBSRP. During JET operations, these bulk tungsten tiles were exposed to deuterium, tritium and helium plasmas. After the end of plasma operations and prior to removal they also were targeted by laser-induced breakdown spectroscopy (LIBS) measurements. Following their removal from JET vessel, divertor tiles were disassembled into their constituent lamellae, and these lamellae were used for the microscopy studies presented in this contribution. Investigated lamellae were extracted from different sections of tile 5. Within this tile, lamellae are arranged in four stacks, denoted A to D, whereby lamellae are arranged toroidally within a stack, and stacks are arranged poloidally, with stack A at the inboard and D at the outboard side.
Scanning electron microscopy (SEM) was used for imaging of plasma-exposed surfaces. Surface imaging was complemented with focused ion beam (FIB) cross-sectioning, whereby the cross-sections of the plasma exposed surfaces were prepared and imaged using SEM. This allowed the studies of morphology and microstructure in the under-surface volumes of tungsten lamellae.
The main focus of the study was on the effects of exposure of tungsten to helium plasma and to high heat flux. During helium plasma operations the outer strike point was located on stack B. Following helium operations, including deuterium-tritium campaigns, the outer strike point was at all times located on stacks C and D, such that the area of strongest interaction with helium plasma was not further disturbed. The lamellae studied here come from stacks A (private flux region), B (where the highest helium ion flux and surface temperature were present) and D (outer SOL and location of the outer strike point during deuterium-tritium operations, where the highest heat flux during high power operations was present). Surface effects of helium plasma exposure in JET environment, as well as microstructural changes, such as grain growth or recrystallization, induced in tungsten by high heat flux, are presented and discussed.
Additional area of study involved surface and sub-surface imaging of the craters produced by LIBS in the tungsten surfaces. Microstructural effects of laser-surface interaction are presented as well.

Speaker: Yevhen Zayachuk (UKAEA)

275 3.028 Surface erosion of tungsten samples under high pulse number ELM-like heat loads simulated using a repeated pulse electron beam source

One of the key challenges in developing a fusion reactor is the erosion of the armor material in plasma-facing components (PFCs). In magnetic confinement devices, the first wall and divertor will be exposed to damaging thermal shocks resulting from plasma transient events such as Edge Localized Modes (ELMs), plasma disruptions and Vertical Displacement Events (VDEs). For ITER, tungsten has been chosen as the primary armor material due to its high melting point, low sputtering rate, and low tritium retention coefficient.

During tokamak operation, repetitive ELM plasma ejections on PFC surfaces will cause pulsed thermal loading, resulting in thermocyclic fatigue in tungsten, which can lead to crack formation. It is believed that artificially increasing the ELM frequency can reduce the energy of a single event, thereby potentially extending PFC lifetime. Consequently, material testing under pulsed thermal loads with varying power, duration, and frequency parameters is a critical task.

At the Budker Institute of Nuclear Physics, an experimental test stand has been developed to study the effects of repeated thermal shocks on material surfaces. A repeated pulse electron beam source is used to simulate heat loads. The beam has a ring-shaped cross-section with an almost uniform current density distribution within the ring. This source can generate beams with varying parameters — current, voltage, pulse duration, and repetition rate — enabling the simulation of thermal loads within a wide range of conditions. For in situ diagnostics, the stand is equipped with an infrared photodiode-based pyrometer and a surface condition monitoring system using a silicon CMOS camera. Post mortem analysis was performed using Scanning Electron Microscopy (SEM) and optical profilometer.

During the experimental campaign, tungsten samples with polished and ground surfaces, and with surface pre-damaged by prior heating were irradiated with various pulsed heat load parameters ($F_{hf} =$ 4–7 MJ$\cdot$m$^{-2}\cdot$s$^{-0.5}$). Samples were exposed up to ~10$^7$ irradiation pulses in each test series. The experiments allowed the determination of threshold irradiation conditions leading to tungsten surface erosion, providing critical data for assessing the material performance under expected fusion reactor transients.

Speaker: Georgii Ryzhkov (BINP SB RAS)

276 3.029 Simulations of tungsten gap bridging under ITER-like disruption conditions

The extreme plasma heat loads arising during disruptions play a major role in determining the lifetime of plasma-facing components (PFCs). In particular, transient surface melting events are of crucial importance, not only because melt displacement constitutes a major PFC erosion mechanism, but also due to the risk of liquid metal filling the gaps between adjacent wall components [1]. In such cases, large eddy currents may short through the gaps during subsequent disruptions, leading to potential mechanical failure [1,2].

Gap filling and bridging have been observed experimentally in several fusion devices and under various conditions [3,4], but extrapolations to future reactors remain highly uncertain, owing to lack of understanding of the main driving parameters. Recent numerical modelling works attempting to address this issue have brought new insights into the temporal dynamics of the bridging process and have been shown to successfully reproduce available experimental data, notably in terms of characteristic melt infiltration depth and overall cross-gap transport [5,6]. Building upon these validation studies, this contribution details the first predictions of gap bridging by liquid tungsten (W) under conditions representative of worst-case, unmitigated vertical displacement events in ITER. Using characteristic heat load and exposure duration ranges available in the literature [1,7], multiphase Navier-Stokes computations are carried out to investigate the formation and mobilization of W melt around a gap. Components on both sides of 0.5-1 mm-wide toroidal gaps, representative of gaps between divertor monoblocks as well as between first-wall panel fingers, are loaded identically. Values of the local magnetic field inclination angle, which affects the intensity of thermionic currents [8], are sampled up to 20° to explore the PFC response at different wall locations, including divertor cassette edges. The results are analyzed through figures of merit such as the time between melting onset and bridge formation, as well as the relationship between bridge thickness and the characteristic liquid layer depth.

[1] R. A. Pitts et al, Nucl. Mater. Energy 42 (2025) 101854
[2] M. Lehnen et al, J. Nucl. Mater. 463 (2015) 39
[3] K. Krieger et al, Nucl. Fusion 58 (2018) 026024
[4] I. Jepu et al, Nucl. Fusion 59 (2019) 086009
[5] L. Vignitchouk et al, Nucl. Fusion 65 (2025) 056013
[6] L. Vignitchouk, Nucl. Mater. Energy 46 (2026) 102048
[7] J. Coburn et al, Nucl. Fusion 62 (2022) 016001
[8] M. Komm et al, Nucl. Fusion 60 (2020) 054002

Speaker: Ladislas Vignitchouk (KTH Royal Institute of Technology)

277 3.030 Experimentally constrained modeling of the energy spectrum and impact angle distribution of wall impacting neutrals and their role in main-chamber erosion

Passive spectral measurements of Balmer-𝛼 are used to constrain DEGAS2 neutral transport simulations that show 20% of the neutral flux to the tiles is above 100eV, and 5% is above 1000eV at the outer strike point in typical DIII-D H-modes with pedestal top temperatures near 1000eV. Multistep charge exchange between ions and recycled neutrals transfers energy and momentum between the two populations and can greatly increase the neutral energy and mean free path. While this allows fueling deeper in the confined plasma, it also leads to higher energy neutrals that drive first-wall erosion and impurity generation. Due to the Doppler shift of the line emission, passive spectral measurements of Balmer-𝛼 using the charge exchange neutral spectroscopy (CENS) system [1] above the lower x-point on DIII-D along with the midplane main ion CER (MICER) system can clearly distinguish emission from these higher energy neutrals despite bright edge emission. Neutrals with energies in excess of the pedestal top ion temperature are often seen, and these energies can exceed 1keV on DIII-D H-mode plasmas. The spectral measurements provide strong experimental constraints and are used in conjunction with filter based measurements from filterscopes [2] and LLAMA [3] to constrain DEGAS2 neutral transport simulations. The spectral measurements are particularly important for quantifying the higher energy neutrals and their impacts on fueling and the first wall. DEGAS2 is monte carlo collisional radiative code that performs calculations in 3D cartesian geometry and includes a comprehensive set of neutral-plasma and hydrogenic molecule-plasma interactions. The experimentally constrained simulations reproduce the 32 separate spectra from CENS and passive MICER systems with a high degree of accuracy. Ionization source rate information along with the spatial and velocity distribution of neutrals in the confined plasma and those impacting the wall of the device are extracted from the simulations. Details of the measurements, simulation workflow and distribution of the neutrals impacting the DIII-D carbon wall will be presented, including initial assessments of their contribution to main-chamber sputtering and impurity production.
[1] S. R. Haskey, Rev. Sci. Instrum. 95, (2024)
[2] J. Herfindal et al. Rev. Sci. Instrum. 95 (2024)
[3] L. Horvath et al. Rev. Sci. Instrum. 95 (2024)

Speaker: Shaun Haskey (Princeton Plasma Physics Laboratory)

278 3.031 FIRST SPECTRUM OF TUNGSTEN’ EMISSIVITY AFTER EXPOSITION IN THE WEST TOKAMAK

In nuclear fusion, the extreme conditions inside of tokamaks, such as the WEST tokamak expose the plasma-facing components (PFCs) to intense heat and particles fluxes up to 10 MW/m². The interaction between the plasma and the PFCs is responsible for the temperature rise of these components, which are actively cooled to dissipate heat. Therefore, real-time temperature monitoring of the components is required to prevent damage. Infrared (IR) thermography is commonly used for this purpose, relying on IR cameras equipped with narrow-band filters centered on a specific wavelength in the IR domain.
However, the use of actively cooled tungsten (W) components in WEST makes temperature monitoring challenging due to the low emissivity of W (ϵ = 0.1). Moreover, emissivity can vary within a single component due to deposition and erosion processes induced by plasma–wall interactions. These processes lead to changes in surface state. This necessitates a detailed understanding of how emissivity evolves inside the machine.
Previous studies have identified specific regions of strong erosion and deposition that result in significant emissivity variations. A decrease in emissivity from 0.12 down to 0.05 has been observed in highly eroded zones, while in deposition zones it increases up to 0.85 [1]. Although these studies provide valuable insights into emissivity variations, they focus exclusively on the wavelengths of the tokamak’s IR cameras, at 3.9 ± 0.250 µm. Note that this wavelength range is used not only because it provides a good signal-to-noise ratio, but also because it highlights various physical phenomena, such as hot spots etc.

The objective of this work is to extend the investigated spectral range to 2.5–15 µm at temperatures up to 400 °C. This will allow to study the evolution of PFC emissivity as a function of surface state and identify specific behaviours or features that could facilitate temperature measurements in the WEST tokamak.
To this end, a dedicated setup equipped with a Fourier-transform spectrophotometer has been developed for studying W PFCs extracted from WEST. After a brief presentation of this setup, the spectra obtained on eroded and redeposited areas will be shown and discussed. In parallel, a characterization of redeposited areas (roughness and chemical composition) has been conducted to correlate the surface state with the emissivity spectra.

[1] J. Gaspar et al., Nucl. Fusion 62 (2022) 096023

Speaker: Estelle ROMULUS (IUSTI)

279 3.032 Reconsidering Chromium for Plasma-Facing Applications in Fusion Reactors

Tungsten (W) and tungsten alloys are presently the leading plasma-facing material (PFM) candidates for future burning plasma devices (e.g., SPARC, ITER), owing to their high melting temperature and strong resistance to sputtering. However, their use carries significant risks, including fusion power degradation from W core contamination and the uncontrolled release of tungsten dust from co-deposited layers [1]. In the absence of a demonstrated, integrated SOL–edge–core operational scenario applicable to fusion reactor tokamak concepts that can reliably mitigate these effects, tungsten may not be a viable PFM choice for sustained reactor operations.
Motivated by the need to reduce tungsten-related risks, we investigate chromium as an alternative plasma-facing material. When actively cooled plasma-facing components and fully detached divertors—both required for reactor operations —are assumed, chromium is shown to perform reasonably well compared to tungsten. We identify an operational SOL plasma regime for chromium as a first-wall material in reactor-class devices, bounded by charge-exchange neutral effects, as expected for mid-Z materials, in which thermal stress, melting, sublimation, and erosion by charged particles do not constitute limiting factors.
We review several favorable properties of chromium relative to tungsten, including reduced neutron activation, radiative power dissipation characteristics intermediate between argon and krypton, strong oxygen gettering, and low tritium retention. In addition, we show that chromium’s relatively high vapor pressure, and the resulting sublimation, may provide a passive mitigation mechanism against uncontrolled heat flux excursions during transient events such as divertor reattachment—an important cost and reliability driver for reactors, e.g. EU-DEMO [2].
Approaches to further expand the operational space of chromium, including surface nano-texturing [3] and alloy development, are discussed. Finally, we emphasize the urgent need for improved pedestal and SOL radial transport models for reactors to accurately predict the operational limits of any alternative to tungsten, as the large engineering margins afforded by tungsten is currently masking the substantial uncertainties in charge-exchange neutral fluxes and far-SOL transport in existing fusion reactor modeling frameworks [4].

[1] N. Fedorczak, et al., Nuclear Materials and Energy 41 (2024)
[2] G. Federici, et al., Nuclear Fusion 64 (2024)
[3] Chang, F. J., D. Nishijima, and G. R. Tynan. Nuclear Materials and Energy 29 (2021)
[4] M. Beckers et al. Nuclear Materials and Energy 12 (2017)

Speaker: Jerome Guterl (Guterl Scientific LLC)

280 3.033 Investigation of Lithium sources and transport with flowing liquid metal plasma facing components

Liquid metals plasma facing component composed either of Lithium, Tin or Gallium are considered for heat exhaust in future fusion power plants. Different concepts from liquid metal in capillary structures or free flowing liquid metal layers have already been applied and considered for different devices [1,2] and are now investigated for future machines such as Renaissance Fusion stellarator concept where a liquid layer of Lithium / Lithium Hydride (Li/LiH) is used as the main plasma facing component. In addition to challenges regarding MHD stability of such layers where electric currents can flow, it is mandatory to study the impact of the contamination of the plasma by eroded and evaporated Lithium.
Following existing modelling with the SOLEDGE2D code [3], a numerical model has been designed to estimate Lithium sources from the wall when exposed to plasma fluxes. This model implements a thermal part to compute Lithium film temperature depending on plasma heat flux deposition, neutron energy absorption and liquid film flow velocity. Given Lithium temperature, evaporated and eroded fluxes are computed and taken into account as sources for edge plasma transport simulations with the SOLEDGE3X code. The impact of the vapor shielding on the heat exhaust is self-consistently taken into account by an iterative procedure coupling SOLEDGE3X transport simulations and Li/LiH wall modelling. This framework is similar to the one recently implemented by the UEDGE/Wall-Li coupling [4] and applied to NSTX. Finally, we attempt to estimate the fuel retention within the liquid layer and discuss the consequences of a low recycling coefficient on Liquid Lithium wall - as observed at LTX [5] - on edge plasma characteristics.

Acknowledgements:
This project was funded within the France 2030 programme (contract DOS0223063/00)

[1] M.A. Jaworski et al., Nucl. Fus. 53 (2013)
[2] R.J. Goldston et al., Phys. Scr. T167 (2016)
[3] L. Balbinot PhD thesis (2022) - https://hdl.handle.net/11577/3480852 (Universita degli Studi di Padova, Gent Universiteit)
[4] M.S. Islam et al., Phys. Plasmas 32 (2025)
[5] D.P. Boyle et al., Phys. Rev. Letters 119 (2017)

Speaker: Hugo Bufferand (CEA)

281 3.034 Boron–Tungsten Coatings for Fusion-Relevant Studies: Plasma Parameters versus Film Properties

The performance, operational safety, and long-term viability of ITER’s magnetic confinement are determined by the interaction of hydrogen isotopes with plasma-facing components. As a full-tungsten reactor environment is now desired, boronization is essential for reducing impurity levels. Boron high affinity for residual gases such as oxygen and nitrogen, lead to improved plasma purity and stability. Nevertheless, under continuous plasma exposure, boronized surfaces are modified by sputtering, erosion, and redeposition processes involving boron, tungsten, and hydrogen isotopes. These dynamic interactions result in the formation of boron-tungsten and boron-deuterium co-deposited layers, which is an important parameter regarding the in impurity retention, hydrogen storage, and long-term wall conditioning in fusion devices. As a result, there is a significant interest in studying these materials with respect to relevant ITER deposition parameters and methods, as well as their retention and release behavior. Therefore, a systematic investigation of boron–tungsten systems under ITER-relevant deposition conditions is necessar to study the mechanisms governing hydrogen isotope retention and release in these materials.
In this study, PVD deposition techniques such as Thermionic Vacuum Arc (TVA), RF magnetron sputtering and High Power Impulse Magnetron Sputtering (HiPIMS) are used to synthesize boron and boron-tungsten-based coatings under controlled conditions. These techniques are selected in order to access different plasma regimes and particle energy distributions relevant to fusion environments. During the magnetron-based deposition processes, plasma conditions are systematically investigated, with particular emphasis on ion flux and ion energy as functions of the applied power and the Ar/D₂ gas ratio. This approach will enable a controlled adjustment of the energetic conditions experienced by the growing films. The resulting boron and boron–deuterium co-deposited layers are characterized with respect to their structural and morphological properties, compound formation, and hydrogen isotope retention behavior. The influence of the deposition conditions on deuterium trapping and release mechanisms within the films is studied. In addition, deuterium release behavior is investigated by thermal desorption spectroscopy, allowing both the determination of release temperatures and the quantification of the total retained deuterium. The study aims to establish correlations between deposition parameters, plasma conditions, and film properties, providing insight into the mechanisms governing hydrogen isotope retention and release in boron-based structures relevant for ITER applications.

Acknowledgement: This work was supported by a grant of the Romanian Ministry of Education and Research, CNCS-UEFISCDI, project number PN-IV-P2-2.1-TE-2023-1140, within PNCDI IV.

Speaker: Corneliu Porosnicu (INFLPR)

282 3.035 Interpretative and predictive analysis of impurity migration in GyM and BiGyM through ERO2.0 Monte Carlo transport code

Plasma-wall interactions (PWI) are crucial in determining the overall performance and life-time of tokamak plasma-facing components (PFCs), particularly in high-performance machines like ITER and future fusion reactors. PWIs can lead to erosion, impurity generation, and fuel retention, negatively affecting plasma confinement and integrity of PFCs. Linear plasma devices, such as GyM [1] and its upgraded version BiGyM (expected operational in 2026), are essential testbeds for investigating PWI under controlled and cost-effective conditions, enabling achievement of ITER-relevant fluences. In this context, modeling tools are required to support the interpretation of experimental data and to establish predictive capabilities for PWI processes.
This work focuses on the application of erosion and impurity-transport Monte Carlo code ERO2.0 [2] to GyM for interpretative analysis, and to BiGyM for predictive studies.
ERO2.0 is used to support the evaluation of the S/XB spectroscopic parameter for tungsten (W), during argon (Ar) plasma discharges in GyM. The S/XB parameter plays a key role in PWIs, as it allows the determination of the erosion flux from photon flux emitted by sputtered particles, measured via optical emission spectroscopy. At low electron densities (<1e17 m-3) and temperatures (<10 eV) as in GyM, accurately estimating the fraction of W atoms leaving the plasma without being ionized—referred to as geometric loss flux (GLF)—not contributing to the S/XB value, is fundamental [3]. Plasma background used as input for ERO2.0 was provided by the SOLPS-ITER code. The simulations were performed using electron impact ionization rate coefficients calculated from Steinbrink’s formula [4], which is better suited for the low-density, cold Ar plasma of GyM, instead of the default values from the ADAS database. ERO2.0 GLF was compared with the value from a simpler analytical model based on the W atom ionization mean free path due to electron collisions. Both approaches agree within a few percent, showing that the GLF accounts for over 90% of the total W flux.
The study will be extended to BiGyM plasma, expected to feature similar temperatures, but densities at least two orders of magnitude higher. Results will be presented at the conference.
[1] A. Uccello, et al., Front. Phys. 11, 1108175 (2023)
[2] J. Romazanov, et. al., IAEA FEC (2021)
[3] A. Cremona, et al., Nucl. Mater. Energy 17, 253 (2018)
[4] T. Schlummer, et. al., Phys. Scr. 2017, 014075 (2017)

Speaker: Francesco Cani (CNR-ISTP)

283 3.036 First results of the application of LIBS for depth profiling and calibration-free analysis of the JET divertor region

Laser-Induced Breakdown Spectroscopy (LIBS) has proven to be an effective technique for in situ characterization of the inner walls of the Joint European Torus (JET) tokamak, particularly for quantifying fuel retention on plasma-facing components (PFCs) without the need for their removal or manipulation. Previous experiments have revealed that large areas of the inner wall—especially the divertor region—undergo progressive co-deposition of erosion by-products, impurities, and unburned deuterium–tritium fuel, with the latter being of critical importance for safety considerations. Recent investigations have focused on the analysis of LIBS spectra acquired at multiple locations across the divertor, showing that the redeposited layers primarily consist of beryllium (originating from first wall components), tungsten (from the divertor region itself), and residual hydrogen isotopes from the unburned fusion fuel. Depth profiling of Be-coated sections has been performed to estimate the ablation rate per laser pulse, while the residual hydrogen isotope content was evaluated through calibration-free (CF) analysis of the Dα/Hα emission lines. Both depth profiling and CF analysis have yielded reliable results; however, the remote operation of the LIBS system has limited the frequency of instrument calibration. This constraint occasionally led to diffraction order misalignments in the collected spectra, resulting in the appearance of intense ghost lines. The problem was mitigated by reverse-engineering the echellograms recorded by the spectrometer into corrected spectra via dedicated reconstruction algorithms, thereby restoring spectral fidelity and minimizing ghost artifacts.
This work presents updated results from depth profiling and CF analyses of some representative areas of the JET divertor by using the newly processed dataset, aiming to eliminate residual errors associated with spectral ghost lines and to enhance the accuracy of LIBS-based diagnostics in tokamak environments.

Speaker: Dr Lidia Baiamonte (ENEA)

284 3.037 Recent results of the SWORD/SPARROW linear plasma devices

The commissioning of SWORD is completed after 7-years of R&D and construction. This machine employs a 3 T superconducting-magnet to confine plasmas and an arc plasma torch to produce intense plasma streams with a particle flux up to several times of 1E24 m-2s-1. The plasma diagnostics include emission spectrum and target probes. Additional Thomson scattering and laser interferometric measurements are under development. The first milestone result of SWORD—a continuous helium plasma discharge with a particle flux greater than 1E24 m−2 s−1 and a duration greater than 1000 s, was obtained in January 2025. The production and measurement of such discharges will be described in detail. Furthermore, the overarching goal of SWORD is to produce plasmas in the strongly coupled regime with a peak particle flux > 1E25 m−2s−1, a peak heat flux > 80 MW m−2 and a beam size > 20 mm simultaneously. Motivation and approach will be explained.

On the other hand, the function of SPARROW is twofold. First, it serves as a test platform for the development of SWORD, such as optimizing the cascaded arc source, quantifying the target Langmuir probe, and practicing plasma diagnostics. Second, we use it to investigate plasma material interactions, including surface modification of tungsten by He plasma exposure, orientation-dependent sputtering of tungsten by Ne plasmas, and He bubbles retarded recrystallization in tungsten.

These two platforms have also catalyzed nationwide collaborations. Examples include the evaluation of 3D printed tungsten alloys with exotic cell structures led by Tsinghua University, monoblocks covered with strongly textured millimeter-scale tungsten grown by chemical vapour deposition led by Beihang University, and the application of tungsten ‘fuzzy’ surfaces for trace detection of macromolecules led by Shandong University.

Speaker: Haishan Zhou (Institute of Plasma Physics, Chinese Academy of Sciences)

285 3.038 Effects of high-power heat loads on sublimation of CFC target in the EHL-2 device environment

The EHL(ENN He-Long)-2 spherical torus is slated for completion by the end of 2026, which is designed to achieve proton-boron fusion in magnetically confined thermal plasmas [1]. The EHL-2 fusion device is planned to employ carbon fiber composite (CFC) as the plasma facing materials (PFMs), taking into account its high conductivity, low Z and resistance to damage due to high heat loads [2]. During plasma operation, the interaction between high-power heat flux and divertor targets can cause the sublimation of CFC [3]. This not only shortens the service life of CFC target but also impacts the steady-state operation of EHL-2.

In this work, the effects of heat loads, materials and cooling conditions on the sublimation of CFC have been investigated in the linear plasma device HIT-PSI [4]. Based on these experimental parameters and results, a detailed numerical simulation study has been conducted using the lattice Boltzmann method [5]. This study provided the temperature distribution and sublimation depth of CFC under various conditions. The simulation results show a good agreement with the measurements obtained on HIT-PSI. Both the experimental and simulation investigations offer valuable references for the design of CFC target on the EHL-2.

[1] Y. M. Wang, K. Li, Z. Huang, et al. Plasma Sci. Technol. 2024 https://doi.org/10.1088/2058-6272/ad9f27
[2] A. T. Peacock, M. Merola, M. A. Pick, et al. Phys. Scr. T128 (2007) 23-28
[3] B. Bazylev, G. Janeschitz, I. Landman, et al. Phys. Scr. T128 (2007) 229-233
[4] T. Huang, Q. Y. Nie, M. Wang, et al. Appl. Sci. 12 (2022) 10501
[5] R. Z. Huang and H. Y. Wu J. Comput. Phys. 315 (2016) 65-83

Speaker: Shuyu Dai (Dalian University of Technology)

286 3.039 Deuterium Retention in Tungsten After Deuterium-Nitrogen Co-bombardment

Plasma facing components in fusion devices are subjected to very high heat fluxes, leading to the potential for material degradation and melting. In the case where PFCs are made out of high Z materials like tungsten, the subsequent contamination of the plasma from wall degradation can become a major problem leading to a shutdown of the fusion reactions. To prevent the heat fluxes from damaging the divertor tiles, it may be necessary to seed gases in the edge plasma, radiating away the heat and cooling down the plasma before it makes contact with walls.

While nitrogen has good radiation characteristics in the edge plasma temperature range, it has the added complication that it will interact chemically with both the hydrogenic fuel and any metal surfaces. The production of tritiated ammonia will complicate the tritium fuel cycle, while interactions with the tungsten surface may lead to enhanced erosion and changes in how hydrogen is trapped in the surface. With limitations on how much tritium can be produced in a fusion reactor, along with regulations governing tritium handling and inventory, it is then crucial to understand how much of it is retained in the PFCs and how much could be released during thermal cycling.

It has previously been observed that simultaneous nitrogen and deuterium co-bombardment can result in a higher retention of deuterium in tungsten [1]. In the current study, we have broadened the parameter range of the previous study and have extended the scope of our surface analysis. As in [1], we observed an increase in deuterium retention in tungsten with the amount retained increased by an order of magnitude in the temperature range ~ 500 to 600K. This increase occurs for N+/D+ ratios as low as 1% and is accompanied by large-scale blistering of the surface.
XPS analysis of the co-bombarded surfaces indicate the formation of a tungsten-nitride layer, both at room temperature and at 575K. However, we believe that a change in the nature of the tungsten-nitride layer from predominantly WN at lower temperature to W2N at higher temperature [2], affects the ability of deuterium to diffuse to the surface.

[1] M.E. Goodland, J.W. Davis, "Influence of nitrogen on deuterium retention in tungsten under sequential and simultaneous irradiation", Nuclear Materials and Energy 26 (2021): 100928
[2] K. Schmid, et al., “Interaction of nitrogen plasmas with tungsten”, Nuclear Fusion, 50 (2010) 025006.

Speaker: Steven Thériault (University of Toronto)

287 3.040 Deuterium retention in radiation- and plasma- induced defects and co-deposited layers in W and W-Cr-Y alloy

The fuel in thermonuclear and hybrid fission/fusion reactors will be a deuterium-tritium (DT) mixture. Plasma-facing materials will be exposed to both stationary and high heat pulsed plasma loads, as well as helium and neutrons, arising from the DT reaction. Such extreme operating conditions can lead to degradation of materials due to erosion, formation of plasma- and radiation- induced defects, and hydrogen isotope embrittlement. To ensure the safe operation of a fusion reactor, it is necessary to be able to predict a stability of material properties under irradiation and an amount of radioactive tritium retained in materials.
Tungsten (W) is envisaged as reference plasma-facing material for ITER. Because tungsten oxidation and helium embrittlement are suppressed in W-Cr-Y alloy, this alloy can be an alternative choice for the first wall in future reactors. In this paper, the effect of dynamic change of surface impurities (B and O), temperature and development of plasma- and radiation- induced damage on D retention in W-based materials is discussed. In order to study stability of W and W-Cr-Y alloy under irradiation, radiation- and plasma- induced defects and surface impurities were characterized by positron annihilation spectroscopy, scanning and transmission electron microscopy, atomic probe tomography, and energy dispersive X-ray spectroscopy. D accumulation was studied by thermal desorption spectroscopy.
It is shown that plasma-induced damage, produced by high fluxes of particles and heat, leads to similar increase in the D concentration as radiation-induced damage. However, total D retention is determined by D retention in radiation defects because they distributed over all material thickness while plasma-induced damage occurs only near the irradiation surface. Coating of boron (B) increases D retention in the material. The binding energy of D to defects in B coating is higher than in W and W-Cr-Y alloy, and also higher than the binding energy of D to defects caused by high-temperature plasma. Although, the D concentration in radiation-induced damage is comparable with the D concentration in layers deposited together with boron and D, it is expected that the dominant contribution to the total D retention will be the D retention in radiation-induced defects due to the larger surface of the first wall and bulk retention. Radiation damage has less influence on the D retention in W-Cr-Y alloy compared to pure W. Moreover, the Cr and Y alloying elements suppress the formation of dislocation loops and pores. Modelling was used for predictions of the D retention for ITER.

Speaker: Olga Ogorodnikova (Moscow Engineering Physics Institute)

288 3.041 Effects of chromium and rhenium on hydrogen isotope retention in tungsten after neutron irradiation

Tungsten (W) is a primary candidate of plasma facing material (PFM) of a fusion reactor because of its high melting point and good thermal conductivity. However, W inevitably suffers damage from 14 MeV neutrons produced in fusion reactions, leading to the formation of a large number of irradiation-induced defects. These defects can trap hydrogen isotopes and substantially increase tritium retention. Therefore, it is necessary to investigate how to reduce tritium retention in neutron-irradiated W.

In this study, Re and Cr were chosen as alloying elements to produce W-5Re and W-0.3Cr alloys. Neutron irradiation was performed for disk samples of W, W-0.3Cr, and W-5Re at 563 K to about 0.07 dpa. Then the samples were exposed to D plasma at 563 K in a linear plasma device to a total fluence of 1×10²⁵ D⁺/m². After exposure, D retention was examined by means of thermal desorption spectroscopy (TDS) in which the samples were heated to 1473 K at a rate of 0.5 K/s. To understand the influence of alloying elements on the annealing process of defects, the samples after the first TDS measurements were subjected to a second round of D plasma exposure and TDS measurement. Positron annihilation lifetime spectroscopy (PALS) was performed on all three materials to examine the concentrations and size distributions of vacancies and their clusters.

The D retention in W and W-0.3Cr increased by an order of magnitude after neutron irradiation. The positron lifetime in these materials indicated the formation of vacancies and relatively large vacancy clusters. Nevertheless, W-5Re showed far smaller extent of increase in both D retention and positron lifetime. The defects induced by neutron-irradiation in W-0.3Cr and W-5Re were almost fully annealed out during 1st TDS measurements, and the positron lifetime and D retention returned to the values comparable with the non-irradiated samples. On the other hand, large vacancy clusters remained in W even after annealing at the 1st TDS measurements, and D retention in the second round test was still clearly larger than that of non-irradiated sample. These results showed that Re enhanced vacancy annihilation under irradiation at 563 K and suppressed increase in D retention, and both Cr and Re promoted vacancy recovery under post-irradiation annealing.

Speaker: Mr Qicong Chen (Tohoku University)

289 3.042 Scaling Laws and Application of a Neutral Gas Evaporation Shielding (NGES) Model in HL-2A Pellet Injection Experiments

The penetration depth of pellets mainly depends on the ablation rate, which further influences fueling efficiency. In this work, based on the neutral gas shielding model [1], we proposed a neutral gas evaporation shielding (NGES) model, in which the ablation cloud radius was evaluated self-consistently to account for the real-time shielding effect of ablated gas on the incoming heat flux. To validate the model, we calibrated it using pellet injection experiments on HL-2A by examining the pellet parameters. The injection velocity, pellet size, and injection location were discussed. Background plasma parameters, including electron density and temperature, were also taken into account.
In the HL-2A tokamak, the core plasma typically operates within a well-defined parameter range. The central electron density generally lies in the range of , while the core electron temperature varies between 0.7 and 1.8 keV. Based on these background plasma conditions, the calculated characteristics of the pellet ablation cloud show a temperature of approximately 2–4 eV and a density of about . We further compared the predicted pellet penetration depth and Hα signals from the model with experimental measurements on HL-2A. These comparisons led to a scaling law under the device conditions. The simulation results show that the ablation rate trends differ from experimental observations by less than 10%, showing the feasibility of the NGES model. Under a wide range of plasma conditions, the ablation model with self-consistent evolution of the ablation cloud radius was compared with the NGPS model. The results exhibit a better agreement with the experimental observations. With further development, the model can be extended to include the rocket effect. The NGES model developed based on HL-2A pellet injection experiments provides a preliminary ground for further investigation of the rocket effect and its impact on fuel pellet trajectories.

[1] Parks P B and Baylor L R Phys Rev Lett 94 1–4

Speaker: Dazheng Li (FZJ-IFN1)

290 3.043 Dynamics of Deuterium Retention in Boron Thin Films on Tungsten During High-Flux Plasma Exposures in DIONISOS

Understanding tritium (T) and deuterium (D) retention in boron-coated tungsten (B/W) plasma-facing components is essential for next-generation fusion devices such as SPARC, ITER, and future commercial reactors, because wall retention sets the in-vessel T inventory and influences D/T recycling and wall pumping, with direct implications for safe operation and sustained plasma performance. In this work, we present the first time-resolved measurements of D uptake and release in PVD (physical vapor deposition) boron thin films deposited on polycrystalline W during low-energy ($<100$ eV) D plasma exposures at fluxes up to $\sim10^{22}$ m$^{-2}$s$^{-1}$ in the DIONISOS linear plasma device, representative of tokamak first-wall conditions. Measurements were performed in operando using 3 MeV $^3$He Nuclear Reaction Analysis (NRA) during plasma exposures.

The D retention in the B/W sample exhibited a four-phase temporal evolution: during plasma exposure, the D inventory initially increased with a steep slope, then decreased, then increased again, and finally decreased once more after the plasma was shut off. This behavior is attributed to physical and chemical erosion of the B layer, diffusion of D into the W substrate, and subsequent outgassing after the end of the plasma exposure. These time-resolved measurements provide insight into D recycling by showing how boronized W surfaces can act as a source and sink for D during tokamak-relevant plasma operation. From these measurements, erosion rates, erosion yields, and D/B ratios were inferred.

Additional experiments vary the initial thickness of the B layer to study how the dynamics of D retention change as the time required to fully erode the layer varies. The sample temperature is also varied, as it affects the dissociation of B-D and B-D-B bonds, chemical erosion, and D diffusion. In addition, we compare PVD laboratory made B/W samples with B/W samples prepared in tokamaks using glow discharge boronization (GDB) to assess how deposition methods and film history influence D retention behavior.

Speaker: Joey Demiane (MIT - PSFC)

291 3.044 Plasma-Driven Hydrogen Permeation Properties of Tungsten Deposited Films Fabricated by Argon Plasma

In the realization of future fusion reactors, the strict management of tritium inventory and the minimization of tritium permeation from plasma-facing components (PFCs) into the coolant systems are critical issues. These challenges directly impact both the radiological safety of the power plant and the fuel efficiency of the fusion cycle. Tungsten (W) is currently the leading candidate material for PFCs due to its robust thermal and physical properties. However, establishing effective permeation barriers against the high flux of energetic particles anticipated in the reactor environment remains a high-priority technological objective. Unlike gas-driven permeation, plasma-driven permeation (PDP) is governed by complex surface dynamics, predominantly determined by the re-emission rate of hydrogen back to the plasma side.

This study aims to develop W coating materials capable of significantly suppressing PDP and to clarify the relationship between fabrication process parameters and hydrogen transport properties. W coatings were deposited on both nickel and W substrates using radio-frequency argon plasma sputtering. The argon background pressure during the deposition process was selected as the primary control parameter, varied systematically (e.g., from 30 Pa to 100 Pa) to modulate the kinetic energy of sputtered particles and, consequently, the microstructural growth of the films. The hydrogen permeation performance was evaluated by exposing the samples to an Inductively Coupled Plasma hydrogen source.

The experimental results demonstrated a strong dependence of the hydrogen permeation flux on the sputtering gas pressure. It was observed that W films deposited under higher argon pressure conditions (100 Pa) exhibited significantly superior barrier performance compared to those fabricated at lower pressures (30 Pa). Specifically, the permeation flux, normalized to account for differences in deposited mass, was reduced by approximately an order of magnitude in the high-pressure samples.

We interpret this behavior as likely being associated with changes in the film microstructure. High-pressure sputtering conditions are generally known to promote the formation of a porous, columnar structure, corresponding to Zone 1 in Thornton's structure zone model. It is inferred that such a structure, if present, would increase the effective surface area compared to denser films. In the context of plasma-driven permeation, a larger surface area is expected to facilitate the surface recombination of hydrogen atoms into molecules. This enhanced recombination would promote the desorption of hydrogen back into the vacuum chamber, potentially reducing the net flux diffusing through the bulk material.

Speaker: Kentaro Masuta (Kyushu University)

292 3.045 Damage induced by low-energy plasma and MeV ions and its effect on deuterium permeation and retention in tungsten-based materials

The integrity of plasma-facing components (PFCs), particularly tungsten (W), against synergistic irradiation damage, specifically from high-energy particles, high-flux plasmas, and high heat loads, remains a critical challenge for future magnetic confinement fusion reactors. This research investigated the combined effects of high-flux, low-energy plasma, and high-energy particle beams on advanced W materials. We established a novel synergistic irradiation experimental platform that coupled a MeV radio frequency quadrupole (RFQ) accelerator with a low-energy, high-density linear plasma device. This unique setup allows for the simultaneous exposure of PFC materials to fusion-relevant conditions.
This study employs three distinct W materials chosen for their varying microstructures: single-crystal W (SCW), commercial polycrystalline W (PCW), and oxide dispersion strengthened W (ODS-W). The samples were synergistically irradiated using a mixed deuterium (D) and helium (He) plasma and high-energy hydrogen (H) ions. The experimental parameters included D/He plasma fluences ranging from 6 × 1024 m-2 to 2 × 1026 m-2, sample temperatures from 773 K to 1373 K, and incident ion energies from 40 eV to 90 eV. The energy of the high-energy hydrogen ions was 0.75 MeV, with associated irradiation dose ranging from 0.01 dpa to 0.1 dpa. The microstructures of the original materials and irradiated samples were thoroughly characterized using Electron Backscatter Diffraction (EBSD), Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM).
Initial results revealed that the D permeability of unirradiated ODS-W was higher than that of pure W. After He plasma irradiation, the formation of surface fuzz structures significantly increased D permeability, an effect that correlated positively with fuzz thickness and the permeation temperature. Regarding high-energy H ion effects, D retention in annealed W after 0.01 dpa proton irradiation was comparable to unirradiated samples, but increased by approximately one order of magnitude after 0.1 dpa irradiation, with the TDS spectra exhibiting multiple distinct desorption peaks.
We conducted further synergistic irradiation experiments, and well report results on the influence of irradiation damage under these various experimental parameters on deuterium permeation and retention behavior.

Speaker: Minyou Ye (University of Science and Technology of China)

293 3.046 The effect of high-energy proton irradiation on deuterium permeation and retention in tungsten-based materials

Tungsten (W) is the leading candidate material for plasma-facing components (PFCs) in future magnetic confinement fusion reactors, such as ITER and DEMO, due to its excellent properties. The PFCs will face extreme operating environments, including intense fluxes of high-energy particles, plasmas and heat, which will cause significant radiation damage and affect fuel retention.
This research performed high-energy proton irradiation experiments on a MeV radio frequency quadrupole (RFQ) accelerator and investigated the effects of proton irradiation on the deuterium (D) permeation and retention behavior in tungsten-based materials, including ITER-grade polycrystalline W, manufactured by Advanced Technology & Materials Co. Ltd Inc., China (ATW), 0.5 wt.% ZrC dispersion strengthened W (W-ZrC), and 2 wt.% Y2O3 dispersion strengthened W (W-Y2O3). The energy of the high-energy hydrogen ions was 0.75 MeV, with associated irradiation dose ranging from 0.01 dpa to 0.1 dpa. The samples were irradiated at room temperature and 773 K.
Microstructural analysis using transmission electron microscopy (TEM) revealed a high density of dislocation loops in the ATW after proton irradiation, indicating significant radiation damage that creates new trapping sites. Correspondingly, the D retention in ATW irradiated to 0.01 dpa was comparable to the unirradiated sample. However, the retention increased drastically by approximately one order of magnitude after 0.1 dpa irradiation, with thermal desorption spectroscopy (TDS) exhibiting multiple desorption peaks linked to various trap types. Furthermore, D permeation measurements showed that in the high-temperature regime with 1073 K and 1123 K, the D diffusion coefficient in the proton-irradiated ATW was slightly lower than that of the pristine material, a finding consistent with the enhanced trapping efficiency of the irradiation-induced defects. The irradiation damage and its effect on D permeation and retention in W-ZrC and W-Y2O3 was also examined. These results are crucial for accurately predicting fuel inventory, tritium consumption, and material lifetime in future magnetic confinement fusion devices.

Speaker: Zhe Liu (University of Science and Technology of China)

294 3.047 The effect of surface-near helium on deuterium retention in tungsten and EUROFER

It was shown experimentally for tungsten, that He retained close to the surface influences transport and retention of hydrogen isotopes (HI). Namely, experiments using He seeded D plasmas, showed that the addition of He leads to reduced blistering accompanied by reduced D retention [1].
Recently, we have performed a systematic series of D exposures in tungsten where He was pre-implanted near the surface [2]. 20 MeV W irradiation was performed before or after He implantation to create defects within the first 3 µm. The defects created by W ions were used to trap D that was penetrating through the He surface layer. Using 3He nuclear reaction analysis (NRA) made it hence possible to quantify D transport beyond the He layer. D and He depth profiling showed increased D retention at the depth where He was implanted, but D retention in the bulk was reduced five times as compared to a He-free reference sample. Transmission electron microscopy showed nanometer-sized bubbles where He was implanted. From this we concluded that the observed D retention in He irradiated samples is due to D trapping at He bubbles and that the He containing microstructure enhances D re-emission. The same methodology was applied to EUROFER97 to clarify the effect of surface-near helium on deuterium transport into and retention in the bulk. To quantify the influence on D uptake at the surface, He was implanted into EUROFER97 samples close to the surface with 1 keV ions with different fluences and at different temperatures. Samples were then exposed to a low flux, low energy (100 eV/D) D ion beam at 370 K. One He-free W-irradiated EUROFER97 reference sample was also exposed only to low energy D ions for comparison. The defects created by W ions trap penetrating D and make it hence possible to quantify D transport below the He layer using NRA. Measured D depth profiles show reduced D uptake in the He-irradiated samples with major D retention near the surface where He is implanted. The D uptake is reduced by a factor of two compared to He-free sample. Surface analysis of the sample with the largest He fluence of 5.5×10^{21} He/m2 showed He bubbles near the surface. Results will be discussed and compared to tungsten.
[1] M. Baldwin et al. Nucl. Fusion 51 (2011) 103021; Nucl. Fusion 57 (2017) 076031
[2] Markelj et al. Nucl. Mater. Energy 45 (2025) 101981

Speaker: Sabina Markelj (Jožef Stefan Institute)

295 3.048 Modelling of Hydrogen Isotopes Exchange in Self-Damaged Tungsten

Plasma-wall interactions in tokamaks lead to transport and retention of hydrogen isotopes, especially tritium, in the plasma exposed material causing safety issues. Cleaning procedures are envisaged in ITER to recover the tritium [1] using pure deuterium (D) operation to trigger isotope exchange. This work presents isotope exchange simulations and experimental validation of the model, which can be used to evaluate the efficiency of such procedure.
To mimic damage caused by neutrons, tungsten is irradiated with MeV W ions (self-damage), creating multi vacancies and self-interstitial atoms which are the kind of defects created by neutron induced cascades. This material is exposed to D to study fuel retention in damaged W. Isotope exchange experiments has been done in self-damaged W by a D/H exposure [2]. To model this data, a single occupancy trap model is compared to a multi-occupancy trap model based on energies data from DFT calculations [3] [4], which has been used to simulate isotopic exchange in undamaged tungsten (W) [5]. This physics is here implemented in the FESTIM transport code [6]. One model includes two traps with single occupancy and the other includes one trap with two occupancy levels, allowing for another isotope exchange process. Simulated D profiles are compared to experimental depth profiles to quantify the effects of the multi-level trap.
The simulations show for both models good accordance with the experimental results at the temperature of 600K. The multi-occupancy model shows a slightly faster isotopic exchange than the single-occupancy traps model. The difference between both models increases with decreasing temperature. This is the expected behavior since higher temperature helps detrapping in the single occupancy model. This difference is also being investigated at higher temperature to see the effects on the model.

[1] Loarte, ITER Physics chapter 4: power and particle exhaust, Nuclear Fusion (2007)
[2] S. Markelj et al., Journal of Nuclear Materials, vol. 469, p. 133‑144, febr. 2016
[3] N. Fernandez et al., Acta Materialia, vol. 94, p. 307‑318, aug. 2015
[4] J. Hou et al., Acta Materialia, vol. 211, p. 116860, june 2021
[5] K. Schmid et al., Journal of Applied Physics, vol. 116, nᵒ 13, p. 134901, oct. 2014
[6] R. Delaporte-Mathurin et al., International Journal Of Hydrogen Energy, vol. 63, p. 786-802, apr. 2024

Speaker: Mr Jonathan DUFOUR (CEA)

296 3.049 Irradiation damaging of EUROFER97 to very large damage doses: lattice strain and deuterium retention

EUROFER97 is the European Reduced Activation Ferritic-Martensitic (RAFM) candidate steel to be used as a structural material in future nuclear fusion devices. Neutron irradiation will degrade the mechanical properties, setting limits in terms of operational temperature and maximum allowed dose. It is anticipated that the European DEMO will utilize a first blanket with a 20 dpa damage limit in the first-wall and then switch to a second set of blankets with a 50 dpa damage limit. In addition to mechanical properties, tritium retention is a crucial consideration for any plasma-facing as well as structural material. In general, tritium retention for EUROFER97 is expected to be small. A very recent study used 20 MeV tungsten (W) ion beam irradiation as a surrogate for the displacement damage that neutrons will cause [A. Theodorou et al., Nucl. Mater. Energy 38, 101595 (2024)]. After, W irradiation, a low-temperature deuterium (D) plasma was used to decorate the created defects. Nuclear Reaction Analysis (NRA) with $^{3}$He was then employed to measure the trapped D within and beyond the damage region. The study showed that the retention of D is substantially increased by the displacement damage. However, post-irradiation annealing at 350°C recovered all radiation-induced defects. As no data was available so far for relevant irradiation doses and temperatures we adopted the methodology of the previous study, modifying it in four essential aspects: Firstly, rather than post-irradiation annealing of a microstructure that was irradiated at room temperature, high temperature irradiation was applied. Secondly, a continuous, broad beam was applied. Thirdly, irradiations were conducted with damage doses ranging between 0.7 and 100 dpa with dose rates ranging from $10^{-5}$ to $10^{-3}$ dpa/s. Fourthly, X-ray diffraction (XRD) analysis was applied to measure the lattice and micro strain induced by the ion irradiations.
XRD analysis revealed increased tensile lattice strain within the ion range, compared to the non-irradiated reference sample for all irradiation temperatures up to 400°C. Post-irradiation annealing at 600°C resulted in defect recovery and strain relaxation. NRA gives a very similar level for D retention within the damage region as compared to previous studies involving 0.6 dpa irradiations with scanned beam. D retention returns back to the level of pristine EUROFER97 when irradiation damaging is performed at 300°C. The present experiments support the initial assumptions about irradiation-induced defect densities that were made to predict tritium loss in a DEMO first wall in [K. Schmid, Nucl. Fusion 65, 026039 (2025)].

Speaker: Thomas Schwarz-Selinger (MPPL)

297 3.051 Research Progress on Supersonic Molecular Beam Injection for Burning Plasma Devices

Supersonic Molecular Beam Injection (SMBI) is an independently developed plasma fueling technology for tokamak devices in China. When high-pressure gas passes through a Laval nozzle into a low-pressure environment, the gas velocity is increased to supersonic speeds, which offers higher fueling efficiency compared to conventional gas puffing. The supersonic molecular beam injection (SMBI) system as a fueling method on EAST tokamak has become primary feedback fueling tool with higher fueling efficiency [1-2].
However, higher parameters of future burning plasma will reduce the fueling efficiency of SMBI. Experimental studies have shown that the formation of clusters in the supersonic molecular beam can further enhance the fueling efficiency. A low-temperature SMBI system has been designed, which uses a liquid nitrogen immersion cooling method to pre-cool the gas source temperature to ~150 K in advance, in order to promote the generation of beam clusters. Experimental results of LT-SMBI on EAST demonstrate that, compared with RT-SMBI, LT-SMBI achieves deeper fueling penetration depth and higher fueling efficiency.
This report meticulously examines the effects of the nozzle throat size, expansion section length, and expansion section angle on the beam. The results show that the most concentrated axial number density distribution is achieved at around 40°, and appropriately increasing the throat size and the length of the expansion section can improve fueling efficiency. At the same time, for burning plasma that requires tritium injection into the plasma, the use of a tritium gas source at a pressure greater than one atmosphere poses safety risks. We have studied the variation in particle fueling efficiency of low-pressure SMBI by reducing the pressure of the SMBI deuterium gas source to around 0.8 bar. Through the study of low-temperature and low-pressure supersonic molecular beam systems, technical support has been provided for the development of supersonic molecular beam injection technology for burning plasma in the future.

[1] G.L. Xiao et al., A review of supersonic molecular beam injection for plasma fueling and physical studies in magnetic fusion devices, Reviews of Modern Plasma Physics (2023)
[2] Xingwei Zheng，Jiangang Li，Jiansheng Hu，Density limits investigation and high density operation in EAST tokamak, 2016 Plasma Physics Control Fusion 58 055013

Speaker: Bin Cao (ASIPP)

298 3.052 Divertor Fuel Injection System for Detachment Experiments in KSTAR

This paper presents the design and construction of a Divertor Fueling System (DFS) developed for the advanced KSTAR operation under diverted discharge configurations aimed at enabling mainly high-density plasma operation and detachment control. The fueling system essential for plasma initiation, density regulation, radifo-frequency (RF) heating coupling, vacuum-wall impurity flushing, radiative-cooling experiments, and mitigation of plasma disruption has been designed with each component tailored to its functional role.
The DFS is built on the gas-injection infrastructure originally installed for the main gas injection system. Its mechanical assembly comprises gas supply, branching, storage, injection, and port-side distribution subsystems, while the control architecture includes data logging and valve actuation. For rapid fueling, ten high-speed piezoelectric injection valves have been employed. The valves are actuated via a digital valve controller that receives real-time command signals from the Plasma Control System (PCS), allowing precise and real-time control. Each piezo valve uses a disk-bender piezo element driven by bias voltage in range of 0–250 V, enabling proportional gas influx according to the applied voltage. The valve design achieves opening/closing response times below 2 ms and delivers a maximum flow exceeding 500 Torr·ℓ/s.
Additionally, the system segregates gas species and injection zones (ports), by a number of manifolds enabling multiple ports, multi-zone gas injection into the divertor region, thereby ensuring not only stable and precise gas delivery but more detailed study of the detachement control. Annual measurements of injected gas amounts per species are recorded and provided, enabling versatile use across various plasma experiments. The implemented design and construction is expected to meet the demanding requirements for high-density plasma fueling and detachment control in KSTAR divertor operations.

Speaker: Mr Jae-in Song (KFE)

299 3.053 Density Limits on the KSTAR Plasmas

The KSTAR tokamak is now under transition period of the plasma facing materials from carbon to tungsten: the carbon tiles as the bottom divertors are replaced with tungsten mono-blocks before 2023. The remaining PFCs such as passive stabilizers, top divertors and limiters will be coated tungsten on the existing carbon tiles for the upcoming 2027 experimental campaign. Accordingly, the normal densities with moderate or zero gas fueling but neutral beams tend to increase. However, high densities with the considerable amount of fuel injections are not clearly increased compared than those of carbon wall era. The density signals used hereafter are basically line-averaged ones $\overline{n_e}$ obtained by tangential two-color interferometer (TCI). The maximum achievable density was up to 0.8 of Greenwald density ($n_{GW}$) as a record number and the $n_{GW}$ has never been achieved in the KSTAR tokamak both under the full carbon PFCs or the recent bottom W divertors. This is due to the mainly two reasons which are combined together : 1) the volume of the KSTAR plasmas (12 m$^3$) are very small compared to the size of the vacuum vessel (110 m$^{3}$). Whenever the H-mode discharge is changed back to L-mode even with strong gas puffing to push the density up to the Greenwald limit, the density drops abruptly down to below 0.5 $n_{GW}$ and the MARFE instability pops up causing immediate major disruptions, and 2) hence the density limit of the KSTAR is inevitably the same term to the H-mode density limit (HDL) which is critically influenced by the total heating power magnitude to sustain the H-mode over the power threshold. In this regards, there will be two ways to achieve the Greenwald density on KSTAR : 1) to reduce the vacuum vessel size to sustain the density as high as possible even after the H-L back-transition like other machines e.g. JET and ASDEX-Upgrade, 2) to enhance the total heating power to sustain H-mode until the density reaches $n_{GW}$.

Speaker: June-Woo Juhn (Korea Institute of Fusion Energy)

300 3.054 Retention modelling for stellarator reactor design based on W7-X

The tritium fuel cycle is an important aspect for the design of a fusion reactor, since the initial on-site amount of tritium must ensure self-sufficient reactor operation with required fusion power and tritium breeding efficiency, complying at the same time with safety limits on in-vessel tritium accumulation. Retention of hydrogenic species inside plasma-facing components (PFCs) represents a challenge for the fuel cycle: although ion and neutral fuel fluxes impinging on the PFCs only penetrate some nanometres into the wall upon impact, diffusive processes of light fuel atoms can drive significant amounts of fuel deeper into the wall structure when constant bombardment by the fuel is considered. For a reactor-class device, with some hundred square metres of PFCs exposed to the plasma and charge-exchange neutral fluxes, accumulation of tritium trapped at material defects, especially in the presence of neutron-induced material damage represents a significant safety concern and may cause issues related to economic operability of the device.
FESTIM [1] is a simulation code designed for fusion applications, which calculates the diffusive transport and trapping of hydrogenic species in materials like tungsten (W), which is the preferred material for PFCs in high-power fusion devices. In this work, the FESTIM code is applied to assess the retention of tritium in a hypothetical full-tungsten fusion device with Wendelstein 7-X (W7-X) stellarator geometry. The approach follows the HISP framework [2]. Taking into account 3D-resolved wall plasma and heat fluxes from an existing EMC3-EIRENE solution for a W7-X plasma, a set of 1D FESTIM simulations evaluates the local fuel retention for selected representative wall locations. The total tritium inventory in the fusion device after $10^6$ seconds of plasma operation is estimated by scaling the local results with the corresponding total surface areas. This result can be used to upscale to larger geometries of potential fusion power plants.

[1] R. Delaporte-Mathurin et al., “FESTIM: An open-source code for hydrogen transport simulations”, Int. J. Hydrog. Energy 63 (2024)
[2] K. Dunnel et al., “Hydrogen Inventory Simulations for PFCs (HISP) in ITER”, presented at the 31st IEEE Symposium on Fusion Engineering (SOFE2025), June 23-26, 2025, Cambridge, MA USA

Speaker: Sebastian Rode (Forschungszentrum Jülich GmbH, Institute of Fusion Energy and Nuclear Waste Management – Plasma Physics, 52425 Jülich, Germany)

301 3.055 Hydrogen isotope permeation and retention in pure and dispersion-strengthened tungsten: from pristine microstructures to helium-induced fuzz surfaces

The transport and retention of hydrogen isotopes is of vital importance for the realization of future commercial fusion reactors because it is closely related to plasma operation, fuel recycling, and radiation safety. Tungsten (W) is a leading plasma-facing material, and its performance can be significantly enhanced by the incorporation of small amounts of ultrafine oxide particles (such as La2O3, Y2O3, and ZrO2) or carbide particles (such as ZrC and TiC) , which suppress dislocation motion and grain boundary migration. However, systematic studies on hydrogen isotope permeation and retention in such dispersion-strengthened W materials remain limited. Furthermore, the plasma-material interactions present another critical operational challenge for fusion devices. Under irradiation by low-energy, high-flux helium (He) plasma, the W surface forms a nanostructured tendril-like layer known as fuzz, which can significantly modify near-surface transport pathways and thus influence hydrogen isotope permeation behavior. To date, the fuzz growth behavior on dispersion-strengthened W and its subsequent impact on deuterium (D) permeation have not yet been thoroughly explored.
This study systematically investigates the hydrogen isotope permeation and retention behavior of pure W, one carbide dispersion-strengthened W (W-ZrC), and three oxide dispersion-strengthened W materials (W-La2O3, W-Y2O3, and W-ZrO2) by means of D2 gas-driven permeation experiments and thermal desorption spectroscopy following static gas-phase D2 charging. Subsequently, the surface fuzz growth behavior of these materials under low-energy, high-flux He plasma irradiation was examined using the linear plasma device CLIPS. He plasma irradiation experiments were conducted over a fluence range from 6 × 1024 m-2 to 2 × 1026 m-2 at a He flux of 2.8 × 1022 m-2s-1, a sample temperature of 1193 K, and an incident He ion energy of 90 eV. The fuzz morphology and its fluence-dependent evolution were characterized by scanning electron microscopy (SEM) and focused ion beam (FIB) cross-sectioning. Transmission electron microscopy (TEM) was further employed to investigate the internal structure of the nanofibrous tendrils and the role of dispersed particles in fuzz growth. In addition, the influence of fuzz structures formed at different irradiation fluence on D permeation behavior was also examined.

Speaker: zeshi gao (University of Science and Technology of China)

302 3.056 Theoretical modeling of hydrogen transport in boron structures

Boronization plays a fundamental role in impurity control and plasma conditioning in fusion devices. With tungsten now implemented as the primary plasma-facing material in ITER, new constraints have emerged regarding impurity transport, plasma stability, and hydrogen isotope retention. One of the main motivations for using boron is the high affinity for oxygen and other impurities, a requirement essential to meet for ITER operational regimes: minimizing impurity release during plasma start-up and enabling accurate monitoring of tritium inventory on the plasma-facing components [1]. During plasma operation, ions and other plasma species bombard the boronized surfaces, leading to sputtering, erosion, and redeposition of boron, tungsten, and hydrogen isotopes. These interactions result in the gradual formation of boron–deuterium and boron–tungsten composite layers, which significantly influence impurity retention and overall plasma performance. However, there are not many studies in the literature that describe hydrogen isotopes behaviour in boron structures in crystalline and amorphous form, which makes H isotopes retention and release energies very hard to predict.

Hydrogen transport in boron structures is investigated using ab initio methods based on density functional theory (DFT) and the SIESTA package, for both crystalline and amorphous structures. An energy-landscape-based approach, Landscape Flooding Algorithm (LFA), is constructed and applied to identify saddle points and determine the most probable pathways of hydrogen within the boron network. In addition, hydrogen transport is modeled as a stochastic Markov process, where transition probabilities between metastable sites are governed by the computed energy barriers. The time-dependent probability distributions of hydrogen positions within the material is found by solving the Master equastions. This multi-scale approach describes the influence of local structural variations on diffusion behavior and gives informations on hydrogen retention and migration in boron-rich layers formed during plasma operations. Diffusion coefficients are further computed by calculating the mean squared displacement (MSD) based on time-probabilities.

[1] A. Bortolon, V. Rohde, R. Maingi, E. Wolfrum, R. Dux, A. Herrmann, R. Lunsford, R. McDermott, A. Nagy, A. Kallenbach, et al., Real-time wall conditioning by controlled injection of boron and boron nitride powder in full tungsten wall asdex upgrade, Nuclear Materials and Energy 19 (2019) 384–389.

Acknowledgement
This work was supported by a grant of the Romanian Ministry of Education and Research, CNCS-UEFISCDI, project number PN-IV-P2-2.1-TE-2023-1140, within PNCDI IV.

Speaker: Bianca Solomonea (National Institute for Laser, Plasma and Radiation Physics, Bucharest, Romania, Doctoral School of Physics, Faculty of Physics, University of Bucharest, Magurele-Ilfov, Romania)

303 3.057 LIBS-Based In-situ Quantification of Fuel Retention in Boron-Coated Plasma-Facing Components

Fuel retention in plasma-facing components (PFCs) is critical for plasma control and safety in fusion devices. Laser-Induced Breakdown Spectroscopy (LIBS) offers a promising approach for in-situ monitoring of hydrogen isotopes, yet its accuracy is challenged by the minimal spectral shift between deuterium (D) and hydrogen (H) lines. In this study, graphite tiles from the EAST tokamak were first used to evaluate the effects of ambient gas and pressure on spectral resolution. It is confirmed that vacuum conditions (10⁻³ Pa) and helium at 100 Pa enabled high resolution and sustained signal intensity. Based on these findings, boron films of varying thicknesses (300, 800, 1200, and 1600 µm) were deposited via chemical vapor deposition onto tungsten substrates. These samples were irradiated at room temperature using a linear plasma device with D fluences of 6.30×10²³ and 9.45×10²³ ions/m² (1–1.5 hours), mimicking D implantation in PFCs. Under optimized conditions (over-focused lens position, 300 ns detection delay), distinct H/D spectral separation was achieved with preserved D signal intensity. Post-irradiation analysis revealed tungsten sputtering from the substrate during plasma injection, corroborated by TDS quantification. A support vector machine (SVM) algorithm was successfully developed to establish a quantitative relationship between boron film thickness and D content, demonstrating the potential of LIBS for precise, in-situ fuel retention diagnostics in fusion environments.

Speaker: Hao Sun (xian jiaotong university)

304 3.058 Plasma Fueling and Particle Exhaust Control in a fusion power reactor

Abstract
The fusion power plant operation with a fuel cycle including a direct internal recycling aims to increase the fuel burnup fraction and to reduce the tritium inventory. The expected accumulation of protium and fuel imbalance requires control of plasma fueling and particle exhaust. The formation of protium and tritium in the reactor chamber in concomitant to DT fusion reactions and the release of tritium from the first wall can lead both to an accumulation of protium followed by the fuel dilution and the fuel imbalance in the reactor chamber followed by the power dip. During relatively long reactor operation, its first wall could be the main source of neutral gas feeding the plasma. Moreover, in these cases the tokamak wall contains the amount of hydrogen isotops, which is significantly larger in many cases by orders of magnitude than that in the plasma.
In this work, we estimate the accumulation of protium in the reactor chamber due to fuel recirculation, when the exhaust gas mixture returns after purification back to the reactor chamber, depending on the separation fraction of the metal foil pumps and the duration of the burning pulse. It has been found that the acceptable concentration of protium is limited by the requirements for a self-sustaining reaction to occur, and strongly depends on wall conditioning. It is shown that in order to achieve maximum burn, in the case of fuel imbalance, a lower tritium concentration will be required in the fuel supplied from an external source to the reactor chamber.

Speaker: Dr Yuri IGITKHANOV (KIT)

305 3.059 Disruptions triggered by fueling pellets on EAST tokamak

A new fueling pellet injection system has been developped based on GM-cryocooler on EAST recently. In 2025 EAST experimental campaign, many deuterium pellets have been injected into the high confinement mode plasmas successfully with the new system. Both minor and major plasma disruptions have been observed after the outer midplane pellet injections. In the case of minor disruption, a rapid density decay occurs immediately after the density enhanced by pellet. But the plasma is not terminated and restored to the state before the pellet injection subsequently. As a result, there is no significant change in overall density evolution after several pellet injections. When larger or neon dopped deuterium pellets are injected, a major disruption offen occurs. Sawtooth crashes are triggered after one big pure deuterium pellet and a major disruption has happened after some pellet injections. Some Ne~2% doped D2 pellets have been also injected into the plasma. During the this process, major disruption is triggered directly after one or two doped pellet injections. These results suggest that the proper pellet parameters should be considered to avoid the trigger of disruption.

Speaker: Jilei Hou (Institute of Plasma Physics, Hefei Institutes of Physical Science, Chinese Academy of Sciences)

306 3.060 Effect of defects on deuterium retention in 3C-SiC coated with tungsten thin films

3C-SiC is a leading candidate material for fusion applications due to its high-temperature
stability, low swelling, and good radiation resistance. SiC-based composites are being considered
as high-temperature, low-activation structural material substitutes, and, because of the good
thermal compatibility between SiC and tungsten (W) and the low tritium (T) permeability of SiC,
CVD 3C-SiC is also of interest as a tritium permeation barrier (TPB) for first-wall components.
However, the introduction of defects could increase the solubility of hydrogenic gases in SiC,
reducing its effectiveness as a TPB and leading to higher retention. A similar phenomenon is
observed in W: increased deuterium (D) retention is observed in tungsten that has been ion
irradiated prior to plasma exposure as compared to pristine tungsten.

To investigate this issue further, an experimental campaign is being developed on the DIONISOS
linear plasma device which is coupled to the CLASS ion accelerator, enabling simultaneous D
plasma exposure and ion irradiation, thus simulating fusion-relevant conditions. The campaign
focuses on bulk W and 3C-SiC coated with a $\sim 100\,\mathrm{nm}$ W layer (W--SiC) and is structured in
three stages. In the first stage, pristine W and W--SiC samples are exposed to D plasma (ion flux
$\phi_{\mathrm{ion}} \approx 10^{21}\,\mathrm{m}^{-2}\,\mathrm{s}^{-1}$) for $\sim 5$ hours at $500\,^{\circ}\mathrm{C}$ to establish a baseline for retention and erosion in the
absence of pre-existing damage. In the second stage, W and W--SiC samples are pre-damaged to
$\sim 1$~dpa at $600\,^{\circ}\mathrm{C}$ using W and Si ion beams, respectively, to generate a saturated population of
irradiation defects before repeating the same plasma exposure conditions. In the third stage, W
and W--SiC are subjected to simultaneous plasma exposure and ion irradiation at $500\,^{\circ}\mathrm{C}$, with
plasma and beam parameters matched to those used in the pristine and sequential cases. The third
stage approximates conditions where gas implantation and defect production occur together, as
expected in a fusion power plant environment.

After exposure, deuterium retention and depth profiles will be quantified using nuclear reaction
analysis (NRA), while changes in the W layer thickness and erosion will be characterized by
Rutherford backscattering spectrometry (RBS).

Speaker: Keshav Vasudeva (MIT PSFC)

307 3.061 Towards clean target exposures in MPEX using innovative helicon window architectures to reduce impurity sourcing

The Material Plasma Exposure eXperiment (MPEX) is a linear plasma device expected to be operational by 2028. MPEX uses a novel, high power (≤ 200 kW) helicon plasma source combined with electron and ion-cyclotron heating schemes to enable divertor-relevant plasma parameters at the target for extremely long pulses (≤ 1x106 seconds) with independent control of electron and ion temperatures. Operation of this high-power helicon source in Proto-MPEX generated impurities due to rectified sheath-induced erosion of the helicon window, which in turn were deposited downstream on to the target complicating the local plasma-material interactions (PMI) studies. To mitigate this issue, two strategies have been developed and implemented at the plasma source region1: 1) coating the plasma-facing surface of the silicon nitride (Si3N4) helicon window with a high-Z oxide material lowering the effective sputtering yield; and, 2) applying a Faraday shield between this window and the helicon antenna to reduce the rectified sheath voltage thereby lowering the incident ion energy impacting the helicon window surface.
Tantalum oxide (Ta2O5) was selected to coat the inner surface of the Si3N4 window due to both low electrical conductivity and low sputtering yield to deuterium. Initial testing of this coated window without a Faraday shield in Proto-Lite (with no vacuum conditioning, ne ~ 4×1019 m-3, Te ~ 2 eV) has greatly reduced impurity production from window erosion with no traces of Ta at the target surface. This is in contrast with previous Proto-Lite operations with an uncoated window showing ~ 400 nm thick impurity deposition due to window erosion under similar experimental conditions (i.e., fluence ~ 1024 m-2). PISCES-A results showed that improving the vacuum level should further lower the erosion of the coated window and increase its lifespan2. The novel Faraday shield uses a pyrolytic graphite coating (~ 100 Ω/sq) on the inner surface of a quartz window enclosing the helicon window, with cooling water flowing in the gap. Implementing this Faraday shield has required unique coating methods and engineering design. Nonetheless, COMSOL modeling shows that the rectified sheath voltage is reduced by a factor of 201, and initial tests are underway to ensure minimal effect on helicon source operation.
These strategies are promising in achieving clean target exposures in MPEX for the maximum expected pulse length, ensuring a successful PMI research platform.

References:
1Rapp et al., IEEE Transactions on Plasma Sci., 52, 9 (2024).
2Dhamale et al., Plasma Phys. Control. Fusion, 67, 095001 (2025).

Speaker: Dr Gayatri Dhamale (Oak Ridge National Laboratory)

308 3.062 Impact of impurity powder injection on the first wall material transport in LHD

In the Large Helical Device (LHD), boron powder injection experiments have been conducted using an impurity powder dropper (IPD), developed by PPPL, for real-time wall conditioning. A pronounced effect of boron powder injection has been observed on the line emission intensity of iron, the main constituent element of the stainless steel type 316L first wall material in LHD. The emission intensity of highly charged iron ions (Fe XXIV) radiating in the main plasma is significantly reduced by boron injection, in some cases to less than one-tenth of its original value. In contrast, the emission intensity of low-charge iron ions (Fe VIII, Fe IX) radiating in the scrape-off layer (SOL) shows little or no change, suggesting that the generation of iron impurities remains largely unaffected. These observations indicate that boron powder injection alters the transport of the first wall material into the main plasma. Such effects on iron transport have also been observed with the injection of carbon powder and lithium granules.

In fusion devices with metallic walls, it is desirable to minimize the penetration of wall materials into the main plasma as impurities. Understanding the physical mechanisms by which impurity powder injection reduces impurity contamination may contribute to impurity control in future fusion reactors. In LHD, the divertor plates are made of carbon, while the first wall material is stainless steel. With regard to the transport of iron originating from the first wall, previous studies have examined the effects of radial electric field formation in the peripheral region of the main plasma, the balance between frictional forces due to plasma flows toward the divertor in the scrape-off layer (SOL), and thermal force arising from temperature gradients, as well as turbulent transport. In this presentation, results obtained by varying the background plasma conditions to examine the effect of impurity powder injection on iron transport will be shown, and the mechanisms by which impurity powder injection suppresses the penetration of iron impurities into the main plasma will be discussed.

Speaker: Suguru Masuzaki (National Institute for Fusion Science)

309 3.063 Experimental and Numerical Characterization of ICRF-Induced Plasma-Wall Interaction in the Regime of Propagative Slow Waves

When the density at the Ion Cyclotron Range of Frequencies (ICRF) antenna limiter’s edge falls below the lower hybrid (LH) resonance density (S = 0 in Stix dielectric tensor), the slow wave (SW) can propagate in front of the antenna. This wave carries large parallel electric fields that can strongly enhance the sheath potentials on the antenna limiters, thereby increasing the sputtering yield of ions striking the wall. Due to the presence of a cold plasma resonance and the SW’s short wavelength, this regime poses a major modeling challenge. Such conditions may be encountered in ITER [Colas et al., JNM (2025)] or with other foreseen ICRF systems such as the WEST travelling-wave array [Ragona et al., FED (2025)].

To document this low-density regime, WEST experiments were carried out in conditions where the SW was propagative in front of an active ICRF antenna. Using a reciprocating emissive probe connected magnetically to the antenna, we measured for the first time the DC plasma potential, VDC, over a radial scan of the LH resonance layer. VDC peaks when the density at the antenna limiter’s edge is near the LH resonance density, but never exceeds typical values of a few hundred volts. From the plasma-wall interaction standpoint, this regime can be highly favorable: as the particle fluxes are lower and the sheath potentials similar to standard operating conditions at the same antenna voltage, the local tungsten sources at the ICRF antenna and all other outer-wall objects become nearly undetectable. By contrast, sputtering in the divertor is primarily determined by the sputtering yield. Core impurity contamination is likewise significantly lower when the antennas are located far from the separatrix, both with and without ICRF power, and the radiated power fraction is reduced. Despite modest coupled powers at high clearance, good ICRF heating is maintained as the edge density drops below S = 0, and no deleterious operational consequences are observed across all relevant core plasma metrics when operating the antenna from this low-density region.

To help interpret these measurements, the 2D SW sheath interaction model by Myra and D’Ippolito [PRL 2008] was generalized to include arbitrary magnetic field incidences, additional dielectric tensor components, and both resistive and capacitive contributions to the sheath RF impedance. This simplified model is qualitatively consistent with the measured VDC trends if the incoming SW electric field is kept fixed while the density or wall inclination angles are varied.

Speaker: Raymond Diab (Massachusetts Institute of Technology)

310 3.064 Influence of impurity injection location on tokamak plasma performance

In tokamaks, intense heat fluxes strike the divertor targets, risking damage and core plasma contamination by eroded wall material. Future reactors must therefore operate in a detached regime, where heat loads and plasma temperatures near the walls are greatly reduced [1]. This can be achieved by impurity seeding, which promotes radiative cooling and momentum losses in the boundary plasma. However, minimizing impurity penetration into the core is critical to preserve fusion performance [2]. In this work, a lower single-null, L-mode, nitrogen seeded discharge on TCV is repeated while changing the seeding valve across six different poloidal locations, including private flux region (PRF), high-field side (HFS), low-field side (LFS) and ceiling (TOP). Results reveal that the access to a detached regime, diagnosed via divertor Langmuir probes and the CIII emissivity front, is mostly insensitive to the seeding location. On the contrary, core contamination (estimated via $Z_{eff}$) shows strongly different behaviors, especially when comparing PFR and TOP seeding, with the latter leading to $50$% higher $Z_{eff}$. Interestingly, HFS seeding from above the divertor baffle yields lower core contamination than LFS seeding from below the baffle, hinting at a potential role of flows in determining impurity redistribution. Bayesian inference is applied to multispectral visible imaging data [3,4] to infer 2D maps of nitrogen concentration in the divertor region and findings are compared to SOLPS simulations. These results highlight the potential to optimize seeding strategies and impurity content quantification for future devices.

[1] P C Stangeby 2018 PlasmaPhys.Control.Fusion 60 044022
[2] T. Pütterich et al 2019 Nucl.Fusion 59 056013
[3] A. Perek et al 2022 Nucl. Fusion 62 096012
[4] B.L. Linehan et al 2023 Nucl.Fusion 63 036021

Speaker: Riccardo Ian Morgan (SPC-EPFL)

311 3.065 Integrated Core-Edge Modelling of Tungsten Using SOLEDGE-3X and Application to WEST Experiment

The WEST tokamak has been used to study long-pulse scenarios in a full Tungsten environment. Its contamination in the core, which reduces performance, as well as the location of erosion/deposition sites are important. It continues to be difficult however, to get accurate estimates of sputtered fluxes and Tungsten content in any given pulse. Modelling that aims to answer this question should ideally be integrated, both in terms of core-edge as well as main ion-impurity.

To this end, the SOLEDGE-3X code was recently extended in order to enable it to perform core-edge integrated simulations of all relevant species. The plasma wall interaction includes the sputtering and prompt redeposition of Tungsten using analytical models that capture high-energy tail sputtering as well as the sputtered energy distribution. The edge-SOL transport is modelled in 2D, using anomalous diffusivities. In the core meanwhile, a 1D model is used, coupled to QLKNN-10D, a fast turbulence model for transport, as well as an analytical neoclassical model for the transport of Tungsten. Using this framework, a number of applications have been shown.

A comparison was made with the long pulse scenario WEST #58245. The flat-top averaged core density and temperature profiles were reproduced using the integrated framework, as well as the density at the strike points measured using Langmuir probes. Oxygen was used as a proxy impurity, with the source being from the wall as opposed to from the inner boundary as in previous work. The radiated fraction of Tungsten was then looked at and found to be in the experimentally measured flat-top range. The lower strike points and baffle were the main contributors in terms of sputtering sources, with the baffle found to be less effectively screened. Comparisons are being done to bolometry and visible spectroscopy at the strike points. Further applications of the model are ongoing, in particular with respect to a study of the sensitivity of the input power to the radiated fraction.

D. Tskhakaya and M. Groth, J. Nuc Mat 463 (2015)
L. Cappelli et al., PPCF 65, no. 9 (2023): 095001
K. L. Van De Plassche et al., PoP 27, no. 2 (2020): 022310
Stephens, C.D. et al, Journal of Plasma Physics 87, no. 4 (2021)
P. Maget et al, PPCF 62, 105001 (2020)
D. Fajardo et al, PPCF 64, 055017 (2022)
S. Di Genova et al 2024 Nucl. Fusion 64 126049
J. Bucalossi et al 2024 Nucl. Fusion 64 112022

Speaker: Naren Varadarajan (CEA)

312 3.066 Erosion and deposition patterns on ITER-grade actively-cooled divertor in WEST high-fluence campaign: analysis and comparison with SOLEDGE3X and ERO2.0 modeling results

A major issue for next step devices is the control of plasma wall interaction, both for keeping the material erosion compatible sufficient lifetime of the components as well as for mitigating core contamination by high Z impurities and consequently the reduction of plasma performances. In this respect, WEST experiments supported by numerical modeling are particularly relevant to progress in the physical understanding of the complex interplay between erosion patterns, impurity migration and efficiency of plasma screening. In this contribution we will present modeling and experimental results concerning deuterium high-fluence experimental campaign which was conducted in the WEST tokamak. In 2023, the ITER-grade actively-cooled divertor has been exposed to ITER-relevant (1.0 x 1027 part.m-2) particle fluences by repeating the same attached plasma scenario (divertor Te ~ 20 eV), with a total cumulated plasma time of 10790 s. Impurities sources have been monitored by visible spectroscopy, showing high content of boron and carbon all along the campaign especially on the high-field side where deposited layers are observed. Using visible spectroscopy data (for WI 4009 Å radiance) and flush-mounted Langmuir probe measurements (to obtain ne and Te dependent local S/XB coefficients) one can estimate tungsten gross erosion flux up to 1.5 × 1019 part.m−2.s−1, leading to around 0.7 μm of net erosion near the outer strike point, assuming the high fluence campaign plasma time and relatively low 75% prompt redeposition rates. These estimations are in contrast with recent results from post-mortem analysis indicating higher values (around 7 μm of net erosion on the inner strike point) after both, foregoing and high-fluence campaign. To investigate the observed erosion and redeposition patterns, several background plasma models have been simulated with SOLEDGE3X-EIRENE, constrained by experimental measurements, with and without the inclusion of drifts. Plasma backgrounds are simulated with proxy light impurity (O) with various concentrations (1 to 3%) to match experimental conditions (particle flux and electron temperature at the edge). The impact of drifts is important to reproduce flux asymmetries between inner and outer sides and target plasma parameters. W erosion and migration is then computed with ERO2.0 using such plasma backgrounds. The simulated erosion depth at both strike-points is barely the same with and without drifts and for each used oxygen concentration (around 4 to 9 µm), consistently with post-mortem measurements. The far inner redeposition area is observed only when background plasma models are computed with drifts are taken into account.

Speaker: Alexis Huart (CEA)

313 3.067 A $^{15}$N tracer injection experiment to study long range impurity transport in W7-X

The complex 3D magnetic topology of the island divertor in Wendelstein 7-X (W7-X) produces feature rich erosion and deposition patterns on the divertor plates and strongly affects impurity transport in the edge plasma region. While crucial for stellarator optimization towards reactor-relevant configurations, modelling impurity transport is challenging and requires thorough code validation. Here, the gold standard is tracer particle experiments: isotope-labeled impurities are injected at known positions and quantities into repeated, identical discharges to establish erosion, deposition, and migration patterns on plasma-facing surfaces. Components or samples are then removed and undergo ex-situ post-mortem analysis after which the results are compared to model predictions.

Therefore, to explicitly study long-range migration in W7-X, a dedicated tracer injection experiment with isotope separated $^{15}$N was performed: a total of $\sim 1.5 \times 10^{21}$ $^{15}$N atoms were injected via the divertor gas injector module. The $^{15}$N was then collected on graphite samples mounted on the multipurpose manipulator which intersects the same magnetic island as the injection but is separated by a toroidal angle of $\sim 70^\circ$). For comparison with previous $^{13}$CH$_4$ tracer experiments~\cite{Kawan2024}, similar magnetic configurations and plasma parameters were chosen: standard configuration $n_{e,\text{line int.}} \sim 6 \times 10^{19} \text{m}^{-2}$, $T_e \sim 2.3$ keV, and 3.5 MW of pure ECRH heating. The injections started 1 s after plasma initialization and continued until the ECRH power was ramped down. Altogether 27 nearly identical discharges with injection were performed consecutively, for a total plasma duration of 350 s. The radiative power $P_{\rm rad}$ increased from $\sim 1$ MW to $\sim 1.5$ MW during the injection in the initial discharges; a moderate legacy effect due to the repeated nitrogen injection was observed as the initial radiative fraction steadily increased to $\sim 2.5$ MW in the later discharges. Due to a scheduled opening of the plasma vessel right after the experiment, additional plasma facing components were retrieved and it is planed to quantify also their surface $^{15}$N concentration.

This contribution summarizes the experimental procedure and gives an overview of the current state of the post-mortem analysis, especially the measurements of $^{15}$N on the collection samples by nuclear reaction analysis (via the $^{15}$N(p,$\alpha)^{12}$C reaction). Furthermore, a comparison of the $^{15}$N tracer experiment with the previous $^{13}$CH$_4$ tracer experiment is presented, and implications for future optimization of stellarator edge configurations are discussed.

[1] C. Kawan et al., Nuclear Material Energy 39 (2024) 101675.

Speaker: Timo Dittmar (FZJ)

314 3.068 Investigations of impurities and pellet injection in EAST and HL-3 tokamaks

ITER has switched to full tungsten wall configuration[1].High-Z tungsten dust originated from the Plasma-Surface Interactions(PSIs) may result in the degradation of plasma discharges, the H-L mode back transition or even disruption. Meanwhile, low-Z impurity pellets such as lithium or boron ones are strong candidates for ELM control with impurity injection. The small or grassy ELMs are triggered before the formation of large type-I ELMs. Consequently, the plasma disruptions can be avoided and the high-confinement, long-pulse discharges are achieved.
In this work, the transport of tungsten impurities, originated during the dust ablation process as well as the PSIs, are investigated with STRAHL code. The corresponding core energy dissipation due to radiation during impurity transport is also assessed. Furthermore, the lithium impurity pellet injections are modeled with the NDS-BOUT++ and the BOUT++ transport module. The penetration and the size distribution of the lithium pellet on its trajectory are simulated with the updated ablation module and compared to experiment diagnostics. The fueling pellets are investigated with PAM code to optimize EAST pellet fueling efficiency, considering the ∇B-induced plasmoid drift. In addition, the pellet ablation trail in tokamak plasma provides significant identifier for tokamak safety factor diagnostic. We will also report our recent work on tokamak diagnostics.
Our simulation results[2] show that the sub-micron sized tungsten dust originated from the EAST upper-divertor region is capable of penetrating through the Separatrix. For impurity pellet injections, the lithium pellet is capable of penetrating ~ 20 cm into the HL-3 plasma when injected horizontally from the low-field-side midplane with the initial velocity of 100 m/s. The plasma pressure in the pedestal region increased ~ 25% due to the lithium pellet ablation and ionization[3]. On fueling pellets, we find pellet penetration contribute more to the deep pellet deposition than the ∇B-induced plasmoid drifts in low temperature plasma, while deep pellet fueling in reactor relevant high temperature plasma has to rely on plasmoid drifts[4]. Finally, our developed stereo CCD system for tokamak safety factor diagnostic proves to be robust and accurate with the average difference of 6.8% compared the traditional MSE diagnostic in EAST discharges[5].

[1] C. Angioni, Nuclear Fusion 65, 062001 (2025).
[2] Z. Liu, etc., Physics of Plasmas 28, 122503 (2021);
[3] Z. Liu, etc., Physics of Plasmas, Under Review
[4] J. Zhang, J. L. Hou, Z. Liu, etc., Nuclear Fusion 64, 076012 (2024).
[5] C. Liang, Z. Liu*, etc., Review of Scientific Instruments 95, 043502 (2024)

Speaker: Zhuang Liu (Soochow University)

315 3.072 Tungsten Impurity Effects on Effective Charge and Radiation Fraction in KSTAR Plasmas

Tungsten impurities in tokamak plasmas introduce significant challenges due to their high radiative efficiency. Following the recent update of the ITER baseline, the importance of understanding tungsten impurity generation and transport has increased significantly. In KSTAR, the installation of a tungsten monoblock cassette divertor in 2023 has opened a new operational regime for ITER-relevant plasma-surface interaction studies. This study provides a statistical assessment of impurity behavior with tungsten divertor in KSTAR. Changes in carbon and tungsten source levels are assessed across 2021-2024 KSTAR campaigns. The effective charge (Zeff) and total radiated fraction (frad) behavior are investigated in terms of carbon and tungsten impurity levels. Subsquently, the impact on plasma performance and confinement is also explored. A comprehensive set of diagnostics is employed for impurity measurement in this study. Filterscope are used to measure the line emissions of neutral tungsten (W I, 400.9 nm) and C2+ ions (C III, 464.7 nm) in the plasma boundary, and visible spectroscopy measures the spectral lines of other light impurities. Vacuum ultraviolet (VUV) spectroscopy and X-ray crytsal spectroscopy (XICS) are used to assess the highly charged tungsten ions in the plasma core. Infrared imaging video bolometer (IRVB) provides 2D radiation profiles and the total radiated power. In addition, this study provides the averaged Zeff, derived from the line-integrated visible bremsstrahlung emissivity, line-averaged electron density, and core electron temperature. General trends before and after tungsten divertor installation are firstly investigated, followed by comparison between plasma configurations such as lower and upper single null configurations. Impurity behaviors in L-mode and H-mode are analyzed with various NBI and ECH heating powers. A focused analysis is dedicated on H-mode discharges with the tungsten divertor, where ELMs are known to be a primary source of tungsten sputtering. The tungsten source levels in inter- and intra-ELM phases are compared. Statistical analysis of their behavior and tungsten core accumulation is performed with respect to ELM frequency. The discharges with ELM mitigation by RMP are also included in statistics. This 0-D parameter avoids detailed profile fitting and reconstruction, and subsquently allows robust statistics over a large number of discharges. This study possibly contributes to the control of tungsten impurity and radiated power in both present partial- and future full-tungsten wall environments in KSTAR.

Speaker: Juhyeok Jang (Korea Institute of Fusion Energy)

316 3.073 On the sputtering and transport of the intrinsic carbon impurity in HL-3 ELMy plasma

In magnetic confinement fusion devices, the interaction between plasma-facing components (PFCs) and plasma leads to significant challenges in material integrity and plasma performance, necessitating a detailed assessment and understanding of sputtering and impurity transport behavior [1]. HL-3, a medium size tokamak, has achieved its critical milestones-a fusion triple product surpassing 0.65 × 10^{20} keV s/m^3 [2]. Notably, this device employs carbon-based materials for its divertor targets and first wall. This study investigates the sputtering yields and transport mechanisms of intrinsic carbon impurities in ELMy plasma on HL-3 tokamak.

This work is performed with the JOREK hybrid kinetic-fluid framework [3-5], where the main plasma (deuterium plasma) is treated by the reduced MHD model and the transport of carbon impurity is traced with full-orbit kinetic model. The carbon impurities originated from both physical and chemical sputtering are considered and the relative fractions of each are quantified [6]. In ELMy plasmas, the heat and particle fluxes on the divertor target are characterized by intense, transient pulses associated with ELM bursts, which exerts a profound influence on the sputtering process of intrinsic carbon impurities. To assess this phenomenon, time-dependent sputtering yields are calculated, demonstrating the significant enhancement on carbon influxes during ELM burst. To elucidate the reciprocal impact of carbon impurities on plasma behavior, the influences of carbon impurity on divertor power exhaust and evolution of MHD modes are investigated through a direct comparison of simulation results from scenario incorporating intrinsic carbon impurity and the other excluding them entirely.

References:
[1] K. Krieger, et al., Nucl. Fusion 65(4), 043001 (2025).
[2] W. Zhong, W. Chen, and X. Ji, The Innovation, 101167 (2025).
[3] M. Hoelzl, et al., Nucl. Fusion 61(6), 065001 (2021).
[4] S.Q. Korving, Physics of Plasmas 30(4), 042509 (2023).
[5] Z. Liang, Physics of Plasmas 32(1), 012502 (2025).
[6] B. Rainer, and E. Wolfgang, Sputtering by Particle Bombardment (Spinger, Berlin, 2007).

Speaker: Dr Zhe Liang (Dalian University of Technology)

317 3.074 Impact of MHD activity on impurity radiations and energetic electrons dynamics in ADITYA-U tokamak

In tokamaks, the energetic electrons are predominately generated during the LHCD and ECRH-assisted scenarios contributing to the non-inductive current drive and heating of the plasma, respectively. However, In the presence of an external electric field, the energetic electrons in plasma can be accelerated to very high energy, becoming “runaway” electrons. If left unchecked, these electrons could severely damage the plasma-facing components on any tokamak as large as ITER. Hence, studying the dynamics of energetic particles is crucial for the safe operation of large-scale fusion devices like ITER.

In the ADITYA-U tokamak, the low energy hard X-ray (LHXR) having an energy range of ~20-200 keV is primarily produced by the suprathermal electrons generated largely during lower hybrid waves injection. In contrast, the high energy hard X-ray (HHXR) having an energy range of ~1-3 MeV produced by the runaway electrons (REs) are always present in the plasma discharges and are modulated by the Magnetohydrodynamic (MHD) activity. The modulation of hard X-ray (HHXR, ~1-3 MeV) by the MHD instabilities such as sawtooth oscillations and drift tearing modes has been extensively studied in the ADITYA-U tokamak as well as other tokamaks [1][2][3][4].

We observe MHD-activity-driven periodic modulation in both low- and high-energy hard X-ray (LHXR and HHXR) signals in LHCD-assisted plasma discharges on the ADITYA-U tokamak, with a clear phase difference between the two. The threshold magnetic field perturbation required to induce this periodic modulation is found to differ for LHXR and HHXR, indicating a distinct response of different electron populations to MHD activity. In this work, we aim to identify the underlying physical mechanisms responsible for these observations, thereby improving our understanding of the interaction between MHD modes and energetic as well as suprathermal electrons in tokamak plasmas. In addition, MHD activity is observed to influence impurity radiation signals (e.g., C III, O II), and a detailed analysis of these effects in ADITYA-U discharges will also be presented.

References:

[1] Harshita Raj et al 2018 Nucl. Fusion 58 076004
[2] Magnetohydrodynamic instability modulated runaway electron losses in the ADITYA-U tokamak by S. Patel et al submitted to Nucl. Fusion
[3] O. Ficker et al 2017 Nucl. Fusion 57 076002
[4] J Kamleitner et al 2015 Plasma Phys. Control. Fusion 57 104009

Speaker: Komal Yadav (GNOI)

318 3.075 Simulation study of the integrated effect on tungsten contamination due to the toroidal field direction, divertor target geometry and divertor operation regime

Tungsten (W) is considered as the primary candidate material for plasma-facing components in the divertor of future fusion reactors owing to its superior physical properties. However, as a high-Z material, W can must be limited to an extremely low level in the core plasma, i.e. the severe contamination must be avoided. The tungsten contamination depends on its sputtering source from the divertor target, and the following transport process including leakage from the divertor and entry into the core. In our previous work [J. Guo et al., Nucl. Fusion 63 (2023) 126033] for the divertor geometry with both inner and outer vertical target, it was found that, due to the effect of drifts, the effective W sputtering source and the W transport toward the core region could be both enhanced, which thus resulted in a increase of core W concentration by about one order of magnitude. The influence due to the outer target geometry has been further studied [H.X. Ding et al., Nucl. Mater. Energy 41 (2024) 101754], with a horizontal outer target, although the sputtering source is still increased due to the drift effect under favorable toroidal field, the core W contamination can be reduced by one order of magnitude.
Recently, Chen et al. [Nucl. Fusion 64 (2024) 126001] indicate that, under unfavorable toroidal field, the divertor plasma distribution and impurity radiation pattern can be dramatically affected due to the formation of the potential well near the X point in private flux region. Wang et al. work [Plasma Sci. Technol. 25 (2023) 115102], the potential well can be formed under a relatively high radiative impurity seeding rate for the EAST configuration. Although in our previous work, the W contamination is higher under unfavorable toroidal field for the case of vertical outer target, the situation is still not clear when the potential well is formed, which is considered to increase the impurity retention dramatically. In this work, to have a comprehensive understand on the integrated effect on W contamination due to the toroidal field direction, divertor target geometry and difference operation regime, the simulation study is further conducted based on our previous work by scanning the radiative impurity seeding rate for divertor geometries with the vertical and horizontal outer target and under favorable and unfavorable toroidal field directions. Detailed results will be reported on the conference.

Speaker: Mr Hongxin Ding (School of Nuclear Science and Technology, University of Science and Technology of China)

319 3.076 Initial experiment of an internal coil device for impurity transport study

Experimental investigation of impurity transport in the scrape-off-layer (SOL) and divertor region of tokamaks is crucial for developing accurate impurity transport models. Since various divertor configurations are proposed for future fusion reactor, flexibility in magnetic field configuration is required for experimental devices. Some tokamaks provide various divertor configurations such as double-null, negative-triangularity, snowflake, super-X. Using smaller tokamaks will accelerate experimental research on impurity transport by improving diagnostic accessibility. However, stable formation of divertor configuration usually requires feedback control of plasma position and shape, which is technically difficult in smaller tokamaks due to the shorter timescale.
In this presentation, a smaller toroidal device for divertor configuration experiment is proposed. An internal coil, installed along the magnetic axis instead of the plasma current, generates poloidal flux. Then divertor configuration is achieved without feedback control of plasma. A scrape-off-layer experiment device using an internal coil, SOLEIL, has been designed and constructed in Nagoya University. Major and minor radii of an internal coil are $R = 0.18$ m and $a = 0.03$ m. The internal coil is suspended by insulator covered stainless steel rods from a center post. The radii of surface of the center post and inner wall of cylindrical vacuum vessel are $R = 0.075$ m and 0.3 m, respectively. Flat divertor plates are installed at $Z = \pm 0.18$ m. Triangularity $-0.4 < \delta < +0.4$ and elongation $\kappa < 1.8$ are roughly achievable shape parameters the limits imposed by electromagnetic forces and thermal constraints.
Plasma was produced by electron cyclotron resonance using a 2.45 GHz microwave. Langmuir probe measurements were conducted. The electron temperature is about 10 eV. In the horizontal plane, the electron density exhibited a radially decaying profile within the SOL.
Development on spectroscopic measurement system is also progressing. Argon ions are planned for impurity; expected charge states and wavelength of those line emissions are evaluated using collisional radiative models. Then, up to doubly charged argon ion ($Ar^{2+}$) will be distributed in SOL for above mentioned initial experiment. Required plasma parameter for higher charge states is also discussed.

Speaker: Atsushi Okamoto (Nagoya Univ.)

320 3.077 INVESTIGATION OF PLASMA-WALL INTERACTION AND IMPURITY TRANSPORT BEHAVIORS OF HL-3 PLASMAS

Plasma-wall interaction (PWI) during steady phases and transient events such as disruptions and edge localized modes (ELMs) is a critical issue in the magnetically confined plasmas, mainly because the particles and heat fluxes from plasmas need to be controlled to prevent damage to the plasma facing components (PFCs), as well as the impact of impurity influx on the main plasma performance should be mitigated. An extreme-wide angle view diagnostic with field of view (FOV) of 〖120〗^°, allowing us to observe the inner and outer walls, the divertor region, the lower hybrid current drive (LHCD) antenna and neutral beam injection (NBI) port in the visible range, was installed on the HL-3 tokamak for PWI investigation. New insights into the physical processes in the plasma edge are gained with this diagnostic by measuring the 3D emission profiles in the main chamber as well as in the divertor region. In the paper, hot spots during the PWI induced by the bulges on the divertor graphite tiles and exposed tile edges were experimentally observed, which causes the increase of impurity influxes. Bright objects (UFOs) moving through the plasma and occasionally inducing major disruption are recorded, together with the magnetic field lines development by their luminous trajectories. The online boron powder injection from the top of vacuum vessel was implemented in order to improve the wall conditions, where the interactions between the plasma edge and puffed boron powders are observed. That the boron gathering and glowing in plasma edge probably by the impurity screening mechanism and the almost simultaneous radiation (mainly the bremsstrahlung) enhancement in the plasma center because of the impurity pollution occurs. Actually, a low recycling level and high confinement phase was concurrently achieved by the online boron powder injection. The high spatial and temporal resolution imaging capability also leads to advances in the characterization of ELMs along with their impact on PFCs. During type-I ELMs, main interaction is on the lower divertor and upper baffle, indicating the enhanced particle recycling during the ELM pulse energy deposition. Zeff during one ELM event is increased, indicating the transiently increased impurity influx because of PWI. Nonetheless, Zeff is basically decreased during the H mode period, which demonstrates the impurities are expelled from the plasma by the type-I ELMs. The more the ELMs frequency is, the more impurity is expelled from the plasma.

Speaker: Liang Liu (Southwestern Institute of Physics)

321 3.078 Comparison of Erosion Behaviors with Carbon and Tungsten targets in Spherical Tokamak EHL-2 Using ITCD and EMC3-EIRENE codes

Proton-boron-11 (p-11B) fusion has increasingly gained attention as a promising alternative to conventional deuterium-tritium (D-T) fusion due to its reduced neutron yield and more favorable fuel availability [1]. Although numerous current devices have successfully conducted fusion experiments using p-11B fuels, several critical scientific and technical challenges need to be resolved to enhance its feasibility [2]. In contrast to D-T fuels, B ions, given their high charge state and large mass, can lead to severe erosion of divertor targets. This, in turn, has the potential to restrict device longevity and undermine operational stability. To address these challenges, the EHL-2 spherical tokamak is proposed as a valuable platform to validate the physics and engineering solutions related to p-11B fusion.
In this study, the Monte-Carlo impurity transport code ITCD [3, 4] and the edge transport code EMC3-EIRENE [5] are utilized to investigate the erosion behaviors of carbon (C) and tungsten (W) divertor targets in the EHL-2 device. Firstly, the EMC3-EIRENE code first generates the background plasma and particle fluxes of H and B near the divertor targets. Based on this information, the ITCD code is utilized to study the (re-)erosion/(re-)deposition dynamics of the C and W targets. Through the establishment of an integrated framework for impurity erosion and transport, this work makes significant contributions to the selection of wall materials for the EHL-2.
[1] Margarone D, et al 2022 Applied Sciences 12 1444
[2] Magee R M, et al 2023 Nat Commun 14 955
[3] Gao G D, et al 2023 Plasma Physics and Controlled Fusion 65 035015
[4] Liu Y L, et al 2025 Journal of Fusion Energy 44 4
[5] Feng Y, et al 2004 Contributions to Plasma Physics 44 57-69

Speaker: Mr Zihao Gao (Dalian University of Technology)

322 3.079 Impurity Transport from Plasma–Wall Interaction as a Trigger for Accelerated Mode Disruptions in Tokamaks

Plasma disruptions have been an inherent challenge in tokamak operation since their inception, as they lead to a rapid loss of confinement and the sudden deposition of large amounts of energy onto plasma-facing components, often resulting in severe material damage. Achieving long plasma confinement times is a primary objective of future fusion-grade tokamaks, however, disruptions undermine this goal by terminating confinement on very short time scales. In this work, we identify and analyze a previously unreported class of disruption, referred to as Accelerated Mode Disruption (AMD). Experimental observations indicate that AMD is closely associated with a rapid increase in core radiation. The disruption is triggered by enhanced transport of impurity species from the plasma edge into the core region, where their accumulation leads to a significant rise in radiative losses. This process results in the formation of a hollow electron temperature profile, destabilizing the plasma and ultimately causing the disruption.
A detailed investigation is presented on plasma-wall interactions, impurity generation at the edge, inward impurity transport, and their accumulation in the core. The role of impurity-induced core radiation in modifying the temperature profile and driving the onset of AMD is examined. These findings provide new insight into disruption mechanisms and highlight the critical role of impurity control, i.e the plasma wall interaction in maintaining stable plasma confinement in future tokamak devices.

Speaker: Mr Soumitra Banerjee (Institute for Plasma Research)

323 3.080 Rapid exhaust operational space studies employing AI workflows: the low-aspect ratio EU-DEMO as a use case

Reactor-scale divertor design demands rapid yet physics-consistent exploration of exhaust operational spaces. Fusion power plants (FPPs) beyond ITER require fully integrated scenario development that ensures both robust core-plasma performance and compliance with engineering limits on plasma-facing components (PFCs), particularly with respect to heat fluxes and wall erosion. Central to this challenge is accelerating the evaluation of key divertor quantities - target fluxes, detachment behaviour, neutral dynamics, and pumping efficiency - while retaining the predictive capability of SOLPS-ITER for upstream conditions, i.e., separatrix density, impurity concentration and plasma fuelling. We present an AI-enabled workflow designed to meet this need by combining SOLPS-ITER model databases with data-driven surrogate modelling.

The workflow employs the SOLPS-NN surrogate model trained on an extensive set of wide-grid SOLPS-ITER simulations incorporating the advanced fluid neutral (AFN) model. These surrogates reproduce the dominant plasma–neutral physics governing the transition from attached to detached conditions and provide orders-of-magnitude speedup relative to full SOLPS-ITER simulations including EIRENE neutral kinetics. A key enhancement is the use of transfer learning to increase surrogate fidelity: AFN-based SOLPS-NN models are retrained on a smaller set of high-fidelity SOLPS-ITER simulations with EIRENE, enabling the surrogate to capture kinetic-neutral effects - such as non-Maxwellian transport, molecular processes, and detailed ionisation–recombination balance - without requiring full kinetic simulations across the entire parameter space. In principle, active-learning methods can further refine the training set by automatically targeting regions where the physics becomes strongly nonlinear or where surrogate uncertainties are elevated.

Using the low-aspect-ratio EU-DEMO (DEMO-LAR) configuration as a test case, we demonstrate how the workflow efficiently maps the exhaust operational space and identifies regimes compatible with power-handling constraints. Comparisons with reduced exhaust models indicate where simplified approaches remain valid and where kinetic-neutral physics leads to significant deviations. Post-processing of SOLPS-NN 2D plasma and neutral profiles enables rapid generation of heat- and particle-flux maps on all PFCs (e.g., for wall-erosion assessments across many scenarios) and supports fast synthetic diagnostics suitable for digital-twin environments (DTE). Finally, preliminary integration of SOLPS-NN into the JINTRAC modelling framework is underway, enabling initial tests of computationally efficient core–edge coupling schemes with dynamic boundary-condition exchange, also suitable for future Pulse Design Tools (PDT).

Speaker: Sven Wiesen (DIFFER - Dutch Institute for Fundamental Energy Research, De Zaale 20, 5612 AJ Eindhoven, Netherlands)

324 3.081 Evaluation of Additively Manufactured Tungsten and Cold Spray Tantalum Coating Performance through DiMES Plasma Exposures at DIII-D

Fabrication of first wall components for Fusion Pilot Plants (FPPs) through conventional powder metallurgy routes may be unfeasible due to cost and durability concerns. Instead, technologies such as additive manufacturing could be used to produce near net shape components with specifically-tailored microstructures. Cold spray deposition is another strategy that can be used to produce thick (>100µm) refractory metal coatings on a variety of substrates, which could provide a viable path for the large-scale manufacture of resilient plasma-facing components (PFCs) [1].

To investigate the performance of these materials under fusion-relevant conditions, additively manufactured pure tungsten, cold spray pure tantalum coatings – approximately 200 μm thick on 316 stainless steel substrates – and conventionally manufactured pure W and Ta specimens were exposed to plasmas using the Divertor Material Evaluation System (DiMES) in the lower divertor at the DIII-D tokamak. Deuterium plasma exposures produced incident heat fluxes of 1.4-1.6 MW/m2 (inter-ELM) and 3.5-4 MW/m2 (ELM) during 6 shots of H-mode. Specimens were characterized with laser scanning confocal microscopy (LSCM), scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS) before and after exposure.

The surface roughness (Sa) of all specimens increased post plasma exposure, with cold spray tantalum coatings also exhibiting growth of existing, and formation of new, surface pits, approximately 30µm in size. SEM imaging of conventionally and additively manufactured tungsten specimens shows the presence of surface particles, with diameter <1µm, which do not appear to contain carbon, according to EDS analysis. However, all specimens did exhibit carbon deposition in fiducial markers and other sub-surface features. SEM imaging also reveals the formation of surface cracking in the additively manufactured tungsten specimens, which will be investigated further through electron backscattered diffraction (EBSD). XRD reports no noticeable changes between pre- and post-exposure for all specimens. XPS shows the presence of surface oxides on all samples, which are observed to change in stoichiometry post exposure, potentially due to contamination, sputtering and/or D implantation [1].

These findings represent a key step toward validating additively manufactured and cold-sprayed coatings as viable solutions for next-generation plasma-facing components in fusion reactors.

This work was supported by US DOE under DE-NA0003525, DE-AC02-09CH11466, DE-FC02-04ER54698, DE-AC05-00OR22725, DE-AC52-07NA27344 and DE-SC00020428.

[1] Ialovega et al. ‘Initial study on thermal stability of cold spray tantalum coating irradiated with deuterium for fusion applications‘ Phys. Scr. 98 (2023) 115611

Speaker: Charles Hirst (University of Wisconsin-Madison)

325 3.082 Experimental investigation on the flux-dependence of He clustering in polycrystalline and single-crystalline W

Ab-initio modeling of helium (He) behavior in tungsten (W) predicts clustering mechanisms, self-trapping and trap mutation, but to date there is little experimental evidence for isolated He self-trapping in a defect-free lattice. In order to investigate these mechanisms and the influence of intrinsic defects in W on them, the effects of He implantation into polycrystalline (poly-W) and single-crystalline (sc-W) W samples were studied in a set of exposures, where several parameters were systematically varied: He energies ranged between 50 and 200 eV, the flux was varied between $10^{17}$ and $10^{19}\,\, \mathrm{He/m^2s}$ and the fluence ranged from $10^{19}$ to $10^{23}\,\, \mathrm{He/m^2}$. Before exposure, the samples were polished carefully and probed for open-volume defects by positron annihilation spectroscopy, and then exposed to low-temperature He plasma at 292 K surface temperature. The He energies 100 eV and 50 eV were chosen for most He exposures including sc-W samples to avoid kinetic defect introduction. Elastic recoil detection analysis (ERDA) was used to measure absolute He areal densities after exposure. To derive high-resolution He concentration depth profiles, ERDA measurements were performed successively in combination with a stripping method using anodic oxidation.
The sc-W data show a flux-dependent threshold behavior: While there is no measurable He signal at lower fluxes, significant He retention is observed at a flux of $1\times10^{19}\,\, \mathrm{He/m^2s}$. This flux-dependence indicates that self-trapping dominates the He retention. Exposed with the same parameters, the poly-W retention data exhibit the threshold-behavior between the fluxes $1\times 10^{17}$ and $1\times10^{18}\,\, \mathrm{He/m^2s}$. A comparison of He retention after low-flux ($10^{17}\,\, \mathrm{He/m^2s}$) exposure of electron-beam irradiated sc-W with undamaged sc-W and poly-W eliminates pre-existing open-volume defects as the cause for the differences in He interaction between the undamaged materials. Impurities in the material emerge as the most plausible promoters for He clustering mechanisms and He trapping in poly-W. The significant influence of impurity defects in He clustering and trapping is not only supported by the differences in total He retention between poly-W and sc-W samples at medium-to-low fluxes but also by a difference in He trapping beyond the implantation depth found in He depth profiles of poly-W and sc-W.

Speaker: Annemarie Kärcher (Max Planck Institute for Plasma Physics, 85748 Garching, Germany (IPP), Technical University of Munich, Germany (TUM))

326 3.083 Reactor-Scale Island Divertor Simulations of W7-X-like Configurations

The island divertor concept in Wendelstein 7-X (W7-X) is the leading candidate for future stellarator reactors for handling heat and particle exhaust [1]. However, to date the study of this concept has primarily focused on experimental scale and shown that achieving detachment and high-recycling is more complex than standard tokamak poloidal divertors due to the complex 3D magnetic geometry. Only a few results at reactor-scale exist and these indicate a significant separation of the heat and particle flux channels [2], not observed in W7-X. Further study of heat and particle transport at reactor-scale is necessary to elucidate if this is problematic/beneficial for efficient heat/particle exhaust, respectively. To bridge the gap between experiment and reactor-scale performance, we simulate 3 different scales (1x, 2x, 4x) of W7-X using EMC3-Eirene [3]. The largest scale (4x) roughly corresponds to reactor-scale. For the W7-X standard magnetic configuration, we examine how increasing device size alters the divertor performance of the island divertor. We analyze how device size impacts location shifts of the ionization source and the downstream density and temperature trends in the island divertor for achieving detachment. For 2x scale, detachment is observed with pressure weighted averaged downstream temperatures $\leq10\text{eV}$ at separatrix densities $>3\times10^{-19}\text{m}^{-3}$. Additionally, we simulate Neon impurities to assess if reactor-scale W7-X leads to Ne concentrations compatible with reactor operation [4]. This analysis helps determine if density build-up and impurity retention for W7-X-like geometry is sufficient or needs improvement for a reactor. Finally, the heat and particle flux deposition on the scaled up island divertor configurations are calculated. Shifts in the heat flux distribution are seen with increasing separatrix density. To characterize the scrape-off layer (SOL) power exhaust and heat transport, we apply the approach developed in [5] for estimating the different SOL transport channel widths and examine how these vary with device size. Estimation of SOL transport channel widths has been instrumental for tokamak divertor configurations to connect the main heat transport mechanisms with the resulting divertor heat flux distribution, and is an area of ongoing work for stellarators [5]. This overall reactor-scale analysis will help guide island divertor improvements for future stellarator reactors.
[1] A. Bader et al, J. Plasma Phys. (2025)
[2] Y. Feng, J. Nucl. Mat. 438 (2013) S497-S500
[3] Y. Feng et al, Contrib. Plasma Phys. 44 (2004) 57-69
[4] E. Sytova et al, Nucl. Mater. Energy 19 (2019) 72-78
[5] A. Kharwandikar PhD Dissertation (2025)

Speaker: Kelly Garcia (Universität Greifswald)

327 3.084 Numerical study on surface heat load and erosion for limiter design in JA DEMO

A limiter is a plasma-facing component in a fusion reactor that protects blanket modules from excessive surface heat load and high-energy particles interaction. It achieves this by protruding from the first wall (FW), effectively shadowing the blanket. The limiter extends continuously in the poloidal direction, except near the divertor, and is installed in each 90-degree toroidal segment. To withstand high neutron irradiation, its plasma-facing surface uses tungsten monoblocks (W-MBs) with reduced activation ferritic/martensitic (RAFM) steel cooling pipes. However, RAFM’s low thermal conductivity limits the heat removal to $4.6\ MW/m^2$, significantly lower than ITER divertor T-MBs with copper alloy pipes.
FW surface heat load is a critical design factor. Excess localized heat must be avoided, making heat load evaluation and surface shape optimization essential. The incident angle of magnetic field lines intersect the FW strongly influences charged particle heat flux. A simple constant-curvature surface can cause localized high heat flux. To address this, a simulation code was developed using a decay length approach to evaluate heat flux along magnetic field lines, incorporating connection length for particle shadowing. Based on plasma operation scenario, simulations indicate that a 30 mm limiter protrusion effectively shadows the blanket. The limiter surface shape was optimized theoretically to achieve a uniform and wide heat load distribution. The peak heat flux on the limiter during flat-top phase is calculated to be $2.4 MW/m^2$, fulfilling the cooling capability.
Further design development included coolant flow path and remote maintenance strategies. Component lifetime is another key factor, requiring erosion analysis of the plasma-facing surface. We picked up a set of reference parameters: electron temperature 5 eV, argon density $ 3 \times 10^{15} /m^3 $ by extrapolating a divertor transport simulation result of JA DEMO. The estimated eroded depth per year was not significant; however, it can easily change due to the high sensitivity of sputtering yield to the electron temperature and the impurity transport. A sensitivity study on heat load and erosion by varying plasma parameters will be presented in detail. This work highlights the correlation between limiter shape, surface heat load and erosion characteristics.

Speaker: weixi chen (National Institutes for Quantum Science and Technology)

328 3.085 Detachment via Impurity Seeding, Heat Exhaust and Erosion in DIII-D in 3D Fields

EMC3-EIRENE modeling of detachment of 3D scenarios at DIII-D with resonant magnetic perturbations applied show that neon as a seeded impurity is able to semi-detach the ITER similar shape at comparatively lower separatrix impurity densities than seeded nitrogen. This impurity buildup is observed on the low-field side at the DIII-D shelf, leading to detachment of the far scrape-off layer (SOL), and remaining semi-attached at the strike line near the lower pumping gap on the floor. In the near SOL, there is neutral pressure buildup occurring for main ion species (D) densities with deep detachment observed for higher separatrix densities (5x1019 m-3), though not within the observed ELM suppression window. The simulations here are in support of an experimental campaign in which extrinsic impurities were used to study the impact on the SOL and exhaust of such species in ELM suppressed scenarios at DIII-D. In order to efficiently pump out these impurities, it is necessary to identify global transport dynamics for them as well as identifying safe operational limits for the wall components. The data collected during this campaign is used to denote differences in particle transport shown in these simulations and those collected with coherent imaging spectroscopy, charge exchange recombination, neutral pressure measurements and helium beam measurements. In addition, a coupled analysis to the ERO2.0 shows the effects of including helium in the simulations on plasma wall interactions, with He2+ being the main contributor to erosion (from other He charge states) on the divertor targets at DIII-D. This has been carried out with a new capability from EMC3-EIRENE, where charge state resolved fluxes can be directly mapped onto target surfaces to better account for the spatial distribution of the species in the 3D equilibria. A similar analysis is currently under investigation for other seeded impurities (i.e. neon, nitrogen), as the detached regimes achieved in the edge can significantly reduce the physical sputtering yield of carbon at the targets.

This work was funded by Department of Energy, Office of Fusion Energy Science, DE-SC0020284, DE-FC02-04ER54698, DE-AC52-07NA27344 and DE-AC02-09CH11466.

Speaker: Marcos Navarro (University of Wisconsin-Madison)

329 3.086 Performance Analysis of Additively Manufactured Tungsten through L-mode and H-mode DiMES Plasma Exposures at DIII-D

Fabricating fusion pilot plant (FPP) first-wall components through additive manufacturing (AM) could enable cost-effective, scalable production of complex, functionally graded components. In addition, optimized AM processes can produce materials that retain, and in some cases improve, the key thermo-mechanical properties needed in an FPP, relative to conventionally processed tungsten.

To evaluate whether additively manufactured tungsten (AM-W) can match or exceed bulk-W performance, AM-W, produced by laser powder bed fusion (LPBF; ~95% theoretical density, increased porosity relative to bulk W), was exposed to L- and H-mode plasmas using the Divertor Materials Evaluation System (DiMES) in the lower divertor of the DIII-D tokamak. The specimens were characterized by scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), laser confocal microscopy, X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS) before and after exposure. Plasma conditions consisted of steady-state inter-ELM heat fluxes of 1.1-1.3 MW/m2 in L-mode and 1.4-1.6 MW/m2 in H-mode, with ELMs reaching 3.5-4.0 MW/m2. Results were benchmarked against conventionally manufactured pure tungsten reference samples irradiated under identical conditions.

Initial laser confocal microscopy surface analysis using the arithmetic mean height (Sa) showed AM-W experienced a greater increase in surface roughness after L-mode exposure than bulk-W, which exhibited minimal change. AM-W and bulk-W saw similar increases in surface roughness post H-mode plasma exposure, exceeding those from L-mode. XRD analysis shows no observable changes in both bulk-W and AM-W pre- and post-exposure, in L-mode and H-mode. XPS analysis indicates metallic W and stable WO3 pre-exposure, and similar trends post-exposure, with changes in oxide composition due to sputtering and contamination, observed in both L- and H-mode. EDS shows carbon deposition on both bulk-W and AM-W specimen features, expected from DIII-D operations. SEM indicates no surface crack evolution in AM-W samples post-L-mode exposure, some surface crack evolution in AM-W post-H-mode exposure. Surface analysis showed general robustness of the additively manufactured tungsten across experimental conditions at DIII-D.

Deuterium retention in the AM-W will be further determined through thermal desorption spectroscopy (TDS), using a custom-built facility being developed at UW-Madison. These findings represent a key step towards validating AM-W as a potential next-generation material for FPP first-wall applications.

This work was supported by US DOE under DE-NA0003525, DE-AC02-09CH11466, DE-FC02-04ER54698, DE-AC05-00OR22725, DE-AC52-07NA27344 and DE-SC00020428.

Speaker: Dylan Kohler (University of Wisconsin-Madison)

330 3.087 A Systematic Study on Detonation-Sprayed Tungsten/Steel First Wall Structures: From Process Development to Performance Evaluation

Following the decision by ITER in its new baseline to replace beryllium with tungsten in the first wall, the fabrication of large-area curved plasma-facing tungsten components has attracted growing attention. Low-cost, high-efficiency manufacturing of tungsten (W) first wall structures for breeding blankets holds significant importance for future large-scale tokamak fusion experimental reactors. Among various coating processes, thermal-spray-based tungsten coating techniques have regained research interest after years of limited activity. Over the past five years, our laboratory has been dedicated to studying and improving the detonation spraying process to produce tungsten coatings suitable for ITER first wall applications.
This report systematically introduces a coating-based tungsten/steel first wall structure developed in our laboratory, along with a series of studies on its high-heat-flux performance and post-treatment modifications. The process begins with depositing a pure tungsten coating onto a steel substrate, which exhibits excellent comprehensive properties including high density, purity, bonding strength, hardness, and thermal shock resistance. Subsequently, a high-purity, high-density iron (Fe) interlayer is deposited between the tungsten coating and the steel substrate to further enhance performance under high heat flux (HHF) conditions. The key thermomechanical properties of both the tungsten and iron coating materials have been thoroughly characterized.
Based on this characterization, finite element analysis was used to optimize the thicknesses of the tungsten and iron coatings in the W/Fe/steel first wall structure, ensuring optimal performance under steady-state HHF conditions. Following optimization, a first wall module with the refined thickness design was fabricated and successfully tested under 1,000 cycles of 1 MW/m² HHF loading as well as under prolonged steady-state HHF conditions.
Furthermore, systematic research was conducted on heat treatment to improve the performance of the tungsten coatings. For example, annealing was found to enhance the crack initiation resistance of the as-sprayed tungsten coating but reduce its crack propagation resistance, whereas recrystallization treatment resulted in a stable equiaxed grain structure, comprehensively improving crack resistance.
Together, these advances demonstrate that the detonation spraying method is a viable and efficient approach for manufacturing and in-situ repairing tungsten/steel first wall structures for breeding blankets, offering notable advantages in both effectiveness and performance.

Speaker: Prof. Zongxiao Guo (GNOI)

331 3.088 Application of the kinetic ion transport model EMC3-EIRENE- KIT for W transport in full-3D wide-grid H-mode scenarios for the ITER SRO and DT-1 phases

With the transition to a full-tungsten (W) machine, detailed understanding of the W erosion source and transport processes are key to assessments of the main plasma performance. In work conducted under the auspices of the ITER Scientist Fellow Network, the capability of the recently upgraded kinetic ion transport model (KIT) of EMC3-EIRENE [1] for W transport with the full shaped 3D wall geometry (including port structures) is demonstrated. For an EMC3-EIRENE plasma background of a neon-seeded ITER H-mode scenario from the DT-1 campaign (c.f. M. Jia et al, this conference) an artificial W source is used as a proxy for physical sputtering in the KIT model to assess the core penetration of W.

A second application addresses the Start of Research Operation (SRO) phase of ITER with a focus on W sputtering and kinetic transport in a pure deuterium plasma background with Resonant Magnetic Perturbations (RMP) for Edge Localized Mode suppression (c.f. J. Van Blarcum et al, this conference). Here, particular attention is given to the interaction between the non-axisymmetric scrape-off-layer (SOL) lobe structures and the inertially cooled Temporary First Wall (TFW) of ITER. Employing EMC3-EIRENE-KIT, RMP-induced W sputtering and subsequent kinetic transport to assess the resulting W penetration into the confined core region is presented.

Discrepancies seen in a previous benchmark [2] of EMC3-EIRENE-KIT against the ERO2.0 code [3] were traced back to inaccuracies in the magnetic-field input provided to ERO2.0. After correcting the magnetic input data, the two codes show nearly perfect agreement under comparable model conditions. New benchmark results are presented and provide a validation basis for applying EMC3-EIRENE-KIT to reactor-relevant impurity transport. A fully self-consistent coupling of impurity fluxes between EMC3 and EIRENE is under development.

Acknowledgement
The authors gratefully acknowledge computing time granted through JARA on the supercomputer JURECA [4] at Forschungszentrum Jülich.

References
[1] D. Harting et al., Nucl. Mater. Energy 33 (2022) 101279. https://doi.org/10.1016/j.nme.2022.101279
[2] D. Harting et al., Nucl. Mater. Energy 42 (2025) 101887. https://doi.org/10.1016/j.nme.2025.101887
[3] S. Rode et al., Contrib. Plasma Phys. e202100172 (2022). https://doi.org/10.1002/ctpp.202100172
[4] Jülich Supercomputing Centre. (2021). JURECA: Data Centric and Booster Modules implementing the Modular Supercomputing Architecture at Jülich Supercomputing Centre Journal of large-scale research facilities, 7, A182. http://dx.doi.org/10.17815/jlsrf-7-182

Speaker: Derek Harting (FZJ - Forschungszentrum Jülich GmbH)

332 3.090 The impact of concurrent displacement damage on retention for high-flux D2-10%He plasma-exposed tungsten

Previous work utilizing deuterium (D) found that while displacement damage increases D retention in tungsten (W) due to generated D trapping sites [1-3], helium (He) pre-treatment or seeding reduces D retention in pristine W for low-energy D since He nanobubbles act as a diffusion barrier to D [2,4]. He pre-treatment or seeding also reduces D retention in pre-damaged W such that under certain conditions when both effects are considered, retention is the same as for pristine W exposed to pure D [2,3]. This work investigates how He bubbles and D retention are affected by displacement damage that occurs simultaneously to plasma exposure.

Bulk W samples were exposed to (i) pure D plasma, (ii) mixed D-10%He plasma, or (iii) D-10%He plasma during simultaneous W$^{6+}$ damaging. For all, D ion flux was kept at 4.2$\times$10$^{22}$ m$^{-2}$, D/He ion energy at 75 eV, damage rate at 6.3$\times$10$^{-5}$ dpa/s, W$^{6+}$ energy at 20 MeV, and duration at ~3600 s, while the sample temperature $T_{\mathrm{s}}$ was varied from 550-750 K. An additional sample was exposed to 1 MeV instead of 20 MeV W$^{6+}$ to reduce the peak damage depth from 1.35 μm to 35 nm, and hence to focus the damage to the near-surface He bubble layer while minimizing the damage in the underlying bulk. Post-exposure characterization included scanning electron microscopy, transmission electron microscopy, nuclear reaction analysis, and will include thermal desorption spectrometry.

Up to 300 nm diameter blisters were observed at $T_{\mathrm{s}}$ = 550 and 650 K for both pure D and D-He samples (in contrast to [4]), while simultaneous damage suppressed blister formation (in agreement with [5] for pre-damaged W). At $T_{\mathrm{s}}$ = 650 K, D retention was increased by two orders of magnitude with 20 MeV damage versus without, while retention peaked at shallower sample depths for 1 MeV damage. In view of all, results point to simultaneous displacement damage as having a measurable impact on He-induced diffusion barriers and damage-induced trapping, which in turn, has important implications for predicting hydrogen-isotope retention in tungsten under fusion-relevant conditions.

[1] W.R. Wampler and R.P. Doerner, Nucl. Fusion 49 (2009) 115023
[2] V.Kh. Alimov et al., J. Nucl. Mater. 420 (2012) 370-373
[3] Q. Bai et al., Nucl. Fusion 59 (2019) 066040
[4] M. Miyamoto et al., Nucl. Fusion 49 (2009) 065035
[5] S. Wang et al., J. Nucl. Mater. 508 (2018) 395-402

Work is supported by US-DOE-FES (DE-SC0022528) and NIFS Collaboration Research program (NIFS25KIET012).

Speaker: Marlene Patino (University of California, San Diego)

333 3.093 Integrated modeling of open-surface liquid metal divertors

Liquid metal (LM) divertor conditions for a pilot-plant class tokamak are simulated using a multi-component integrated model which considers plasma transport, LM magnetohydrodynamics and heat transfer, and formation and transport of lithium released from the divertor surface. Fast-flow LM divertors have several intrinsic advantages over solid surface concepts due to their self-replenishing nature and low atomic charge, avoiding long-term erosion and unacceptable core radiation issues that are faced by tungsten divertors. On the other hand, lithium can still lead to unacceptable core ion dilution and particles emitted via evaporation and sputtering should ideally be confined to the near-SOL to avoid accumulation on first wall components. Limiting these effects leads to maximum liquid lithium and plasma temperatures to keep evaporation and sputtering to tolerable levels. To better understand the lithium sourcing, transport, and impact on the plasma, integrated models are required.
Building on previous open-surface LM divertor modeling using SOLPS-ITER [1,2], a new workflow includes the OpenEdge code [3] for a self-consistent description of sputtering, evaporation, and droplet transport applied. The integrated model is applied to the CAT tokamak [4], modified to have flat target plates consistent with a fast-flow liquid lithium divertor concept. OpenEdge is a multi-purpose Monte Carlo code which traces neutral and charged particles against a prescribed plasma background considering Lorentz forces, plasma collisions, and for this problem mass and size evolution due to evaporation. Several characteristic lithium macroparticle mechanisms are considered, including mist (~um scale), droplets (~mm scale), and ligaments (~cm scale), supported by two-phase flow simulations. Depending on the size and velocity of the macroparticles, time-dependent simulations may be used to describe the effect of the emission on the background plasma state. Initial results of the coupled workflow and an assessment of compatibility the LM divertor concept for CAT will be presented.
[1] M.S. Islam et al 2024 Nucl. Fusion 64 056036
[2] S. Smolentsev S. et al 2022 IEEE Trans. Plasma Sci. 50 4193
[3] A. Diaw et al OpenEdge: A collaborative, multi-purpose direct simulation Monte Carlo for plasma simulation in magnetic fusion environments, submitted to Computer Physics Communications
[4] R.J. Buttery et al 2021 Nucl. Fusion 61 046028

Speaker: Jeremy Lore (Oak Ridge National Laboratory)

334 3.094 Optimization of Additive Manufactured Tungsten Lattice Structures for Liquid Metal Infiltration in Plasma-Facing Components

Keywords: Magnetic confinement Fusion, liquid metal, capillary structures, lattice, tungsten, additive manufacturing

One of the main challenges in the development of a magnetic confinement fusion reactor is the durability and maintenance of internal components, particularly the divertor, which is exposed to the highest heat and particle fluxes. To withstand these extreme conditions, plasma facing components (PFCs) in ITER rely on actively cooled tungsten monoblocks. However, these monoblocks show several durability limitations notably, including localized melting, cracking and erosion caused by sputtering. These promote the exploration of alternative solutions that could improve maintainability & availability of future fusion power plants.

Liquid metals (LMs) represent a particularly promising solution for PFCs, as they help mitigate several limitations of solid materials. Their continuous renewal provides self-healing capabilities, while local evaporation contributes to heat removal. Moreover, because atoms in a liquid are already mobile and disordered, LMs are largely unaffected by displacement-damage mechanisms that degrade solid materials. Convective motion of the LMs also promotes efficient and uniform heat dissipation. Among the concepts developed so far, Capillary Porous Systems (CPS) infiltrated with low-melting-point metals such as lithium and tin have attracted significant attention. Capillary forces stabilize the liquid metal within the porous structure, reducing plasma contamination. Nevertheless, precise control of pore geometry and manufacturing constraints remain major challenges.

In this work, we investigate the fabrication of tungsten lattice structures via Laser Powder Bed Fusion (LPBF) to achieve an architectured porosity suitable for tin infiltration and retention. The first phase focuses on optimizing process parameters through the fabrication of bulk cubes. Performance comparison was based on measurement of relative density and cracks observation. This step is essential to ensure the manufacturability of future porous structures, which are particularly sensitive to sintering defects. The resulting samples were characterized in terms of density and microstructure. In a second step, the wettability of tin on tungsten was assessed. These results provide a solid foundation for the fabrication of fully architectured tungsten lattice structures able of being infiltrated by liquid metal and functioning as CPS.

Speaker: Matthieu Spinosi (Institut Jean Lamour IJL UMR CNRS 7198, Université de Lorraine, 54011 Nancy, France)

335 3.095 SONIC simulation studies of neutral-neutral collision and extended thermal force, and effects on JA DEMO detached divertor and impurity control

SONIC divertor code has simulated the divertor performance of power exhaust by seeding impurity (Ar) and He exhaust in the detached divertor for JA DEMO design (1.5 GW-level fusion power), where exhaust power, fuel and He particles at the core-edge boundary of 250 MW, 1x$10^{22}$ D/s and 5.3x$10^{20}$ He/s, respectively, were given [1]. Recently, NEUT2D and IMPMC codes (kinetic MC modellings for neutral and impurity transport, respectively) incorporated (i) neutral-neutral elastic collision (NNC) database [2], where collision rates and momentum exchange rates are evaluated from differential cross-section database for $D_{0}$-$D_{0}$, $D_{0}$-$D_{2}$, and $D_{2}$-$D_{2}$, and it is rather theoretical expression compared to those in EIRENE, and (ii) an extended kinetic thermal force (TF) model for the impurity transport, including a collisionality parameter and heat flux limiter in the conventional formula as a collisionality-dependent term, which expects reduction in TF in the high $T_{i,e}$ SOL and edge [3]. Effects on the divertor detachment and impurity concentrations in SOL and edge were investigated in the reference series of the JA DEMO design with similar total radiation fraction of $f_{rad}$ = $P_{rad}/P_{sep}$ ~0.8.
NNC affected distributions of neutral and molecular pressures ($P_{D0}$, $P_{D2}$) in the private and sub-divertor regions. Outer peak heat load at the divertor target ($q_{target}^{out}$, produced in attach plasma region) was reduced from 6.5 to ~4 $MW/m^{2}$ with increasing $P_{D0}$+$P_{D2}$ from 2.1 to 3.5 Pa and reducing the local $T_{e,i}^{div}$ from 35 to 10 eV. These peak-$q_{target}^{out}$ values were systematically smaller than those without NNC model since $P_{D0}$+$P_{D2}$ became larger. At the same time, comparing to result without NNC model, $D_{0}$ penetration in the sub-divertor was reduced, and $P_{D2}$ was increased by the factor of two. Extended TF-model also affected detachment in the outer divertor, which was extended to the upstream and radial directions. As a result, Ar concentration in the SOL ($c_{Ar}^{sol}$) was increased and $c_{Ar}$ inside the separatrix ($c_{Ar}^{edge}$) was increased from 0.6-0.8% to ~1%. The latter value was larger than the design reference (0.6%). $c_{He}^{sol}$ was also increased, but $c_{He}^{edge}$ was similar than the design reference (0.7%). Operation window of $f_{rad}$, $D$-gas puff rate will be investigated to satisfy requirements of the JA DEMO divertor design such as $q_{target}^{out}$, $c_{Ar}^{edge}$ and $c_{He}^{edge}$ values.
[1] N. Asakura, et al., Nucl. Mater. Energy 26 (2021), 100864.
[2] S. Tokunaga, et al., PSI22, P.3.105, May 2016, Rome, Italy.
[3] Y. Homma, et al. Nucl. Fusion 60 (2020) 046031.

Speaker: Dr Nobuyuki Asakura (National Institutes for Quantum Science and Technology (QST))

336 3.096 Simulation of disruptions and small-ELMs regimes by a CW laser over different advanced tunsgten materials at the OLMAT high heat flux facility

Tungsten is the preferred candidate as plasma-facing material for future power plants like DEMO. However, current W materials produced by industry like ITER-Grade Tungsten (IGW) will most likely not meet their requirements. New advanced armor concepts are being studied to withstand those steady state and transients heat loads [1]. One of the methods to simulate the heats loads expected in a device as DEMO is by a high-energy fiber laser as in OLMAT High Heat Flux (HHF) facility, described in [2]. Laser methods are very well suited as a fast and cheap screening test, and the only possibility to simulate certain heat loads.
In this work a small overview of the optimization of the high-energy fiber laser system will be first given. The installation has not been straightforward due to its high energy: laser beam shape characterization to find a flat-top profile, special viewports, beam dump for reflections, laser absorption quantification, etc.
The second, and main part, of this work is to compare different types of advanced W materials —2 types of Wf/W, WCrYZr self-passivating alloy, W-W2C— against IGW from 4 suppliers/batches as a reference material. They have been subjected to heats loads simulating disruptions (150-1000 MW/m2) at the edges of the armor (critical part), and small-ELMs regimes (50-300 MW/m2, 50-2000 Hz, up to 107 pulses) at the surface. Most materials were not damaged if the small-ELMs regimes were under 200 Hz and 106 pulses (being recent Wf/W the best suited), but were damaged with disruptions of >450 MW/m2 at the edges, but not if impacted at the surface (being IGW the more resilient against thermal shocks). However, the cracks at the edges were nucleated/originated by small cracks produced during fabrication at all non-polished sides (only the surface was). This indicates that in a future reactor there must be found a compromise in at least a rough polishing to eliminate those cracks and the high cost of it. Finally, future upgrades to overcome some of the problems found will also be presented.
[1] F.Maviglia et al. Nucl. Mat. Ener. 26, 100897 (2021)
[2] D.Alegre et al. J. Fus. Energy. 39, 411 (2020)

Speaker: Daniel Alegre Castro (Laboratorio Nacional de Fusion. CIEMAT. Madrid)

337 3.097 Testing of VPS and PVD Tungsten Coatings on Steel in WEST for the ITER Temporary First Wall

Since the introduction of the new ITER baseline [1], significant efforts were made in selecting the Tungsten (W) based armour materials for the inertially cooled Temporary First Wall (TFW). The TFW provides an environment for ITER’s initial operation campaign “Start of Research Operations (SRO)” that is more forgiving to disruptions while mimicking the final actively cooled First Wall in terms of materials and geometry.
The TFW design considers a mixture of potential plasma-facing material types, including bulk W, W heavy alloy, and W coating on a stainless-steel substrate. While bulk W will be used in heavily loaded areas during thermal transients, alloys and coatings have been proposed for inboard, including start-up area and outboard TFW panels.
The W-coated tiles need to withstand steady heat fluxes of up to 0.4 MW.m-2. The operating temperatures are 100-300°C between discharges, 500-600°C for nominal conditions during discharges and up to 800°C as design temperature limit on locally loaded area [2].
In view of the testing of W coating technologies namely Vacuum Plasma Spray (VPS) and Physical Vapor Deposition (PVD) in tokamak conditions, three coated tiles were installed on two inner guard limiters in WEST and tested under successive limiter plasma discharges. These tiles were instrumented with thermocouples and the surface temperature is assessed with infrared thermography. In total, 24 limiter discharges of ~11-12 s duration each were performed on the ITER TFW tiles in WEST with a plasma current between 500 and 700 kA.
A model constrained by thermocouple measurements provides surface temperatures up to ~300°C during experiments while infrared thermography measurements are between ~400 and ~700°C. Visual inspections performed before and after the WEST experiments through a window reveal no evolution of the coating. The present study describes the outcome of the experiments with more refined data and the latest results from Langmuir probes and visible spectroscopy on the gross erosion of the different W coatings.

[1] R.A Pitts; A. Loarte; T. Wauters et al., “Plasma-wall interaction impact of the ITER re-baseline ☆,” vol. 42, no. December 2024, 2025, doi: 10.1016/j.nme.2024.101854.
[2] T. Hirai, Poster presentation at PFMC 2025 conference

Speaker: Marianne Richou (CEA)

338 3.098 Investigating Divertor and PFC Safety Limits for Planned NSTX-U Experiments Using UEDGE

Managing extreme heat and particle fluxes to divertor targets remains a major challenge in spherical tokamaks. In this work, we use UEDGE simulations to investigate three interconnected topics relevant to NSTX-U operation: (1) graphite plasma-facing components (PFCs) performance, (2) lithium PFCs vapor shielding and its dependence on upstream plasma conditions, and (3) the impact of snowflake (SF)-type magnetic configurations on divertor heat handling.

The 2D heat-transport model (Wall-Li) is coupled self-consistently with UEDGE to monitor target surface-temperature evolution over discharge times up to 5 s, representative of NSTX-U pulses. Graphite target simulations show a monotonic rise in surface temperature with increasing deposited heat flux; for a wide range of input-power conditions, the surface temperature exceeds the graphite sublimation threshold (≈1400 °C) once peak heat fluxes surpass ≈7 MW/m². This occurs despite enhanced impurity radiation (≈2 MW radiated for a 10 MW input power), indicating that graphite radiation alone is insufficient to accommodate projected NSTX-U heat loads.

This work also explores lithium vapor shielding using a self-consistent coupling between UEDGE and Wall-Li, which models lithium sourcing based on local plasma and surface conditions. A broad parameter scan over input power and core density shows that lithium can form a dense vapor layer near the divertor target that provides strong passive thermal regulation: the surface temperature remains below ≈700 °C even as core power is increased. However, core lithium accumulation increases with surface temperature, with core lithium concentration rising sharply once the surface temperature exceeds ≈600 °C. Increasing core density enhances main ion–impurity friction, which suppresses upstream lithium transport, confines lithium closer to the target, and thereby extends the temperature window over which effective shielding occurs.

The modeled equilibrium for the SF configuration is analyzed using an “umbrella-type” transport model, which approximates the enhanced cross-field transport associated with the churning mode (CM) of plasma convection near the X-points. Power-scan simulations (4–10 MW) indicate that CM-driven transport substantially broadens the divertor heat-flux profiles and increases power deposition to the outer wall, while significantly reducing peak heat loads compared to a uniform radial-transport model applied to the same SF geometry. Analytical variation of the field-line incidence angle up to 90° further shows that the peak divertor heat flux remains below 7 MW/m² even at the highest power-loading conditions.

*This work was performed under the auspices of the U.S. Department of Energy (DOE) Office of Fusion Energy Sciences under contract No. DE-AC52-07NA27344.

Speaker: Dr Vlad Soukhanovskii (Lawrence Livermore National Laboratory)

339 3.099 Influence of inter-pulse delay in double-pulse laser-induced breakdown spectroscopy on the detection of implanted deuterium in tungsten

Laser-induced Breakdown Spectroscopy (LIBS) is a method for the analysis of material composition and is used for Plasma-Wall Interaction studies. Especially for the detection of implanted fuel, deuterium (D) and tritium (T), in plasma-facing components (PFCs) in a fusion reactor, its in-situ capabilities are promising. In the case of tungsten (W) PFCs, the expected low levels of implanted fuel will pose challenges to the technique in early operation phases, when neutron damage and associated increased retention is low.
Detection relies on the line emission of excited atoms in laser-induced plasma (LIP) plumes after ablation. Under vacuum conditions, a major contribution to the limit of detection (LOD) is essentially the total energy present in the LIP. A vacuum is required e.g. to study in operando in the vessel. Therefore, schemes need to be applied to enhance the energy content in the LIP. A laser pulse energy increase of the LIBS setup increases line emission as a larger volume is ablated, but it also reduces the depth-resolution of the LIBS measurement, thereby shifting its balance.
Another way to increase the total energy in the LIP is to absorb laser light from a second laser pulse in the established LIP [1]. In this double-pulse configuration, a second laser pulse is focussed into the LIP with a given delay relative to the first laser pulse, which initiates the LIP. Differences in LIP species mass cause differences in expansion speed in vacuum [2]. This difference leads to a different optimum Inter-Pulse Delay (IPD) for enhancement of hydrogen or tungsten line emission.
The IPD is set by an adjustable optical delay stage that delays the second laser pulse by varying the optical beam path length (20-160ns) as both pulses are generated simultaneously by a Nd:YAG laser (τ=35ps; 1st pulse: 355nm, 21mJ ;2nd pulse: 1064nm, 5.3mJ). The first laser pulse impacts normal to the sample surface (self-damaged tungsten, plasma loaded with 1 at.% D [2]) in a vacuum chamber (p=2*10-7mbar). The second laser pulse has an incident angle of 5°. A custom-made Littrow spectrometer measures Balmer-α line emission of the LIP and is thus used to detect hydrogen isotopes. A signal enhancement factor of up to 1.95 was measured at an IPD of 92ns. This optimum hydrogen IPD is different from the optimum of tungsten line emission.
This work is part of the project SyrVBreTT, funded by the BMFTR under grant no. 13F1011G
[1] https://doi.org/10.1016/j.sab.2006.09.003
[2] https://doi.org/10.1063/5.0211493

Speaker: Erik Wüst (Forschungszentrum Jülich GmbH, Institute of Fusion Energy and Nuclear Waste Management – Plasma Physics)

340 3.100 Boron-Enhanced W–Ta–V–Cr–Ti High-Entropy Alloys as Advanced Plasma-Facing Materials for Fusion Reactors

To improve the performance of plasma-facing components, various advanced concepts of W materials are being developed. In particular, tungsten-based refractory high-entropy alloys (HEAs) in the W–Ta–V–Cr–Ti system are attracting attention due to their excellent radiation resistance and strength at high temperatures. In this work, the intentional addition of boron to the W–Ta–V–Cr–Ti high-entropy alloy system is explored. Boron can significantly enhance plasma-facing performance by reducing sputtering, suppressing oxygen and other impurities, and lowering hydrogen recycling, while also promoting improved microstructural stability through boride formation.
The microstructure and properties of radiation-damaged samples of pure W and HEA alloys were compared. To simulate the damage morphology, 15 MeV W ion irradiation was conducted at the UK Dalton Cumbrian Facility at 700 °C. The microstructure investigations show that the BCC HEA phase was reinforced with TiC and TiB2-based precipitates formed during sample fabrication. Phase segregation under irradiation was examined using both grazing-incidence X-ray diffraction (GIXRD) and atom probe tomography (APT). The BCC phase was relatively stable, while the ceramic reinforcements showed significant irradiation-induced swelling. Minor formation of Cr clusters in the BCC HEA phase after irradiation was also detected using APT. Nanoindentation was used to investigate the irradiation hardening, showing an increase of 3-3.5 GPa and 1.5 GPa for the HEA and pure W, respectively. For plasma–material interaction studies, the damaged samples were exposed to deuterium plasma under two conditions: low-fluence/low-temperature irradiation at UKAEA and high-fluence/high-temperature irradiation at DIFFER. Deuterium retention and release behaviour were assessed using thermal desorption spectroscopy (TDS).

Speaker: Dr Zori Harutyunyan (Imperial College London,London, UK)

341 3.101 Assessing the tightly baffled long-legged divertor (TBLLD) concept in TCV

The Swiss Plasma Center is upgrading the TCV tokamak to test a tightly baffled, long-legged divertor (TBLLD), a novel concept that promises to enhance power exhaust capabilities with minimal modification to the magnetic configuration [1,2].
Simulations using the SOLPS-ITER code indicate that a TBLLD can improve TCV’s power exhaust capability by an order of magnitude compared to the unbaffled configuration [2]. Tight baffling sustains a high poloidal neutral density gradient, thereby, increasing the neutral density in front of the divertor target and enhancing volumetric power dissipation. In addition to a lower detachment threshold, extended leg length and tight baffling provide a large detachment window and a mechanism for passive detachment front stability. The neutral cushion, futhermore, provides a reservoir of potential energy that can temporarily buffer transient loads.
The simulations informed the design of a proof-of-principle TBLLD for the outer TCV divertor. A straight, vertical divertor design enables diagnostic access via TCV’s reciprocating divertor probe array, while maintaining engineering simplicity. Compatibility with neutral beam heated, high-power plasma scenarios constrains the baffled leg length to 0.34m. A trade-off between predicted plasma plugging and excessive recycling at the outer baffle yield a divertor width of 0.11m. The ability to expand the polodial flux along the divertor leg provides a means to vary both. The required gas tightness limits diagnostics access. Foreseen are poloidally distributed Langmuir probes, thermocouples, pressure gauges, and spectrometric lines of sight, to provide measurements of target fluxes, neutral density distribution, position and dynamics of the detachment front, which are critical to assess the TBLLD concept.
The main concerns for the proof-of-principle are the open inner divertor, which may limit the benefits of the closed outer divertor, and potential self-baffling of dense divertor plasmas. Recent SOLPS-ITER simulations also identified thermo-electric currents resulting from vastly different conditions in the inner and outer divertors as a potential project limitation. These concerns are addressed through modelling and, ultimately, through experiments.
A dedicated experimental campaign with the proof-of-principle TBLLD is planned for 2026. Following a successful validation of the TBLLD concept, a second phase of upgrades will optimise the baffle geometry, extend the exhaust solution to the inner divertor, address particle exhaust, through pump ducts at the top of the TBLLD, and integrate the plasma exhaust solution with an attractive core plasma scenario.
[1] M.V. Umansky, et al., Phys. Plasmas 24 (2017) 056112.
[2] G. Sun, et al., Nucl. Fusion 63 (2023) 096011.

Speaker: Holger Reimerdes (École Polytechnique Fédérale de Lausanne (EPFL), Swiss Plasma Center (SPC), Lausanne, Switzerland)

342 3.102 Development and exhaust performance of the X divertor on MAST Upgrade

We report on the development of the X divertor configuration [1] on MAST Upgrade and the reduction in heat and particle fluxes at the divertor targets this configuration achieves over a conventional divertor configuration. Using the Toksys [2] and TED [3] frameworks, we design magnetic equilibria with increased poloidal flux expansion at the target compared with a conventional divertor configuration and apply feed-forward changes to poloidal field coil currents to achieve the designed magnetic configurations in Ohmic, beam-heated L mode and beam-heated H mode plasmas, in double null and lower single null.

A strategy of zeroing the radial and vertical magnetic fields Br and Bz just below the strike point produces high target poloidal flux expansion but is very sensitive to plasma evolution and cannot be sustained with only feed forward control. Moving the field null further from the target reduces the peak flux expansion but gives a more robust scenario. Designing solely for target flux expansion without imposing a field null constraint at all produces similarly-robust scenarios with lower peak flux expansion. Adding strike point position feedback after forming the X divertor maintains acceptable flux expansion for much longer durations, but only for lower poloidal flux expansion scenarios.

In Ohmic and NBI-heated L mode plasmas using the field-null strategy, we achieved high peak flux expansion but only in the far SOL, with modest flux expansion at the strike point. These scenarios also showed increased interaction of the SOL with the divertor entrance and a corresponding increase in upstream neutral influx and radiated power. While peak heat and particle fluxes were reduced by 1/3 and 1/2 respectively, it was not possible to isolate the impact of the divertor geometry from the impact of the increased wall interaction at the divertor entrance. Moving the location of the peak flux expansion closer to the strike point while sacrificing the degree of peak flux expansion was more effective at avoiding excessive wall interaction. This scenario achieved a 10%-30% reduction in peak parallel heat flux and earlier detachment onset in N2-seeded H mode plasmas compared with the conventional divertor, with the effect being more pronounced in lower single null than in double null.

[1] M. Kotschenreuther et al Phys. Plasmas 20, 102507 (2013)
[2] H. Anand et al 2024 Nucl. Fusion 64 086051
[3] O P Bardsley et al 2024 Plasma Phys. Control. Fusion 66 055006

Work supported by EPSRC EP/W006839/1 and US-DOE DE-AC05-00OR22725.

Speaker: Jack Lovell (Oak Ridge National Lab)

343 3.103 Towards dynamical system identification simulations using multisine perturbations for MAST-U with SOLPS-ITER

The safe and controlled exhaust of heat from magnetically confined fusion plasmas requires having dynamic information about the system. On present-day reactors, this information is obtained using system identification experiments, which involve perturbing the gas injection rate and observing the Scrape-Off-Layer (SOL) response in the frequency domain. However, this strategy is cumbersome for future reactors, as accidental disruptions or reattachment of the SOL plasma could lead to catastrophic damage to the device. Therefore, numerical dynamic models of the exhaust plasma, validated against experiments, are needed. To this end, we present a multi-sine gas injection perturbation approach that uses high-fidelity time-dependent SOLPS-ITER simulations, with a grid extending to the vessel wall, to quantify the plasma response. Comparison to an existing set of experiments on MAST-U is done to validate this approach. The simulation grid was generated using the Grid Optimization and Adaptation Toolbox (GOAT), and both the Advanced Fluid Neutral (AFN) model and the kinetic neutral model Eirene available in SOLPS-ITER are used. New boundary condition types were implemented in the code to enable multi-sine gas injection perturbations in SOLPS-ITER. In steady state, qualitatively similar behaviour to previous SOLPS-ITER simulations with impurities and kinetic neutrals is observed, allowing comparison between our model and experimental data at similar operating conditions. The comparison in the frequency domain shows that the AFN SOLPS-ITER simulations predict much faster response times to divertor fuelling perturbations than the experiments, around a factor of 10 at 15 Hz, likely in part due to the fluid approximation used for the neutrals. Another possible cause of the discrepancy is the effect of remote vessel regions in MAST-U, some of which are not included in the computational grid. These regions can act as neutral gas reservoirs which introduce additional timescales to the system. Therefore, an additional boundary condition type was implemented, which models these reservoirs as 0D particle balance equations. When including the reservoir models for remote vessel regions, the response time increases, but is still too low compared to experiments by a factor of 4 at 15 Hz. Simulations employing Eirene for kinetic treatment of neutrals show better agreement with the experiments, i.e. within a factor of 2 at 15 Hz. In conclusion, this work presents a first step towards using the SOLPS-ITER code package for synthetic system identification using multi-sine perturbations.

Speaker: Stijn Kobussen (DIFFER, Dutch Institute for Fundamental Energy Research)

344 3.104 First L-Mode experiments in Alternative Divertor Configurations in ASDEX Upgrade

In a future fusion reactor, the power crossing the separatrix must be at least 100$\,$MW [1], with a power fall-off length $\lambda_q$ in the order of millimeters [2]. This presents a significant challenge to the divertor target plates, which cannot be solved without strong radiative losses in the divertor volume, leading to detachment. Thus, high impurity concentrations are required in the divertor region, which could be detrimental to the core.

One possible solution to reduce the impurity concentrations needed for detachment are alternative divertor configurations (ADCs). They augment the classic single null (SN) divertor through an increase in the connection length $L_c$ from the outboard midplane to the divertor target, e.g. by the introduction of another X-point in the divertor plasma.

The longer $L_c$ results in an effective increase in radial transport, as well as a greater volume for plasma neutral interactions. The combined effects were shown previously to reduce the required neutral pressures and impurity concentrations for reaching similar detachment depths as in the SN configuration, both experimentally [3] and in simulations [4].

At the tokamak ASDEX Upgrade, the upper divertor was recently equipped with additional coils and a cryopump to study a variety of ADCs, for the first time in a machine with fusion-relevant heating power and a tungsten wall [5]. First experiments explored, among others, the X-divertor (XD) with increased flux expansion at the target, and the low-field-side snowflake minus (LFS SF$^-$) with a secondary X-point in the LFS scrape-off layer.

This contribution presents the first L-Mode experiments with ADCs in ASDEX Upgrade, where feedback-controlled density ramps were performed in the SN, XD and LFS SF$^-$ configuration. The detachment onset could be determined from Langmuir probe data.

Close to the primary strikepoint, the plasma was detached throughout XD and LFS SF$^-$ discharges, in contrast to the attached SN. Further away from the separatrix, however, no difference in detachment behavior depending on edge density or divertor neutral pressure was found between the configurations. These results are interpreted using the plasma and neutral transport code SOLPS-ITER. In addition, a comparison to the typically observed detachment behavior in H-mode is presented.

[1] R. Wenninger et al., NF 2016
[2] T. Eich et al., NF 2013
[3] C. Theiler et al., NF 2017
[4] T. Lunt et al., NME 2019
[5] I. Zammuto et al., FED 2025

Speaker: Felix Albrecht (Max Planck Institute for Plasma Physics, Garching, Germany and Technical University of Munich, TUM School of Natural Sciences, Physics Department, 85748 Garching, Germany)

345 3.105 UEDGE modelling of drift effects on target heat flux of snowflake divertor in HL-3

In the tokamak, managing heat and particle exhaust through the divertor is crucial for long-term plasma confinement. The snowflake (SF) divertor [1] is an effective solution for mitigating deposited heat flux. Both experiments and simulations have shown significant reduction of deposited heat flux in SF divertors compared to standard divertors [2,3]. Experimental studies suggest that E×B drifts are important for redistributing heat flux in SF− configurations, but a quantitative analysis is still lacking [4]. Previous simulations studies often overlook drift effects due to computational challenges, and few simulations have directly compared with experimental data. Given these research gaps, it is essential to investigate the impact of drift effects on deposited heat flux in snowflake divertors. HL-3 device [5], with its strong magnetic field and precise plasma control, provides an ideal platform for generating snowflake divertors. The UEDGE[6], which can simulate two X-point snowflake geometries and include drift effects, is a suitable tool for this study. In this work, full-drift UEDGE simulations are performed for the snowflake-minus discharge (SF−, #6590) on HL-3. The simulated upstream electron density and temperature, as well as the target electron density and temperature profiles are compared with those of the corresponding experimental measurements, showing well agreement. A comparative study between the SF− divertor configuration (#6590) and the standard divertor configuration (SD, #6587) is carried out to elucidate the role of drift effects in reducing the deposited heat-flux density in the SF− configuration. In addition, the influence of magnetic-field direction on deposited heat flux will be investigated in the SF− divertor configuration, to elucidate the drift-driven heat flux mechanisms in the region between the primary and secondary X-points.

Key words: snowflake divertor, deposited heat flux, UEDGE simulation, drifts, HL-3

References:
[1] Ryutov D D and Soukhanovskii V A 2015 Physics of Plasmas 22 110901
[2] Soukhanovskii V A et al 2018 Nucl. Fusion 58 036018
[3] Wu H S et al 2026 Nucl. Fusion 66 016032
[4] Tsui C K et al 2021 Nucl. Fusion 61 046004
[5] Duan X R et al 2024 Nucl. Fusion 64 112021
[6] Rognlien T D et al 1999 Physics of Plasmas 6 1851–7

Speaker: Mingzhou Zhang (School of Physics, Dalian University of Technology, Dalian, China)

346 3.106 Avoidance of large ELMs through impurity seeding

Mitigating the heat load of type-I edge-localised modes (ELMs) via impurity seeding has been investigated in single-null plasmas of the Tokamak à Configuration Variable (TCV). The measurements exploit TCV’s fast diagnostics to resolve ELMs, from the plasma edge to the divertor targets: the ELM energy loss is measured with a diamagnetic loop (DML, acquisition frequency of f=10kHz); the radiation evolution is inferred with absolute extreme ultraviolet diodes (AXUV, f=200kHz), and both inner and outer target heat flux profiles are evaluated with infrared thermography (IR, f≥10kHz). The heating is mainly provided by neutral beam injection, while electron cyclotron heating is used to reduce core contamination by impurities. A systematic scan of neon, argon, and nitrogen seeding rates was performed, covering from unseeded discharges to ones where impurities caused a radiative collapse. Comparing DML and IR data indicates that dissipative processes lower the ELM energy by only 10-25% before the ELM crash. The findings agree with previous experiments and modelling in JET, where injecting argon in the divertor did not yield considerable radiation during ELMs [1]. Nevertheless, in TCV, strong seeding modified the pedestal profiles, reducing the ELM energy loss, which substantially decreased inner and outer peak heat fluxes—up to 40% with neon, 60% with argon, and 90% with nitrogen. In the nitrogen case, the plasma stored energy diminished by 11% during seeding. Nonetheless, the energy confinement time, the core effective ion charge $Z_{eff}$, the core electron temperature and the pedestal electron density were not affected, while the radiation around the X-point intensified, core $n_e$ rose by 30%, and the pedestal $T_e$ lowered by about 35%, transitioning the ELM regime from type-I to type-III. Furthermore, in AUG, JET, and EAST, an increase in confinement time was observed simultaneously with a reduction of the ELM energy loss [2-4], by using nitrogen (AUG) and neon (JET, EAST). The TCV, AUG, JET, and EAST results indicate that, although impurities cannot prevent the burn-through of high-energy transients, they can modify edge profiles to avoid the formation of large ELMs—a feature of direct interest for future fusion reactors.

[1] J. Rapp et al. Nucl. Fusion 44 312–319 (2004)
[2] M. Komm et al. Nucl. Fusion 63 126018 (2023)
[3] C. Giroud et al. EPS Conference on Plasma Physics (2022)
[4] K. Li et al. Nucl. Fusion 63 026025 (2023)

Speaker: Martim Zurita (EPFL - SPC)

347 3.107 Modelling kinetic electron effects for transient heat-loads in the detached MAST-Upgrade Super-X divertor

We present the modelling of detachment burn-through scenarios with a new exhaust model developed in ReMKiT1D [1], a framework for 1D fluid-kinetic modelling and collisional-radiative interactions, and the 2D fluid modelling suite SOLPS-ITER. Starting from a "1D" SOLPS case, a detached plasma background representing the MAST-Upgrade Super-X divertor is reproduced within the ReMKiT1D exhaust model, in both a fluid "benchmark" case and kinetic electron case. The latter introduces non-Maxwellian dynamics not captured in SOLPS that can enhance fast electron transport during energetic transient events and modify interactions with the detachment gas cloud [2, 3, 4]. Detachment burn-through is then modelled by applying heat pulses of different energies to the background, comparing the two cases from the ReMKiT1D exhaust model with SOLPS. Complementary interpretations of detachment burn-through from both codes will improve our basic understanding - such as the apparent lack of detachment front pressure balance in 2D SOLPS not seen in 1D fluid codes [5] - and to improve the accuracy and predictive capabilities of the exhaust model in ReMKiT1D.

Preliminary simulations have shown that the required transient energy to burn through the detachment front has a high sensitivity to the timescale of wall recycling. In addition, the neutral pressure downstream of the front is highly sensitive to the energetic efficiency with which neutrals reflect from the walls.

This work has been funded by the EPSRC Energy Programme [grant number EP/W006839/1].

References
[1] Mijin et al. Comp. Phys. Comm. 300 (2024)
[2] Power et al. Nucl. Fusion 63 086013 (2023)
[3] Tskhakaya et al. Contrib. Plasma Phys. 48(1-3), 89-93 (2008)
[4] Chankin et al. Plasma Phys. Control. Fusion 60 (2018)
[5] Dudson et al. Plasma Phys. Control. Fusion 61(6) (2019)

Speaker: Sid Leigh (UKAEA)

348 3.108 Introduction of cross-field drifts in SOLPS-ITER simulations of STEP with fully tracked Ar impurities

Introduction of cross-field drifts in SOLPS-ITER [1] simulations of STEP [2, 3] with fully tracked Ar impurities was found to re-distribute the impurities in the simulation volume differently from the main ions, improving the representation of Ar in the plasma from the first drift simulations for STEP [4], where Ar was proxied as a fixed fraction of the main ion content, and complementing earlier drift-free simulations with evolving Ar [5—7]. The drift terms were activated in a connected double-null configuration with Ar seeded into the outer divertors for detached initial conditions.

The Ar content and the radiated power inside the separatrix were found to increase by approximately 50% and 30%, respectively, with the drifts active. At the separatrix, the Ar ion concentration increased from 3% to 4% of the electron density at the outer midplane, posing a risk for potential degradation of the upstream conditions from the point of view of core-edge compatibility. Strong accumulation of Ar by almost an order of magnitude was observed also in the inner lower divertor.

The activation of drifts resulted in an up-down asymmetry in the power entering the divertor volumes with the upper divertors receiving more power according to a 53:47 ratio. The direction of the asymmetry is opposite to that in [4] but follows the earlier experimental observations in DIII-D [8], MAST [9] and Alcator C-Mod [10], being favourable for the foreseen easier maintenance access into the upper divertor. Due to the notable increase in the power entering the outer upper divertor leg, the 2—5-eV temperature fronts were found to shift significantly closer to the divertor target, but at the target detached conditions were preserved by simultaneous increase in the Ar content and radiation in the divertor volume.

[1] S. Wiesen et al., Journal of Nuclear Materials 463 (2015) 480—484
[2] H. Wilson et al., in Commercialising Fusion Energy, IOP Publishing (2020) 8-1—8-18
[3] H. Meyer, 29th IAEA Fusion Energy Conference, London, UK, 16.—21.10.2023
[4] J. Karhunen et al., Nuclear Fusion 64 (2024) 096021
[5] R.T. Osawa et al., Nuclear Fusion 63 (2023) 076032
[6] R.T. Osawa et al., Nuclear Fusion 64 (2024) 106007
[7] S.L. Newton et al., Nuclear Fusion 65 (2025) 096026
[8] C.J. Lasnier et al., Nuclear Fusion 38 (1998) 1225
[9] G.F. Counsell et al., Plasma Physics and Controlled Fusion 44 (2002) B23—B37
[10] D. Brunner et al., Nuclear Fusion 58 (2018) 076010

Speaker: Dr Juuso Karhunen (VTT)

349 3.109 3D Synthetic Bolometry and Radiation Asymmetry Analysis for ITER SPI-Mitigated Disruptions

Accurate assessment of radiated power during Shattered Pellet Injection (SPI)–mitigated disruptions is essential for ensuring the protection of ITER plasma-facing components [1].
In this work, we extend the synthetic bolometry framework for ITER by incorporating a full 3D treatment of the diagnostic geometry, including inner apertures and sub-collimators, using the CHERAB framework.
This development enables more realistic modeling of the line-of-sight transmission and provides improved synthetic bolometer signals based on 3D JOREK simulations of SPI scenarios [2].

A series of dual Ne/H SPI simulations, obtained from the ITER Integrated Modelling & Analysis Suite (IMAS) disruption database, is used to evaluate the diagnostic performance. Geometrical response matrices are generated through ray-tracing on unstructured JOREK meshes, and the foil bolometer’s temporal response is modeled to assess its ability to follow the rapid variations during the thermal quench (TQ).
To enhance the accuracy of total radiated power estimation, we apply a weighted-sum method, which reduces systematic underestimation observed in simpler summation techniques [3].

We further investigate ITER’s capability to capture toroidal radiation asymmetries, which may arise during SPI due to localized shard ablation and MHD-driven redistribution.
By computing the Toroidal Peaking Factor (TPF) from the 3D emissivity and comparing it with the TPF inferred from synthetic bolometer signals, we assess how effectively the bolometer system can retrieve toroidally varying radiation patterns. The results indicate that, despite the limited number of toroidally separated cameras, the diagnostic retains the potential to resolve significant non-axisymmetric radiation structures, although local peaks may still be underestimated in regions of high TPF.

Finally, we evaluate the heat load on the bolometer foil itself during the intense, short-duration radiation bursts of the TQ phase.
When the measured time response is included, the resulting temperature rise remains within the diagnostic design requirements, demonstrating that the bolometer can tolerate even the strongest radiation spikes predicted by the SPI simulations.

These findings support the robustness of the ITER bolometer system and provide improved tools for analyzing 3D radiation behavior in SPI-mitigated disruptions, contributing to future disruption mitigation strategy optimization.

[1] M. Lehnen et al., J. Nucl. Mater. 463, 39 (2015)
[2] D. Hu et al., Nucl. Fusion 64, 86005 (2024)
[3] G. Partesotti et al., Nucl. Fusion 65, 16035 (2025)

Speaker: Koyo Munechika (ITER Organization)

350 3.110 Development and validation of an XPR model in the ASDEX Upgrade flight simulator integrated modelling framework

Efficient power exhaust is a critical challenge for fusion power plants. The so‑called X‑point radiator (XPR) has been proposed as a promising concept to redistribute and dissipate heat loads via impurity radiation.

While detailed edge‑plasma simulations provide valuable insights, they are computationally intensive and therefore not ideally suited for rapid design iteration or extensive parameter scans. A promising alternative is offered by a flight-simulator-type framework such as Fenix [1], which incorporates the equilibrium solver FEQIS [2] and the plasma transport solver ASTRA [3] to simulate plasma behaviour and equilibrium, as well as a simulated control system based on the Plasma Control System Simulation Platform [4]. This combination enables simultaneous modelling of the interaction between the reactor’s control system and the plasma dynamics.

ASTRA is equipped with physics modules that allow it to capture more complex plasma phenomena. In this work, we extend the reduced modelling approach for the X‑point radiator developed in [5]. By imposing a pressure balance between upstream and X‑point parameters, main‑ion and impurity densities at the X-point are obtained. A simple model based on a detachment qualifier [6] is used to estimate the amount of neutral particles at the X-point. Moreover, an approach to predict ELM suppression phases is implemented.

The model is validated in Fenix against ASDEX Upgrade discharges with nitrogen seeding. The reduced XPR model successfully predicts the formation of the XPR and the evolution of its position. Furthermore, experimental radiation measurements are reproduced, and the temporal evolution of the simulated density and temperature profiles shows good agreement with experimental data.

[1] P. David et al 2025 Open Plasma Sci. 1 3
[2] E. Fable et al 2025 Open Plasma Sci. 1 2
[3] G. V. Pereverzev and P. N.Yushmanov 2002 ASTRA. Automated System for Transport Analysis in a Tokamak. IPP Report Nr. 5-98
[4] M. L. Walker et al 2014 Fusion Engineering and Design 89 518-522
[5] U. Stroth et al 2022 Nucl. Fusion 62 076008
[6] A. Kallenbach et al 2015 Nucl. Fusion 55 053026

Speaker: Fabian Solfronk (IPP)

351 3.111 Preparation of Boundary Digital twin for SPARC boundary physics operations

The SPARC tokamak, a high field (12 T), high current (8.7 MA) machine designed to achieve an energy gain Q of 11 in H-mode with DT fuel, is currently under assembly and will be expected to be in operation in 2027. As the expected heat-flux width for SPARC is in the range between 0.3 and 0.6 mm, the power exhaust is extremely difficult due to upstream steady-state parallel heat fluxes of about 10 GW/m2. As such, it is essential that power exhaust is optimized during operational planning with interpulse feedback on performance. The Boundary Digital Twin (BDT) is a novel toolchain that couples physics and engineering models to meet this challenge.

For reliable calculations of the power exhaust defining the operating window for SPARC, the required inputs for predictions can be provided by simulation workflows like SOLPS-NN or HEAT which provide the 2D radiation pattern or the parallel heat flux to the targets. In order to accommodate the requirements for interpulse timescales and for PFC integrity lifetime tracking, the specialised HEAT code is used.

In addition, a set of synthetic instrumentations (shunts, IR, TCs, bolometry) have been created for comparison with experimental measurements. These comparisons will help to identify incomplete or new physics details. An important aspect is the energy balance and 2D/3D energy distribution, which is essential for a correct assessment of PFC integrity and for the energy gain mission. To reliably perform this task in the challenging environment of SPARC, a synergy of IR and thermocouple measurements, two independent approaches for determining tile-resolved energy distribution loads and the global energy balance, is being implemented. For the complex 3D data visualization, a graphics render engine is used in addition to standard plotting libraries. We present the Boundary Digital Twin, the roadmap for development, and its planned application as an operational tool for SPARC.

Speaker: Andreas Redl (Commonwealth Fusion Systems)

352 3.112 Control of the radiated power using bolometers and the impact of observed asymmetries on the radiated power proxy in Wendelstein 7-X

The high heat fluxes in magnetic fusion devices pose an immediate threat to their plasma facing components (PFCs). These fluxes can be mitigated by injecting low to medium Z impurities [1]. These impurities stimulate radiation emission in the edge of the plasma, which dissipates power volumetrically and reduce the heat fluxes to the PFCs. Efficient control of the radiated power ($P_{rad}$) requires accurate representation of its magnitude and dynamics, as well as reliable real-time diagnostics [2]. Especially in 3D devices, such as the Wendelstein 7-X stellarator (W7-X), determining a robust $P_{rad}$ proxy is a challenge due to its asymmetries observed in the past [3]. This contribution reports on the system identification experiments performed on W7-X and on observed asymmetries
of the $P_{rad}$ proxies calculated from different bolometer systems for different actuator (valve) locations.
W7-X utilizes piezo-electric valves [4] to regulate the amount of injected gaseous impurities. Furthermore, a wide angle bolometer camera [3] offers a real-time-capable $P_{rad}$ estimation and streaming to the control system, with the rest of the bolometer systems [3][5][6] providing radiation data from different cross-sections.
Asymmetries are observed in locations with and without plasma-wall-interaction (PWI) which reduce with increasing densities and with decreasing input power. The asymmetries while seeding drop when the plasma transitions to detachment. Additionally, asymmetries between stellarator equivalent locations show a configuration dependence. These observations allow to calculate a 𝑃𝑟𝑎𝑑 proxy considering the toroidal radiation distribution. This will be addressed in future work.
Despite the asymmetric $P_{rad}$ behaviour, system identification experiments provided the data for designing controllers for different seeding scenarios. Nitrogen and neon actuators are utilized with the controllers for performing $P_{rad}$ scans, enabling studying plasma parameters at different radiation levels. In addition, stable detachment was feasible for a variety of magnetic configurations and during high power scenarios.

[1] A. W. Leonard, Plasma Phys. Control. Fusion 60 044001 (2018)
[2] R. Pintelon et al, System Identification: A Frequency Domain Approach (IEEE Press, 2001)
[3] G. Partesotti et.al, Rev. Sci. Instrum. 95, 103503 (2024); doi: 10.1063/5.02
[4] M. Griener et.al, Rev. Sci. Instrum. 88, 033509 (2017), https://doi.org/10.1063/1.4978629
[5] D. Zhang et al, Rev. Sci. Instrum. 81, 4 (2010) https://doi.org/10.1063/1.3483194
[6] G. Partesotti Rev. Sci. Instrum. 96, 063503 (2025) https://doi.org/10.1063/5.0261413

Speaker: Anastasios Tsikouras (MPPL)

353 3.113 Divertor Closure, fueling locations and impurity seeding in MAST-U Tokamak

The effect of divertor closure, fueling locations and impurity seeding is presented for the MAST-U tokamak. SOLPS-ITER simulations for MAST-U H-mode plasmas reveals that changing D2 fueling location significantly impacts the detachment conditions. Under lower outer divertor fueling, the total power loss at the lower outer divertor is higher compared to the midplane fueling.
The closed divertor shows higher neutral trapping and power dissipation under both midplane and lower outer divertor fueling compared to the open divertor configuration. This demonstrated the role of divertor closure to enhance dissipation and enabling detachment at lower upstream density in agreement with results reported in the DIII-D tokamak ( L. Casali et al. Contrib. Plasma Physics 2018, L. Casali et al. Nucl. Fusion 2020). The combination of LD fueling and a closed divertor provides the most favorable conditions for detachment. A lower amount of nitrogen is needed in the closed divertor compared to the open divertor to achieve detachment conditions due to higher impurity compression. Impurity flows, ionization source locations and impurity enrichment calculations demonstrate the mechanism of nitrogen leakage out of the divertor which is more pronounced at higher seeding rates in the open divertor geometry. These studies are consistent with impurity seeding studies at DIII-D (L. Casali et al. Nucl. Fusion 2022) highlighting a similar role of divertor closure in conventional and spherical tokamaks.
This work is supported by US Department of Energy under DOE Grant DE-SC0023381 and Nuclear Regulatory Commission under NRC Grant # 31310022M0014.

Speaker: Livia Casali (GNOI)

354 3.114 Role of Radial Electric Field on Heat Flux Scrape-off Layer (SOL) Width in ADITYA-U Tokamak

The scrape-off layer (SOL) heat-flux width, λ$_q$ , is a key parameter for controlling heat loads on plasma-facing components in tokamaks. Recent studies suggest that the radial electric field E$_r$ and its associated flows and shear, plays a significant role in shaping SOL transport and the resulting heat-flux profile. Enhanced shear suppresses turbulence and can narrow the SOL width, whereas weaker or reversed shear increases cross-field transport and broadens λ$_q$ [1]. Experimental studies have correlated measured edge electric fields with changes in SOL conditions in devices such as ASDEX Upgrade [2], and biasing experiments using Langmuir probe and Doppler reflectometry diagnostics have demonstrated the impact of modified E$_r$ on low-frequency flows and turbulence in the edge/SOL region [3–5].
Recently broadening of heat flux SOL width is observed in ADITYA-U tokamak with increased cross-field diffusion [6]. In this present work, we investigate the correlation between the radial electric field and the heat-flux scrape-off layer (SOL) width. Experiments involving the application of an external electric field using a biased electrode were carried out. The radial electric field and its associated shear found to influence cross-field transport strongly in the edge/SOL region by modifying turbulent transport mechanisms. The results show that the application of a negative radial electric field leads to a significant broadening of the SOL width. This scenario will be also simulated using the BOUT++/UEDGE code, and comparison between the experimental observations and numerical results will be obtained. These findings suggest a promising approach for controlling the heat-flux SOL width in future tokamak devices.

References:
[1] T. Eich et al, Nuclear Materials and Energy 25 (2020) 100795.
[2] A. Loarte et al., Journal of Nuclear Materials 390–391 (2009) 899–902.
[3] G. Birkenmeier et al., Plasma Physics and Controlled Fusion 57 (2015) 014018.
[4] A. D. Liu et al., Physical Review Letters 103, 095002 (2009).
[5] J. Dong et al., Plasma and Fusion Research 5, S2014 (2010).
[6] SK Injamul Hoque et al, to be submitted in peer review journal (2025) [https://doi.org/10.48550/arXiv.2508.00339].

Speaker: SK Injamul Hoque (Institute for Plasma Research)

355 3.115 The importance of drifts and radial transport assumptions in 2D modelling of the access to detachment and XPR in WEST

Until lately, codes such as SOLEDGE3X [1] or SOLPS [2] have been routinely used without including drift effects, although past studies have shown their significance [3,4]. Thanks to the continuous development and improvements in the numerical treatment, the inclusion of drifts has now become more common and proves to be important in detachment studies [5,6]. In this contribution, we evaluate their impact on WEST dissipative divertor regimes using the SOLEDGE3X code.

Recent modelling efforts [7] show that the inclusion of drifts is necessary to obtain an agreement between simulation and experiment in WEST in attached conditions, where the poloidal ExB flux directed from the HFS to the LFS leads to a strong reduction of the outer divertor temperature that would otherwise be strongly overestimated. Based on these simulations, here we continue to explore the role of drifts in the access to detachment. It is observed, that when a nitrogen seeding is introduced, the seeding rate required to obtain a detached divertor on the HFS is significantly lowered by the drift effects. Due to an accumulation of nitrogen on the HFS, driven by the drifts, the detachment becomes strongly asymmetric, as commonly observed, with the outer target detaching less easily [8].

In the course of these studies, it has been identified that, when the drifts are taken into account, the detachment onset and XPR behaviour are sensitive to radial transport assumptions. An accurate resolution of radial transport coefficients in the pedestal, supported by reflectometry measurements, is essential to reproduce the experimental-like detachment. The radial transport around the X-point influences the feasibility of achieving a simultaneous detachment at both targets within the XPR stability limits. The results also suggest that the XPR formation and stability might depend on the transport regime, with reduced transport in the pedestal having a favourable effect on the XPR stability and control. With constant transport coefficients, typically assumed in simulations of WEST L-mode plasmas [8], the radiation and the density fronts cross the separatrix inwards of the X-point and drifts drive the cold high-density region into an unstable MARFE before a radiation collapse occurs and before the outer target detaches. Reduced transport in the pedestal slows down the propagation of the cold high-density region, directing it towards the X-point. A negative potential well develops, similar as in [9], modifying the distribution of the ExB flux around the X-point, allowing a subsequent detachment of the outer target.

Speaker: Eva Havlickova (IRFM, CEA Cadarache, France)

356 3.116 Characterizing SOL Heat Transport in Wendelstein 7-X with SMoLID

The island divertor is the leading plasma exhaust concept in stellarators, and evaluating its reactor relevance is a key objective of the Wendelstein 7-X (W7-X) experiment [1]. From a power exhaust perspective, maintaining steady-state divertor heat loads below 10 MW/m² is essential for safe and sustained device operation. However, a robust framework for quantitatively characterizing the 3D transport and facilitating reliable extrapolation to reactor-relevant conditions has been lacking for stellarators.

In this contribution, we present SMoLID (Simple Model for Loads in Island Divertor) – a novel framework for interpreting target heat flux patterns and the underlying transport mechanisms [2]. The central idea is that the 3D transport can be decomposed into distinct channels, each associated with a specific topological region of the island scrape-off layer (SOL), namely, main SOL, private flux and target shadowed regions and characterized by a representative width ($\Lambda_W$, $\Lambda_D$ and $\Lambda_S$ respectively) reflecting the balance between parallel and cross-field transport processes. The heat transport in the main SOL is described by a 3D power-carrying layer, with its width $\Lambda_W$ (analogous to the $\lambda_q$ metric in tokamaks) inferred from target heat flux using an adapted Wagner-Eich formalism [3]. For the standard magnetic configuration of W7-X, $\Lambda_W$ values at the outboard midplane (bean-shaped cross-section) are typically of the order of 1 cm under attached divertor conditions.
Furthermore, we apply SMoLID to the Infrared thermography observations at W7-X [4, 5], to develop empirical scaling laws for the transport widths as functions of some key operational and plasma parameters. Notably, $\Lambda_W$ exhibits a weak dependence on the power entering the SOL (∼ P$_{SOL}^{-0.1}$). However, we observe that $\Lambda_W$ tends to exhibit a positive dependence on plasma density, with the scaling strengthening significantly above a line-integrated density threshold of approximately 7 × 10¹⁹ m⁻².
Finally, we demonstrate the utility of SMoLID for divertor design. In particular, the concept of transport channels is leveraged to obtain a heat load compatible closed island divertor geometry for W7-X. Overall, the SMoLID framework opens new avenues for advancing the power exhaust research in island diverted stellarators.

[1] M. Endler et al, Fusion Engineering and Design 167 (2021) 112381
[2] A. Kharwandikar, PhD Thesis. University of Greifswald 2025
[3] T. Eich et al, Phys. Rev. Lett. 107 (2011) 215001
[4] Y. Gao et al, Nuclear Fusion 59.6 (2019) 066007
[5] S. Thiede et al, 2025 submitted to Rev. Sci. Instrum.

Speaker: Amit Kohinoor Kharwandikar (MPPL)

357 3.117 Revisiting the ITER $Q_\text{DT} = 10$ SOLPS-4.3 database with SOLPS-ITER with drifts and currents

The design of the ITER divertor was established via guidance of extensive scoping studies of the baseline burning conditions at $Q_\text{DT} = 10$, conducted with the SOLPS-4.3 boundary code without cross-field drifts or currents [1, 2]. Since those scoping studies, the development of the SOLPS-ITER code package, launched by the ITER Organization in 2015, has enabled numerically robust and computationally feasible inclusion of these drift and current terms in ITER-scale simulations [3 – 6].

In this work, this SOLPS-ITER capability is leveraged to revisit the previous SOLPS-4.3 database with drifts and currents to address their impact on the operational space and on the conclusions drawn in the previous studies. The work is focused on the baseline burning conditions at $Q_\text{DT} = 10$, assuming fuel and impurity injection from the top of the machine and pumping underneath the dome with 100 MW of input power to the computational domain, consistent with the original SOLPS-4.3 database [1]. An assumption that does differ from the SOLPS-4.3 database is that the new ITER baseline with the full tungsten wall is assumed in these SOLPS-ITER simulations [7]. While the original gas injection and pumping configurations are retained in this study, a complementary research contribution in this conference addresses the impact of a higher fidelity approach, including sub-divertor structures, fuel injection from the sub-divertor area, and neutral bypasses around the divertor cassette [8].

The overarching observation in this study is that while the drift and current terms do lead to visible changes in the overall plasma solution, they do not fundamentally change the overall operational space established in the earlier studies. In these ITER-scale simulations at $Q_\text{DT} = 10$ conditions with the heat flux width of about $\lambda_q \sim 3.5$ mm, the impact of drifts is a correction term that does impact the absolute values, especially approaching reattachment, but not the overall conclusions of the earlier scoping studies.
[1] H.D. Pacher, et al. J. Nucl. Mat. 463 (2015) 591-595.
[2] R.A. Pitts, et al. Nucl. Mat. Ene. 20 (2019) 100696.
[3] S. Wiesen, et al. J. Nucl. Mat. 463 (2015) 480-484.
[4] E. Kaveeva, et al. Nucl. Fusion 58 (2018) 126018.
[5] E. Kaveeva, et al. Nucl. Fusion 60 (2020) 046019.
[6] A.A. Pshenov, et al. Nucl. Mat. Ene. 42 (2025) 101851.
[7] A. Loarte, et al. Plasma Phys. Control. Fusion 67 (2025) 065023.
[8] A.A. Pshenov, et al. This conference.

Speaker: Aaro Järvinen (VTT)

358 3.118 Plasma-Assisted Growth and Stability of Thick Boron Layers on Tungsten Under Fusion-Relevant Conditions

The extreme conditions in ITER demand a carefully selected first wall material. Tungsten, which replaced the originally planned beryllium, offers improved resilience and reactor relevance but introduces new challenges in impurity management. A thin boron coating (~100 nm) has been proposed to mitigate these issues [1]. However, during ITER operation, erosion and redeposition processes are expected to lead to the erosion, migration, and redeposition of boron, forming layers several micrometres thick [2]. The behaviour of these layers under plasma loading and during venting is not yet understood and may contribute to unwanted dust generation and tritium retention.

Studies have been carried out using the Magnum-PSI and Upgraded Pilot-PSI linear plasma devices to investigate these issues. In ITER, boron deposition will be carried out via helium glow-discharge with diborane gas. Due to the toxicity of diborane an alternative in-situ deposition to grow thick, fusion relevant boron layers is required. We have conducted scoping studies of several deposition techniques, including hydrogen plasma chemical erosion, powder-injection ablation, laser-assisted evaporation, and plasma-assisted pulsed laser deposition (PA-PLD), where laser-ablated material is entrained in the plasma and deposited onto the substrate. Among the investigated techniques, laser evaporation and PA-PLD demonstrated the highest potential to produce thick ITER-relevant layers suitable for future studies.

Boron, oxygen and deuterium co-deposited layers with thicknesses between 0.5 and 5 μm grown by PA-PLD have been created on a Tungsten substrate kept at 400 K, and subsequently were exposed to ITER first wall and divertor relevant plasma conditions in Upgraded Pilot-PSI by stepwise increase in power density, between 0.5 and 3 MW/m2, until flaking occurred. Layer integrity and flaking behaviour was monitored using infrared and ultra-fast camera imaging, while layer composition and thickness were quantified via ion beam analysis. This integrated methodology enables controlled studies of layer stability and adhesion. Layer stability was found to be higher for rougher substrates and thinner layers. In addition, the effects of different surface temperatures and growth rates are discussed.

[1] R.A. Pitts, et al., J. Nucl. Energy, 42 (2025) 101854
[2] K. Schmid, T. Wauters, J. Nucl. Energy 41 (2024) 101798

Speaker: Cas Robben (DIFFER)

359 3.119 Lithium impurity transport in fusion devices

Recent developments in advanced nuclear fusion reactors consider the use of solid plasma-facing components (PFCs), typically made of tungsten. However the emergence of new compact nuclear fusion reactor concepts, presented as more viable for commercial applications, can lead, due to the reduced plasma wetted surface, to heat fluxes much higher than the 15MW/m$^2$ estimated for the ITER divertor. Consequently, the periodic replacement of PFCs will become more frequent due to faster structural degradation and erosion. To overcome this issue, several concepts consider the use of liquid metal, generally lithium (Li) or tin (Sn), as a plasma-facing material.

The physics of liquid lithium PFCs differs significantly from that of solid tungsten. Simulations performed with a 1d3v Particle-in-Cell code show that a high evaporated flux of lithium from the wall, once ionized and interacting with the magnetized plasma, can substantially modify both the sheath and presheath structures. These modifications alter the potential drop and influence in-sheath redeposition. The study further identifies four key parameters governing these boundary conditions: (i) the ratio of lithium influx to hydrogen-isotope outflux, (ii) the ionization mean free path, (iii) the magnetic-field strength, and (iv) the magnetic-field inclination.

If Li impurities are not redeposited due to the sheath and presheath electromagnetic field, they enter the Scrape-Off layer and are subjected to a competition between parallel (to B) and perpendicular transport. Therefore lithium can cross the separatrix and penetrate the core plasma. As a light species, Li is mainly transported by turbulence, with neoclassical contributions playing only a minor role. Using the global gyrokinetic code GYSELA, a new method was developed and benchmarked to determine impurity turbulent-transport coefficients, enabling the separate evaluation of diffusion, thermodiffusion, and pure convection. In the absence of a transport barrier at the core edge, the study finds impurity-transport trends similar to those observed for helium, with a much stronger outward thermodiffusive flux than in the case of tungsten. Introducing a transport barrier, however, reduces turbulence intensity, and thus turbulent transport, while generating a dominant inward convective contribution inside the barrier due to strong $E \times B$ shear. An analysis of the peaking factor shows that steady-state lithium accumulation in the core results from a balance between pure convection and thermodiffusion, underscoring the crucial role of the transport barrier in limiting Li contamination of the core plasma.

Speaker: Mr Romain Avril (Université de Lorraine, Institut Jean Lamour, UMR 7198 CNRS)

18:30 Conference Dinner

(boats leave at 19:00)

Friday 22 May

Review Talk: Morning session

R1

360 R5 Alternative divertor research – from physics studies to reactor implementation

Handling plasma exhaust without compromising core performance remains one of the central challenges for fusion energy. Research on alternative divertor configurations (ADCs) has progressed rapidly in recent years, moving beyond exploratory studies [1,2] toward demonstrations of substantial benefits [3–5]. Reported advantages include order-of-magnitude reductions in target heat flux, extended operational windows for detachment, improved active feedback control, and enhanced resilience of the detached state to transients such as ELMs - all while preserving good core conditions. Particularly strong exhaust performance has been achieved with the X-Point Target Divertor [3] and the Super-X Divertor [4,5], supporting their prospective implementation in SPARC/ARC and STEP, respectively. Additional key benefits have further been projected for tightly-baffled, long-legged divertor configurations, which improve neutral confinement and detachment stability while remaining compatible with both conventional and alternative magnetic geometries [6–9]. Extrapolation capabilities of these concepts have been strengthened through higher-power experiments and advances in modelling, with state-of-the-art transport simulations now routinely covering arbitrary geometries, including multiple divertor X-points, and incorporating drift effects. In parallel, the need for ADCs in next-step devices has become increasingly clear: they are essential for compact, high-field designs such as SPARC/ARC and STEP, while recent findings that even modest, strategic divertor modifications can yield significant benefits make ADC research equally relevant and timely for divertor optimization in more conventional approaches such as DEMO and CFEDR. Beyond specific designs, ADC research also improves the general understanding of power exhaust, providing validation of models and operational insights directly relevant to ITER. This talk will review the rapid progress in both experimental demonstrations and theoretical understanding of ADCs, highlighting the emerging physics basis for optimized divertor solutions that balance exhaust performance with engineering complexity.
[1] C. Theiler et al., Nucl. Fusion 2017
[2] V. Soukhanovskii, Plasma Phys. Control. Fusion 2017
[3] K. Lee et al., Phys. Rev. Lett. 2025
[4] K. Verhaegh et al., Nature Comm. Phys. 2025
[5] B. Kool et al., Nature Energy 2025
[6] M. Umansky et al., Phys. Plasmas 2017
[7] M. Wigram et al., Nucl. Fusion 2023
[8] G. Sun et al., Nucl. Fusion 2023
[9] J. Yu et al., Nucl. Mater. Energy 2024

Speaker: Christian Theiler (EPFL-SPC)

Oral: Morning session

361 O27 Negative Triangularity as a Reactor-Relevant Scenario: High-Performance Detached Plasmas on TCV Leveraging Alternative Divertor Configurations

On TCV and DIII-D, negative-triangularity (NT) plasmas have demonstrated good, reactor-relevant energy confinement while remaining in L-mode, thereby avoiding major H-mode power-exhaust issues such as ELMs and the L–H power threshold. On TCV, Ohmic studies found that NT plasmas were harder to detach than similar positive-triangularity (PT) plasmas [1], ascribed, in part, to a lower scrape-off layer (SOL) width [2]. NT Ohmic experiments on the same machine also showed that Alternative Divertor Configurations (ADCs) enhanced detachment processes, particularly with the formation of an X-point radiator [3]. Detachment of all targets was achieved at core line-averaged densities, where, with a conventional divertor, only outer target cooling is observed. In this contribution, we extend the NT power exhaust studies to integrated scenarios with auxiliary heating, closer to being reactor-relevant, focusing on divertor-core coupling. Specifically, this work focuses on the Snowflake (SF), a divertor with a secondary X-point and the X-divertor (XD), which has increased poloidal flux expansion. The combination of ADCs and negative non-X-point triangularity (δNXP) led to a L-mode scenario at high input power featuring H-mode-like performance (H98>1) together with a detached divertor. At δNXP~-0.3, an SF scenario with 550 kW of NBI heating and an average core density of 4.5×10^19 m-3 (fG ~0.4) has been developed. In stationary conditions, βN ≈ 1.6 and H98 reaches 1.4, approximately 10% higher than a TCV positive triangularity (PT) H-mode reference with 1.3MW NBI. Nitrogen seeding led to the formation of an X-point radiator and to the detachment of the outer targets. At similar core-performances, the NBI-heated PT ELMy H-mode achieved detachment only between ELM burn-throughs. In the NT-SF scenario, seeding reduced the core performances (H98 and βN) by ~10%, correlated with core dilution, with Zeff increasing from 2 to 2.4. In comparison, the PT reference scenario’s performances are affected by ~15%. Further use of beta-control allowed for mitigating the effect of seeding on the core performances while maintaining a detached divertor. Experimental insights of these findings, enabled by TCV's extensive diagnostic coverage, are presented, contributing to understanding the conditions in which NT can be detached with a reactor-relevant core.

[1] O Février et al 2024 Plasma Phys.Control.Fusion 66 065005
[2] R. Morgan et al 2025 Nucl.Fusion 65 106030
[3] G. Durr-Legoupil-Nicoud et al, in prep

Speaker: Garance Durr-Legoupil-Nicoud (EPFL)

Invited Talk: Morning session

362 I22 Physics basis of stationary power exhaust in the X-Point Target divertor configuration

This work investigates stationary power exhaust in the X-Point Target (XPT) divertor, combining experiments in the TCV tokamak with SOLPS-ITER simulations. Power exhaust is a major challenge for magnetic confinement fusion: future reactors will face intense heat fluxes channelled through a narrow scrape-off layer onto divertor targets, exceeding material tolerances if unmitigated. Detached operation will therefore be essential, requiring divertor concepts that dissipate power and momentum via impurity radiation and plasma–neutral interactions while preserving core performance.
The XPT introduces a secondary X-point in front of the divertor target, splitting the divertor leg into two branches. It is predicted to reduce target loads and broaden the detachment window relative to a Lower Single-Null (LSN) configuration. Recent TCV results confirmed this improved XPT performance up to Lengyel metric levels, characterising the detachment challenge, relevant to future reactors. However, extrapolation and magnetic design optimization require understanding the mechanisms driving the XPT advantage and validating modelling tools.
Ohmic L-mode density ramps and nitrogen-seeded experiments with auxiliary heating in TCV are compared against SOLPS-ITER simulations. These simulations include realistic geometry, multiple impurities, kinetic neutrals, drifts, currents and improved sheath conditions, using an optimized setup validated across several TCV divertor configurations.
The simulations reproduce the reduced peak target power loads in the XPT relative to the LSN, while radiated power remains similar, indicating a dominant role of non-radiative processes, particularly radial transport. Radial transport in the XPT is dominated by macroscopic drifts due to two geometric features near the secondary X-point: nearly toroidal field lines that enhance sensitivity to cross-field transport and a low-potential secondary private flux region generating an ExB convective cell. This causes a strong particle redistribution, consistent with spectroscopy, and reduce target density in one branch depending on toroidal field direction. In simulations this yields a higher peak target temperature in that branch compared to the LSN, a trend not observed experimentally. Code experiments explore this discrepancy by indirectly testing kinetic effects, turbulent transport and performing geometry scans to account for reconstruction errors. The latter reveals strong sensitivity, within fractions of a power decay length, to the separation between the main and secondary separatrices. The simulated temperature rise is mitigated by impurity seeding, consistent with experiments, and correlates with a distinct ionization pattern that allows seeded species to better access high-power flux tubes in the XPT.
These results advance the understanding of the XPT physics and assess SOL modelling capabilities and remaining challenges.

Speaker: Massimo Carpita (SPC-EPFL)

09:30 Coffee Break

Postersession 4: Tracks F, G and I

363 4.072 Liquid metal erosion module for SOLPS-ITER code and its application to T-15MD lithium divertor simulations

Self-replenishing liquid metal coatings of the divertor targets are sometimes suggested as a possible solution to the power exhaust problem of future fusion tokamak-reactors. Lithium (Li), tin (Sn) and their alloys are the most promising candidates. Simulation of scrape-of-layer (SOL) plasmas in such configurations is a necessary step towards development of the corresponding divertor design.

We have implemented a new liquid metal erosion module in SOLPS 5.2 (SOLPS-ITER) code. The physical model is the same as reported in [1,2] for our SOLPS 4.3 implementation. All essential processes, namely, physical sputtering, thermal sputtering, evaporation, and prompt redeposition are taken into account. In contrast to the other SOLPS-ITER implementation reported in [3], where fluid approximation was used for simulation of the eroded material flow, we use a complete kinetic description of eroded neutral particles. This difference can be important for both Sn and Li in detached divertor scenarios as well as the erosion of the target regions far from the strike points where ionization of eroded material is not instantaneous.

Compared to the older version [1,2] a new model for prompt redeposition is implemented. Unlike the previous implementation that used Fussmann equation for estimation of the redeposition efficiency, a new equation taking into account the Debye sheath is used [4]. Some technical improvements were also made making the module more universal.

Using the erosion module, we investigate divertor performance and Li flow in SOL of T-15MD tokamak with Li covered divertor targets. For the first time, the simulations are performed including drifts and currents in the SOL. Their effects on Li transport and distribution in the SOL are discussed. We also consider in more detail the forces acting on Li flow. We notice that, in contrast to conventional impurity transport models, Li self-pressure force can be comparable to the friction and thermal forces in regimes with pronounced Li shielding.

[1] E.D. Marenkov et. al, Nucl. Fusion 61 (2021) 034001
[2] E.D. Marenkov et. al., Plasma Physics and Control Fusion, 64 (2022)115006
[3] G.F. Nallo et. al., Nucl. Fusion 62 (2022) 036008
[4] E. D. Marenkov, F. R. Kolesov. Physics of Plasmas, 32 (2025) 5

Speaker: Evgeny Marenkov (MEPhI)

364 4.026 Revisited reaction probabilities for atomic and molecular hydrogen for modelling cold divertor plasmas

The plasma parameters in the divertor region of fusion devices are highly relevant for characterizing particle and power exhaust in attached and detached regimes or during the injection of impurities. Due to reduced temperatures compared to the core plasma, both atomic and molecular hydrogen can be present in the divertor plasma and, depending on the operational scenario, also isotopes like deuterium, tritium or mixed molecular species.

Reaction probabilities for atomic and molecular hydrogen are essential input for understanding the physics of the divertor plasma and for determining its properties, typically based on transport codes like EIRENE or on collisional radiative models. One input of such codes are probabilities of collisional processes. These can be defined by cross sections or rate coefficients where cross sections are preferable as they enable calculations for non-Maxwell energy distribution functions. Input cross sections are needed for an energy range from the excitation threshold up to at least around 100 eV for accurately describing the different divertor regimes. Particularly critical is the energy range close to the threshold as cross sections for these energies can be determined only by comprehensive quantum-mechanical methods.

Several data bases are used since many years for the needed reaction probabilities, for example [1,2]. These data bases are far from being perfect, in particular for low temperatures; often used are rate coefficients resulting from semi-empirical methods or scaling laws and consequently attributed with a high uncertainty. The collisional radiative models Yacora H and Yacora H$_2$ are used for testing and benchmarking a revisited set of input data, based on recently available atomic cross sections calculated using the CCC method [3] and molecular data from the MCCC method [4]. The presentation introduces the set of input data and the benchmark results. Additionally discussed is the current status of developing molecular models for the isotopes of hydrogen. Widely used for determining cross sections for the isotopes are simple scaling rules; replacing these by specific isotope-dependent cross sections is highly desirable in order to reduce the uncertainty of the model results.

[1]: D. Reiter, “The data file AMJUEL: Additional Atomic and Molecular Data for EIRENE”, FZ Jülich, 2000
[2]: R. Janev et al, “Collision Processes in Low-Temperature Hydrogen Plasmas”, FZ Jülich 2003
[3]: I. Bray et al, Comput. Phys. Commun. 85 (1995) 1
[4]: M. Zammit et al, Phys. Rev. A 95 (2017) 022708

Speaker: Dirk Wünderlich (IPP Garching)

365 4.085 EMC3-EIRENE simulations of boron injection effect on heat loads in EAST wall conditioning experiments

Heat loads on divertor plates in EAST H-mode plasma experiments with boron powder injection for real-time wall conditioning has been investigated by three-dimensional (3D) Edge Monte Carlo transport code EMC3-EIRENE. The simulated profiles of electron density and temperature are consistent with experimental measurements by edge reciprocating Langmuir probe installed on EAST with the help of the scanning study of anomalous particle and energy transport coefficients. In order to acquire mitigation of plasma-material interactions, boron powder injected at different poloidal locations has been investigated to evaluate its influence on heat loads on the lower divertor plates. It is found that boron powder injected at the lower inner and outer strike points gives rise to a toroidally asymmetric profile of heat loads on the lower in- and out-board divertor targets, respectively. While the boron powder injected at upstream SOL region leads to a symmetric distribution of heat loads on the lower divertor targets. The deposition pattern of boron ions shows a sector-like structure on the lower divertor targets of EAST with the downstream boron injection, whereas the toroidal uniform and lobe-like profiles of heat loads are achieved for boron injection at upstream SOL region. The 3D effects of the boron radiation on the heat loads distribution have been performed by using field line tracing technique, which indicates that boron impurity injected at the lower strike points can radiate more power and result in a lower heat loads distribution compared with boron injected at upstream SOL region.

Speaker: Mr Jiashuo Liang (Northeast Agricultural University)

366 4.087 FORGE – A tool for the optimisation of divertor magnetic geometries by simulated annealing

Advanced divertor configurations (ADCs) are a promising pathway towards a reactor-relevant exhaust solution in tokamaks. These configurations are characterised by a number of features of the divertor’s magnetic geometry, such as an increased poloidal connection length, enhanced flux surface flaring and the presence of additional x-points in the divertor region. ADCs have been posited to improve detachment access and stability [1], with these configurations having been studied on a number of devices [2,3]. Recent experiments on the medium-sized tokamak MAST-U have observed the benefits to detachment in ADCs [4].

To produce ADCs, a set of poloidal field (PF) coils situated close to the divertor region are typically required. When projecting towards a power-producing reactor, this poses a design integration challenge. Such coils require space for mechanical supports, cooling, and radiation shielding, with space at a premium in reactor class devices. In order to investigate the feasibility of producing ADCs using the smallest possible number of divertor shaping coils, a new code, FORGE (FORGE Optimises Reactor Geometries to improve Exhaust), has been created.

FORGE adjusts the currents in a given PF coil set in order to minimise a cost function related to both the engineering cost of the coils, as well as features of the magnetic geometry of the divertor such as the resultant connection length. This optimisation is carried out using a simulated annealing approach, which promotes an efficient exploration of the space of possible divertor configurations. This exploration is carried out in such a way as to not require the Grad-Shafranov equation to be resolved each time a new configuration is produced. In doing so, FORGE is able to maintain the magnetic geometry of the core plasma from an initial given equilibrium, enabling a decoupling of the design of the divertor’s magnetic geometry from the production of a desired core plasma scenario.

[1] C. Cowley et al 2022 Nucl. Fusion 62 086046
[2] C. Theiler et al 2017 Nucl. Fusion 57 072008
[3] J R Harrison et al 2024 Plasma Phys. Control. Fusion 66 065019
[4] Verhaegh et al., Commun. Phys. 8, 215 (2025)

Speaker: Mr Chris Marsden (Tokamak Energy Ltd.)

367 4.027 Sheath-driven redeposition of sputtered and evaporated impurities in Liquid Metal based plasma-facing components

Plasma-facing components (PFCs) primarily made of tungsten face significant lifetime limitations due to neutron embrittlement, dust formation, local melting and cracking, which threatens the high availability required for future fusion power plants [1,2]. Liquid-metal (LM) based PFCs have emerged as a promising route to overcome these limitations thanks to their intrinsic replenishment capability, providing so-called self-healing behaviour. Two main LM concepts have been explored since the 1970s: (i) direct flowing LM along plasma-facing walls [3], particularly in the divertor, and (ii) Capillary Porous Systems (CPS), where LM is retained inside a porous network by capillary forces to prevent splashing [4]. Both approaches rely on low-melting-point metals such as Li or Sn, while CPS additionally requires a refractory matrix, typically tungsten.
However, plasma–surface interactions may still lead to sputtering or evaporation of both low-Z and high-Z species, resulting in possible plasma contamination. The net impact depends on the fraction of emitted impurities that are redeposited on the surface, a process strongly influenced by the grazing magnetic field and the electric field structure within the sheath [5]. Quantifying this redeposition fraction, and distinguishing the physics associated with sputtering-driven versus evaporation-driven emission, is therefore essential for assessing LM-based PFC viability.
In this work, different edge-plasma conditions in density and temperature are simulated using a home-developed 1D–3V Particle-in-Cell code that self-consistently computes the electric potential in the sheath and pre-sheath. Impurity test particles (W, Sn, Li) are injected at the wall and their trajectories are tracked until redeposition on the emitting surface or loss toward the plasma core.
The simulations enable a systematic comparison of redeposition rates as a function of (i) the emission process—sputtering versus evaporation, which produce different initial velocity distributions and therefore ionization mean free paths, (ii) plasma conditions, (iii) incident particle energy, and (iv) the ion species responsible for sputtering. This analysis will make it possible to identify the combinations of plasma edge parameters and emission mechanisms that favour impurity return to the surface rather than escape, thereby helping to determine the operating windows in which LM-based divertor components can function without unacceptable plasma contamination.

[1] R. Pitts et al, Nucl. Mater. Energy 20, 100696 (2019)
[2] Y Corre, the WEST team et al. Phys. Scr. 96, 124057 (2021)
[3] L.E. Zakharov, Phys. Rev. Lett. 90, 045001 (2003)
[4] J.G.A. Scholte et al, Nucl. Mater. Energy 37, 101522 (2023)
[5] RJ. Guterl et al, Nucl. Mater. Energy 47, 100948 (2021)

Speaker: Jerome Moritz (Université de Lorraine, Institut Jean Lamour, UMR 7198 CNRS, Campus Artem, 2 allée André Guinier, 54011 Nancy, France)

368 4.073 Validation of nitrogen-seeded XPR simulations and comparison with other impurities on ASDEX Upgrade using JOREK

The X-Point Radiator (XPR) regime is a promising exhaust solution for future large-size tokamaks, featuring a cold, dense, highly radiative region above the X-point, inside the confined region. Such regimes have been achieved experimentally on several tokamaks, with different seeded impurities, and ELM suppression is seen when the XPR reaches a threshold height [1]. Modeling with SOLPS-ITER [2], JINTRAC [3], and GRILLIX [4] has successfully reproduced several features seen in experiments. However, existing simulations have not been able to self-consistently study the dynamical interplay between the XPR and the pedestal which results in ELM suppression. For this an MHD code with realistic scrape-of layer (SOL) physics is needed.

In this contribution, we present the current status of modeling of the XPR regime on the ASDEX Upgrade (AUG) tokamak with the visco-resistive nonlinear MHD code JOREK [5]. The longer-term objective of this work is to extend the modeling to ITER provided the AUG validation is favourable. A two-way coupled particle-in-cell model treats neutrals and impurities kinetically, including their interactions with the background fluid plasma [6,7].
Recently this coupling has been extended to the reduced MHD, two-temperature model of JOREK, allowing for benchmarking with other codes and experimental validation. The validation work focuses on the stationary XPR features and ELM suppression. Nitrogen seeded simulations are compared to SOLPS-ITER and AUG experiments, further modeling is presented using different impurity species such as argon and neon.

[1] M. Bernert, et al Nuclear Materials and Energy 43 (2025): 101916.
[2] O. Pan, et al Nuclear Fusion 63.1 (2022): 016001.
[3] S. Q. Korving, et al EPS (2025).
[4] K. Eder, et al, Nuclear Fusion 65.9 (2025): 096029.
[5] M. Hoelzl, et al Nuclear Fusion 64.11 (2024): 112016.
[6] S. Q. Korving, et al Physics of Plasmas 30.4 (2023).
[7] S. Q. Korving, et al Physics of Plasmas 31.5 (2024).
[8] H. Zohm, et al Nucl. Fusion 64 112001 (2024)

Speaker: Máté Szűcs (Max Planck Institute for Plasma Physics)

369 4.086 Simulation study of SF- divertor particle dispersion and detachment in HL-3 by SOLPS-ITER

The divertor is one of the most crucial components of a tokamak device, serving two primary functions: heat exhaust and helium ash removal. In future fusion reactors, the parallel power flux entering the divertor is expected to exceed 1 GW/m². The divertor is expected to experience extremely high thermal loads in future fusion reactors. Therefore, the development of an advanced divertor is essential to reduce the heat load on the divertor targets. Extensive experimental studies have demonstrated that the snowflake divertor (SF-) can effectively reduce the target heat flux. This reduction can be attributed mainly to two mechanisms: (1) The extremely weak poloidal magnetic field near the second X-point decreases the projection angle between the magnetic field lines and the divertor targets, and (2) The SF- spreads the incident particle flux over a larger area, increasing the plasma–divertor contact surface.
The HL-3 tokamak is capable of flexibly realizing multiple snowflake configurations, including snowflake-plus (SF+) and snowflake-minus (SF−) [1]. Experiments have verified that the SF- effectively disperses the particle flux and reduces the local heat load on the divertor targets. However, HL-3 experimental results also showed that although the open SF-divertor provides strong flux spreading, it exhibits relatively weak particle recycling and low neutral pressure, which are unfavorable for achieving both detachment and effective pumping [2]. Thus, in this work, based on HL-3 SF- experiments, we employed SOLPS-ITER simulations to investigate the physics of the open SF-divertor. The main topics include [3]: (a) the mechanism of target particle flux double peaks and particle flux spreading, and (b) the impact of drifts on particle transport and detachment behavior for SF-. This work provides a solid theoretical foundation for the application of the SF- divertor in future reactors.

Speaker: Hailong Du (GNOI)

370 4.009 Simulation study of the influence on pedestal instabilities due to impurities seeded by supersonic molecular beam injection

Impurity injection is an effective way to control the burst of edge localized modes (ELMs), which have to be avoided due to the severe damages on the plasma-facing components in future fusion reactor. In recent EAST experiments, it has been demonstrated that, large ELMs can be mitigated or even suppressed by neon (Ne) seeding utilizing supersonic molecular beam injection (SMBI) when the Ne density reaches a certain level in the pedestal region. Cao et al. [Plasma Sci. Technol. 27 (2025) 115105] have attributed the ELM suppression to the open of particle transport channel. In this work, to investigate the underlying mechanism on the enhanced particle transport due to the Ne injection via SMBI, simulations are performed using the CLT (Ci-Liu-Ti, the Chinese name of the magnetohydrodynamics (MHD)) code.

CLT code is widely adopted to study the MHD instabilities such as tearing modes, kink modes and so on. For investigating the pedestal instabilities, the ion diamagnetic drift effect is included in the physical model. Furthermore, to simulate the process of Ne SMBI injection, an impurity injection module is developed. Based on the EAST-like equilibrium generated by the high-accuracy free-boundary equilibrium solver CLT-EQ (CLT-EQuilirium), the CLT simulations are performed with the Ne injection from the outer mid-plane. It is found that, due to the Ne injection, a n = 1 mode can be triggered in the pedestal region, which leads a significant enhancement of the plasma transport. The pressure gradient and current density in the pedestal region are thus reduced, causing a reduction of ELM size by around 30%. The n = 1 mode is considered as the tearing mode according to the analysis of the magnetic structure and drift frequency. By scanning the Ne injection flux, it’s found that ELMs can be mitigated when the maximum Ne concentration in the pedestal region reaches ~5%. In summary, the results of CLT simulations are qualitatively consistent with the EAST experimental observation with Ne SMBI injection, and the enhancement of plasma transport and further mitigation of ELMs could be attributed to the trigger of the n = 1 tearing mode in the pedestal region due to impurity injection.

Speaker: Yuchen Xu (University of Science and Technology of China)

371 4.010 Development of the three-dimensional simulation of scrape-off layer plasma transport code using finite volume method

It is critical to control the heat load onto the divertor target for the future fusion reactor. While the resonant magnetic perturbation (RMP) is proved to be an effective way to control the transient heat load due to the burst of edge-localized modes (ELMs) [1], the distribution of the inter-ELM divertor heat load will be significantly affected simultaneously [2] due to the three-dimensional (3D) magnetic perturbations. To understand the effect 3D fields on the transport of scrape-off layer (SOL) plasma, the 3D code EMC3-EIRENE [3] based on Monte Carlo method is widely used. On the other hand, pronounced effects on the SOL plasma have been found due to the drifts, which can be well simulated in the 2D code SOLPS-ITER [4] based on the finite volume method. To simulate the SOL plasma with both effects of 3D field and drifts, one possible solution is to develop a 3D code based on finite volume method and find a proper way to introduce the 3D field effect.

In this work, we will report our progress on the development of the 3D SOL plasma transport code based on the finite-volume method. According to the 3D fluid equations of continuity, parallel momentum and ion and electron internal energy, which are derived in a similar way of SOLPS-ITER equations without the restriction of toroidal symmetry, the 3D code is developed using C++ based on the finite volume method. The numerical calculations are performed on the collocated grid, where the convection term is discretized using a hybrid scheme, the diffusion term is discretized using a central difference scheme, and the time differential term is discretized using a first-order fully implicit scheme. The 3D code is preliminarily tested based on an RMP experiment of EAST. The simulation results showed that additional particle flux peaks appeared on the divertor target plate, which reflects the 3D field effect reasonably.
[1] T.E. Evans, et al., Nat. Phys. 2 (2006) 419-423.
[2] J-W. Ahn, et al., Plasma Phys. Control. Fusion 59 (2017) 084002.
[3] Y. Feng, et al., Contrib. Plasma Phys. 44 (2004) 57–69.
[4] X. Bonnin, et al., Plasma Fusion Res. 11 (2016) 1403102.

Speaker: Jiafeng He (University of Science and Technology of China)

372 4.011 Simulation study of the direct impurity effect on the ELM dynamics using BOUT++ six-field two-fluid model

The huge heat load onto divertor is a crucial issue in fusion reactor. While the injection of radiative impurity is necessary for promoting divertor detachment, it can also affect ELM behaviors essentially [1]. While the effects of impurities on the linear instability, such as radiative cooling, profile regulation and fuel dilution, have been studied, there is still a lack of the comprehensive understanding for the complicated ELM behaviors observed in experiments [2, 3].
In this work, the dynamical effect of impurities on pedestal stability, ELM evolution and turbulence transport is systematically studied of using BOUT++ six-field two-fluid module. Different from the general BOUT++ simulations, the evolution of toroidal axisymmetric (n = 0) parallel current is included [4]. It is found that the magnetic flutter particle flux is overestimated without self-consistent evolution of the n = 0 component. Furthermore, with evolving n = 0 parallel current, the amplitude of parallel current perturbation is significantly reduced and the current sheet formed during the nonlinear ELM simulation can be dissipated self-consistently by electron momentum transport without additional hyper-resistivity [5].
Combined with the impurity model developed by Li et al. [6], the simulations with direct impurity effect show that, ELM suppression can be achieved only when including the effect on vorticity distribution. It is also found there is an impurity density “threshold” for ELM suppression, which is qualitatively consistent with experiment findings [3]. The impurity effect on the ELM dynamic process is further explored by varying both the impurity mass ratio and the pedestal electron temperature. According to the difference in the “two-stage burst” feature [4] (the first stage is related to peeling-ballooning modes and the second stage is related to the nonlinear triggered drift-tearing mode [7]), the two-dimensional parametric plane can be divided into four regions, and the different effect on the ELM evolution can be understood based on the shift of operation point between different regions.

[1] G.S. Xu et al., Rev. Modern Plasma Phys. 7 (2023) 14
[2] X. Lin et al., Phys. Lett. A 431 (2022) 127988
[3] G.L. Xiao et al., Nucl. Fusion 61 (2021) 116011
[4] H. Seto, Physics of Plasmas 31 (2024) 032513
[5] P.W. Xi et al., Phys. Plasmas 21 (2014) 056110
[6] Z. Li, et al. Nuclear Fusion 62 (2022) 076033
[7] A. B. Hassam, Physics of Fluids 23 (1980) 2493

Speaker: Taihao Huang (University of science and technology of China)

373 4.012 SOLPS-ITER as open-source software

The SOLPS-ITER code suite [1,2] is one of the main tools used worldwide for tokamak edge modelling. Launched in 2015 and maintained and distributed by the ITER Organization (IO), it was recently extended to a “Wide Grids” version [3], allowing the plasma solver computational domain to extend to the full vacuum vessel and provide a much more detailed description of the interactions between the SOL plasma and the first wall. It remains the main workhorse for ITER design and operation planning activities, in addition to its use in the ITER Member States for simulating existing experiments and predicting future device performance.
To accompany the development of the private fusion ventures landscape, the IO has triggered a large effort to make as many of its software and knowledge sources publicly available as possible. Among these, of course, is SOLPS-ITER. It is thus now available as open-source software, on GitHub, along with its underlying components and some library dependencies that were not previously available. These include the B2.5 plasma fluid solver, the Eirene Monte-Carlo neutral kinetic solver [4], the Carre grid generator [5], the Uinp case build-up tool, and many analysis and post-processing visualization tools, along with the support libraries GR for graphics and MSCL for input data file handling.
This presentation will provide details on what features are available as part of the open-source distribution, how to make the best use of them and how to contribute and participate in further development of the code. The open-source code is benchmarked against an ITER throughput scan and other reference cases from the literature.

[1] S. Wiesen et al., J. Nucl. Mater. 463 (2015) 480.
[2] X. Bonnin et al., Plasma Fusion Res. 11 (2016) 1403102.
[3] W. Dekeyser et al., Nuclear Materials and Energy 27 (2021) 100999.
[4] D. Reiter et al., Fus. Sci. Technol. 47 (2005) 172.
[5] R. Marchand et al., Computer Phys. Comm. 96 (1996) 232

Speaker: Xavier Bonnin (ITER Organization)

374 4.013 Combined effects of plasma current and impurity dynamics on pre - access and stability divertor parameters for XPR in WEST

The X-Point Radiator (XPR) regime provides a promising avenue for controlled detachment upstream of the divertor targets, enhancing radiative cooling and reducing divertor heat fluxes [Loarte et al., 1998]. Its stability and controllability depend on scenario parameters, including the safety factor $q_s$ and thereby the plasma current $I_p$. Since next-step fusion devices are expected to operate at higher plasma currents, a deeper understanding of the impact of $I_p$ on pre-detachment target conditions is essential for assessing plasma sensitivity to radiation and for optimizing core performance.

A preliminary study on the role of $I_p$ in the access and stabilization of pre-detachment conditions in the WEST tokamak—based on simulations with and without drifts has already demonstrated its influence on divertor conditions. In the absence of drifts, purely geometrical considerations linked to the connection length confirm that increasing $I_p$ hinders detachment, leading to a hotter divertor with higher heat fluxes concentrated in a narrower SOL. In contrast, when drifts are included, the same increase in current results in divertor cooling and heat-flux redistribution. Under these conditions, the adverse effects of higher plasma current appear mitigated, suggesting potentially beneficial implications for next-step devices.

Using SOLEDGE3X-EIRENE simulations with nitrogen seeding (motivated by WEST observations where nitrogen injection triggers XPR formation [Rivals et al., NME 2024]), we assess how impurity radiation modifies the $I_p$ dependence of XPR access and stability, and also compare our findings with WEST’s extensive experimental database.

Acknowledgements
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, under award(s) DE-SC0023100

Speaker: Virginia Quadri (University of Tennessee Knoxville)

375 4.014 Feasibility study of the hydrogen recombination front measurement by analyzing the Zeeman effect on the chord-integrated Ar I line spectrum at 912 and 965 nm in JT-60SA

In this study, we develop a diagnostic technique to detect detachment onset. The technique utilizes a characteristic feature of the atomic line emission profile in the detached plasma, namely the emergence of atomic line emission peaks near the hydrogen recombination front. Our aim is to establish a simple and robust technique suitable for future fusion reactors. The recombination front is defined as the position where the hydrogen recombination flux reaches its maximum, and the electron density also peaks at this position. Consequently, the intensities of hydrogen and impurity atomic emission lines also exhibit local maxima. Our concept is to measure these peak positions with Zeeman spectroscopy, in which the line emission peak position along a given chord can be measured from the Zeeman splitting observed in the chord-integrated spectra. The use of near-infrared emission lines enhances the relative magnitude of the Zeeman splitting compared with the Doppler broadening [1, 2].

Numerical analyses were performed on the detached plasma data with Ar and Ne injection for JT-60SA calculated using the integrated divertor code SONIC [3]. Our previous work demonstrated that detachment onset can be detected with the deuterium Paschen-α line at 1875 nm. In the present study, we examine the feasibility of applying this technique to two relatively bright Ar I lines at 912 and 965 nm. Compared with Paschen-α, these Ar I lines have narrower line widths due to smaller Stark and Doppler broadenings, which enables clearer observation of the Zeeman splitting. A virtual viewing chord which is nearly parallel to the outer separatrix and passes through the hydrogen recombination front was assumed. Chord-integrated spectra were synthesized using excited argon atom densities calculated with a collisional–radiative model [4, 5], considering both Zeeman splitting and Doppler broadening. From the synthesized spectra, we evaluated the line emission peak positions and found that they are close to the recombination front.
[1] T. Chatani, et al., Sci. Rep. 12, 15567 (2022).
[2] M. Murakumo, et al., Rev. Sci. Instrum. 96, 043501 (2025).
[3] K. Shimizu, et al., Nucl. Fusion, 49, 065028 (2009).
[4] J. Vlcek, J. Phys. D 22, 623 (1989).
[5] H. Akatsuka, Adv. Phys. X 4, 257 (2019).

Speaker: Tsunehiro Morita (Kyoto University)

376 4.015 Electron density re-increase associated with a recombination-process transition during neutral gas injection in a GAMMA 10/PDX divertor simulated plasma

Injecting impurity gases into the divertor region to enhance volumetric recombination and form detached plasmas is a promising approach to mitigate intense divertor heat loads. Understanding the underlying physics of divertor detachment is important for controlling heat and particle fluxes in the divertor while maintaining core plasma performance.
Molecular activated recombination (MAR) has relatively high reaction rates even at moderately high electron temperatures (Te). In contrast, electron–ion recombination (EIR) becomes dominant at Te ≪ 1 eV, resulting in a shift of the dominant recombination process along with the decrease in Te [1, 2].
So far, we have utilized the divertor simulation experimental module (D-module) of a tandem-mirror, GAMMA 10/PDX, and observed two phases of emission patterns during the formation of a detached plasma by injecting molecular hydrogen into the end-loss plasma flowing into the V-shaped tungsten target plates of the D-module [3].
As the neutral gas pressure (pn) in the D-module increased, Te decreased; a strong Hα emission associated with MAR was observed [4], and the emission region moved upstream as the electron density (ne) decreased. A further increase in pn (> ~ 5 Pa) led to the second phase, which is characterized by a drastic rise in emissions from hydrogen atoms in higher excited states, indicating the occurrence of EIR. At that point, Te was calculated to have dropped to ~ 0.1 eV using the Boltzmann plot method [3]. Notably, during this second phase, despite the low Te, a sharp increase in ne up to ~ 5 x 10^18 m^-3 was measured by the microwave interferometer, which then gradually decreased until the end of the plasma discharge.
In the presentation, the mechanism of the shift in recombination processes from MAR to EIR and the re-increase in ne will be discussed by comparing the results of molecular and atomic emissions obtained by a spectrometer with multiple lines of sight, complemented by newly implemented Mach-probe measurements between the V-shaped target plates.
This work was partly supported by JST SPRING Grant Number JPMJSP2124, JSPS KAKENHI Grant Numbers 22H01198, 23K22469, and NIFS Collaboration Research program (NIFS23KUGM174, NIFS23KUGM186, NIFS25KFFT001).
[1] K. Verhaegh et al., Nucl. Fusion 63 (2023) 016014.
[2] J. Shi et al., Physica Scripta 98 (2023) 115605.
[3] S. Takahashi et al., Nuclear Materials and Energy 43 (2025) 101945.
[4] M. Sakamoto et al., Nuclear Materials and Energy 12 (2017) 1004–1009.

Speaker: Mr Satoshi TAKAHASHI (Plasma Research Center, University of Tsukuba)

377 4.016 Impact of the Lyman alpha discrepancy on He EDGE2D-EIRENE simulations

An understanding of the behaviour of the D or He fuel used in tokamak discharges is necessary for modelling edge and divertor transport. Despite there being well-established models describing the emission from the H-like fuel, poor agreement is found between JET line-of-sight measurements and Collisional-Radiative (CR) models used to predict their line intensities. Lawson et al. (2024) compare measurements of He II intensities with the ADAS and CHEM CR models and find that the modelled Lyman alpha intensity is underestimated (x3-5) compared with the other Lyman intensities. Similar discrepancies are found in the majority of H and D discharges. Although allowance can be made for different temperature regions falling within the spectrometer's line-of-sight, it is nevertheless supposed that the dominant emission for all lines in the Lyman series originates from the same spatial location. However, evidence from H-fuelled discharges suggests different spatial locations for the Lyman alpha emission as opposed to the higher series members and agreement is obtained for the He II measurements when two distinct emission regions are supposed, a higher temperature region corresponding to the ionization front and a low temperature region dominated by recombination.
A long-standing discrepancy between measurements of the divertor radiated power and that predicted by transport simulations has been noted by among others Groth et al. (2013), Jarvinen et al. (2015) and, most recently, by Rees et al. (2026). The simulated radiated powers underestimate the measurements by typically ~x2. Exploratory EDGE2D-EIRENE simulations of a He-fuelled discharge are used to investigate applying constraints to the simulations in order to reproduce the two observed emission regions. The simulations are particularly sensitive to the electron power loss term (Lawson et al., 2018), which includes the losses due to radiation, and additional constraints are applied by varying the magnitude of this term. He II is ideal for this study in that the intensity ratios are reproducible, with near-constant line intensity ratios (Lawson et al., 2024).
Groth M et al. 2013 Nucl. Fusion, 53, 093016
Jarvinen A et al. 2015 J. Nucl. Mater., 463, 135
Lawson K D et al., 2018, EPS conf., Prague
Lawson K D et al., 2024, PPCF, 66, 115001
Rees D et al., 2026, This conference

Speaker: Kerry Lawson (UKAEA)

378 4.017 Describing the island divertor physics via simplified models

The island divertor is the leading exhaust concept in stellarators. Used in W7-X, it has proven to provide density control, stable detachment, and impurity screening [1]. Although a regime with enhanced recycling has been observed in W7-X, moderate neutral pressures of up to 0.18 Pa have been measured so far [2]. Higher recycling conditions are necessary to ensure the reactor-relevance of the island divertor in terms of pumping efficiency.
Simplified models are used in this work to understand the leading order physics and investigate the recycling behavior in simplified island divertor geometries. One of these simplified models, the two-point model, has been found to successfully represent the tokamak SOL [3]. This model has been extended into a stellarator two-point model (STPM) by Feng [4]. We show that the model predicts a detrimental "diffusion-limited" transport regime which suppresses high recycling. This contribution extends the STPM to include correction factors: a convected power fraction, a target-localized dissipated power fraction, and a more general parametrization of the momentum loss factor.
We validate the STPM against EMC3-Eirene simulations, concentrating on the island power-carrying-layer. We elucidate the importance of the correction factors and show that when extracted from the simulations, a reasonable agreement is found between both models. This provides the ability to do two-point-model formatting [3] which allows to gain significant physics insight. We show how the volumetric losses and the diffusion-limited regime interact to limit recycling and identify a threshold parameter to avoid this detrimental transport regime.

We also show the impact of typical island 3D features (target shadow region TSR) leading to non monotonic parallel teperature profiles and strongly localized presure variations in the divertor interaction region.

[1] M Jakubowski et al 2021 Nucl. Fusion 61 106003
[2] V Haak et al 2023 Plasma Phys. Control. Fusion 65 055024
[3] D Moulton et al 2017 Plasma Phys. Control. Fusion 59 065011
[4] Y Feng et al 2006 Nucl. Fusion 46 807

Speaker: Nassim Maaziz

379 4.018 Effect of magnetic field gradient on detached plasma formation in a linear diverter simulator TPDsheet-ICR

In the DEMO reactor, the enormous heat flux released from the core to the divertor plasma reaches several times that of ITER. Therefore, research is being conducted on advanced divertors aimed at reducing heat flux by increasing the plasma wetted area under divergent magnetic field and generating detached plasma. It is extremely important to investigate the effect of plasma flow acceleration due to magnetic field gradients on the formation of detached plasma using advanced divertors with divergent magnetic field (magnetic nozzles). The present study elucidates the formation process of attached plasma and detached plasma when the magnetic field gradient near the end-target is varied. Our research group, previously, developed a linear divertor simulator (TPDsheet-ICR) that applies ion cyclotron resonance heating (ICRH) to a high-density sheet plasma (approximately 10¹⁹ m⁻³) to increase the ion temperature. Utilizing this divertor simulator, we have successfully observed the process of transitioning from attached to detached plasma in experimental conditions [1,2]. In this experiment, a high-speed camera with an Arbba-prism was used to measure the two-dimensional emission intensity distribution in the Balmer series (Hα, Hγ) and the Fulcher-band wavelength range (600–640 nm) emitted from the plasma in the divergent magnetic field region. The plasma electron density and temperature were measured using a Langmuir probe. The experimental results revealed that, in the configuration of the divergent magnetic field, an increase in the magnetic field gradient resulted in an increase in electron temperature and caused the transition from attached plasma to detached plasma to occur at higher gas pressures. Furthermore, an increase in the ICR power resulted in an elevated electron temperature, thereby inducing the transition from a detached to an attached plasma. The results of the two-dimensional emission intensity distribution of the Balmer series and the Fulcher-band, along with the electron temperature, clarified the effect of the magnetic field gradient on detached plasma formation.
[1] A.Tonegawa,et al,Fusion Eng. & Des,203(2024)114441.
[2] A.Tonegawa,et al,Nucl.Materials & Energy,41(2024)101802.

Speaker: Akira TONEGAWA (GNOI)

380 4.019 UEDGE simulations with drifts of snowflake divertor experiments at MAST-U

A snowflake divertor (SFD) is formed by bringing a second X-point into the vicinity of the first, resulting in four divertor legs instead of two. SFD experiments have provided evidence of enhanced transport across the region close to the X-points, resulting in redistributed exhaust power across divertor legs and reduced peak heat fluxes at the targets [1]. Two leading candidates for this transport are 1) the ‘churning mode’, a loss of toroidal equilibrium close to the X-points leading to plasma convection in the poloidal plane [2], and 2) strong ExB drifts across the X-points driven by large poloidal gradients [3]. Our previous work has focused on the churning mode [4], while here we focus on the effect of drifts.

NBI-heated, H-mode SFD experiments at MAST-U have been modelled with the edge transport code UEDGE, with and without self-consistent electromagnetic drift effects included. The secondary X-point was around 10cm from the first in these experiments, and its relative position was varied to capture the three unique topologies of SFD: in the private flux region, on the high-field side SOL, and on the low-field side SOL. The experiments featured both upper and lower divertors and so four X-points in total, but UEDGE can only simulate up to two X-points. To overcome this, we simulate a lower-divertor-only scenario by stitching together flux surfaces at the midplane. We can then reverse the toroidal field direction to approximate conditions in the upper divertor.

Anomalous transport coefficients $D_n$ and $\chi_{e,i}$ are tuned in UEDGE to match experimental measurements, and we find that simulations with drifts require around 50% lower $D_n$ and $\chi_{e,i}$. In the lower divertor, ExB drifts are responsible for reduced peak heat fluxes at the primary strike points (SPs) by 20-40% and increased activation of the secondary SPs. In the upper divertor we find that the drift-driven transport across the X-points is reduced. By varying the core heating power boundary conditions, we find that the effect of drifts on transport to the secondary SPs decreases at higher heating, in contrast with predictions of the churning mode.

This work was carried out under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC5207NA27344.

[1] Ryutov & Soukhanovskii, PoP 22 (2015)
[2] Ryutov et al., Physica Scripta 89 (2014)
[3] Canal et al., Nuclear Fusion 55 (2013)
[4] Power et al., Physics of Plasmas 32 (2025)

Speaker: Vsevolod Soukhanovskii (LLNL)

381 4.020 Analysis and validation of k-model for anomalous transport with data from the TCV-X23 discharge

The ad-hoc description of anomalous transport remains a dominant uncertainty in mean-field plasma edge modeling. In this contribution, we further develop an improved anomalous transport model based on self-consistent time-averaging of the turbulence equations, assuming electrostatic interchange turbulence [1]. This model solves an additional transport equation for the turbulent kinetic energy k, and computes anomalous diffusivities self-consistently from its solution. An analytically exact source of k inherently introduces ballooning in the model. The plasma conductivity leads to fast parallel transport of k. We further refine the model by time-averaging the nonlinear sheath conditions: electron current fluctuations, determined by electron temperature and electric potential fluctuations, impact the average fluxes through the sheath in the mean-field model.
To validate the model, we perform SOLPS-ITER simulations [2] of the TCV-X23 discharge [3] with fully extended grids [4] built with the GOAT grid generator [5]. A D-only plasma is assumed. We include drifts and use the advanced fluid neutral model [6]. We compare the performance of the new k-model with the standard approach using fixed transport coefficients.
Both models are calibrated to the experimental data through parameter optimization [7], and reproduce the midplane and target data with similar accuracy. Despite the ballooning source of k, the transport coefficients in the near SOL are fairly uniform for this case, as a result of fast parallel transport of k. The model does predict a significant increase in k, and hence perpendicular transport, towards the low-field side, qualitatively consistent with a filamentary transport picture. As a result, main chamber loads predicted by both models differ significantly, even for nearly equal upstream and target profiles.The self-consistent transport coefficient variation of the k-model leads to somewhat earlier transition into detachment in a density scan, and more rapid increase in decay lengths. We demonstrate the impact of the averaged sheath boundary conditions on target and wall fluxes.
[1] R. Coosemans et al., J. Plasma Phys. 90 (2024) 905900202.
[2] S. Wiesen et al., Journal of Nuclear Materials 463 (2015) 480–484; X. Bonnin et al., Plasma and Fusion Research, 11 (2016) 1403102.
[3] https://gitlab.eufus.psnc.pl/tsvv3/tcvx23
[4] W. Dekeyser et al., Nuclear Materials and Energy 27 (2021) 100999.
[5] S. Van den Kerkhof et al., submitted to Contrib. Plasma Phys.
[6] N. Horsten et al., Nucl. Fusion 57 (2017) 116043.
[7] S. Carli et al., Contrib. Plasma Phys. (2021)e202100184.

Speaker: Wouter Dekeyser (KU Leuven, Department of Mechanical Engineering)

382 4.021 SOL Width Dependence on Separatrix Parameters in ST40 and NSTX

The Scrape-off Layer (SOL) power fall-off length, $\lambda_q$ is used to characterize the divertor heat load in tokamaks and extrapolating how it will behave under reactor relevant conditions is of utmost importance for future fusion reactors. Analyzing the link between $\lambda_q$ and separatrix plasma quantities such as the electron density, n$_{e, sep}$ and temperature, T$_{e, sep}$ has led to a better understanding of the underlying, competing physical processes occurring in the edge plasma that control $\lambda_q$. The SepOS framework [1] is used to analyze data from two spherical tokamaks (ST’s). ST40 is a compact, high magnetic field (B$_T$ ≤ 2.1 T) ST designed, built, and operated by Tokamak Energy Ltd in the United Kingdom [2], and NSTX (B$_{T}$ ≤ 0.55 T) operated by Princeton Plasma Physics Laboratory in the United States. ST40 can operate up to 1 MA of plasma current, I$_p$ and ≤ 1.8 MW of neutral beam injected (NBI) power in mostly disconnected double null (DN), diverted, H and L-mode discharges. NSTX [3] operated with up to 7.5 MW of NBI power in lower single-null (LSN), diverted, H-mode discharges. Analysis of combined datasets from each ST shows a proportional relationship between q and the separatrix electron pressure, P$_{e, sep}$ similar to findings from Brunner [4]. However, because of the strong dependence on T$_{e, sep}$, multiple analysis techniques are used including Spitzer-Harm parallel thermal conductivity [5] in SOL and the reverse two-point modeling [6] to determine T$_{e, sep}$.

This work was supported by the U.S. D.O.E contract DE-AC05-00OR2272.
`
[1] T Eich, et. al. Nucl. Fusion. 60 (2020) 056016
[2] M. Gryaznevich, O Asunta and the Tokamak Energy Ltd Team, Fus. Eng. Design. 123 (Nov 2017) 177-180
[3] M Ono, et. al. Nucl. Fusion. 41 (2001) 1435
[4] D Brunner, et. al. Nucl. Fusion. 58 (2018) 094002
[5] T Eich, et. al. Nucl. Mater. Energy. 42 (2025) 101896
[6] J H Nichols et al, Plasma Phys. Control. Fusion 66 (2024) 045013.

Speaker: Travis Gray (ORNL)

383 4.022 Influence of Near-Surface Reactions of Small-Angle V-Shaped Target on Detached Plasma Formation in the GAMMA 10/PDX Diverter Simulation Module

Detached plasmas are essential for controlling heat and particle fluxes to the plasma-facing components of magnetic fusion devices. We have studied the fundamental processes during detached plasma operations in the divertor simulation experimental module (D-module), using a variable-angle V-shaped target plate at the end-loss region of the tandem mirror plasma device GAMMA 10/PDX [1,2]. Recently, different configurations of diverter shapes have been examined, including compact diverter structures to maintain core plasma volume [3] and methods to spread heat loads using a long-leg diverter [4]. In both cases, the ratio of the target plate area in contact with the plasma and/or the surrounding wall area to the plasma volume is high. Molecular Dynamics simulations have indicated that hydrogen atoms and molecules produced from the diverter plate are excited [5], and, because they are highly reactive, their atomic and molecular reactions near the surface increasingly influence the formation of detached plasma.
So far, we have observed that increasing H₂ pressure expands the hydrogen molecular activated recombination (MAR) area near the target plate. Recently, further increases in gas pressure have led to strong emissions from highly excited states [2]. These findings emphasize the importance of surface reactions, including the mutual neutralization of H₂+ and H, and the formation of vibrationally excited H₂. Therefore, atomic and molecular processes are considered to involve reactions on the target and wall surfaces that generate excited molecules. Plasma-gas-surface interactions are also crucial to the formation of detached plasma. To accurately understand the reaction process, further issues include the need for detailed measurements of the energy states of hydrogen atoms and molecules, especially near the target plate.
In this study, we adjusted the small angle (~15 deg) of the V-shaped target plate to enhance the influence of the surface. We investigated spectroscopic measurements near the V-shaped target, with particular focus on hydrogen molecular emission. The dependence on the distance from the target plate was also examined. These experiments involved heating the V-shaped target plate from room temperature up to 573 K to observe changes in the surface reaction. In this presentation, we will discuss these results and show the impact of different strike point positions.

This work was partly supported by JSPS KAKENHI Grant Numbers 22H01198 and 23K22469 and the NIFS Collaboration Research program (NIFS23KUGM174, NIFS23KUGM186, NIFS25KFFT001).

Speaker: Prof. Naomichi Ezumi (University of Tsukuba)

384 4.023 Development of a Fluid Code for Simulating Plasma Transport in MPS-LD linear plasma device

Linear plasma devices employ analogous transport mechanisms to those in tokamaks, The plasma, constrained by magnetic fields, flows axially along field lines toward the target plate, enabling effective simulation of tokamak divertor conditions. The MPS-LD, owing to its simple structure and flexible parameter control, serves as an important experimental platform for investigating boundary plasma transport and divertor physics[1]. To achieve accurate divertor condition simulation, electron density and temperature near the MPS-LD target plate must be sufficiently increased. However, radial particle diffusion and energy losses impede parameter enhancement at the target. Therefore, a deeper understanding of boundary transport mechanisms is essential. Magnetic field configuration control is an effective approach to address this issue. Specific field configurations can significantly suppress radial losses, thus increasing particle and energy deposition on the target. Due to the spatial limitations of experimental diagnostics, numerical tools are essential for complementing experimental data and systematically simulating plasma transport processes. However, mainstream simulation codes (such as SOLPS-ITER [2]) face limitations in geometric adaptability and computational efficiency, making it difficult to achieve fast and flexible simulations for linear plasma devices. Therefore, it is imperative to develop dedicated simulation codes tailored to linear devices. To address this need, we developed LiFT (Linear Device Fluid Transport), a numerical code based on the 2D Braginskii equations, where both plasma and neutrals are modeled as fluids. LiFT employs a fourth-order finite-difference scheme for spatial discretization, significantly improving simulation accuracy while effectively suppressing numerical dissipation. Furthermore, an adaptive time-stepping algorithm is implemented to enhance computational efficiency. For hydrogen discharge experiments, the model validation was conducted sequentially, first for the plasma-only module and then for the coupled plasma-neutral fluid module. By comparing our simulations with BOUT++ [3] results and MPS-LD experimental data, the reliability of the code was confirmed. Based on this validation, LiFT was applied to systematically investigate the effects of diffusion and thermal conduction coefficients on radial transport and target plasma parameters. Further studies examined three magnetic field configurations (expanding, uniform, and magnetic mirror) and their impact on transport dynamics. Compared to a uniform field, the magnetic mirror configuration effectively reduces radial losses, resulting in higher electron temperature at the target.
Keywords: MPS-LD device; Boundary plasma transport; Braginskii equations
[1] Sun C et al 2021 Fusion Engineering and Design 162 112074.
[2] Alberti G et al 2023 Nucl. Fusion 63 026020.
[3] Dudson B D et al 2009 Computer Physics Communications 180 1467–80.

Speaker: Xiuxin Tang (GNOI)

385 4.024 Filamentary transport in small-ELM regime with conventional and snowflake divertor configurations in MAST-U tokamak

Filamentary transport in the small-ELM H-mode regime, with typical loss in the stored energy of ~ 1 – 2%, is investigated in conventional and snowflake divertor (SFD) configurations in MAST-U with fast framing $D_α$ filtered cameras. Small, frequent edge localized modes (ELMs) are important for future fusion devices as they avoid large transient heat fluxes due to type-I ELMs and still prevent impurity accumulation.

In conventional divertor (CD) configuration with double null (DN) plasmas, strong pre-cursors before a small-ELM are observed in the form of filaments. These field aligned filaments are localized on the high field side (HFS) near the upper X-point. The filaments originate near the magnetic flux surface $Ψ_Ν$ ~ 0.98 and move radially outwards. A strong correlation is observed between the occurrence of small-ELM and the connection of HFS filaments to the X-point. This phenomenon is observed in wide range of plasma current (600–1000 kA), plasma density ($3–6 \times 10^{19} m^{-3}$), normalized beta (1.5–3), and is independent of fueling location. Conditionally averaged analysis of inter-ELM profiles reveal that the edge electron temperature profile does not change significantly; however, there is a clear buildup of electron density near the separatrix just before an ELM (within 70–99% of the ELM cycle), localized on the HFS. Similar peaking is observed in the carbon density profile near the edge. This suggests that particle source asymmetry likely influences filamentary transport and ELM behavior.

Filamentary transport in SFD plasma during small-ELMs is seen to be slower (~ 400 μs) compared to CD operation (~100 μs). Further, there is a significant delay (~ 1 ms) in the transport of ELM filaments between the outer and inner divertor legs. Turbulence correlation reveals that the ELM filaments are disconnected between the outer and the inner divertor legs and connected between the inner divertor leg and the secondary null region in SFD during small-ELMs. Slow propagation of ELM filaments is partly explained by an increase in the connection length during SF operation (~ up to 3 times). Redistribution of heat and particles due to an additional transport mechanism such as the churning mode in the secondary null of SFD likely leads to a difference in the transport between outer and inner divertor legs.

Prepared by LLNL under Contract DE-AC52-07NA27344.

Speaker: Tanmay Macwan (Lawrence Livermore National Laboratory, USA)

386 4.025 The impact of the isotope mass on divertor detachment and pedestal structure in DIII-D

Dedicated H-mode hydrogen experiments has been conducted at DIII-D and are compared to a similar deuterium dataset under a range of divertor conditions, revealing significant effects of the isotope mass on divertor conditions and pedestal behavior. H-mode confinement is found to be systematically lower in hydrogen and experiences a more pronounced degradation with increased fueling. Hydrogen is found to have a higher ne,sep at similar fueling rates. When comparing hydrogen and deuterium discharges under similar divertor regimes, the hydrogen cases are shown to have a higher electron density and lower electron temperature in the outer divertor. Hydrogen and deuterium density ramps demonstrate a 20% higher ion current and of a 17% higher upstream density at detachment onset in hydrogen. The trends of divertor conditions with isotope mass through detachment are studied through a 2 Point Model analysis and detailed 2D interpretive SOLPS-ITER modeling. Both approaches are able to reproduce key experimental trends. Highlighted in the modeling are a significant role of the ion sound speed, the momentum balance, and carbon sputtering and radiation. An analysis of the particle source in the 2D modeling confirms a higher core particle source for hydrogen, consistent with the experimental hydrogen pedestal structure. However, deuterium is shown to have a lower ionization Mean Free Path (MFP) at the target than in hydrogen at the same upstream density, contrary to expected scaling with isotope mass. Extending the modeling with tritium confirms the experimental and modeling trends in detachment onset, which is found to scale inversely with mass. However, implementing a tungsten wall removes most of the mass dependence. An experimental analysis of the pedestal stability and neutral fueling demonstrates that neither of these effects are sufficient to explain the pedestal differences between isotopes nor the hydrogen pedestal structure respectively. These results indicate that increasing ion mass is likely beneficial for divertor performance in future devices, reducing heat and particle fluxes as well as the ne,sep required for detachment.

Acknowledgements
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under award(s) DE-SC0023100 and DOE GA PO-4500093207

Speaker: Ray Mattes (University of Tennessee Knoxville)

387 4.028 About the relevance of molecular processes in divertor plasmas: Which processes determine $T_{\rm vib}$ at Magnum-PSI?

Divertor detachment is a promising solution to the heat exhaust problem in future fusion devices by reducing the heat load onto divertor target plates$^1$. During the detachment process a neutral gas layer builds up in-between the plasma and the divertor plates through recombination reactions.

Molecular assisted recombination (MAR) is discussed to be among the dominant volume recombination processes, whereby several multi-step channels based on hydrogenic molecules are distinguished$^2$. Thereby, the effective reaction rate for MAR strongly depends on the vibrationally excited population densities of molecular hydrogen H$_2$ (or its isotopologues) that can typically be characterized by the vibrational temperature $T_{\rm vib}$.

This contribution investigates $T_{\rm vib}$ in divertor-like H$_2$ plasmas by combining optical emission spectroscopy (OES) measurements with interpretation via collisional-radiative (CR) models based on the Yacora code for atomic$^3$ and molecular$^4$ hydrogen applying state-of-the-art reaction probabilities.

OES measurements (including the molecular Fulcher-$\alpha$ band and the atomic Balmer lines) as well as Thomson scattering and two-photon absorption laser induced fluorescence (TALIF) are performed at the linear divertor plasma simulator Magnum-PSI$^5$ for plasma conditions with $T_{\rm e} \leq 3.5$ eV and $n_{\rm e} \approx 10^{19} - 10^{20}$ m$^{-3}$. From the spectroscopic measurements first $T_{\rm vib}$ is deduced and secondly the molecular and ionic particle densities are determined applying a CR model for atomic hydrogen (Yacora H), while considering the results from Thomson scattering and TALIF as constraints. Subsequently, these plasma parameters are used as input for the CR model Yacora H$_2$($X^1$,$v$), which is aimed for self-consistent $T_{\rm vib}$ predictions, in order to benchmark the model and to study, to which extent which processes determine $T_{\rm vib}$ and thus the effective MAR rate.

$^1$ S.I. Krasheninnikov et al., Journal of Nuclear Materials 241-243 (1997) 283-287
$^2$ N. Ohno, Plasma Physics and Controlled Fusion 59 (2017) 034007
$^3$ D. Wünderlich et al., Journal of Quantitative Spectroscopy and Radiative Transfer 240 (2020) 106695
$^4$ R.C. Bergmayr et al., European Physical Journal D 77 (2023) 136
$^5$ M.J. van de Pol et al., Fusion Engineering and Design 136 (2018) 597-601

Speaker: Richard Christian Bergmayr (IPP Garching)

388 4.030 Asymmetric island localized profile shifts and their implications for scrape-off layer diagnostic comparability in Wendelstein 7-X

The Wendelstein 7-X (W7-X) stellarator experiment uses the island divertor concept for power and particle exhaust. In the most studied configurations, the scrape-off layer (SOL) includes a chain of five islands (m/n = 5/5) around the core. To characterize the SOL, different toroidally separated diagnostics view separate islands. The five-fold toroidal symmetry is typically used to combine measurements of separate quantities from different islands.
The gas puff imaging (GPI) diagnostic is a SOL turbulence diagnostic, imaging electron density/temperature fluctuations via H$_α$ emission of a localized gas puff at high spatio-temporal (mm/μs) resolution in W7-X [1]. It observes complex drift flow patterns in the islands [2] not yet included in any validated code. These flow patterns and SOL profiles are strongly dependent on the magnetic geometry of the island, and discharge parameters such as power and density.
GPI measurements show sudden shifts of light emission and fluctuation parameters, indicating corresponding changes in plasma parameters. These transitions are not reflected in global core plasma parameters. In addition, divertor heat flux measurements and further SOL diagnostics viewing different islands show that these sudden transitions often are isolated to just one of the five islands. This raises the question under which conditions the assumption of five-fold symmetry applies to processes and physical quantities in the SOL. In this contribution, we investigate the prevalence and asymmetry of these profile shifts, as well as their implications for the comparability of physical quantities in different islands.

Speaker: Floris Scharmer (IPP Greifswald)

389 4.031 First Integration of SOLPS-NN into JINTRAC for Fast Core-Edge Tokamak Simulations

Achieving reliable power production in future tokamak reactors requires operating
regimes that simultaneously deliver high fusion plasma performance and maintain
divertor and first-wall particle and heat-fluxes within material limits. The plasma core
and the scrape-off layer (SOL) regions cannot be decoupled for predictive scenario
development. Integrated modelling frameworks self-consistently resolve core–edge
interactions but still require some external information for example on anomalous SOL
transport. Within the JINTRAC suite [1] such simulations remain computationally
constraint by the numerical cost of detailed edge/SOL solvers such as EDGE2DEIRENE. This limitation prohibits systematic parameter exploration, uncertainty
quantification, and the broad optimisation studies required for advanced scenario
development.
A promising strategy to overcome this bottleneck is the use of fast surrogate models
trained on SOL simulations [2]. In this work, we integrate SOLPS-NN—a neural-network
surrogate trained on a uniformly sampled multi-parameter SOLPS-ITER simulation
database in JET geometry [3,4]—into JINTRAC for the first time. This approach replaces
the conventional edge fluid solver with a rapid predictor of SOL and divertor plasma
conditions, enabling time-dependent core-edge simulations at drastically reduced
computational cost, effectively reducing the problem to core-solver timescales.
We present the technical implementation of the investigated JINTRAC–SOLPS-NN
coupling approaches, including the choice of exchanged quantities, coupling interface,
and numerical procedures used to connect the surrogate-based SOL model to the core
solver. The resulting performance is assessed through a series of initial tests on a JET
scenario that is consistent with the SOLPS-NN training dataset, providing a controlled
setting to evaluate the behaviour of the coupled system. Simulations obtained with the
different coupling variants are compared against reference JINTRAC/EDGE2D-EIRENE
results to examine the ability of the surrogate-based framework to reproduce expected
physical trends. These studies provide initial insight into the accuracy, robustness, and
limitations of this accelerated integrated-modelling approach.
Based on these findings, we outline the foreseen developments required to enhance the
physics fidelity of the coupling, such as extensions of the communicated quantities and
further validation, and discuss the prospective applications enabled by the achieved
speed-ups.
[1] M. Romanelli et al., “JINTRAC: A System of Codes for Integrated Simulation of Tokamak Scenarios”,
Plasma Fusion Res. (2014).
[2] S. Wiesen et al., “Data-driven models in fusion exhaust”, Nucl. Fusion (2024).
[3] S. Dasbach, “Surrogate models for particle and power exhaust in divertor plasmas”, PhD thesis (2025).
[4] S. Dasbach, S. Wiesen, “Surrogate models for interpolation of tokamak edge plasmas”, Nucl. Mater.
Energy (2023).

Speaker: Rick van Schaik (GNOI)

390 4.032 2D Boundary Fluctuation Characterisation in TCV X-point Radiator Plasmas

To obtain high fusion performance and satisfy the first wall material limit in future fusion devices, increasing effort has been made to develop and investigate high-confinement no-ELM (Edge Localized Mode) and small-ELM scenarios. Among them, recently, the X-point radiator (XPR) regime has been developed in multiple devices, including the TCV tokamak, and shows promising ELM suppression [1,2]. However, relatively little is known about fluctuations and filamentary turbulence at the plasma edge and first-wall for this scenario. In this work, we will present recent experimental observation in TCV to explore the dynamics of the edge and SOL fluctuations in the XPR regime, using Gas Puff Imaging (GPI) and other diagnostics including wall-embedded Langmuir probes, reciprocating probes, bolometry and high-speed cameras [3]. Specifically, we will characterize and compare fluctuations, filament properties, and far-SOL transport in Type-I ELMy and ELM-free phases, at the outboard midplane and X-point. In the Snowflake XPR ELM-free H-mode, a periodic burst in the plasma edge is captured by the midplane and X-point GPI, potentially representing the mechanism replacing the particle transport channel otherwise provided by ELMs. In more recent experiments, an ELM suppression was obtained via nitrogen seeding in NBI and ECRH heated single null plasmas, and a gradual reduction in ELM amplitude was observed in the transition to the ELM-free phase. The general fluctuation features of filamentary turbulence and bursts in this regime will be presented.
[1] M. Bernert et al. 2021 Nucl. Fusion 61 024001
[2] H. Reimerdes et al. 2024 Nuclear Materials and Energy Volume 41, 101784
[3] N. Offeddu, C. Wüthrich, W. Han, et al. 2022 Rev. Sci. Instrum. 93, 123504

Speaker: Yinghan Wang (EPFL Swiss Plasma Center)

391 4.033 A Propagator-based Multi-level Monte Carlo Method for Kinetic Neutral Species in Edge Plasmas

We propose and investigate a new multi-level Monte Carlo scheme for numerical
solutions of the kinetic Boltzmann equation for neutral species in edge plasmas [1]. In
particular, this method explicitly exploits a key structural property of neutral particle
dynamics: the prevalence of frequent collisions for which the outgoing velocity is
determined by local plasma parameters. Using this property, we derive a multi-level
algorithm based on collision event propagators and show, both analytically and through
numerical experiments, that it reproduces the results of standard Monte Carlo methods.
We further demonstrate that, in the context of coupled plasma-neutral edge simulations
employing correlated Monte Carlo, the proposed scheme retains trajectory correlation to
machine precision as the system evolves, whereas conventional methods exhibit rapid
decorrelation. These results indicate that the propagator-based multi-level Monte Carlo
scheme is a promising candidate for use in fully implicit Jacobian-free Newton-Krylov
(JFNK) solvers for coupled plasma-neutral systems.

[1] Parker et al., https://www.arxiv.org/abs/2512.09334

*Work supported by NSF Mathematical Sciences Postdoctoral Research Fellowship
Award No. 2303102 and by the U.S. DoE, at LLNL under DE-AC52-07NA27344.

Speaker: Greg Parker (UC Berkeley)

392 4.034 Interpretive modeling of asymmetric electron-ion transport during ELMs with absorbing divertor targets

‌A new plasma expansion model with absorbing walls is developed to simulate the ELM transport in the Scrape-off Layer (SOL), and the modeling results can well explain the time difference between the electron temperature (Te) and ion saturation current (Jsat) peaks measured by divertor Langmuir probes on EAST. The plasma expansion model was initially used to study the expansion of plasma bunch into vacuum [1] and has been applied to describe the parallel ion transport in the SOL during ELMs [2,3]. However, the original model solves the Vlasov equation in infinite space which neglects the presence of divertor targets. By treating the divertor targets as absorbing walls, we derive a new analytic solution for the plasma expansion model to describe the boundary condition during ELMs. The new model reveals that the electron motion exhibits oscillatory behavior and predicts a higher electron incident flux on divertor targets with eliminating the backward flow across targets. Meanwhile, in the early stage of an ELM burst, energetic electrons undergo more intense parallel transport than ions, which results in a time interval (~0.1 ms) between the electron temperature and ion saturation current peaks measured by Langmuir probes. Due to the higher intra-ELM electron flux than ion flux, 500 V bias voltage is required in Langmuir probes to repel electrons originating from the pedestal top of EAST. A lower bias voltage of Langmuir probes may significantly underestimate the ion flux on divertor target, especially at the initial phase of an ELM burst. Considering the steady state background plasma, the new model can give the evolution of Te and Jsat on the divertor targets during the intra-ELM phase. Modeling results quantitatively agree with the Langmuir probes data for both type-I and small ELMs. ‌Compared to small ELMs, large ELMs exhibit more pronounced electron-ion transport differences and thus a more obvious time interval between Te and Jsat peaks due to stronger source perturbations. Furthermore, the early-arriving ELM-induced electrons create an enhanced sheath potential drop, providing additional acceleration to subsequent ions. This effect is more pronounced during large ELMs due to the higher incident electron flux, thereby enhancing tungsten erosion rates at the initial phase of an ELM burst.

References:
[1] D. S. Dorozhkina, et al. 1998 Phys. Rev. Lett. 81.2691
[2] W. Fundamenski, et al. 2005 Plasma Phys. Control. Fusion 48 109
[3] Guoliang Xu, et al. 2021 Nucl. Fusion 61 086011

Speaker: Jinheng Zhao (ASIPP)

393 4.036 Investigation of Machine Learning based neutral transport model in UEDGE

In the boundary of magnetic fusion devices, plasma is strongly coupled with recycling
neutrals. Consequently, an integrated model for fusion boundary transport must encompass
neutral physics. A recently proposed propagator-based approach has the potential to
dramatically enhance the accuracy and efficiency of neutral transport modeling in plasma,
across the range of neutral collisionality regimes [1]. However, the cost of calculating the
propagator may still be prohibitive for efficient coupling with the plasma in an integrated
time-dependent model. To address this, a machine learning (ML) algorithm has been
developed for approximating the propagator. This allows for fast and accurate calculation
of neutral density for a given plasma background and neutral sources [2]. A salient feature
of the ML-predicted propagator is its smooth dependence on the plasma parameters. This
property facilitates the integration of the ML neutral model with plasma models using
Newton-based time integration methods. The present study is focused on implementation
and investigation of the ML propagator-based neutral transport algorithm in the edge
plasma transport code UEDGE. The propagator is calculated using the Monte Carlo neutral
transport code DEGAS-2 for a large number of samples corresponding to different plasma
profiles. These samples are used to train an ML model that approximates the propagator.
The performance of the resulting neutral calculation algorithm, integrated with the UEDGE
plasma model, is investigated for explicit and implicit time-stepping, using ODE and
Newton-Krylov solvers in the SciPy library.

Speaker: Maxim Umansky (LLNL)

394 4.037 Observation of the Recombination Front Position and Its Movement in the Linear ECR Plasma Device NUMBER

In detached divertor experiments, impurity gas injection used to achieve detachment has been reported to cause core plasma cooling near the X-point and, in some cases, radiative collapse. These adverse effects have been attributed to the approach of the recombination front toward the X-point, highlighting the need to establish a technique for controlling the recombination front position.
Measurement of spatial profile along a magnetic field line under limited observation ports is an issue in the recombination front study. We have developed an estimation method of spatial density and temperature profiles using single observation port. Applicability of the method is demonstrated with electrostatic probe diagnostics in a linear plasma device NUMBER.
NUMBER consists of a 2 m vacuum chamber. Helium plasma is generated using electron cyclotron resonance (ECR) by injecting microwaves (2.45 GHz, 6 kW) through a quartz window at z = 0 m. Plasma flowing along the magnetic field lines (z-axis) undergoes volumetric recombination when additional neutral gas is supplied from behind the endplate installed at z = 1.98 m. Double probes are installed at z = 1.38, 1.53, and 1.68 m, enabling measurement of the axial plasma distribution with 150 mm spacing.
In order to estimate spatial profile, a forked probe is installed at z = 1.68 m, which is capable of measuring the axial gradients of electron temperature and density with a high spatial resolution of 40 mm. We derived a new method for estimating the z-profile using forked probe. This technique can reconstruct the z-profile by utilizing the pressure dependance of electron temperature and density together with the pressure dependance of their local axial gradients.
The measurements revealed that, in the central region of the plasma, the electron temperature decreases monotonically with increasing z, with typical values of a few eV. In contrast, the electron density exhibits a “rollover” structure: it increases with z, reaches a peak value of 2.0 × 10¹⁸ m⁻³, and then decreases. This structure was confirmed using both the double probe and the forked probe, supporting applicability of the estimation method. The rollover position provides a useful indicator for identifying the recombination front location.
Furthermore, by adjusting the balance between neutral gas injection and pumping, we conducted experiments under different neutral pressure distributions and observed the resulting movement of the recombination front. This finding suggests that controlling the neutral pressure distribution can be effectively applied to control the recombination front position.

Speaker: Kota TAKEDA (Nagoya University)

395 4.038 Investigation of Hβ/Hα Ratio during the Plasma current Ramp-down in the ADITYA-U Tokamak

Investigation of hydrogen Balmer series emissions, such as Hα, Hβ and Hg lines, is essential for diagnosing the edge plasma region, studying the fueling efficiency and neutral penetration, and understanding the plasma detachment. Monitoring of Hα and Hβ and their ratio may lead to investigating various atomic and molecular processes happening in the tokamak edge and divertor region of the tokamak, as the population mechanism of the upper level of these two transition are different, which in turn results in different intensity ratios of Hα and Hβ at various stages of temporal evolution of the tokamak plasma. In the ADITYA-U tokamak, which has an plasma edge electron density of 0.5 – 1.0 ´ 1018 m-3 and a temperature of 7 - 15 eV, temporal evolution Hα and Hβ emissions are routinely monitored along various lines of sight passing through the plasma from many toroidal locations. The intensities of Hα, Hβ, and the Hβ/Hα intensity ratio are analyzed with the variation of plasma current and its ramp-down rate during the current flat-top and current ramp-down phase just before plasma current termination for many discharges. The statistical analysis reveals a correlation between the intensity ratio and plasma current and the rate of change of plasma current before the plasma termination. It is seen that the Hβ/Hα ratio becomes higher during the current ramp-down phase as compared to the flat-top phase in ADITYA-U tokamak plasma. Not only that, but the ratio increases with decreasing the plasma current. Detailed investigation through the estimation of Hα and Hβ intensities by considering the excitation and recombination processes of hydrogen atoms and molecules. It shows that higher Hβ emission and its ratio with Hα are mainly due to the contribution from recombination processes during the ADITYA-U toakamk’s current ramp-down phase.

Speakers: Dipexa Modi (Pandeet Deendayal Energy University, India), Utsav Rajvanshi (Institute for Plasma Research Gandhinagar, INDIA)

396 4.039 The Importance of Plasma-Neutral Interaction; Er formation, H-mode transition, and Intrinsic Rotation from arc discharge to tokamaks and space plasmas

In the plasma physics, the momentum exchange by charge exchange reactions between plasma ions and boundary neutrals has been underestimated. Lacking of understanding this important mechanism is one of main causes of the complication in the field of plasma surface interaction. When there is magnetic field, perpendicular component of momentum exchange can generate current and the electric field, which is induced by the process of Gyro-Center Shift (GCS). This electric field formation has been observed in the arc discharges, earth’s ionosphere including black aurora, and the edge region of fusion plasmas, especially at the High confinement mode (H-mode) [1]. The perpendicular component of momentum change in the charge exchange reaction is the cause of the intrinsic rotations observed in tokamaks [2]. Overall friction between plasma and neutral generates turbulence, and when the Reynolds number which is the ratio of inertia force to the viscosity force is lower than critical value, the turbulence is suppressed and H-mode transition takes place [3]. The GCS current equation, the turbulence induced diffusion equation, and the Reynold number equation of tokamak boundary will be presented as well as many applications of these equations.

[1] K. C. Lee, Phys. Plasmas 24 112505 (2017)
[2] K. C. Lee, S. G. Lee, Current Appl. Phys. 55 16 (2023)
[3] K. C. Lee, Plasma Phys. Control. Fusion 51, 065023 (2009).

Speaker: Kwan Chul Lee (KFE)

397 4.040 Study of intrinsic toroidal rotation in the Scrape-Off Layer of inboard-limited Aditya-U plasmas

Coupled plasma-neutral transport simulations are performed for the scrape-off layer of the ADITYA-Upgrade tokamak to investigate experimentally observed reversals of plasma rotation with increasing edge density [1,2]. The simulations are carried out using the three-dimensional EMC3-EIRENE plasma-neutral code and incorporate a toroidally continuous and poloidally localized high-field-side belt limiter in a moderately circular magnetic equilibrium. For relatively high upstream density and input power conditions, the simulations recover the formation of mutually counter-propagating toroidal plasma flows localized in the upper and lower regions of both the SOL and the adjacent closed-field-line region.

The origin of these flows is traced to strong poloidal density and pressure gradients induced by enhanced recycling at the inboard localized limiter. The dominant drive for parallel flows is therefore identified as the residual component of the Reynolds stress arising from pressure gradients in the SOL. The pressure modulation exhibits, to lowest order, an up-down antisymmetric standing structure consistent with a poloidal mode number mp=1, resulting in a pair of counter-streaming flows and a reversal of flow direction across the midplane, consistent with intrinsic rotation [3].

A quantitative evaluation of the right-hand-side parallel momentum source associated with the divergence of the residual Reynolds stress is planned. This includes direct calculation of the gradient-driven residual stress contribution to the toroidal flow velocities. By establishing this framework in a medium-sized, circular tokamak, the present study provides a physics-based basis for understanding intrinsic rotation generation in larger fusion devices, where similar edge asymmetries and recycling-driven pressure gradients are expected to play an important role.

Speaker: Arzoo Malwal (Institute for Plasma Research)

398 4.041 Characteristics of deuterium molecule spectrum during divertor detachment processes in EAST

Detachment is an important method for controlling the thermal load on the target plate in the high-confinement mode in tokamak, a process accompanied by complex atomic and molecular physics. Using spectroscopic diagnostics, the evolution characteristics of the molecular spectrum during detachment were systematically studied.
Experiments showed that in detachment experiments with Ne impurity injection, the Dδ/Dα ratio initially increased sharply, indicating the enhancement in volume recombination. As the neutral pressure increased, the D₂ spectrum first increased and then weakened, with its peak shifting from the strike point to the far scrape-off layer (SOL) region. Simultaneously, the D-atom spectrum was significantly enhanced in the near-target region. This suggests that prior to detachment, the neutral gas pressure was dominated by molecules, whereas after detachment, atoms contributed more substantially. The attenuation of the molecular spectra is attributed to the enhancement of molecular dissociative attachment and ion conversion processes, which increased the population of excited-state atoms while reducing the molecular population.
In the density ramp experiment, the inner and outer target plates exhibited different molecular spectral distributions: the D₂ spectrum at the inner target extended into the SOL above the strike point, while its peak at the outer target was located in the corner region at the junction of the vertical and horizontal plates. This difference may be related to geometric confinement and the longer ionization mean free path at lower temperatures. The degree of detachment between the inner and outer targets also influenced the molecular spectral distribution.
The spatial distribution of the D₂ spectrum differed significantly under different toroidal magnetic field (Bt) directions. Under favorable Bt, the D₂ peak at the inner target was close to the strike point. In contrast, under unfavorable Bt conditions, it extended into the far SOL region, a consequence of changes in ion drift direction.

Heading

Speaker: Bingcheng Qi (Institute of Plasma Physics,Chinese Academy of Sciences)

399 4.042 SOLPS-ITER and BOUT++ Iterative Coupling (SIBIL) to self-consistently inform 2D anomalous diffusion coefficients

A novel method of coupling SOLPS-ITER to a six-field two-fluid model using BOUT++ is reported in this work. Turbulence-informed simulations of transport in edge plasmas are crucial for correct predictions of divertor particle and heat flux. This coupling bridges the multi-spatiotemporal, multi-physics challenge in plasma edge simulations. The model utilized in BOUT++ includes peeling-ballooning physics to simulate the effects of ELMy H-mode plasmas [1]. This 3D BOUT++ simulation was coupled to the fluid plasma and Monte-Carlo neutral 2D transport code SOLPS-ITER for self-consistent transport at the plasma edge. Radial fluxes in BOUT++ have been successfully used to inform poloidally and radially varying anomalous diffusion coefficients in SOLPS-ITER. A converged H-mode SOLPS-ITER simulation is used as an initial state with the same physical space simulated in both SOLPS-ITER and BOUT++ simulations. The BOUT++ grid has significantly higher radial resolution to account for increased fidelity of edge turbulence modelling, and the SOLPS-ITER grid has significantly higher resolution in the divertor. The density, temperature, and pressure profiles from a steady-state SOLPS-ITER simulation are used to define the BOUT++ initial state by means of a radial then poloidal spline fit in (psi,theta) space. The plasma state is then simulated in BOUT++ such that the particle and energy fluxes can be extracted to inform the anomalous diffusion coefficients in SOLPS-ITER. The plasma state is then iteratively passed between the codes until the steady state solutions in SOLPS-ITER cease significantly changing. The impacts of the profile mapping on globally conserved quantities are reported. This new capability enables self-consistent transport at the plasma edge, thereby eliminating the diffusion coefficient profile as a user-defined input to match experimental results. Additionally, this coupling enables the inclusion of effects arising from transients and the interplay between the pedestal structure, edge turbulence and divertor-plasma solutions.

[1] T.Y. Xia, X.Q. Xu, and P.W. Xi. Six-field two-fluid simulations of peeling–ballooning modes using BOUT++. Nuclear Fusion, 53(7):073009, May 2013.

Work supported by the U.S. Department of Energy, under Award(s) R011382908 (DOE - SCIDAC), NRC 31310022M0014.

Speaker: Benjamin Taczak (University of Tennessee - Knoxville)

400 4.043 Particle and Impurity Behavior in Edge Plasma of ADITYA-U tokamak in presence of MHD Activity

In tokamak devices, the edge plasma region is strongly affected by neutral particles. Understanding the interaction between neutral gas and tokamak plasma is crucial for controlling edge plasma dynamics. Gas puffing is commonly used for plasma fueling and density regulation. However, sometime gas puffing may leads to either increase or decrease magnetohydrodynamic (MHD) activity. In ADITYA-U tokamak, the interaction between MHD activity and fueling is investigated using visible spectroscopic diagnostic in tokamak plasmas. The visible spectroscopic diagnostics is based on optical fiber, interference filter and photo-multiplier tube detector and measure temporal evolution of H, O1+ emission at 441.9 nm and C2+ emission at 464.7 nm. In ADITYA-U tokamak, it is seen that gas puffing leads to a rapid increase in Hα emission, followed by a gradual decay, eventually almost returning to its initial level in the discharges with weaker MHD activity. However, when gas is puffed in the plasma having strong MHD activity, Hα emission reduces significantly compared to discharges without MHD activity. This reduction depends on the amplitude of the MHD activity. A similar analysis is performed for impurity emissions such as C2+ and O1+, and it is observed that the changes in impurity emission are significantly weaker as compared to the Hα emission. The variation of the intensities of the H and impurity emissions with the MHD activity are investigated to understand the role MHD in the particle and impurity behavior at the plasma edge of ADITYA-U tokamak discharges with gas puffing.

Speaker: Dipexa Modi (Pandit Deendayal Energy University)

401 4.044 Improved Density Build-Up in Stellarator Island Divertors

Stellarators are a promising candidate for steady-state fusion power plants, and recent achievements on Wendelstein 7-X (W7-X) have further strengthened this prospect. A key determinant of reactor viability is the performance of the divertor, which governs impurity control and particle exhaust. While W7-X has demonstrated excellent heat-flux mitigation and steady-state capabilities, it has not yet achieved one of the central divertor performance metrics: robust particle exhaust enabled by strong divertor density build-up. Present-day stellarator island divertors typically show significantly weaker density build-up compared to tokamak divertors, limiting their ability to operate in regimes with efficient neutral compression and stable exhaust.
Previous modeling has shown that improved density build-up can be achieved in closed island divertors. However, while this behavior has been demonstrated experimentally in open tokamak divertor geometries, comparable density enhancement has not yet been achieved in open stellarator island divertors. Nevertheless, the stellarator 2-point model predicts that, under suitable conditions, similar performance should be attainable in stellarator open divertors. Demonstrating enhanced density build-up in the open island divertor is therefore crucial both for validating the stellarator reactor concept and for guiding the design of future optimized devices.
In this work, we investigate plasma pressure conservation and the resulting divertor density build-up using EMC3-EIRENE simulations of a next-generation stellarator experiment. Parameter scans are performed spanning W7-X–like to reactor-relevant regimes, systematically varying the heating power and cross-field transport coefficients. Our results identify clear operational boundaries at which enhanced pressure conservation along magnetic field lines—from upstream to downstream—emerges, enabling substantially improved divertor density build-up. The mechanism responsible for this transition is analyzed, along with its implications for achieving reactor-relevant particle exhaust in stellarators.
This study provides quantitative guidance for the design of a new stellarator experiment featuring higher heating power and reduced cross-field transport, capable of experimentally demonstrating improved density build-up in an open island divertor. Realizing such performance would represent a critical step toward establishing the stellarator as a competitive pathway to a fusion reactor.

Speaker: Sergei Makarov (Proxima Fusion GmbH, Flößergasse 2, 81369 Munich, Germany)

402 4.045 Inline Deep Surrogates for Accelerating SOLPS-ITER Simulations

Edge plasma simulations with SOLPS-ITER are expensive and sensitive to initialization, especially when spanning wide parameter ranges or tailoring to specific discharges. We present a reduced-modeling workflow that combines (i) warm starts via nearest-neighbor initialization from a KD-tree [1] of converged runs and (ii) learned surrogates for rapid prediction. The KD-tree selects a converged neighbor whose terminal state seeds the new run, reducing runtime relative to cold starts. For prediction, we train two complementary families of surrogates on DIII-D datasets [2] spanning gas puff, core density, cross-field transport, and diffusivities: (1) one-dimensional profile surrogates at the outer midplane and targets, using an ensemble fully connected network that achieves high R2 with typical relative errors ≲ 20% on held-out cases; and (2) two-dimensional map surrogates for poloidal fields and sources, using a U-Net [3] for pixel-aligned regressions and a variational autoencoder [4] to learn a low-dimensional latent representation that enables fast sampling and anomaly detection. To limit costly sampling in low-value regions, we employ an adaptive sampling approach [5] that targets high-variance areas and iteratively retrain the models. Finally, we demonstrate the initial inline use of the surrogate within the SOLPS-ITER loop to reduce selected EIRENE calls under physics-based constraints. The combined approach accelerates scenario generation, reduces convergence risk, and enables real-time inference in edge-plasma studies.

[1] J. L. Bentley, Commun. ACM 18, 509–517 (1975).
[2] J. D. Lore et al., Nucl. Fusion 63, 046015 (2023).
[3] O. Ronneberger et al., Lect. Notes Comput. Sci. 9351, 234–241 (2015).
[4] R. Anirudh et al., Proc. Natl. Acad. Sci. U.S.A. 117, 9741–9746 (2020).
[5] A. Diaw et al., Nat. Mach. Intell. 6, 568–577 (2024).

Speaker: Abdourahmane Diaw (Oak Ridge National Laboratory)

403 4.046 Neutral transport surrogate model development for coupling to plasma simulation

The transport of neutral atoms and molecules produced from gas puff and the wall recycling process is a critical aspect of the core-edge integration challenge. Because of this, predicting the evolution of edge plasma profiles from simulation requires coupling to self-consistent predictions of the neutral gas. The Monte Carlo method is the most widely-used kinetic method for doing so. Coupling neutral and plasma physics remains at best a performance bottleneck or even, heretofore, unfeasible when long timesteps are desired with implicit solution methods [1]. As a result, kinetic neutral physics remains often neglected in implicit plasma solvers in favor of more efficient and differentiable fluid models. Algorithmic and software enhancements continue to be developed to address this issue [2], and we present a machine learning approach to train a deterministic surrogate model from stochastic simulation data. The Monte Carlo neutral transport solver DEGAS2 is used to efficiently and robustly predict kinetic neutral profiles against a background of more than 70,000 UEDGE simulations of KSTAR [3]. A UNet model is trained on this data to predict for a wide range of plasma profiles. This model is not only very efficiently evaluated, but so are derivatives with respect to the evolving plasma profile, unlocking the feasibility of fully kinetic and implicit neutral coupling to plasma simulation. The departure of the kinetic to fluid results is examined, as is the importance of neutral-neutral scattering in the former. The path toward training a mesh-agnostic model is explored, as is direct implementation of such models in implicit plasma simulations.

This work is supported by USDoE contracts DE-AC02-09CH11466, DE-AC52-07NA27344.

[1] I. Joseph, et al. “On coupling fluid plasma and kinetic neutral physics models.” Nuc. Mat. & Energy. 12:813 (2017)
[2] G. J. Wilkie, P. K. Romano, R. M. Churchill. “Demonstration of OpenMC as a framework for atomic transport and plasma interaction.” Plas. Phys. & Cont. Fus. 67:055046 (2025)
[3] B. Zhu, et al. “Latent space mapping: Revolutionizing predictive models for divertor plasma detachment control.” Phys. Plasmas 32:062508 (2025)

Speaker: George Wilkie (PPPL)

404 4.047 Optimising the island divertor for a stellarator power plant: resilience of magnetic structures and coil optimisation

The most mature stellarator divertor concept is the island divertor (such as used by Wendelstein 7-X [1]), in which a spatially large magnetic island chain diverts exhausted plasma onto plasma-facing components (PFCs). Despite its maturity, there remain fundamental open questions about how island properties affects edge transport and divertor performance, and how this understanding can be integrated into stellarator design [2, 3, 4]. This work addresses two aspects of island divertor optimisation: (1) the physics determining the properties of island chains in optimised stellarators (for example, the island size and phase, and how this varies as the equilibrium changes) and (2) coil optimisation schemes to target properties of the island chain.

We first present theory relating the properties of the magnetic field to O- and X-points of island chains, using Wendelstein 7-X as a paradigmatic example. We also present fast, user-friendly computational methods to allow automated topological analysis of stellarator magnetic fields. The second part of the talk applies these insights to stellarator coil optimisation schemes. We present a proof-of-principle scheme to design modular coils for an optimised stellarator equilibrium of the “SQuID” variety [5], showing how edge island structures can be explicitly targeted. Implications for plasma-facing component and integrated stellarator power plant design are discussed.

References
[1] Renner, Hermann, et al. "Physical aspects and design of the Wendelstein 7-X divertor." Fusion science and technology 46.2 (2004): 318-326.
[2] Feng, Y. "Up-scaling the island divertor along the W7-stellarator line." Journal of Nuclear Materials 438 (2013): S497-S500.
[3] Lion, J., et al. "Stellaris: A high-field quasi-isodynamic stellarator for a prototypical fusion power plant." Fusion Engineering and Design 214 (2025): 114868.
[4] Davies, R., et al. "Stellarator divertor optimisation for a Stable Quasi-Isodynamic Design (SQuID): magnetic topology, divertor plates and baffle design." 51st EPS Conference on Plasma Physics. European Physical Society, 2025.
[5] Goodman, Alan G., et al. "Quasi-isodynamic stellarators with low turbulence as fusion reactor candidates." PRX Energy 3.2 (2024): 023010

Speaker: Robert James Davies (MPPL)

405 4.048 Upgrade of voltage sweeping system for Langmuir probe to measure ne, Te profile in the tungsten divertor region on KSTAR

A Langmuir probe (LP) system, designed to measure basic plasma parameters and their poloidal profiles, was installed in the upgraded tungsten divertor of KSTAR [1]. Since the KSTAR 2023 campaign, temporal ion saturation currents have been measured in the lower divertor region by applying a negatively DC-bias of -240 V to the LP using batteries, with a DAQ sampling frequency of 200 kHz. As the ion saturation current alone is insufficient to determine plasma parameters such as electron density(ne) and electron temperature(Te), a voltage-sweeping method was adopted to obtain IV characteristics over time. The applied voltage waveform is generated by an SDG1032X function generator and subsequently amplified using a KEPCO BOP100-4D bipolar power supply. The corresponding current is measured across a 10 Ω resistor placed between the LP tip and the power supply. Single voltage-sweeping LP data were successfully obtained during the KSTAR 2024 campaign by applying a 1 kHz triangular waveform ranging from -100V to +100V.
From the KSTAR 2025 campaign, ten (LPs) were operated with voltage sweeping to measure of ne and Te at certain regions near the inner and outer strike point. The LP measurements at the scrape-off layer(SOL) are useful characterizing divertor detachment and underlying physics. When deuterium and nitrogen gases were injected into H-mode plasmas with the newly installed W divertor, deuterium fuelling increased the ELM frequency, thereby promoting tungsten flushing and reducing core radiation. In contrast, nitrogen seeding led to increased core radiation despite lower tungsten sputtering, indicating changes in transport. Further analysis of target electron density and temperature measured by a voltage sweeping LP will be performed [2]. The results will be compared and discussed with the tungsten optical emission spectroscopy (OES) diagnostics in the divertor region.

Acknowledgements
This work was supported by the R&D Program of the “KSTAR Experimental Collaboration and Fusion Plasma Research (EN2503)” and Grant Nos. RS-2022-00155917 and NRF2021R1A2C2005654, Grant Nos. RS-2022-00155960 trough the Korea Institute of Fusion Energy (KFE) funded by the Korean Ministry of Science and ICT.

References
[1] Yegeon Lim, et al., "Development of a fixed Langmuir probe system for newly installed tungsten monoblock lower divertors in KSTAR." Fusion Engineering and Design 205 (2024): 114552.
[2] Junghoo Hwang, et al., “Experiment investingation of deuterium and nitrogen-seeded H-mode plasma in KSTAR with new W divertor.”, FEC(2025), in progress.

Speaker: Eunnam Bang (Korea Insutitute of Fusion Energy)

406 4.049 Investigating the Sensitivity of CARS Diagnostics for Measuring Rovibrational Populations of Hydrogen in Divertor-relevant Plasmas

One of the biggest challenges of a reliable fusion reactor is the handling of large heat and particle loads on the divertor wall. Key to reducing these loads is plasma detachment, in which a large range of processes occur between the plasma and the neutral background [1,2].
Molecular processes dominate the plasma dynamics, which is why molecules are often studied in divertor research [1,2]. However, reliable rovibrational measurements on the electronic ground state are lacking for divertor-relevant plasmas, causing for example MAR (Molecularly Assisted Recombination) to be poorly understood. We will use the active laser spectroscopy technique named CARS (Coherent Anti-Stokes Raman Scattering) to measure these hydrogen populations.

This contribution will focus on two parts. First, an investigation on the optimal form of CARS is performed for use in the low-pressure hydrogen environments of the divertor region, using ns-pulsed lasers. To this end, three distinct CARS schemes are tested in the lab: collinear CARS, BOXCARS, and collinear polarization CARS. The investigation focuses on the lowest pressure each scheme could measure, including the sensitivity of each scheme when measuring with a large nitrogen background pressure. The first results of these tests will be presented. Most sensitive at low pressures are collinear CARS and collinear polarization CARS, enabling density measurements in the order of 10^{18} m^{-3}. Collinear polarization CARS is the most sensitive diagnostic when detecting species at a large background pressure.

The second part of this contribution shows results of collinear CARS applied to the linear plasma device Magnum-PSI, which is able to generate divertor-relevant plasma conditions (ne ≈ 10^{21} m^{-3}, Te < 5 eV). Spatially-resolved measurements of the lower rovibrational population are ongoing and the first preliminary results will be presented.

[1] A. Loarte, et al., Nucl. Fusion 47 S203 (2007). DOI: 10.1088/0029-5515/47/6/S0
[2] S.I. Krasheninnikov and A.S. Kukushkin, J. Plasma Phys. 83 (2017). DOI:10.6100/IR58304

Speaker: Kay Schutjes (DIFFER)

407 4.050 Assessment of the fluid transport model for neutral atoms in Hermes-3 in non-axisymmetric geometries

A notable recent success of the island-diverted stellarator Wendelstein 7-X is the achievement of stable detached operation. However, central processes involved in detachment access and detached operation of stellarators still need to be fully understood to explain the observed phenomena and improve the design of future machines.
Shedding light on these mechanisms is the aim of the simulation tool Hermes-3, which is based on the BOUT++ framework and is able to use the flux-coordinate independent (FCI) approach. Drift effects for non-axisymmetric domains are currently added to the code, opening the road to never before made investigations. Yet, the neutral dynamics essential for high density operation, high recycling regimes and detachment remain to be verified for stellarator geometries.
This contribution presents the status of the neutral verification effort of the Hermes-3 FCI implementation. We quantitatively compare the diffusion and convection of neutrals resulting from the numerical calculations with and without charge exchange collisions, and match them to analytically expected values in a 3D slab geometry. We check the charge exchange momentum transfer from background ions onto the neutrals for physical consistency.
In a second stage test we plan to further substantiate credibility of the Hermes-3 neutral model by also including field line curvature and surface interaction in a large-aspect-ratio approximated limiter configuration. The resulting neutral fluxes observed in Hermes-3 are then compared to solutions obtained by the well-established EMC3-EIRENE code.

Speaker: Annika Lisa Stier (MPPL)

408 4.051 Implementation of core-edge coupling using ETS-PAF and SOLPS-ITER

When modelling the behaviour of current and future tokamaks, the core plasma is often treated separately from the edge/SOL, with different classes of codes used, with weak on non-existent coupling between the models. This work covers efforts to ensure consistency between the modelling of these separate regions by coupling a code suite describing the core plasma (ETS-PAF, a 1.5D core transport code including heating sources and transport models) with another code suite describing the edge plasma (SOLPS-ITER, a 2D plasma / 3D neutrals transport code modelling the behaviour of the edge and divertor plasma).

This work builds on earlier work coupling older version of the two codes (ETS6 and SOLPS5.0) which was paused because the underlying coupling infrastructure was replaced, and significant development of a wide grid version of SOLPS_ITER was planned.

The coupling relies on one of the pair of value and flux being calculated by the core code and then passed to the edge code as a boundary condition for the edge equations, one or more time-steps of the edge code being performed, and then the other member of the pair being passed from the edge code to the core code and then used as the boundary condition for the core equations where one or more time-steps would be performed.

Initial results of the coupling will be presented, discussing continuity of profiles as well as particle and energy conservation across the coupling surface.

Speaker: David Coster (MPG-IPP)

409 4.052 Towards Divertor and Heat Exhaust Modelling for Quasi-Isodynamic Stellarators with JOREK

Heat exhaust remains a critical challenge for stellarator-based fusion reactors, from experimental machines like W7‑X to reactor candidates such as SQuID. Although significant progress has been made in tokamak divertor physics, including promising regimes such as detachment and X-point radiators (XPR), the extension of these solutions to the inherently three-dimensional stellarator geometry involve substantial challenges, both for the computational stellarator modelling and obtaining experimental detachment.
This contribution presents recent advances in terms of improvements to divertor and heat-exhaust modelling capabilities for quasi-isodynamic (QI) stellarators using the non-linear extended-MHD code JOREK, which only recently started being used to study detachment and similar phenomena. The primary technical challenge lies in implementing sheath boundary conditions (SBC) that accurately capture plasma-wall interactions in the complex non-axisymmetric 3D stellarator geometry. Unlike tokamaks, where 2D axisymmetric approximations are sufficient, stellarators require a full 3D treatment of the magnetic field structure and its intersection with divertor surfaces.
Our approach involves developing and validating a projection matrix that maps the fluid plasma onto the 3D divertor wall grid. To ensure code validation, we first benchmark against known tokamak solutions by simulating 2D axisymmetric cases within the 3D stellarator framework. Once validated, the 3D SBC implementation will be one of the key enablers of systematic studies of plasma detachment physics in stellarator divertors.
This work directly addresses power-exhaust requirements for future stellarator reactors and creates synergies with ongoing research on island divertors. By extending JOREK's proven capabilities to stellarator geometry, we aim to provide a validated tool for investigating detachment scenarios and radiative divertor solutions that are essential for stellarator-based power plants. Initial results from validation studies and preliminary 3D simulations will be presented, along with planned investigations of detachment physics in QI stellarator configurations.

Speaker: Elias Waagaard (MPPL)

410 4.053 Global Fluid Simulations in the W7-X Divertor: Development and Initial Benchmarking

For the design of future experimental devices as well as for the first fusion devices, reliable simulation of the plasma in the scrape-off layer (SOL) is of significant importance. The current state of the art plasma code for stellarator geometries, EMC3-Eirene, is however showing significant disagreement compared to experimental observations [1]. The discrepancies might be explained by missing physics in the EMC3-Eirene physics model, such as magnetic field errors or drifts. This work utilizes the Hermes-3 code [2] – a flexible fluid model which simulates an arbitrary number of ions and neutrals. Additionally, recent developments that allow the simulation of stellarator geometries, including island divertors, will be discussed. Hermes-3 can use the flux coordinate independent (FCI) approach [3, 4] for parallel derivatives, which is the underlying method used in this work.

Recent developments in the Hermes-3 code for stellarators will be presented, as well as Wendelstein 7-X simulations using a model comparable to EMC3. These will subsequently be compared to EMC3-Eirene simulations.

[1] D. Bold et al., Nuclear Fusion 62 (2022): 106011
[2] B. Dudson et al., Comp. Physics Commun. 296 (2024) 108991.
[3] F. Hariri et al. Physics of Plasmas 21 (2014).
[4] B. Shanahan et al. Journal of Physics: Conference Series. 775 (2016) 012012

Speaker: David Bold (IPP Greifswald)

411 4.054 Coupling of the neutral code Eirene and the gyrokinetic code GENE-X.

The full-f, gyrokinetic edge and scrape-off-layer turbulence code GENE-X has been coupled with the Monte Carlo neutral code Eirene. A kinetic treatment of the neutrals was chosen for the improved accuracy over fluid neutrals. This is particularly important for low collisionality regimes expected in reactors. It is impractical to call Eirene every GENE-X timestep, so the neutral source won’t update on turbulent timescales. The coupling uses a similar method as SOLPS-ITER: densities, temperatures, velocities, and particle fluxes are passed to Eirene; ion and electron particle, momentum and energy sources are passed from Eirene to GENE-X. Currently, the output from Gene-X is averaged toroidally when passed to Eirene and the Eirene output is copied to the different poloidal planes in GENE-X, but future work will change Eirene to a 3D grid. Initial results for a DIII-D case with and without neutrals will be presented.

Speaker: Jonathan Roeltgen (ExoFusion)

412 4.055 Experimental Determination of Diffusion Coefficients near a Magnetic X-Point in a Partially Ionized, Partially Magnetized Plasma

Analyzing charged particle transport near a magnetic X-point in the presence of neutral species is crucial for understanding and controlling tokamak plasmas. To study plasmas in a similar environment, a steady-state, low-temperature plasmas are generated in the MAgnetic X-poInt siMUlator System (MAXIMUS) [1] using a DC-heated cathode. Langmuir probes are utilized to obtain spatially resolved electron energy distribution function (EEDF) profiles near the magnetic X-point under various plasma conditions. From the electron density and temperature measured around the magnetic X-point, the collisional source and sink terms for the continuity equation are calculated using the ionization cross section [2] and the recombination rates [2,3]. The differential term in the continuity equation is also evaluated from the electron density profile, assuming Fick’s law and a steady-state condition. Consequently, the diffusion coefficient in Fick’s law can be estimated. In this work, we present experimentally determined diffusion coefficients around the magnetic X-point in MAXIMUS under various plasma conditions.

This work was supported by National Grant no. RS-2022-00155917 and NRF-2021R1A2C2005654.

References
[1] Lim, Yegeon, et al. "New low temperature multidipole plasma device with a magnetic X-point and its properties." Plasma Sources Science and Technology 29.11 (2020): 115012.
[2] Bogaerts, Annemie, Renaat Gijbels, and Jaroslav Vlcek. "Collisional-radiative model for an argon glow discharge." Journal of applied physics 84.1 (1998): 121-136.
[3] Li, C. Y., Y. Z. Qu, and J. G. Wang. "State-selective radiative recombination cross sections of argon ions." Journal of Quantitative Spectroscopy and Radiative Transfer 113.15 (2012): 1920-1927.

Speaker: Hoiyun Jeong (Korea Advanced Institute of Science & Technology)

413 4.056 Progress in edge turbulence modelling with self-consistent neutrals in SOLEDGE3X

Mean-field fluid edge codes are widely used for planning operations of magnetic fusion devices. Turbulence, a key factor in edge transport, is represented in these codes via diffusion coefficients tuned to match the Scrape-Off Layer (SOL) power fall-off length to existing experimental scaling laws, which are derived from attached plasmas. However, experimental results indicate that high-density regimes significantly impact SOL transport [1], limiting the applicability of the existing scaling laws when modelling high-density regimes. High edge density will be necessary for the operation of high-performance devices, in order to achieve enough plasma-neutral interactions to protect the plasma-facing components. To further inform the choice of diffusion coefficients in mean-field simulations of high-density regimes, turbulence and neutrals should be modelled self-consistently. In this work, we demonstrate edge fluid simulations with the SOLEDGE3X 3D turbulence code including a self-consistent 3-moment fluid neutrals model as derived by Horsten et al. [2], applied to TCV geometry with a long outer divertor leg [3]. This model is a significant improvement from the previous 1-moment fluid neutrals model implemented by Quadri et al. [3]. In mean-field simulations, the model is shown to reduce error between the plasma sources computed by fluid neutrals and EIRENE kinetic neutrals, and results in a less deeply detached plasma which is in better agreement with experimental results. First results of 3D turbulent simulations with the 3-moment fluid neutrals model concur with the results of the 1-moment fluid neutrals model, with detachment being associated with an increase in the SOL transverse transport and leading to a broadening of the SOL power fall-off length. Turbulent structures in the outer divertor leg are found to be qualitatively similar to experiment, with the size and velocity of the fluctuations depending on the divertor density regime.

[1] T. Eich et al. 2020 Nucl. Fusion 60 056016
[2] N. Horsten et al. 2017 Nucl. Fusion 57 116043
[3] V. Quadri et al. 2024 Nucl. Mater. Energy 41 101756

Speaker: Bridget McGibbon (CEA)

414 4.057 Predictive SOLPS-ITER simulations of tokamak COMPASS Upgrade: progress from structured grids to Wide Grids

Faithful simulations of the tokamak edge plasma, particularly in view of detachment access, are crucial for the design of future fusion reactors. The COMPASS Upgrade tokamak, currently under construction at IPP CAS, Prague, will feature reactor-relevant magnetic fields and target energy fluxes. [1] Previously, predictive SOLPS-ITER [2] simulations have been presented of two COMPASS-U H-mode scenarios, ITER-like #24300 and high-performance #5400. [3,4] Performed using the "structured grid" SOLPS-ITER version, these simulations suffered the drawback that protruding divertor baffles prohibited the construction of a sufficiently wide SOL grid. As a result, a large fraction of the input power (27 % and 47 %, respectively) was deposited on the outer boundary of the computational region, effectively escaping the simulation. It was suspected that this missing power may be responsible for the observed partial detachment, which occurred in spite of the lack of impurity seeding.

In this contribution, we present new simulations of these H-mode scenarios, performed using the SOLPS Wide Grids version [5], where the plasma domain covers the entire edge plasma up to the first wall. Power balance assessment and loss factor calculation show that the new solutions differ significantly from those achieved with the older code version. The plasmas are attached with high peak heat loads (increased from 12 MW/m2 to 25 MW/m2 on the outer target of scenario #24300) and low pressure losses. This is in line with sheath-limited conditions achieve at similar separatrix parameters on Alcator C-Mod, the closest tokamak relative of COMPASS Upgrade. [6] Power injected into the simulation region is recuperated entirely as incident on the divertor targets or the first wall in the form of plasma heat, neutral heat or radiation. Our results underscore the possible positive impact in switching to the Wide Grids SOLPS-ITER version and the importance of impurity seeding for target heat load mitigation in COMPASS Upgrade H-modes.

[1] P. Vondracek et al, Fusion Engineering and Design 169 (2021) 112490
[2] S. Wiesen et al, Journal of Nuclear Materials 463 (2015) 480-484
[3] I. Borodkina et al, SOLPS-ITER predictions for power and particle exhaust in COMPASS Upgrade tokamak, poster contribution to the 26th International Conference on Plasma Surface Interactions in Controlled Fusion Devices, 2024, Marseille, France
[4] M. Komm et al, Nuclear Fusion 64 (2024) 076028
[5] W. Dekeyser et al, Nuclear Materials and Energy 27 (2021) 100999
[6] B. Lipschultz et al, Fusion Science and Technology 51 (2017) 369-389

Speaker: Kateřina Hromasová (Institute of Plasma Physics of the Czech Academy of Sciences)

415 4.058 Buffering of transients via fuelling and impurity seeding in a detached MAST-U Super-X divertor

Transient events such as ELMs, sawteeth and H-L back-transitions are inevitable in some form in next step reactor scale tokamaks. Such transients lead to heat fluxes on divertor target tiles which risk cracking and further damage [1, 2]. Mitigation of these fast transient heat fluxes has been shown to be possible in part by increasing neutral gas pressure in the detachment cloud [3, 4] and by impurity seeding in the scrape off layer [5]. Demonstration and advanced understanding of buffering capability for small transients would facilitate the consideration of small ELM regimes for future plasma scenarios.
In this work progress on analysis of transient buffering by detachment and impurities at MAST-U is shown. Ultrafast divertor spectroscopy has been utilised to measure how the speed of burn-through of the detachment front changes with detachment pressure, transient energy and burn-through state. These measurements are compared to SOLPS-ITER simulations of transient burn-through speeds. Further attempts are made to measure the atomic and molecular neutral density in the detachment cloud which is expected to relate to its buffering capacity.
Recent studies on MAST-U are presented which show the effectiveness of nitrogen seeding in buffering transients in a tightly baffled Super-X divertor, with attempts to quantify the reduction in heat flux achievable without significant degradation of core performance. The buffering of ELM and sawteeth instabilities is compared, as well as the buffering performance of the Super-X compared to the conventional divertor.
[1] Eich et al., Nucl. Mater. Energy, 12 84-90, (2017)
[2] Linke et al., Nucl. Fusion, 51 073017, (2011)
[3] Federici et al., Nucl. Fusion, 64 126068, (2024)
[4] Flanagan et al., Nucl. Fusion, 65 116031, (2025)
[5] Komm et al., Nucl. Fusion, 63 126018, (2023)
Acknowledgements
This work was supported by the Engineering and Physical Sciences Research Council (EPSRC) (Grant Number EP/S022430/1) and has been part-funded by the EPSRC Energy Programme (Grant Number EP/W006839/1).
This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No. 101052200—EUROfusion). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the European Commission can be held responsible for them.

Speaker: Mr Jack Flanagan (University of Liverpool)

416 4.059 Impact of inner leg total flux compression on divertor detachment in MAST-Upgrade

In a fusion power plant it is paramount to control the heat flux reaching the targets. Most of the exhaust heat flux is directed to the outer leg, that hence is often the main focus, but for increasing machine size a correct characterization of the heat flux on the inner target is important, even more for spherical tokamaks as the inner target wetted area is more limited.
Previous studies on MAST-U have shown that the inner leg detaches with a sharp movement from the target to a region near the X-point, while it detaches gradually on the outer leg [1]. The detachment onset and front movement sensitivity is predicted by the detachment location sensitivity (DLS) model [2,3]: if the magnetic field reduces towards the strike point (outer leg) the front movement should be gradual from target to X-point. Conversely (inner leg), the front should move sharply. The experiments confirmed these behaviors.
The aim of this study is to investigate the impact of changing the inner leg configuration on the detachment process. We varied the inner leg orientation from horizontal to ~60 degrees while maintaining the same core shape, and made it almost vertical allowing for more significant changes (negative triangularity). In these experiments the infrared video bolometer (IRVB), that can reconstruct the total emissivity profile in the divertor with high spatial resolution [4], was the prime diagnostic. Preliminary evidence suggests that even an inner leg close to vertical causes a sharp transition from target to X-point and detachment starts at lower upstream density for a more horizontal inner leg, both in agreement with the DLS model. It also appears that the radiation at the target elongates when the DLS profile becomes locally marginally stable, and that the X-point radiation elongates towards the inner target for increasing poloidal length of the inner leg, pointing to a dominant role of neutral dynamic for the location of the cold radiating regions. Experiments in lower single null rather than balanced double null have been performed to obtain a higher and more reliable heat flux on the inner leg, with results currently being analyzed.
We also decreased the density from an already detached plasma, observing the reattachment process. Reattachment seems to be also characterized by a sharp movement of the front, similar to detachment. Further analysis and experiments will provide deeper insight into the physics of inner leg detachment and the factors governing its stability.

Speaker: Fabio Federici (ORNL)

417 4.060 Preliminary X-Point Radiator Experiments on ST40

Recent experiments on ST40 have begun to explore the X-point radiator (XPR) regime in H-mode discharges, with modelling support via SOLPS-ITER. Previous studies on TCV, AUG, WEST, and JET suggest that the XPR regime offers an operating space which reduces divertor heat loads while maintaining good core confinement through the introduction of a stable impurity radiation front above the X-point [1,2,3,4].

Early observations suggest as the radiative region forms, the radiation front crosses the X-point and resides on the core side, cooling the plasma edge significantly. This radiative region appears to form a banded emission structure which consists of impurity radiation then hydrogenic radiation, like what has been observed on AUG [1,2]. This appears to coincide with the detachment of the outer divertor leg. Compared to non-XPR discharges, these plasmas exhibit lower divertor heat fluxes (confirmed by IR thermography, and Langmuir probe measurements).

Initial spectroscopic analysis and SOLPS-ITER modelling suggest that the impurity radiation front mainly consists of sputtered carbon, which has been compressed towards the X-point via high field side (HFS) and low-field side (LFS) D₂ fuelling, and appears to lead to the formation of this radiative region.

Extrinsic seeding with lithium (via impurity powder dropping) and neon (via LFS puffing) in this regime are also under investigation. IR thermography, and Langmuir probe measurements suggest that this seeding further reduces divertor heat loads. SOLPS-ITER modelling with a lithium neutral particle source positioned on the upper outer divertor walls, suggests that lithium can be localised within the divertor region while avoiding upstream dilution and maintaining core performance. Overall, this regime appears to be promising for long-pulse operation by reducing erosion and power loading on plasma-facing components, without degrading core conditions.

References:

[1] – O. Pan, et al. “SOLPS-ITER simulations of an X-point radiator in the ASDEX Upgrade tokamak”, Nucl. Fusion,  63  016001 (2023).

[2] – M. Bernet, et al. “The X-Point radiating regime at ASDEX Upgrade and TCV”, Nuclear Materials and Energy, Volume 34, 101376 (2023)

[3] – N. Rivals, et al. “Experiments and SOLEDGE3X modeling of dissipative divertor and X-point Radiator regimes in WEST”, Nuclear Materials and Energy, Volume 40, 101723 (2024)

[4] – M. Bernet, et al. “Power exhaust by SOL and pedestal radiation at ASDEX Upgrade and JET”, Nuclear Materials and Energy, Volume 12, pp 111-118, (2017)

Speaker: Joseph Bryant (Tokamak Energy Ltd)

418 4.061 Assessment of the H-mode DTT Scenario A in partially detached divertor conditions

The Divertor Test Tokamak facility (DTT) is designed to investigate power exhaust in fusion reactor-relevant conditions. The schedule of the different operational phases foresees an increase in the installed auxiliary power up to the final full power level. In the very early phase A, the standard single Null (A-SN) plasma scenario is defined by reduced parameters: $B_T=3T$, $I_p=2MA$, and $P_{aux}=P_{ECRH}=8MW$. In the A-SN, the H-mode transport is supposed to be achieved in an integrated scenario taking into account both the core plasma target performances and acceptable Scrape Off Layer (SOL) and divertor conditions. On one hand, the threshold on the power crossing the separatrix to operate in H-mode is an increasing function of the average density: a low density value is thus preferable by considering the limited available auxiliary power. On the other hand, the outboard separatrix density value, and in turn the line average density, is determined by the achievement of the roll-over point to get a partially detached divertor plasma.
In this contribution, the analysis of the SOL is presented to assess the compatibility of a partially detached divertor plasma with H-mode confinement in the DTT Scenario A-SN. The study is carried out using the SOLPS-ITER code suite with a pure D plasma. The transport coefficients are set by considering the power decay length obtained according to the methodology presented in [1]. A density scan has been performed to estimate the roll-over density. In the simplest case without drifts, the inner target is characterized by the most demanding conditions in terms of peak target temperature, due to the location of the stagnation point, quite close to the outer target. The inclusion of drift strongly affects this behaviour, leading to a drop in the electron temperature on the inner plate. Indeed, the Bx∇B drift point downward, with a particle flux from the outer to the inner divertor and a change in the stagnation point position towards the outboard midplane. The roll-over is obtained at $f_{GW}=0.9$ on both targets, corresponding to $n_{e,sep,omp}$=5$\times 10^{19}\;m^{-3}$. By considering the scaling for metallic tokamak[2], the threshold power to enter the H-mode confinement is approximately $P_{LH}=6MW$. Therefore, an operational space can be found to operate DTT in H-mode with partially detached plasma conditions in the early first phase, but with a quite narrow window.

[1]L.Balbinot,Nucl.Mat.En, 34(2023)101350
[2]E. Delabie et al, Status of TC-26: l-H/H-L scaling in the presence of metallic walls ITPA Meeting September 2017

Speaker: Dr Giulio Rubino (a ISTP-CNR, via Amendola 122/D, Bari, 70126, Italy)

419 4.062 Development of a Large-Diameter High-Density Hydrogen Plasma Source and Initial Heat-Flux Measurements in Pilot GAMMA PDX-SC

Research on fundamental plasma physics in the divertor region for DEMO divertor design has required divertor simulation experiments using linear devices. Existing linear devices are still unable to fully replicate DEMO-class divertor plasmas. To address this limitation, we have constructed a new linear device, the Pilot GAMMA PDX-SC (PGX-SC), designed to verify high-density plasma sources and heating system for the DEMO divertor simulation test facility. PGX-SC has two superconducting coils that generate a steady mirror magnetic field of up to 1.5 T and uses a hot-cathode arc discharge to produce high-density, large-diameter hydrogen plasmas. We are tackling the generation of large-diameter, high-density plasma, which has been difficult to achieve until now, using a hot-cathode arc discharge plasma source which incorporates a 150 mm diameter LaB6 cathode, five intermediate floating electrodes, and a copper anode. The inner diameters of the floating electrodes and the anode are tailored to follow the magnetic field lines generated by the superconducting coils. The large cathode size, aligned with the magnetic-field geometry, also plays a key role in achieving a large-diameter plasma.
So far, hydrogen plasmas generated by the plasma source have had electron densities of approximately 1019 m-3. To reach the target density of 1020 m-3, further improvements to the plasma source are underway. In recent discharge experiments, the end plate potential has played a crucial role in drawing plasma into the end region, and it is expected to also play an essential role in achieving higher plasma density.
In addition, two heat load measurement systems were installed in the end region of PGX-SC: a calorimeter located in front of the end plate and a thermocouple-based system to evaluate heat load from the end-plate temperature. As an initial result, the heat load to the end plate during ICRF heating was successfully measured. These measurement systems constitute essential experimental infrastructure for future DEMO-class divertor simulation experiments and studies of plasma–wall and plasma–surface interactions.
This presentation reports on the status of plasma source development and initial results from end-plate heat-load measurements.

This work was partly supported by the NIFS Collaboration Research program (NIFS23KUGM174, NIFS25KFFT001), JST SPRING, Grant Number JPMJSP2124.

Speaker: Reina Miyauchi (Plasma Research Center, University of Tsukuba)

420 4.063 Response of JT-60SA radiation profile to gas injection toward radiation feedback control by transient analysis using integrated divertor code SONIC

Feedback control of radiative power from in divertor plasmas by D₂ fuel gas injection, through carbon radiation, is considered for the next operation phase of JT-60SA. The location of gas injection ports can affect the radiation response. For feedback control, the radiation profile must be converted into measurement signals, taking into account the diagnostic fields of view. In JT-60SA, resistive bolometers with both wide and narrow viewing angles per channel will be installed to evaluate divertor radiation power. The bolometer signals provide a linear measure of the total radiated power within the viewing volume [1][2]. This study aims to evaluate the response of radiation profile and bolometer signal with stepped gas injection for feedback control. In following, signal responses to stepped D₂ gas injection from different injection port locations were analyzed using the integrated divertor code SONIC. In JT-60SA, four gas injection ports are available on the poloidal cross section: (a) the upper port, (b) the lower port, (c) the inner-divertor port, and (d) the outer-divertor port. These ports enable injection of fuel or impurity gas into the vacuum vessel. For a low-density attached plasma under conditions where the electron temperature at the strike point exceeds 300 eV, the time evolution of radiation power in the outer and the inner divertor plasma was simulated with the stepped D₂ gas injection with 6.0 ×10^21 /s from four different injection ports. Among the four cases, it was found that a prompt increase in radiation power by injection from the private region port ((c) or (d)) compared to that by injection from the common flux region port ((a) or (b)). Then, it took 40 ms and 80 ms from start of stepped gas injection for the total radiation power in the outer and the inner divertor plasma to become constant, respectively. Considering that it takes 100 ms for the D2 gas to travel from the gas valve to the outlet of the injection port based on JT-60U results [3], it is estimated that it takes 140 ms – 180 ms for the radiative power to increase from the injection signal. Moreover, the time required for the radiative power to be constant depends on plasma density and gas species. To address this, additional simulations with various plasma densities and various gases, such as neon, will be presented.

[1] R.Sano et al., Rev.Sci.Instrum. 89,10E104(2018)
[2] https://www.jt60sa.org/
[3] H. Tamai et al., Fusion Eng.Des. 39–40(1998)163–167

Speaker: Ryuichi Sano (QST)

421 4.064 The effect of fluid neutral momentum on 1D detachment burn-through for STEP

The STEP program run by UK Industrial Fusion Solutions Ltd aims to deliver a tritium self sufficient prototype fusion power plant generating 100 MW net electric power based on the spherical tokamak. The compact spherical geometry of STEP’s SPP-2 design raise significant exhaust challenges, with powers crossing the separatrix reaching Psep ≈ 110−140 MW, corresponding to Psep/R0 ≈ 25−32 MW/m. To overcome these challenges solutions, such as a double null with long out legged divertor, will be employed to achieve sufficient detachment and protect plasma facing components [1]. While these measures are predicted to achieve acceptable steady-state power to the divertor targets, transient power loads from fluctuations in core fusion/radiation power or vertical displacement-induced disconnected double null effects may lead to burn-through of the detachment front, and unacceptable power loading to material surfaces.

High-fidelity (SOLPS-ITER) simulations of transient burn-through remain prohibitively expensive to explore the wide parameter space of reactor scale devices. The main cause of this expense is the kinetic treatment of neutral species. To provide initial screening for STEP and comparison with analytical models [2], it is attractive to adopt the more computationally tractable method offered by reducing dimensionality to 1D and adopting a fluid treatment of the neutral species. In this study, we used the multi-fidelity Hermes-3 fluid code to study power transient burn-through in STEP-relevant 1D exhaust scenarios. We investigated the effect of a rapid rise in power entering the SOL (up to 4× power increase over a timescale of several miliseconds). Simulations found two regimes of front movement: slow burn-through driven by neutral ionisation, and faster compression in which charge exchange exerts a force on the neutrals, shrinking the neutral cloud. This imbalance leads to the detachment front motion tracking the rise-time of the power increase. In contrast, simulations in which the pressure imbalance was artificially removed saw much slower front motion with burn-through dominated by neutral ionisation.

As this phenomenology could be caused by a lack of cross-field transport and the fluid neutral description, we compare our 1D simulations against 2D SOLPS-ITER MAST-U simulations with kinetic neutrals. We then investigate reservoir models [3] to improve the agreement between the 1D fluid neutral and 2D kinetic neutral simulations, both in steady-state and transient power pulse scenarios.

[1]S.S. Henderson et al., Nucl. Fusion 65 016033 (2025).
[2]S.S. Henderson et al., Nucl. Fusion 64 066006 (2024).
[3]G.L. Derks et al., Plasma Phys. Control. Fusion 66 055004 (2024).

Speaker: Mr Lloyd Baker (University of York)

422 4.065 SOLPS-ITER simulations for SPARC advanced divertor configurations

The SOLPS-ITER is employed to model advanced divertor configurations in the SPARC tokamak aiming to explore potential solutions to the power exhaust challenge [1]. The configurations studied include the standard Single Null Divertor (SND), Super-X Divertor (SXD), and X-Point Target Divertor (XPTD).
The outer targets of the SXD and XPTD configurations benefit from their advanced geometries, as the peak heat flux is reduced to approximately 30% of that in the SND case (q_surf~80MWm^-2) without seeding of impurities. Subsequently, simulations were performed with neon and argon impurities injected from the Private Flux Region (PFR). The results indicate that when the impurity seeding rate reaches 5×10¹⁹ neon atom s⁻¹ or 2×10¹⁹ argon atom s⁻¹, the outer targets become detached. However, the inner target faces a severe power exhaust challenge that the peak value of heat load remains ~30 MWm^-2, even as the seeding rate increases to 3×10²⁰ neon atom s^-1.
The simulations indicate that when impurities are puffed from the PFR, even a small seeding rate is sufficient to detach the SXD and XPT outer targets. As a result, a deuterium ion flow develops in the PFR, which direction is from the High-Field-Side (HFS) toward the Low-Field-Side (LFS), driven by the pressure reduction at the outer target. In this situation, most of the injected impurities are ionized within the PFR and are subsequently advected by the strong deuterium flow toward the outer target. Consequently, impurities cannot effectively penetrate the PFR, even puffing at the HFS PFR, and it is hard to enter the HFS SOL region, leading to weak radiation in HFS SOL that the inner target remains attached even if the seeding rate is significantly increased.
In order to explore possible ways for reducing the inner target heat load, a series of numerical simulations are performed, including adjustments to the inner target geometry, modifications to the in–out power sharing, and changes to the impurity puffing locations. A noticeable result is when impurity puffing from the HFS SOL is optimized, both inner and outer target can achieve full detachment. The corresponding seeding rate is ~2×10²⁰ neon atom s^-1 and the resulting q_surf reduced to below 10MWm^-2. In addition, the preliminary results with the activation of E×B drift, which can be expected to drive a flow from LFS to HFS in PFR when B×∇B points towards the lower X-point, are presented.

[1] A.Q. Kuang et al. Journal of Plasma Physics 2020.

Speaker: Dr Haosheng Wu (NEMO Group, Dipartimento Energia, Politecnico di Torino, Corso Duca degli Abruzzi 24, 10129 Torino, Italy.)

423 4.067 Lithium Vapor Cave Performance Uncertainty Due to SOL Cross-Field Transport and Deuterium Recycling Variations

The lithium vapor cave is a detached divertor concept that uses a single Private Flux Region (PFR) baffle to confine a dense lithium vapor cloud, reducing target heat flux while minimizing contamination of the main plasma. Plasma flows are driven by deuterium puffing to control lithium transport from the divertor to the main plasma. Previous SOLPS-ITER modeling demonstrated sufficient outer target heat flux reduction, reducing an unmitigated 90 MW/m$^2$ to less than 10 MW/m$^2$. This was achieved with acceptable lithium concentrations at the last closed flux surface (n$_{Li}$/n$_e$<0.05) across variations in divertor geometry, target recycling, upstream plasma conditions, lithium evaporation, and fueling locations.

Within the SOLPS-ITER code, both cross-field anomalous transport and recycling are set by user-defined, spatially varying coefficients. This work systematically examines the impact of varying both cross-field transport and recycling on lithium contamination predictions. The role of deuterium recycling is considered since deuterium recycling is known to decrease with lithium injection and can strongly influence poloidal divertor flows that determine contamination levels. Simulations indicate that decreasing the radial particle transport by a factor of four increases the prediction for upstream lithium density by a factor of 2.1, when adjusting the external gas puff to maintain a similar outer midplane separatrix electron density. Varying the recycling coefficient over the potentially accessible range from high to low recycling, results in a factor 2.4 variation in upstream lithium concentration, at fixed upstream density, with lower recycling corresponding to lower concentration. Across all recycling regimes and cross field transport assumptions, the lithium vapor cave was successful at target heat flux reduction with sufficient lithium evaporation. The solutions were also always found to improve with greater deuterium gas puffing which resulted in a more favorable combination of impurity forces (mostly friction and thermal forces) acting on the lithium in the SOL.

These results demonstrate that both cross-field transport and recycling assumptions introduce significant uncertainty in lithium concentration SOLPS-ITER predictions for the lithium vapor cave. Bounding these uncertainties provides guidance for divertor design optimization, helping ensure sufficient heat flux reduction while limiting lithium contamination.

This work is sponsored by DOE Contracts No. DE-AC02-09CH11466

Speaker: Eric Emdee (Princeton Plasma Physics Laboratory)

424 4.068 The X-Point radiating regime at JET with mixed impurities

Power exhaust is a crucial issue for future fusion reactors and X-point radiation might provide a valuable solution. X-point radiators (XPR) are usually initialized by impurity seeding and are observed in almost all currently operating tokamaks. In JET they were first observed in 2016. In the recent JET campaigns, including the final DT campaign (DTE3), the XPR was investigated in detail.

Several seed impurities were injected in order to trigger an XPR, such as nitrogen, neon, argon, and combinations thereof. With pure neon or argon seeding the plasma exhibits dithering between H- and L-mode, even at heating powers of up to 26MW. This instability was suppressed by N₂ seeding or mixed impurities. Since N₂ was prohibited in DT operations, Ar+Ne mixtures proved optimal, enabling stable H-mode and XPR formation. The impact of the seed impurities used and the benefits of the mixed seeding are analysed with the help of SOLPS modelling.

For the first time at JET, a movement of the XPR inside the confined region was observed. This movement was tracked by the horizontal bolometer camera, using the algorithm developed at AUG, with the XPR moving mainly between two lines of sight. The XPR location is detected to a sub-channel accuracy of 4mm (with a channel spacing of 8cm) and can be provided in real time to the control system. A PI controller using Ar seeding as actuator (Ne pre-programmed) was implemented, with gains optimized via system identification. The reaction time of the XPR location to a change in the seeding rate is about 1s. Faster external perturbations, such as drops in heating power or pellet injection, are first buffered by a movement of the XPR, and can then be effectively counteracted by the slower control.

The active control of the XPR helped to efficiently establish the same power exhaust conditions when moving from pure D plasmas to DT plasmas. The plasma performance, initially low without seeding (H98≈0.65), did not degrade with impurity injection. Notably, the edge kinetic profiles are, within measurement uncertainties, not affected by the strong seeding, while ELMs become fully mitigated.

The successful demonstration of the XPR in D and DT plasmas at JET, combined with its stable operation, makes a strong case for incorporating XPRs into future fusion devices. This scenario meets several requirements of a reactor, including high power dissipation, control of full detachment, and ELM mitigation, though not at highest confinement yet.

Speaker: Matthias Bernert (IPP Garching)

425 4.069 Detachment of high power discharges in TCV X-point target divertor

Future tokamak reactors are expected to generate divertor heat fluxes that greatly exceed the technological limits of plasma-facing components if unmitigated. While a partially-detached conventional single-null (SN) divertor is currently foreseen for ITER, alternative divertor concepts based on strategic magnetic shaping and neutral baffling are under investigation to improve power exhaust performance [1,2]. Among these concepts, the X-point target (XPT) divertor has shown significant promise in facilitating detachment onset and improving detachment front stability [3]. This work discusses experiments conducted on TCV (R=0.88 m, a=0.25 m, BT=1.4 T) to explore the parameter space of the XPT’s power handling capabilities. A high-power, high-plasma-current scenario was developed with up to 2.5 MW of injected electron cyclotron heating, while sustaining L-mode operation at low core line-averaged densities of 1.5–2 × 10¹⁹ m⁻³ (Greenwald fraction fGW=0.08–0.1). This scenario is enabled by a novel model-based density controller with real-time density profile estimation, together with an XPT divertor shape controller providing feedback control of the X-point separation in flux coordinates. The Lengyel metric, (PsepB/R)/nsep², which characterizes the divertor detachment challenge for a given scenario, approaches values relevant to future fusion reactors. Compared with the SN configuration, the XPT exhibits a substantial redistribution of impurity line emission and radiation during nitrogen seeding—an effect absent in the SN case—indicating an earlier onset of detachment. In addition, the peak parallel heat flux to the outer target is reduced by up to 50% in the XPT configuration without compromising the inner target. Guided by recent SOLPS-ITER simulations, which revealed the significant roles of macroscopic E×B drifts and X-point separation in the physical basis of XPT's heat flux mitigation mechanisms, a new set of field-reversal experiments at the same power was performed on TCV to validate these effects, and comparisons of these features will be presented.

[1] C. Theiler et al., Nucl. Fusion 57, 072008 (2017)
[2] K. Verhaegh et al., Commun. Phys. 8, 215 (2025)
[3] K. Lee et al., Phys. Rev. Lett. 134, 185102 (2025)

Speaker: Kenneth Lee (Swiss Plasma Center, EPFL)

426 4.070 SOLEDGE3X Predictions for Initial Lithium Vapor Cave Divertor Experiments on NSTX-U

The lithium vapor cave is a detached divertor design that aims to mitigate divertor heat flux via near-target lithium evaporation with a private flux region (PFR) baffle to contain dense lithium vapor [1]. The future implementation of a lithium vapor cave in NSTX-U is planned to be a staged process where the early iterations consist of a single, unbaffled tile insert comprising a 15$^\circ$ toroidal sector, designed to serve as a test bed for in-situ lithium evaporator technology, diagnostics, and to provide preliminary physics results. Such a system will produce a toroidally localized lithium vapor source in the divertor region. In preparation for these experiments, multiple elements of the evaporator technology are being developed [2], including the embedded capillary porous system for evaporation and liquid lithium reservoir level sensors. Furthermore, fluid edge modeling is being performed to predict the toroidal extent of divertor heat flux reduction in the single tile tests, to help determine the degree of toroidal coverage required.

The multi-fluid plasma edge code SOLEDGE3X is being used to examine the induced toroidal asymmetries. Preliminary 2D simulations of NSTX-U have been performed to benchmark SOLEDGE3X results with SOLPS-ITER simulations. A reference simulation without lithium injection has been developed where the unmitigated peak heat flux at the low field side (LFS) target is predicted to be ~32 MW/m$^2$ and at the high field side (HFS) target ~16 MW/m$^2$. A lithium source in the PFR evaporating at a rate of $10^{23}$ s$^{-1}$ reduces the peak heat flux to both the LFS and HFS target by about a factor of 2. These initial results indicate that early NSTX-U experiments will need to rely on gas puffing to retain lithium in the divertor, where there will not be baffling to restrict the particle flows. Comparisons with equivalent SOLPS-ITER modeling of the same NSTX-U discharge are ongoing work.

These simulations are being extended into 3D to examine the toroidal extent of the heat flux reduction, and the effect of deuterium gas puffing will be examined as was done in previous SOLPS-ITER simulations, to contain lithium in the lower divertor region [3]. Understanding these SOLEDGE3X simulation results for a toroidally localized lithium vapor source are crucial for understanding initial stages of lithium vapor divertor experiments, as well as informing subsequent installation stages.

[1] Emdee and Goldston NME 41 (2024)

[2] Parsons et al. JFE 44 (2025)

[3] Emdee and Goldston NME 34 (2023)

Speaker: Margaret Porcelli (Princeton University)

427 4.071 Optimization of the DTT Inner Limiter Through PFCflux Heat Load Simulations

In the Divertor Tokamak Test (DTT) facility, the Limiter Inboard First Wall (LIFW) plays a key role in shielding the standard inboard wall modules from direct plasma contact during plasma-limited operation, such as ramp-up and ramp-down phases, negative triangularity configurations, transient events, and accidental scenarios. The limiter design relies on modules that protrude radially toward the plasma and feature a toroidally contoured plasma-facing surface, with the objective of enhancing the spatial distribution of heat loads during wall interaction [1]. Although this design approach is expected to reduce localized thermal peaks, the definition of the most effective limiter shaping is still ongoing and calls for a quantitative evaluation of plasma-driven heat fluxes under realistic DTT operating conditions.
In this work, heat load simulations of the DTT Inner Limiter are performed using the PFCflux code to support the ongoing optimization of the LIFW geometry and to quantify the thermal loads associated with representative plasma scenarios. Magnetic equilibria corresponding to plasma-limited configurations are used as input to model the power flow in the scrape-off layer and its interaction with the limiter surface [2]. The resulting spatial and temporal distributions of heat flux are analyzed for different limiter shapes and plasma positions, with the aim of identifying configurations that minimize peak heat loads while preserving effective plasma protection.
The analysis shows that the limiter geometry, including toroidal shaping and radial protrusion, affects the spatial distribution of heat loads on the limiter surface. The predicted heat fluxes also exhibit sensitivity to the assumed plasma configurations, indicating that both geometric and operational aspects should be considered in the limiter design assessment.
This study represents a continuation of the limiter design activity in DTT, bridging preliminary geometric concepts and engineering validation. By integrating plasma heat load simulations into the design workflow, it contributes to the definition of a reliable Inner Limiter configuration capable of withstanding the demanding operational scenarios foreseen for DTT and future fusion devices [3].

1. De Luca R., et al. (2025). Thermo-structural assessment of the limiter inboard first wall design of the Divertor Tokamak Test facility. Fusion Engineering and Design, 219, 115281.
2. Firdaouss, M., et al. (2013). Modelling of power deposition on the JET ITER like wall using the code PFCFLux. Journal of Nuclear Materials, 438, S536-S539.
3. Romanelli, F.,et al. (2024). Divertor Tokamak Test facility project: status of design and implementation. Nuclear Fusion, 64(11), 112015.

Speaker: Francesca Zarotti (ENEA)

428 4.074 JOREK simulations on the X-point radiator formation and its movement in ASDEX Upgrade

The power exhaust problem in the future large-scale fusion reactors necessitates operational regimes that can avoid extreme heat fluxes onto the plasma-facing components. One promising regime is the X-point radiator (XPR), which relies on a highly radiative, cold and dense plasma volume forming above the X-point, and which can be accessed via impurity seeding. Experimentally, the height of the XPR can be controlled by tuning the seeding rate and heating power [M. Bernert et al., NF 2021].

This contribution presents axisymmetric (2D) simulations of the XPR regime in ASDEX Upgrade (AUG) using the nonlinear MHD code JOREK extended with a kinetic particle framework for the main species neutrals and nitrogen impurities. From the time-dependent simulations, the progression from attached divertors to complete detachment with the XPR formation is shown, highlighting the effects of the neutrals and impurities separately.

After the XPR is well-formed at the height of 6.8 cm, the fueling and seeding rates are adjusted so that the XPR remains stationary. From the stationary case, the seeding rate is then changed to see how the XPR location reacts. By increasing the seeding rate, the XPR moves further inside the main plasma, and a MARFE-like (unstable XPR) scenario is eventually achieved. In the core of the MARFE, the electron temperature is reduced to well below 1 eV. On the other hand, by reducing the seeding rate, the XPR moves vertically downwards and is gradually lost. As the XPR retreats, the ionization front shifts from above the impurity radiation peak to below it, confirming the importance of neutral density for the XPR development [U. Stroth et al., NF 2022].

These simulations show JOREK’s capability of simulating time-varying XPR, which will provide a baseline for the transition to 3D simulations, so the MHD activities and their interaction with the XPR can be studied.

Speaker: Mr Yu-Chih Liang (Max Planck Institute for Plasma Physics Garching)

429 4.075 Radiation Dependence on Divertor Leg Length in Detachment on DIII-D

Experiments performed on DIII-D demonstrate that extending the outer divertor leg allows for expanding the volume for impurity and hydrogenic/fuel radiation as required for dissipation in detachment. This is roughly in line with convective transport estimates although deep detachment can trend toward the peak radiation moving to the X-point. Convective estimates relate the spatial scale for dominant radiating impurities (e.g., C or N on DIII-D) for H-mode conditions with 6 MW of injected power suggest approximately 20 cm of divertor leg length is necessary for adequate dissipation. The radiated power is found to extend along the outer leg at approximately this scale in detachment in small core-volume DIII-D plasmas at 1MA and 2T with an elongated outer leg as measured by CIII (465 nm) emission and total radiated power. With increasing density, the strong peak in CIII radiation at the target pulls away with some target radiation remaining transiently due to ELMs. The radiating volume can ultimately collapse to the X-point pushing the divertor into deep detachment. SOLPS-ITER simulations similarly show an extended radiation pattern along the outer leg at detachment onset and into the detached state with it being pushed deeper toward the X-point at higher density. Future devices may allow for little to no core degradation when coupled to a detached divertor, thus determining the effectiveness of volumetric radiation for dissipation is crucial.
This work was supported in part by the US Department of Energy under DE-AC05-00OR22725, DE-AC52-07NA27344, and DE-FC02-04ER54698.

Speaker: Morgan Shafer (ORNL)

430 4.076 Heat load assessment on European low-aspect ratio DEMO (DEMO-LAR) first wall

The next generation of fusion reactors will face unprecedented conditions in terms of power exhaust from the burning plasma. Quantifying heat loads due to impurity radiation following line and bremsstrahlung emission, as well as recombination, is essential for the thermo-mechanical design of plasma-facing components. This work describes the development of an integrated framework for the assessment of radiative power deposition on reactor first wall, combining core, scrape-off layer (SOL) plasma, and neutral modeling with 3D radiation transport.
Within this framework, an interface was developed capable of loading radiation source data provided by ASTRA/STRAHL or SOLPS-ITER, which are then processed with the Monte Carlo radiation transport code Raysect-CHERAB. As a first example, we discuss here a set of steady-state plasma scenarios, where Xenon has been included as core radiator, while Argon accounts for the SOL contribution. Also, He is present in the plasma composition, which refers to a reactor producing about 2 GW fusion power. The edge volumetric energy source has been evaluated with the SOLPS-ITER code wide-grid version, which allows extending the plasma computation up to the chamber wall. When running coupled with the EIRENE code for neutral gas transport, the neutral load due to high-energy charge-exchange atoms can also be included on the top of radiative calculations. This allows a quite comprehensive assessment of thermal and erosion wall exposure, which might be critical for reactor design. Sources are produced by plasma codes in the form of 2D poloidal maps and then mapped to the full toroidal domain assuming plasma axisymmetry before feeding them as input to Raysect-CHERAB.
As a preliminary assessment we consider the EU-DEMO reactor for a low-aspect ratio configuration (DEMO-LAR) to test our workflow performance. Preliminary calculations suggest a power load onto the wall within the range of $0.08-0.2\,\mathrm{MW\,m^{-2}}$, with peak values localized on the outboard segment of the first wall. Due to the high recycling and detached conditions required for divertor protection, strong localized sources in proximity of target surfaces can lead to peak heat load of $2.30\,\mathrm{MW\,m^{-2}}$ near strike points.
The workflow proposed here aims primarily at the evaluation of radiative loading conditions to directly support the thermo-mechanical design and assessment of the EU-DEMO-LAR first wall. It is sufficiently flexible to allow, in perspective, extensions to other reactor-relevant applications such as the development of spectroscopic synthetic diagnostics.

Speaker: Matteo Maria Robaldo (Politecnico di Torino)

431 4.077 Reduced boundary models for the ITER Pulse Design Simulator

Predicting different key boundary plasma constraints such as peak heat load to divertor targets and first wall, density at the separatrix and impurity influx as a function of the measurable control parameters (i.e. divertor neutral pressure, impurity seeding rate etc.) is crucial to prevent damage to the divertor targets and achieve required core plasma performance in fusion devices. In the case of ITER, experimental scenarios will be designed using a Pulse Design Simulator (PDS) which is currently under development at the ITER Organization. The PDS will operate within the ITER Integrated Modelling Analysis Suite (IMAS) and is being constructed using the third incarnation of the Multiscale Coupling Library and Environment (MUSCLE3) [1]. A key requirement of the PDS is that it be capable of end-to-end scenario design on reasonably rapid timescales and must thus incorporate reduced model actors describing the plasma boundary in connection to a core plasma and relevant measurable control parameters.

Here we examine a variety of candidate boundary models, including Reverse 2-point modelling [2], scaling laws for key divertor and upstream parameters based on the well-established ITER database of high fidelity SOLPS-ITER simulations [3] and neutral networks (SOLPS-NN) [4] trained on the same database . The candidate reduced models are first compared amongst themselves and then tested against new SOLPS-ITER [5] datasets with an improved physics description of the ITER plasma boundary, including simulations for the Start of Research Operations (SRO) phase of the new 2024 ITER re-baseline, under conditions which lie far outside the original training set. These lower power, reduced performance conditions are in fact the most important region to explore from the point of view of PDS development given that they will be the first plasmas which will be encountered in the ITER Research Plan.

MUSCLE3 actors have been generated for each of the reduced models and first steps taken towards their integration into the PDS framework. Early scenario simulation results including these boundary plasma actors for representative SRO cases will be presented.
References:
[1] L.E.Veen et. al.,Computational Science ICSS 2020 425-438 (2020)
[2] P. C. Stangeby, Plasma Physics and Controlled Fusion 60 044022 (2018)
[3] H.D. Pacher et. al., Journal of Nuclear Materials 463 591–595 (2015)
[4] S. Dasbach, Phd Thesis, https://docserv.uni-esseldorf.de/servlets/DocumentServlet?id=69084 (2025)
[5] S. Wiesen et. al., Journal of Nuclear Materials 463 480-484 (2015)

Speaker: Mate Karacsonyi (ITER Organization, CS 90 046, 13067 St Paul Lez Durance Cedex, France; HUN-REN Centre for Energy Research, Institute for Atomic Energy Research, Budapest, Hungary)

432 4.078 Neural Network Prediction of Scrape-Off Layer and Divertor Physics for SPARC

Understanding and controlling the detachment and power-exhaust behavior of SPARC is essential for ensuring divertor survivability in a device characterized by narrow scrape-off layers and high parallel heat fluxes.
We apply SOLPS-NN, a neural network surrogate model trained on SPARC SOLPS-ITER simulations, to explore the SOL and divertor operational space for SPARC.
In this context, we present and focus specifically on the detachment physics and divertor operational limits that the currently available model enables us to explore.
Using targeted scans of fueling, upstream density, and divertor neutral pressure, we analyze the operational domain when SPARC enters partial and full detachment. Our approach enables us to e.g. verify the appearance of how the rollover of particle and heat fluxes emerges in diagnostic signals, but also whether divertor heat-flux limits can be satisfied while maintaining favorable upstream confinement conditions. These two-dimensional reconstructions highlight the structure and movement of the detachment front and reveal inner–outer divertor asymmetries in electron temperature reduction, neutral accumulation, and momentum-loss processes, which can be used to inform fast real time control models.
We further quantify the extent of SPARC’s viable operational space and examine its robustness by scanning pivotal parameters within the surrogate model, i.e. the cross-field transport coefficients. These controlled variations probe uncertainties in predicted behavior, explore sensitivity to heat flux width scaling and assess how modest parameter adjustments might reduce model-reality gaps.
Using the surrogate model to unlock key physics insights, we demonstrate how neural network tools can swiftly map SPARC’s divertor operational space. These methods reveal detachment margins early and steer scenario development toward better protection of plasma-facing components.

Speaker: Stefan Dasbach (DIFFER)

433 4.079 Initial observation of lithium screening by the Scrape-off-Layer in ST40

A lithium vapor divertor is a proposed detached divertor design, predicted to be capable of handling reactor-relevant heat flux without reducing the performance of the confined plasma [1]. Lithium is screened from the main plasma by taking advantage of main ion scrape-off layer flows, controlled via external fuel gas puffing. The induced main ion flows act to increase the friction force acting on the lithium, allowing access to low heat flux divertor solutions with minimal upstream concentrations of lithium in SOLPS-ITER simulations [2]. The fundamental physics of these predictions can be tested in present-day machines via external lithium powder injection.

In the 2025 campaign, an impurity powder dropper (IPD) was installed on ST40, a high-field spherical tokamak operated by Tokamak Energy [3]. To explore the aforementioned mechanism of lithium screening, dedicated experiments injecting lithium via the IPD were conducted to test the penetration of lithium powder and transport of ionized lithium in the plasma. Preliminary analysis comparing the lithium scrape-off layer transport in various plasma scenarios suggest that the injected lithium is strongly screened by the scrape-off layer. Upstream concentration is observed to respond to the difference in gas puff locations and scrape-off-layer conditions. Experimental evidence will be presented along with interpretive analysis by SOLPS-ITER.

References:

[1] R J Goldston et al 2016 Phys. Scr. 2016 014017

[2] E.D. Emdee and R.J. Goldston 2023 Nucl. Fusion 63 096003

[3] McNamara, S. A. M., et al. "Overview of recent results from the ST40 compact high-field spherical tokamak." Nuclear Fusion 64.11 (2024): 112020.

Speaker: Xin Zhang (Tokamak Energy)

434 4.080 Characterizing Attached Alternative Divertor Configurations in Deuterium and Hydrogen on ASDEX Upgrade

Power exhaust remains a key challenge for steady-state operation of future tokamak reactors. More than 90% of the heating power must be dissipated through radiation before reaching the divertor targets to ensure sufficient material lifetime and low sputtering. At the same time, the divertor must enable efficient pumping of fuel and helium neutrals, and prevent impurity accumulation in the plasma core, which would degrade the confinement. Whether the conventional single null (SN) divertor, as planned for ITER, can meet these combined requirements in a reactor environment with sufficient headroom remains uncertain. Alternative Divertor Configurations (ADCs), which employ advanced magnetic field shaping, have been proposed to improve exhaust performance compared to the SN divertor. These ADCs can affect a number of divertor processes. The modified magnetic geometry can result in enhanced cross-field transport due to increased connection lengths, while the larger divertor volume enhances neutral interactions, promoting power and momentum losses and facilitating detachment access.

Over the past years, ADCs have been investigated on several devices, including TCV and MAST-U [1, 2]. Following a major upgrade, ASDEX Upgrade (AUG) is the first tokamak to combine ADC capability with reactor-relevant heat fluxes, a full tungsten wall, and a cryopump, enabling studies under conditions approaching those in future reactors [3]. During the first operational campaign, all planned ADC configurations were successfully realized. Dedicated parameter scans in heating power, fuelling, and nitrogen seeding were carried out for several configurations, including the X-Divertor (XD) and the Low-Field-Side Snowflake Minus (LFS SF–). In both configurations, detachment of the near scrape-off layer was achieved and well characterized. However, to fully assess the potential advantages of ADCs and to understand the governing physical mechanisms, a detailed characterization of attached divertor regimes is required as well.
This contribution presents experimental analyses of attached divertor heat fluxes and plasma conditions in these configurations and compares them with the conventional SN. The study combines infrared thermography, Langmuir probe, divertor Thomson scattering, and bolometry measurements in deuterium and hydrogen plasmas. The results are interpreted using simplified models for heat flux profiles [4] and plasma–neutral transport simulations with EMC3-EIRENE. First findings on attached heat flux distributions in the XD and LFS SF– configurations are presented, offering new insights into transport processes in ADCs.

[1] Theiler, C., et al., NF, 2017
[2] Verhaegh, K., et al., NF, 2022
[3] Zammuto, I., et al., FED, 2025
[4] Lunt, T., et al, NME, 2017

Speaker: Dominik Brida (Max-Planck-Institute for Plasma Physics)

435 4.081 Influence of target chamber geometry on impurity dynamics and neutral pressure in ARC-class divertor legs

Managing power and particle exhaust is a central challenge for a compact, high-field fusion pilot plant, where power-plant-level exhaust in a limited divertor volume leads to severe thermal and particle loading to the material surfaces. Achieving stable, dissipative divertor operation under these conditions (while simultaneously ensuring effective helium removal) requires divertor geometries, magnetic configurations, and pumping arrangements capable of accessing and regulating high-density, strongly radiating regimes with sufficient helium exhaust capability. To rapidly explore this parameter space, the isolated-leg “Box” SOLPS-ITER framework (originally developed for MAST-U [1-3]) is extended to ARC-relevant operating conditions. The Box approach enables controlled comparisons across geometric configurations while maintaining fixed upstream boundary conditions determined from the ARC physics basis [4].

This work introduces several advances to the Box model framework. First, realistic baffle and vessel wall contours are incorporated, together with flexible specification of puffing and pumping locations, enabling targeted studies of neutral closure and divertor pumping efficiency. Second, an updated fixed-fraction radiative impurity model is implemented, and the framework is extended to include, for the first time, the full multi-charge-state impurity model. This higher-fidelity treatment is essential for accurately capturing helium behavior, as its long mean free path and distinct atomic characteristics make its exhaust highly sensitive to local flow stagnation, connection length, and divertor neutral pressure.

Using these expanded capabilities, systematic scans of flux expansion, baffle geometry, impurity concentration and species, and pump placement are performed at fixed upstream conditions with ARC-like parameters of 1e20 m-3 upstream density and 40 MW power to emulate a double-null outer divertor leg [4]. The influence of divertor closure, magnetic geometry, and pump placement on helium behavior is analyzed, and comparisons between the fixed-fraction and full multi-charge-state impurity formulations provide a framework for assessing how different modeling assumptions affect predicted detachment thresholds and radiation characteristics under ARC-class conditions within the context of the extended Lengyel model [5].

Together, these developments establish a workflow for rapid exploration of magnetic and geometric divertor concepts, and identify promising pathways for helium exhaust compatible with ARC’s liquid FLiBe immersion blanket environment.

Supported by Commonwealth Fusion Systems.

[1] Lipschultz, et al. Nuclear Fusion 2016, 56, 056007.
[2] Moulton, et al. Plasma Physics and Controlled Fusion 2017, 59, 065011.
[3] Cowley, et al. Nuclear Fusion 2022, 62, 086046.
[4] Body and Eich, et al. Submitted to Journal of Plasma Physics.
[5] Body, Nuclear Fusion 2025, 65 086002

Speaker: Rebecca Masline (MIT Plasma Science and Fusion Center)

436 4.082 Divertor shaping as a continuum strategy for tackling power exhaust

Fusion energy is accelerating through conventional (DEMO) and alternative compact reactor designs, that are potentially faster and cheaper to build (e.g., ARC, STEP). Power exhaust is a key challenge and a potential show-stopper for all these designs. Recent experiment show the key benefits of strongly shaped Alternative Divertor Configurations (ADCs) [1-5], demonstrating their potential as a power exhaust solution. However, integration of ADCs in a reactor is complex: a compromise between power exhaust benefits and engineering feasibility is required [5].

Initial MAST-U results show a continuum of ADC optimisation through total flux expansion exists: modest, yet strategic, divertor shaping greatly enhanced power exhaust performance [3] and control [4]. To study this continuum, geometries with intermediate strike point position in between the MAST-U Super-X and Conventional geometrieswere compared. These intermediate geometries have less than half the total flux expansion increase of MAST-U’s Super-X divertor: lower than that of STEP [7] and comparable to that of SPARC [8], ARC [9] and the DEMO Super-X Divertor [6].

Crucially, the key benefits of the MAST-U Super-X Divertor over the Conventional Divertor are largely maintained for these geometries.
1. Target heat loads are reduced beyond geometric spreading expectations.
2. The sensitivity of detachment is drastically reduced, improving real-time power exhaust control [4].
3. Reducing the detachment onset without any adverse core impact increases the operational window of the detached regime. This improved core-edge compatibility can enable reactor operation at lower core power losses [6] and/or lower upstream densities/fuelling (relevant for ELM-free scenarios and fuel cycle limitations [7]) and/or lower impurity concentrations.

Studying the physics driving these benefits shows synergies between neutral baffling, poloidal leg length and total flux expansion. This shows power exhaust benefits can be obtained from strategic divertor shaping; consistent with reduced and full models, enabling improving power exhaust with relatively modest additional engineering complexity. This provides lessons on both divertor design – relevant for reactors - and exhaust physics/simulations and control [4] – relevant for ITER.

[1] C. Theiler, et al. This conference. [2] K. Lee. PRL. 2025. [3] D. Moulton. NF, 2024. [4] K. Verhaegh. Comm. Phys. 2025. [5] B. Kool. Nat. Energy, 2025. [6] R. Kembleton. FED 2022. [7] Xiang. 2021, NF; [8] Henderson. NF 2025; [9] Kuang, et al. JPP 2020; [10] Wigram, NF, 2019.

Speaker: Kevin Verhaegh (Eindhoven University of Technology)

437 4.083 Analysis framework for quantitative understanding of power losses during transients within detachment, with 2D inference of plasma parameters during ELM burn-through

Power exhaust remains a critical challenge for ITER, DEMO, and next-generation tokamaks, with unmitigated exhaust fluxes in an ITER-like device estimated at 100MWm−2, for which real-time control is essential. Maintaining detachment is further complicated in reactors by transients, caused by pellet fuelling and MHD perturbations (sawteeth, LH/HL transitions, unsuppressed ELMs, ...), resulting in intense divertor power bursts. These can ’burn-through’ the detached neutral divertor buffer, reattaching it on sub-millisecond timescales, resulting in target melting [1]. Although ELM suppression is essential for reactor-scale devices, such as ITER [1], some transients are unavoidable in steady-state reactor operation; a quantitative understanding of the power losses and processes in the divertor during burn-through of the detached buffer is essential. To characterise this evolution, diagnostics with high temporal resolution for intra-transient analysis are required, paired with analysis tools for a quantitative understanding of plasma-neutral interactions [2-4].
Bayesian analysis techniques for chordally-integrated spectroscopic data was successfully used to infer hydrogenic processes (ionisation, Electron-Ion Recombination, Molecular Activated Recombination / Dissociation) and their power losses in quasi-steady-state conditions [2]. This work prepares these techniques for application to MAST-U’s UltraFast Divertor Spectrometer diagnostic (UFDS) [5], operating at 800 kHz, enabling intra-transient estimates of ionisation sources, associated power losses and recombination. This provides insight into the physics of ELM burn-through, enabling validation of full and reduced [5] models for transient burn-through. Developing reduced analysis techniques to prepare real-time line-integrated sensors for ELM burn-through is considered. However, detachment is inherently a 2D phenomena, requiring camera diagnosis for 2D analysis. With multi-diagnostic Bayesian integrated data analysis techniques, this enables inferences of 2D parameter maps in the divertor (Te, ne, ...) [3,4], using cameras. These, along with fast multi-wavelength imaging (10 kHz), are planned to obtain 2D insight into ELM burn-through.
The aim is to develop tools for quantitatively understanding transient burn-through in the detached divertor, with an emphasis on identifying dominant hydrogenic radiative loss channels and processes in the intra-transient detached neutral buffer, contributing to an improved understanding of transient interactions within a detached divertor volume to improve reactor-scale extrapolations and develop sensors that provide a pre-warning signal for transient burn-through.
[1] Paschalidis, et al. 2024 Nucl. Fusion, 64 126022
[2] Verhaegh, et al. 2024 Plasma Phys. Control. Fusion, 63 035018
[3] Greenhouse, et al. 2024 Plasma Phys. Control. Fusion, 67 035006
[4] Bowman et al 2020 Plasma Phys. Control. Fusion 62 045014
[5] Flanagan et al 2025 Nucl. Fusion 65 116031

Speaker: Dr Kevin Verhaegh (Department of Applied Physics and Science Education, Eindhoven University of Technology)

438 4.084 Neutral Transport in the Vacuum Region and Its Impact on Detachment Physics

The Vacuum Neutral Model (VNM) has been implemented in UEDGE to account for neutral bypassing in the far scrape-off layer (SOL) [1], which lies outside the UEDGE computational mesh. In this work, the model is further extended to incorporate kinetic effects in the neutral flux leaking from the UEDGE boundary into the vacuum region. From a kinetic perspective, the dominant contribution to the outward-going neutral flux originates from ion–neutral charge-exchange interactions occurring approximately one mean free path inside the plasma. This allows the distribution of outward-going neutrals to be reconstructed self-consistently.
The inward-going neutral population consists of recycled particles from ion bombardment at material surfaces and reflected neutrals. While reflected neutrals can be modeled using TRIM-based reflection data, this capability is still under development and is therefore not included here; instead, inward-going neutrals are assumed to follow a half-Maxwellian distribution for simplicity based on the density and temperature obtained from the VNM. The reconstructed neutral distribution is then used to compute particle, momentum, and energy fluxes, which serve as boundary conditions for the neutral continuity, parallel momentum, and temperature equations.
This self-consistent treatment of the outgoing neutral flux is applied not only at the outer boundary of the VNM but also at the target plates. The impact of incorporating these kinetic boundary effects into the fluid neutral model on divertor detachment physics is investigated in this study.
[1] M. Zhao et al. “Vacuum neutral transport model in UEDGE for tokamak far scrape-off layer” 2025 Plasma Edge Theory Workshop, Leuven, Belgium

Speaker: Menglong Zhao (Lawrence Livermore National Laboratory)

439 4.088 Understanding of Power Sharing in the Scrape-Off Layer with respect to TCV results

In this work, we investigate power sharing between the inner and outer divertor of the TCV tokamak under low-density, attached Single Null divertor conditions, that is we check how varying the outer divertor leg length impacts parallel heat transport and target power loading, using a modelling framework that combines the conduction-based approach of Maurizio et al. with the classical Two-Point Model (2PM) and the radial resolution of the Onion-Skin method. To this end, we model a series of experimental TCV discharges using the 2D edge fluid transport solver TECXY, applying experimental boundary conditions to simulate realistic magnetic equilibria and SOL parameters.
The TECXY code, incorporates three regions: Scrape-Off Layer (SOL), Private Flux Region and a part of the pedestal, which for normalized magnetic radius is around 0.9-0.95. The simulations utilize a classical set of Braginskii transport equations for multi-species plasma. The model considers atomic processes such as ionization, recombination, excitation, charge exchange, prompt re-deposition, and sputtering. The transport along field lines is assumed to be classical, with anomalous radial transport coefficients.
Assuming electron heat conduction as the dominant transport channel and neglecting volumetric sources and sinks, we reconstruct the parallel temperature and heat flux profiles along a single flux tube extending from the midplane to the divertor targets.The simulations capture the experimentally observed trends in inner–outer divertor power sharing with increasing outer leg length and validate the predictive capability of the combined analytical–numerical model.
They highlight the influence of inner/outer leg lengths on the asymmetry in power deposition, the extent to which TECXY agrees with experimental results and the comparison with the analytical conduction-based model developed by Maurizio et. al. for Single-Null plasmas.
These results offer valuable insight into geometric control of divertor heat loads and provide a fast, physics-based tool for interpreting and extrapolating experimental observations in future divertor designs.

Speaker: Radhika Mishra

440 4.089 JOREK transient detachment simulations with kinetic neutral collisions: initial results for ASDEX Upgrade

Robust control of the heat and particle exhaust is required for machine protection in future high power magnetic confinement devices such as ITER. Understanding the dynamic behaviour of the plasma boundary is essential both for control design [1] and more generally to improve knowledge of the relevant transient physics. Recently, the inherently dynamic MHD code JOREK [2] has been extended with a kinetic neutral model to improve its fidelity regarding the description of the scrape-off layer and divertor plasma [3]. This provides an opportunity to study time dependent heat and particle exhaust, for reasonable computational times at the ITER scale, and thus to construct a synthetic environment to explore divertor detachment control.

There is, however, still a gap between the kinetic neutral capability in JOREK compared to EIRENE [4], and the dynamics in the model have not yet been validated. To close part of this gap in the neutral capability, neutral self-collisions have been implemented in the form of a direct binary collision algorithm, which has been verified to reproduce diffusion correctly. This model is more accurate than the widely used BGK model (e.g. in EIRENE), which can only reproduce correctly one collisional effect at a time (e.g. viscosity or diffusivity).

To validate the dynamics in the model, dedicated experiments have been performed on ASDEX Upgrade in which a transient detachment state is initiated by a step in the gas fuelling rate. The discharge was an Ohmic deuterium L-mode (AUG shot #42651; P=0.7MW, Ip=0.8mA, Bt=-2.5T, $Γ_{D}=0.22 \mathrm{(initial)}-8.4 \mathrm{(final)}⋅10^{21} e/s$, κ=1.58, $δ_u$=0.33) with extensive diagnostic coverage to provide the best possible set of experimental measurements under the simplest conditions. After initializing the transient, the new equilibrium was reached within one second.

We present a comparison of simulations with the upgraded JOREK model including the impact of the neutral collisions against this experiment. Initial results indicate, as expected, low impact of the addition of the neutral collisions on the behaviour of the plasma in the initial attached phase of the shot. That the simulations are found to be more attached than in the experiment is also expected owing to the lack of molecular interactions in the model which are known to play a major role in detachment onset [5].

[1] Ravensbergen, Nat. Commun. 12(2021) 1105.
[2] Hoelzl, Nucl. Fusion 64(2024) 112016.
[3] Korving, Phys. Plasmas 30(2023) 042509.
[4] Reiter, Fusion Sci. Technol. 47(2)(2005) 172–186.
[5] Verhaegh, Nucl. Fusion 63(2023) 016014.

Speaker: Daniël Maris (DIFFER)

441 4.090 Summary of Boundary Physics Work at CFS in Support of SPARC Operations

Commonwealth Fusion Systems (CFS) is rapidly assembling the SPARC tokamak, with commissioning of many systems already underway. SPARC is a high field (12T), high current (8.7MA) device capable of achieving an energy gain of 11 when operating in H-mode with DT fuel. While the high magnetic field enables high fusion performance, it also results in a narrow heat flux width (between 0.3-0.6mm) due to a poloidal field of ~2.5T at the outer midplane. This narrow heat flux width makes power exhaust a central challenge, as unmitigated H-mode parallel heat fluxes are expected to reach as high as 10 GW/m2.

SPARC utilizes plasma facing components (PFCs) composed of pure tungsten and tungsten heavy alloy (97% W, 2% Ni, 1% Fe), which have been optimized for plasma power exhaust handling. Because the PFCs have the potential to contaminate the core plasma with high-Z sputtered impurities, SPARC will run in a partially or fully detached regime (e.g. XPT, XPR), using Ne, Kr, and Ar as the seeded impurity species. A relatively lean set of diagnostics will be combined with state-of-the-art plasma control techniques to maintain detachment while optimizing for core performance. Later campaigns will prepare for ARC, a 400MWe power plant, by operating SPARC with a neutron compatible diagnostic set.

SPARC edge plasmas are simulated across a wide range of fidelities by CFS and our collaborators. Fast hyperspace scoping is performed using physics models such as the extended Lengyel model and SepOS. Axisymmetric plasma distributions are predicted with a neural network trained on SOLPS runs, which can be connected to the plasma control system. The 3D plasma power exhaust and corresponding 3D PFC state is predicted using the HEAT code. Synthetic diagnostics are used to predict the expected diagnostic signals from thermocouples, cameras, spectroscopy, shunt measurements, and Langmuir probes. Workflows to reconstruct the plasma state from diagnostic data, and to track lifetime consumables (e.g. recrystallization), are under development in preparation for first plasma.

This presentation summarizes ongoing work across the SPARC boundary physics team, along with the challenges expected for operations. It provides a summary of the expected boundary physics regimes that will be observed in SPARC, and outlines the boundary control strategies that will be employed to operate the machine at high performance while preserving the PFCs.

Speaker: Tom Looby (Commonwealth Fusion Systems)

442 4.091 Assessment of power exhaust capabilities on the academic tokamak TRUST with the SOLPS-ITER code package

TRUST (Tuscia Research University Small Tokamak) is a new university-scale tokamak currently under design at the University of Tuscia (UNITUS).The main parameters are as follows [1,2]:R= 0.32 m, a= 0.11 m, A= 2.80, K≈ 1.7, Ip≈ 0.11 MA, Bt≈ 0.77 T, H98≈ 0.8, βpol(1)≈0.11, li(3)≈0.77 and will be operated with an ohmic power of Pohm = 0.3 MW.

TRUST is scheduled to be deployed in three phases, during which upgrades will include modifications to the magnets and power supply systems. At least four configurations will be explored: single null (SN), double null (DN), upper single null (USN), and limiter [2].

In addition to its academic and educational purposes, the main objective of the device is to achieve a high-power flux density on the targets, enabling the testing of different materials, such as metamaterials (currently studied as potential solutions for DEMO limiter applications [3]) under significant thermal stress. The system is designed to allow easy replacement of plasma-facing components, and the high flexibility of the machine makes it possible to investigate liquid metals as test targets.

Recent 1D analyses, performed through a parameter scan of the electron density at the outer midplane (ne) and the power crossing the separatrix (PSOL), indicate that within the expected operational range, and the calculated heat-decay length of the parallel heat-flux density λq ≈ 5 mm, the maximum perpendicular power flux achievable at the targets is ≈ 4 MW/m² and ≈ 2.5 MW/m² for SN and limiter configurations.

Given the academic nature of the project, the compact environment, and the technological constraints, achieving such a power flux sufficient to test innovative materials represents a significant milestone.
The work presented is aimed to verify the 1D scan results using SOLPS-ITER simulations, to study the expected power flux to the targets in greater detail and with higher accuracy. These simulations will also support the design of the device, both in terms of target geometry and magnetic equilibrium, to achieve the highest possible power flux at the target plates.
Simulations were performed using both the 'standard' SOLPS-ITER version [4] and the new extended (wide) grid version [5], in which the plasma computational mesh extends to the entire vessel wall. The differences and similarities between the results obtained using the two code versions are compared and discussed.

[1]S.Carusotti et al,https://doi.org/10.1016/j.fusengdes.2025.115416
[2]S.Carusotti et al,https://doi.org/10.1016/j.fusengdes.2025.115246
[3]D.Paoletti et al,https://doi.org/10.3390/jne3040028
[4]S.Wiesen et al,https://doi.org/10.1016/j.jnucmat.2014.10.012
[5]W.Dekeyser et al,https://doi.org/10.1016/j.nme.2021.100999

Speaker: Christian Avanzato (University of Tuscia)

443 4.092 TALIF on Magnum-PSI to characterize ground state atomic hydrogen in detached conditions

Divertor detachment is the leading candidate for solving the heat exhaust problem in future fusion reactors. Key to understanding detachment is the interaction of the plasma with a background of neutral particles in the divertor region. Collisions of the plasma with these neutral background particles result in a rich range of physical and chemical processes, causing the plasma to dissipate its energy and momentum, and finally to recombine, preventing damage to the wall. Whereas charged particles are routinely diagnosed, information on neutral particles is often missing. The properties of neutral particles can be measured using active spectroscopy. At Magnum-PSI, a linear plasma generator that can simulate the high heat and particle flux conditions of future fusion reactors, two types of active spectroscopy are currently being developed: TALIF and CARS. In this poster an overview of the TALIF diagnostic is presented, see [1] for the CARS developments.

TALIF, Two photon Absorption Laser Induced Fluorescence, is a local measurement of the properties of H atoms in the electronic ground state. The spatially resolved TALIF profiles provide the atomic H density, temperature and flow velocity. TALIF data was taken over a broad range of plasma parameters in both the UPP and Magnum-PSI linear devices, providing a dataset that can be used to validate divertor models.

Calibrated data taken during various degrees of detachment will be presented. Also, two types of TALIF calibration are compared using either Xe or Kr as reference gases. We revisit measurements made by Elliott et al [2]. for the Xe:Kr relative calibration factor, and find a value of 0.181 ± 0.025, which differs from the earlier work by a factor of 5.

[1] K. Schutjes et al., This conference
[2] D. Elliott et al., Rev. Sci. Instrum. 87, 11E504 (2016)

Speaker: Ivo Classen (DIFFER)

444 4.093 Removal of Boron Mixed Contaminant Films on Platinum Mirror

The metallic First Mirrors (FMs) will play a crucial role in most optical diagnostic systems in ITER. As the initial elements in the optical path of diagnostic systems, the FMs will be subjected to deposition of the first tungsten wall materials and to regular boronization, compromising their optical properties. The FMs would thus need periodic cleaning to restore their optical properties, which is expected to be achieved using an in-situ plasma-cleaning technique based on radio-frequency (RF) discharges [1].
Boron and tungsten mixed films were deposited using magnetron sputtering with argon, argon/oxygen, and argon/deuterium atmospheres on a platinum mirror. Films of boron oxide, boron deuterium and boron tungsten were deposited and their chemical composition was measured in vacuo by X-ray Photoelectron Spectroscopy (XPS).
Removal of these films was performed using radio-frequency plasma cleaning with argon, helium, and deuterium gas at 250 and 350V, for 13.56MHz. Argon can erode all contaminant films, but it has a high erosion rate on the substrate. Helium and deuterium sputtered boron, but were ineffective on tungsten. On a platinum mirror, 15nm of boron was sputtered and reflectivity was recovered.
Argon and deuterium mixture also eroded EAST boronized films contaminant, with a partial restoration of the optical properties of the molybdenum mirror.

[1] A P. Shigin et al., Fusion Eng. Des. 64 (2021) 112162

Speaker: Dr Laurent Marot (University of Basel, Department of Physics)

445 4.094 Calibration of LIBS diagnostics operated on remote handling arm in JET

A compact laser-induced breakdown spectroscopy (LIBS) set-up recently operated on a remote handling arm in the JET tokamak after end of operation in a radioactive environment. The picosecond-LIBS set-up has successfully measured the fuel retention and surface composition on different plasma-facing components in-situ at 840 positions [1]. This LIBS set-up uses a spectroscopic plasma observation, which is co-linear with the laser beam. Two identical fused-silica lenses with focal length of 75.3 mm were used to collect and focus the LIBS plasma emission light into a 20 m long 1.5 mm diameter fused-silica fiber. The fiber is terminating into a 7-branch fiber bundle, which distributes the plasma emission light between an Echelle spectrometer, a large etendue Littrow spectrometer, both of which had equipped with ICCD cameras, a compact spectrometer, a compact monochromator equipped with silicon photomultiplier (SiPM) and a filterscope having 3 SiPMs equipped with different bandwidth H$\alpha$ interference filter. The use of lenses provides the necessary compactness of the optical system 340x136x(328+89)mm$^3$. The light collecting lens is also used to focus the laser beam (1064 nm, 800 ps, 10 mJ) on the PFC surface.
Quantitative analysis of LIBS spectra requires calibration of the sensitivity of all available spectral channels. The standard calibration procedure using an extended calibrated light source at the target position is not applicable due to the substantial chromatic aberration of the lens optical system. To account for this effect, the optical transmission of LIBS collecting optics was calculated using ZEMAX software for different light source diameters. Calibrations of the LIBS system along with the 20m long optical fiber were performed with help of available calibration types (absolutely calibrated integrating sphere with halogen light source, deuterium lamp, H$_2$/D$_2$ capillary discharge lamp Hg low pressure discharge lamp, Ne glow discharge
lamp) prior to its installation on the remote handling arm. In addition, several types of samples were used to determine crater size, ablation rate, and conversion factor of Balmer alpha line radiation to deuterium content in tungsten, JET divertor material. After the LIBS set-up was
jacketed and introduced into the tokamak, periodic calibrations were performed by connecting the 1.5 mm diameter fiber bundle tip to the calibration source. The complete calibration procedure and selected quantifications for a divertor and main chamber position with one spectrometer is presented as an example for the successful calibration and quantification. The calibration is applied in adjacent presentations for plasma-wall interaction studies.
[1] https://doi.org/10.1016/j.nme.2025

Speaker: Dr Gennady Sergienko (FZJ)

446 4.095 Direct measurement of heat flux deposition on the first wall from neutral beam shine-through in DIII-D tokamak

Quantification of shine-through (not absorbed by the plasma) heat flux from neutral beam injection (NBI) is essential for protecting plasma-facing components (PFCs). In this work, we present the direct measurement of heat flux deposition from NBI shine through on the DIII-D first wall using newly implemented, high temperature capable Surface Eroding Thermocouple (SETC) sensors [1]. The SETCs were installed on strategically selected high field side center-post tiles with direct line-of-sight to the 30R neutral beam. These tiles are adjacent to the Low Hybrid Current Drive (LHCD) antenna, a component known to be highly sensitive to localized heat loading. The SETCs provide real-time monitoring of surface temperature and incident heat flux, enabling improved protection of the LHCD antenna during high power beam operation.

During the beam-into-gas shots (no plasma), the SETCs measured peak heat fluxes of 20–40 MW/m², corresponding to surface temperature rises exceeding 300 °C in only 8 milliseconds. In plasma discharges, the measured shine-through heat flux shows a strong dependence on the plasma line-averaged density. Higher density plasmas ionize a larger fraction of the injected neutrals, thereby substantially reducing the shine-through power reaching the wall. For discharges with neutral beams at 75 keV injection energy, the measurements indicate that the plasma density of approximately (2–3) × 10¹³ cm⁻³ is required to reduce the shine-through fraction by 50%. The local heat fluxes measured by SETC were quantitatively compared with the simulated power deposition of neutral beams. The shine through power is calculated by the pencil beam attenuation model [2] and the heat deposition is further mapped to the first-wall on the center post, following the beam footprint associated with the NBI divergence. Furthermore, a one-dimensional slab model is used to estimate the first wall temperature from the projected power, enabling a direct comparison between the simulation with the measurement at the SETC locations. The calculated heat flux footprint and the simulated temperature responses agree with the SETC measurements within 20%.

This work demonstrates that the SETC system is robust and reliable to monitor the neutral beam shine-through power. These measurements provide new insight into beam–wall interactions in high-power scenarios and offer a practical pathway for validating NBI deposition models and guiding future wall-protection designs.

Work supported by US DOE under DE-FC02-04ER54698, DE-SC0023378.

[1] J. Ren et.al. Rev Sci Instrum, 93, 103541 (2022), doi: 10.1063/5.0101719
[2] M. A. Van Zeeland et.al. Plasma Phys. Control. Fusion 52, 045006 (2010) doi: 10.1088/0741-3335/52/4/045006

Speaker: Jun Ren (University of Tennessee - Knoxville)

447 4.096 Ex-situ analysis of JET plasma-facing components using laser induced breakdown spectroscopy in support of post-DTE3 in-situ measurements

The feasibility of laser-induced breakdown spectroscopy (LIBS) for measuring fuel retention was first demonstrated in 2024 in a tokamak operating with tritium. This was achieved using a remotely controlled in-situ application at JET [1]. Following the third deuterium-tritium campaign, DTE3, and the in-situ LIBS experiment selected plasma-facing components were removed from the JET vacuum vessel for post-mortem analyses. Prior to the in-situ LIBS experiment at JET the LIBS setup was tested at VTT by analysing JET divertor CFC and bulk Be samples from the main wall exposed in 2011-2016 [2] using the LIBS tool developed at ENEA. The tool consisted of the LIBS enclosure equipped with a sub-nanosecond Nd:YAG laser and focusing optics, which was connected via a 20 m optical fibre to an Echelle type spectrometer with wide spectral range 260-760 nm. The LIBS experiment at VTT focused on the co-deposited layers and on the calibration of ablation rates. The depth profiles clearly distinguished the various layers on the samples.
The LIBS experiment at VTT will be repeated in 2026 mainly because a high resolution Littrow spectrometer used at JET was not available at VTT in 2024. More importantly, the LIBS setup was optimized at FZJ for the experiment at JET after the experiment at VTT and the modifications made to the optical setup changed the focusing of the laser beam on to the sample and the transmission properties of the optical setup. In addition to the Littrow spectrometer, a high spectral resolution Double Echelle Monochromator series in Littrow configuration (DEMON) will also be tested for the separation of the alpha lines of the hydrogen isotopes. The resolution of the DEMON spectrometer is 3-10 times better than that of the Littrow spectrometer at FZJ. The same samples as in 2024 as well as the first DTE3 samples, including from JET limiters and divertor, will be characterised for calibration-free LIBS and for calibration of the ablation rates. This work also presents updated results from depth profiling and CF analyses of the JET divertor obtained during the experiments at VTT in 2024.

[1] J. Likonen et al, First Demonstration of Laser Induced Breakdown Spectroscopy using Remote Handling for In-vessel Analysis of JET Components, Nuclear Materials and Energy 45 (2025) 102021.
[2] J. Ristkok et al., Preparing LIBS for in-situ measurements in JET tokamak: system overview and co-deposited layer thicknesses, Nuclear Materials and Energy 44 (2025) 101968.

Speaker: Jari Likonen (VTT Technical Research Centre of Finland)

448 4.097 Depth-Resolved ps-LIBS Characterization of Boron/Deuterium and Deuterium- Loaded Boron Coatings for Fusion Plasma-Facing Applications

Boron layers are widely employed in magnetic-confinement fusion devices due to their strong oxygen gettering capability and their role in reducing impurity release from plasma-facing components. Accurate characterization of boron film thickness, erosion behaviour using laser ablation, and fuel retention is essential for predictive plasma–wall interaction modelling in ITER- relevant conditions[1]. Recent studies using LIBS have demonstrated quantitative detection of ultrathin boron layers on tungsten substrates and established calibration strategies for high-resolution surface analysis in fusion environments [2], [3]. Building on these developments, the present work investigates thicker and fuel-loaded boron coatings representative of boronized first-wall surfaces.

Boron films of 500 nm and 1000 nm thickness deposited on Si and W substrates, along with corresponding deuterium-loaded variants, are analyzed predominantly by depth-resolved ps-LIBS. The short-pulse excitation enables controlled material removal with reduced thermal diffusion, allowing the evolution of deuterium emission to be quantified as a function of ablation depth. Complementary ion-beam techniques are employed to obtain absolute deuterium areal densities and depth distributions, providing independent reference profiles for evaluating the accuracy of the LIBS-derived depth information[4,5].

Systematic comparison of pristine B and D-loaded B layers enables extraction of the boron ablation rate and identification of modifications to crater formation induced by retained deuterium. The resulting dataset elucidates the interplay between fuel retention and ablation dynamics in boron coatings and supports the establishment of a depth-calibrated, quantitative LIBS framework for fuel-retention diagnostics in fusion devices. This work extends LIBS beyond ultrathin boron films and addresses key analytical requirements for assessing erosion and D retention in future reactor-scale plasma-facing components.

References:
[1] J. Winter et al., Journal of Nuclear Materials, 162–164 (1989) pp. 713–723.
[2] H. Wu et al., Nuclear Materials and Energy, 45 (2025) 102018.
[3] H. Wu et al., Nuclear Materials and Energy, 41 (2024) 101812.
[4] P. Veis et al., Nuclear Materials and Energy, 25 (2020) 100809.
[5] R. Mateus et al., Elemental analysis of divertor marker tiles exposed during the 2018 (C3), 2019 (C4) and 2020 (C5) WEST campaigns, PFMC 2025 conference.

Speaker: K. E. Ramachandran (Comenius University, Faculty of Mathematics, Physics and Informatics, Bratislava, Slovakia)

449 4.098 The case for multiband infrared imaging in fusion devices

A multiband interpretation of infrared emission allows for accounting of spatial and temporal variations in emissivity for metallic armor materials and thus accurate determination of surface temperature and heat flux in a fusion device. Without adequately accounting for true emissivity of material in the imaged scene, the situation where true surface emissivity is less than expected can lead to inference of surface temperatures less than actual. In machines which push the temperature of plasma-facing materials near the limits of recrystallization and melting of high-Z materials, such a situation could be catastrophic for first-wall integrity and machine operation, particularly when there is little flexibility in strike point positioning for high performance plasmas.

In a fusion device, surface emissivity can vary significantly due to changes in temperature, surface conditions/morphology with erosion, and impurities/contamination of the wall material with plasma exposure. To remove the dependence of inferred temperature on material emissivity, the multiband interpretation of infrared radiation relies on calibration of ratios between emission bands rather than their integrated intensity as used in most traditional infrared cameras. First demonstrated in a dual-band approach on NSTX which is appropriate when emissivity is constant with wavelength, expanding to the multiband interpretation of infrared radiation allows compensation of variable emissivity with wavelength and other conditions. The approach presented for multiband detection is by splitting the full image frame into multiple sub-frames, each integrating a separate portion of the observed wavelengths of light. The motivation for such an approach, hardware and interpretive aspects, as well as limitations, are presented.

This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract DE-AC52-07NA27344.

Speaker: Adam McLean (LLNL)

450 4.099 A versatile Langmuir probe for measuring the edge structure of HSX

The Helically Symmetric eXperiment (HSX) is a medium sized stellarator optimized to study confinement properties in a quasi-helically symmetric (QHS) configuration. The QHS configuration of HSX exhibits non-resonant divertor like behavior in its edge region [1]. In addition, HSX is equipped with a set of auxiliary coils leading to a wide range of possible magnetic configurations allowing for the study of island divertors. There are many unanswered questions about the structure and behavior of both island and non-resonant divertors in QHS optimized stellarators. This includes the 3D variation of plasma flows in the edge region, or the width λ_q of the of the power scrape layer [2].
In this work a versatile Langmuir probe is developed, benchmarked, and used to gather first data on HSX. The new probe consists of 2 triple probes at different radial locations, a Mach probe, and a unique implementation of thermocouples for measuring the overall heat flux on the central probe tip. This design allows for instantaneous measurement of spatially dependent ion density, flow velocity, electron temperature, floating potential, and total heat flux, potentially leading to ion temperature measurements not previously possible with Langmuir probes. The new probe was successfully benchmarked in a steady state DC multipole plasma source where both triple probes, and the Mach probe produced ion density and electron temperature values consistent with one another and consistent with previous measurements of the plasma parameters on this apparatus taking into account probe positioning.
First experiments using this probe on HSX demonstrated its robustness against plasma operation and provided first insight into the spatial variation of edge plasma parameters, observed super-thermal electron behavior during the pre-discharge phase, and was able to obtain first measurements of heat flux which will be presented. Moreover, the probe is ready for experiments during the upcoming 2026 experimental campaign of HSX starting in January during which detailed radial scans of the probe, as well as power and density scans of the plasma are foreseen. Results of these experiments will help to answer important questions about both island and non-resonant divertors.

[1] A. Bader et al., Phys. Plasmas 24(3), 032506 (2017)
[2] A. Bader et al., J. Plasma Phys. 91(2), E67 (2025)

Speaker: Alex Klasing (University of Wisconsin, Madison)

451 4.100 Implementation and Evaluation of a Thermo-Mechanical Monitoring System for KSTAR Plasma-Facing Components

The reliability of a thermo-mechanical monitoring system for KSTAR plasma-facing components (PFCs) was evaluated. Displacement sensors and strain gauges were used to analyze mechanical behavior, and thermocouples were employed for temperature measurements of the PFCs. Because the PFCs operate under extreme conditions—such as very high temperatures, strong magnetic fields, and radiation—measurement signals can be distorted. Therefore, it is important to minimize noise and accurately measure only the meaningful signals. Displacement sensors with built-in full-bridge configuration were installed on the passive stabilizer structure. These sensors were configured to measure displacements of the passive stabilizer in the radial and vertical directions from the center of the vacuum vessel up to 30 mm. Encapsulated strain gauge, including a dummy gauges were attached to the support structure of the passive stabilizer using spot welding. Half-bridge and full-bridge configurations compensate for the effects of temperature and magnetic field in the sensor signals. During KSTAR baking operations, the PFC temperature rises up to 250 °C, and the thermal expansion of the passive stabilizer structure was measured as strain and displacement signals. The electromagnetic forces generated by the magnet currents and the plasma current were also measured as strain and displacement signals. We compared the measured signals with the calculation results. Thermocouples measure temperature by directly contacting the graphite tiles or tungsten monoblocks of the PFCs, and their reliability was verified by analyzing the thermal distribution of the PFCs. This paper presents the specifications and installation status of the sensors used for KSTAR PFC monitoring, and analyzes the corresponding measurement results.

Speaker: Mr Youngok Kim (KFE)

452 4.101 Development of a VUV-VIS LIBS system for the investigation of boron spectra and layer lifetimes toward fusion wall boronization

ITER switched its first‑wall material from beryllium to tungsten to avoid beryllium’s toxicity, erosion, and tritium retention, and now relies on boronization to suppress oxygen and impurities. However, boride formation during boronization and plasma operation can enhance fuel retention in deposition layers. In addition, the complex tokamak environment makes post‑mortem analysis challenging, creating an urgent need for in situ, quantitative evaluation of boron layer thickness and lifetime to optimize wall conditioning. Conventional Laser-Induced Breakdown Spectroscopy offers only a few weak boron lines in the visible range, limiting determination of boron content and layer parameters. In contrast, vacuum‑ultraviolet LIBS provides intense boron emission lines in the 150–200 nm range with better spectral separation, while its typical low‑pressure operation further suppresses the continuum background, improving the line‑to‑background ratio compared with VIS‑LIBS.

In this work, a dedicated VUV‑LIBS system was developed to investigate boron spectra relevant to fusion wall boronization. The VUV spectrometer is based on a Seya–Namioka configuration with a 20-100 µm entrance slit and a concave, aberration‑corrected holographic grating (600 grooves mm⁻¹) covering 50–700 nm. A back-illuminated CCD detector (Andor DO920P) was employed for signal acquisition, providing high quantum efficiency in the VUV region. Wavelength calibration was performed in air using six Hg lines in the 300–550 nm range, and then linearly extrapolated to 100–200 nm for VUV operation. From this calibration, the spectral dispersion, wavelength coverage, and pixel positions corresponding to emission lines between 150 and 200 nm were determined. The spectrometer dispersion was approximately 0.18 nm / pixel, with a total spectral window of about 180 nm. The VUV spectrometer is coupled to the vacuum chamber of the previous VIS - LIBS system, sharing the same laser focusing optics and plasma geometry. This configuration enables simultaneous acquisition of VUV and visible spectra from a single laser‑produced plasma plume, providing access to strong VUV boron lines together with high‑resolution hydrogen isotope lines in the visible for plasma parameter and fuel‑retention studies.

Future validation will utilize boron-coated tiles from the EAST tokamak first wall and boron‑containing reference targets. These measurements will focus on detecting boron emission lines in VUV range, optimizing time‑gated detection, and comparing depth profiles from VUV‑LIBS with obtained from conventional VIS - LIBS. This VUV‑VIS LIBS system is expected to enable in situ, quantitative assessment of boron layer thickness and lifetime in fusion devices, providing a critical diagnostic for optimizing wall conditioning in ITER and future reactors.

Speaker: Zhenhua Hu (Institute of Plasma Physics, Chinese Academy of Sciences)

453 4.103 In-situ LIBS measurements of fuel retention and deposit formation on inner wall guard limiter tiles by JET LIBS remote handling arm system

Laser-induced breakdown spectroscopy (LIBS) is a technique developed for determining fuel retention in ITER plasma-facing components (PFCs) [1]. It can also analyse the formation of various layers on PFCs caused by material redeposition eroded from other regions. A series of experiments was conducted at the JET tokamak after the third deuterium-tritium campaign, DTE3, to demonstrate the feasibility of using the LIBS tool for in-situ fuel retention measurements and material migration analysis. [2]. The LIBS instrument, operated on the remote handling arm system MASCOT, included a sub-nanosecond laser with focusing optics inside a compact enclosure [2]. Emission from the LIBS plasma plume was directed into a 20 m optical fibre and then through a multi-channel fibre bundle to various spectrometers. An Echelle-type spectrometer captured spectra over a broad range of 260-760 nm to determine the elemental depth profiles of elements present in the PFC layers [2]. A custom-made Littrow-type spectrometer recorded spectra in a narrow wavelength range of 654-658 nm with good spectral resolution and was used to measure the emission line intensities of hydrogen isotopes T, D and H [3].
LIBS experiments were conducted at 840 different locations on the JET wall and divertor. This study presents experimental results from various locations on the inner wall guard limiter (IWGL), including 2X beam tiles in upper, central and lower poloidal positions, as well as wing and central toroidal positions. The in-situ LIBS results aligned with previous studies using ex-situ characterization methods after the ILW1-ILW3 campaigns [4-5]. On the wings of the tiles, LIBS depth profiles revealed Ni-, Cr-, and W-containing surface layers, which were absent in the central tile regions. The D signal was present on the surface of all investigated positions. On the wings of the tiles, the D signal remained detectable at a deeper level within the tile but was relatively weaker at the surface.
[1] H.J. Meiden, S. Almaviva, J. Butikova et al. Nucl. Fusion 61 (2021) 125001
[2] J. Likonen, S. Almaviva, R. Rayaprolu et al. Nucl. Mater. Energy 45 (2025) 102021
[3] R. Yi, R. Rayaprolu, J. Likonen et al. Nucl. Mater. Energy 45 (2025) 102016
[4] A. Baron-Wiechec, A. Widdowson, E. Alves et al. J. Nucl. Mater 463 (2015) 157
[5] A. Widdowson, J.P. Coad, E. Alves et al. Nucl. Mater. Energy 19 (2019) 218

Speaker: Peeter Paris (UT)

454 4.104 LCIF Diagnostics on RAID for Validation of Collisional-Radiative Models in Tokamak Plasmas

Accurate spectroscopic diagnostics for tokamak and stellarator divertors rely on collisional–radiative models (CRMs), whose predictive power is often limited by uncertain collisional rates [1]. The Resonant Antenna Ion Device (RAID) at the Swiss Plasma Center provides a controlled, steady-state plasma environment relevant to magnetic confinement fusion that is ideally suited for validating CRMs by classical and advanced spectroscopic techniques.
RAID is a linear device based on helicon excitation that produces stable plasmas in Ar, He, H₂, and D₂ with densities up to a few $10^{18}$  m$^{−3}$ (few $10^{19}$ m$^{−3}$ for He/Ar) and electron temperatures up to 10 eV [2,3], representative of divertor and edge conditions. The device is equipped with a comprehensive set of conventional diagnostics [3], and has recently become a platform for advanced active-spectroscopy development. A TaLIF system for atomic hydrogen density measurement [4] demonstrated its relevance for benchmarking of neutral-transport models like SOLPS-EIRENE [5] and has recently been upgraded for investigations of single-laser-pulse H density measurements, with potential application in tokamaks.
Building on this capability, we have recently implemented a laser-collisional induced fluorescence (LCIF) diagnostic based on a wavelength-tunable picosecond laser system. By selectively overpopulating excited atomic states and tracking the population redistribution, LCIF provides direct, in-situ measurements of collisional rates [6]. These rates, traditionally inferred from simulations or particle-beam experiments, are critical inputs for CRMs used to interpret optical emission spectroscopy (OES) measurements in fusion devices. LCIF measurements on RAID enable validation and refinement of CRMs, particularly in the low-temperature He/H/D regime relevant for tokamak divertors, that was shown to be challenging for CRMs [7].
In this contribution, we will present the LCIF diagnostic and first measurement results on RAID. We will discuss how LCIF, OES, and Thomson scattering measurements on RAID will be used in synergy to validate CRMs to improve spectroscopic diagnostic of tokamak and stellarator divertors.

[1] Flom et al., Nuclear Materials and Energy, Vol. 33, 2022
[2] Furno et al., 22 Topical Conference on Radio-Frequency Power in Plasmas, 2017
[3] Jacquier et al., Fusion Engineering and Design, Vol. 192, 2023
[4] Kadi et al., Plasma Physics and Controlled Fusion, Vol. 66, No. 12, 2024
[5] Kadi, L., Phd Thesis, EPFL, 2025
[6] Denkelmann, et al., Journal of Physics B: Atomic, Molecular and Optical Physics, Vol. 32, No. 19, 1999
[7] Linehan et al., Nuclear Fusion, Vol. 63, No. 3, 2023

Speaker: Christine Stollberg (EPFL)

455 4.105 Development of the D-alpha line measurement system for Tungsten Divertor in KSTAR

During the 2023 KSTAR experimental campaign, the lower divertor was replaced with tungsten monoblock plasma-facing components. The transition from carbon to tungsten is expected to substantially modify impurity sources and transport in the divertor plasma. However, prior to this upgrade, KSTAR lacked a dedicated diagnostic for measuring D-alpha emission in the divertor region, limiting detailed investigation of divertor impurity behavior and plasma–wall interactions.
A divertor D-alpha diagnostic system has been designed, fabricated, and installed to enable localized measurements in the KSTAR divertor. The system consists of nine lines-of-sight, with five channels viewing the central divertor and four channels viewing the X-point region. Each channel employs a 0.25-inch collimator, resulting in a spot diameter of approximately 4 mm, constrained by the narrow gaps between divertor blocks. Two identical nine-channel modules were installed at the K-port and G-port, providing toroidally separated measurements with a separation of 90 degrees. Channel alignment and absolute intensity calibration were performed using laser-based alignment and a calibrated integrating sphere, respectively.
The diagnostic utilizes interchangeable narrowband optical filters, allowing flexible measurements of D-alpha emission as well as other impurity line radiation. This system provides a new capability for systematic studies of impurity behavior and divertor plasma characteristics in KSTAR under tungsten wall conditions.
This paper presents the design, fabrication, installation, and initial operational results of the divertor D-alpha diagnostic system.

Speaker: Dr Dongcheol Seo (Korea Institute of Fusion Energy)

456 4.106 In situ LIBS diagnosis of elemental distribution on the surface of divertor in EAST

The online elemental analysis of plasma-facing components (PFCs) is crucial for magnetic confinement nuclear fusion deceives, such as tokamak and stellarator. Elemental distribution directly reflects the conditions of PFCs and processes of plasma-wall interaction (PWI). The laser-induced breakdown spectroscopy (LIBS) diagnostic technology provides a promising method for wall composition monitoring for nuclear fusion deceives. An in situ endoscopic LIBS diagnostic system for the full tungsten divertor in EAST has been developed since the 2021 experimental campaign. This system provides online elemental distributions on the divertor with various discharge parameters and wall conditions. This work focuses on using the in situ LIBS system to study D fuel retention evolution in the tungsten upper divertor region, the impact of wall conditions on the H-D ratio, fuel retention behavior during long-pulse discharges, and short-term retention phenomena. The H/(D+H) decreases from 100% to 17%~24% after the wall conditioning of baking and glowing discharge clean. The D and Li signal intensities measured at different poloidal positions are positively correlated with the EAST plasma discharge duration. In addition, the short-term fuel retention study shows that the dynamic D content on the W divertor decreases after the plasma exposure due to the dominant short retention with the outgassing process. The results provide direct experimental data for in situ research on PWI in magnetic confinement fusion devices.

Speaker: Prof. Cong Li (Dalian University of Technology)

457 4.107 Determining the stoichiometry of tungsten borides (WxBy) using LIBS extended to vacuum UV spectra range

Transition metal borides (TMBs) are a class of compounds known for their exceptional mechanical strength, thermal stability, and chemical inertness. These compounds, formed by strong covalent bonding between transition metals and boron, are widely studied for applications in cutting tools, coatings, and high-temperature environments such as thermal fusion reactors and aerospace components [1]. Among the TMBs, tungsten borides (WxBy) stand out due to their remarkable hardness and oxidation resistance. The mechanical performance of tungsten boride strongly depends on its stoichiometry; studies indicate that stoichiometric phases such as WB5 exhibit greater hardness up to a Vickers hardness of 45 GPa due to their dense covalent B–B and W–B networks, which provide high hardness combined with good thermal and chemical stability [2].
Accurate stoichiometric quantification of such borides poses a major analytical challenge. Techniques like EDS, WDS, and XPS often fail to provide reliable boron quantification because of its low atomic number. One of the solutions is Laser-Induced Breakdown Spectroscopy (LIBS), which is a well-established technique for on-site stoichiometric analysis of the materials. However, conventional LIBS operated in the standard UV – NIR range (220-900 nm) is not effective for B detection as only one B I doublet @ 249.7 nm can be observed. Our experimental setup included a picosecond laser with a pulse width of 250 ps, a repetition rate of 100 Hz, λ=1064nm, an echelle-type spectrometer for the UV-NIR range, and a Seiya-Namioki operating in the VUV range. The electronic temperature (Te) varied between 0.65 - 0.80 eV at the considered gate delay and the Ar gas pressure.
By extending the measurements of LIBS spectra into the vacuum ultraviolet (VUV) region, more B lines could be observed. After extending the measurement to the VUV range, the upper-level energy of B I lines ranges between 4.96-9.54 eV, enabling the evaluation of Tₑ using the Boltzmann plot (BP) from both W and B lines rather than solely depending on W. This enables a more accurate elemental quantification using the calibration-free (CF) approach. An earlier study on quantifying lighter elements, including B, has been carried out by our group [3]. Additionally, the texture of the sample coating will be measured using the texture measurement using XRD.

[1] M. Magnuson et al., 196 (2022) p.110567, Elsevier Ltd.
[2] A. G. Kvashnin et al., J Phys Chem Lett, 9 (2018) p. 3470.
[3] P. Veis, et al, Plasma Sources Sci Technol, 27 (2018) p. 095001.

Speaker: Sanath Shetty (FMPI, Comenius University, Bratislava)

458 4.108 Development and Heat Flux Testing of Actively Cooled, Long Pulse Sample Probe for the WEST Tokamak

A reciprocating, actively water-cooled probe is currently under development to be deployed in the WEST tokamak for long pulse/high fluence plasma-material interaction (PMI) and materials transport studies. The WEST long-pulse probe (LPP) probe head will be composed of tungsten coated onto an additively manufactured advanced Cu alloy, GRCop-42, body. GRCop-42 was selected as the primary structural material for the LPP head due to its desirable thermal and mechanical properties under anticipated operation temperatures (RT - 250$^\circ$C) and its suitability for powder-bed laser melting additive manufacturing, allowing for the development of complex geometry components. Tungsten coatings were chosen both to prevent foreign elemental contamination in WEST tokamak and due to their applicability to next-generation tokamaks such as ITER and SPARC which will utilize primarily tungsten-based plasma facing components. W coatings will be applied by plasma-enhanced chemical vapor deposition utilizing a nickel interlayer to ease dramatic biaxial stresses in the W coating-GRcop interface, stemming from CTE mismatch between the materials.

To qualify these advanced materials and manufacturing processes for long-pulse operation in WEST, e-beam heat flux testing is planned utilizing the High heAt load tESt (HADES) facility at CEA-IRFM on LPP test articles that mimic actual probe flow channel geometry and operation. Test articles will include non-coated GRCop-42 articles and articles with varied thicknesses, 10 and 20 $\mu$m, of W coatings to: 1) Demonstrate delamination/spalling resilience of W-coatings to cyclic thermal stresses beyond what probe materials will endure (~1000 cycles, 10 second, 1.75 MW/m$^2$ pulses) and 2) Characterize coating and substrate material degradation under intense, low-cycle thermal stresses to predict material cycle limits and possible failure mechanisms under fusion-relevant thermal conditions. This work summarizes efforts made thus far to develop the WEST LPP design, e-beam heat flux testing of WEST probe test articles, and the WEST LPP science mission during operation.

Speaker: Lauren Nuckols (Oak Ridge National Lab)

459 4.109 Real-Time IR-Based Hotspot Monitoring and Feedback Protection for Tokamak First Walls Using Machine Learning

To protect against abnormal thermal events on the tokamak first wall and divertor in real time, this work proposes an intelligent monitoring and feedback system based on infrared diagnostics and machine learning. At the model level, a hotspot detection network is fine‑tuned on infrared data starting from visible‑light pretrained weights, enabling robust hotspot localization and estimation of key temperature regions. On the engineering and control side, the server is connected to the EAST Plasma Control System (PCS) via a fibre‑optic link, enabling low‑latency infrared frame transmission and control signal delivery. On this basis, the model outputs are coupled to PCS actuation and impurity puffing, achieving a closed‑loop response from hotspot detection to impurity injection–based divertor cooling within about 200 ms. The system significantly improves the response speed and active cooling capability for first‑wall thermal events on EAST, and suggests a new divertor cooling strategy of relevance to ITER.

Speaker: Zhongfang Guan (GNOI)

460 4.110 Design of a Fast Reciprocating Diagnostic to characterize the plasma boundary in TCV

Controlling heat and particle fluxes on plasma-facing components remains a major challenge on the path toward ITER and future fusion reactors. These fluxes are strongly influenced by the dynamics in the boundary region, where turbulence and plasma flows partly determine first-wall heat and particle loads, as well as impurity transport, ultimately affecting sputtering and core plasma performance [1,2]. Characterizing these processes requires diagnostics capable of providing fast and spatially resolved measurements of, among other, electron and ion temperatures, densities, electric fields, and flows. Reciprocating probes have proven effective in meeting this need, providing spatial profiles across the outer midplane scrape-off layer (SOL) and divertor for such key quantities [3, 4], motivating the development of the Fast Reciprocating Diagnostic (FReDi) for the Tokamak à Configuration Variable (TCV).

FReDi is designed to allow measurements of low-field-side SOL profiles either at the midplane or in the divertor, with interchangeable probe heads, the first of which integrates opposingly oriented Retarding Field Analyzers (RFAs), providing ion temperature and density. Inherent challenges in the design, such as transparency dependence on parallel velocity and space-charge effects [3, 5], which constrain the geometry and set the operating limits for accurate measurements, are evaluated through Monte Carlo simulations of ion trajectories. On the same probe head, a flexible array of Langmuir probes enables the simultaneous inference of electron temperatures, densities, electric fields, and parallel plasma flow.

Designing such an advanced probe is demanding due to the harsh plasma environment in which it must operate. The fast and precise horizontal motion is achieved using a servolinear motor. Other key design considerations include material selection to minimize plasma perturbation, shaping of the probe head to prevent temperature surges above material limits, and characterization of the electronics transfer function to ensure distortion-free measurements.

The finalized design meets these constraints, establishing FReDi as a powerful diagnostic for investigating SOL and divertor turbulence, flows, and heat and particle flux profiles across a wide range of operating scenarios, thereby substantially enhancing TCV’s diagnostic capabilities.

[1] Giacomin, M. & al. (2021), Nuclear Fusion
[2] Pitts, R. A. & al. (2025), Nuclear Materials and Energy
[3] Brunner, D. & al. (2013), Review of Scientific Instruments
[4] Killer, C. & al. (2022), Journal of Instrumentation
[5] Kočan, M. & al. (2008), Review of Scientific Instruments

Speaker: Alysée Khan (EPFL-SPC)

461 4.111 Reactor-relevant sensing of enhanced PSI phenomena via the sub-divertor: an exploratory study

There is renewed interest in using the sub-divertor for diagnosis of the fusion plasma burn process, as this region of the reactor chamber would not require access via the main chamber wall and T-breeding blanket [1]. In the referenced study, it was shown that it would be feasible to control the DT burn by feedback from $^3$He concentration measurement in the sub-divertor, assuming the reactor is using the same $^3$He ICRH heating scheme as used in JET DTE2.
The present study is looking at the viability of sensing plasma surface interaction phenomena that could link to the sub-divertor via enhanced emission of fuel or impurity gases, or molecular compounds thereof. Specifically, cases from two long-pulse devices, WEST and W7-X are explored, in which the mass 3 (amu = 3) reading from a sub-divertor mass spectrometer is correlated to PSI phenomena in the main chamber. In the case of WEST, a rise in mass-3 is correlated to formation of HD at heated plasma surfaces, contribution for more HD molecular radical presence in the mass spectrometer than what might naturally form in the analyzer [2]. In the case of W7-X, where deuterium is not yet deployed, a rise in mass-3 is interpreted as due to formation of H$_3$ under certain PSI conditions, and H$_3$ molecular emission from the island divertor region is being explored via IR spectroscopy to validate this correlation. [3, 4]
The present contribution will include first outcomes from a modeling exploration for the case of WEST, deploying an advanced modeling framework- STRIPE -recently developed to analyze material erosion and the global transport of eroded impurities originating from radio-frequency (RF) antenna structures in full 3D linear and toroidal geometries [5,6]. The aim for this work is to determine the likelihood of HD formed in enhance PSI can reach at least the divertor pumping gap.
*Work supported, in part, by U.S. Department of Energy under Contract No. DE-AC05-00OR22725 with UT-Battelle, LLC.
References
[1] C.C. Klepper et al., 2025 Nucl. Fusion 65 086015; DOI: 10.1088/1741-4326/ade9dc
[2] G. Schlisio et al., "A novel fast mass spectrometer for fusion applications - in preparation", and related contribution #147 this Conference.
[3] E.A. Hodille et al., 2025 Nucl. Mater. Energy 45 101999, DOI: 10.1016/j.nme.2025.101999
[3] E. Gauthier, EPS 1995.
[4] A. Kumar et al 2025 Nucl. Fusion 65 076039; DOI: 10.1088/1741-4326/ade455
[5] A. Kumar et al 2025 EPJ Web. Conf. (accepted for publication)

Speaker: C. Christopher Klepper (Oak Ridge National Laboratory, Oak Ridge, TN 37831-6169, United States of America)

462 4.112 Langmuir probe and infrared thermography measurements of wide and narrow heat flux profiles in the ST40 tokamak

In previous ST40 experimental campaigns, distinct wide and narrow scrape-off layer (SOL) profile widths have been observed [Zhang et al., Nuclear Materials and Energy 41 (2024)]. The “wide” profiles follow traditional scalings built on a multi-device empirical database [T. Eich et al., 2013 Nucl. Fusion 53 093031], while the “narrow” profiles have been found to be an order of magnitude narrower. A comparison between divertor Langmuir probe and infrared (IR) thermography measurements of parallel heat flux yield good agreement for both cases with parallel heat fluxes reaching up to 150 MW/m2 in the narrow case. Upgrades to the ST40 tokamak related to divertor diagnostics by the addition of more Langmuir probes and another IR camera system viewing the lower divertor are also presented.

Speaker: Trenton Brewer (Tokamak Energy Ltd.)

463 4.113 Daily wall condition assessment using ECE measurements during KSTAR wall reference discharge

Maintaining plasma performance stability and enhancing equipment durability during the long-term KSTAR experimental campaign is crucial. To achieve these goals, removing contaminants attached to the inner wall of the tokamak device is essential. Among various wall cleaning methods, Glow Discharge Cleaning (GDC) is one of the most effective techniques. At KSTAR, GDC is typically conducted before experiments, usually in the morning, or periodically. The wall reference shot (ohmic discharge) is used as a baseline to assess the wall condition and compare it with that of the previous day, prior to starting an experiment. In parallel, the electron temperature of the plasma is regularly monitored using Electron Cyclotron Emission (ECE) diagnostics. This presentation outlines a methodology for tracking wall conditions after GDC, employing ECE measurements along with average plasma density data from daily wall reference discharges. Additionally, the effectiveness of the GDC process is assessed weekly by comparing the diamagnetic energy and neutral gas pressure. A comparative analysis of GDC performance under carbon and tungsten divertor environments is also provided.

This work was supported by the R&D programs of “High Performance Tokamak Plasma Research & Development” (EN2501) through the Korea Institute of Fusion Energy (KFE) funded by the Government funds, Republic of Korea.

Speaker: Dr Kyu-Dong Lee (Korea Institute of Fusion Energy)

464 4.114 Investigation of the noise crosstalk associated with bias-voltage swept KSTAR divertor Langmuir probes depending on the waveforms

A suite of Langmuir probes (LPs) installed on the KSTAR lower tungsten divertor, consisting of 52 in the D-port and another 52 in the L-port, has been operated since 2023. The D-port Langmuir probes are battery-biased at -240 V to measure temporal behavior of the ion saturation currents. During the 2024 campaign, a single sweeping Langmuir probe driven by a bipolar supply (-100 to +100 V, 1 kHz, triangular waveform) was operated to acquire the IV-characteristics. To suppress effects of the edge-localized-mode (ELM) on the IV-characteristics, an ELM filter was first applied to the raw signals. The IV characteristics were further processed by applying either a 1 kHz comb filter or a 30 kHz low-pass filter to the ELM-filtered data. Plasma parameters, such as density and temperature, were extracted by fitting the filtered IV-characteristic curves within 50 ms time windows using the Weinlich-Carlson asymmetric double-probe (WC-ADP) model [1]. The resulting temporal trends show good agreement with mid-plane edge ECE (electron cyclotron emission) [2]. From the KSTAR 2025 campaign, nine additional power supplies and associated circuitries have been installed, enabling simultaneous sweeping of the bias voltage (-100 to +100 V) for a total of ten designated Langmuir probes. During the measurements, it has been observed that such a simultaneous sweeping introduces noise crosstalk which is not negligible. In particular, bias-voltages with the triangular waveform produce peaked noises near the local extrema that couple, i.e., crosstalk, into other Langmuir probes. Although triangular bias-voltage provides a uniform dV/dt, the sign reversals at the turning point drive large capacitive (displacement) and inductive currents, producing artificial ringing signals. We, therefore, investigate and compare the noise characteristics under triangular and sinusoidal bias-voltages, and estimate parasitic-R/C leakage and displacement currents.

[1] Rudischhauser, Lukas, et al. "The Langmuir probe system in the Wendelstein 7-X test divertor." Review of Scientific Instruments 91.6 (2020).
[2] Bong, Seungmin, et al. "Newly designed Langmuir probe system at the tungsten lower divertors in KSTAR." 9th Asia-Pacific Conference on Plasma Physics MF2-9-I2 (2025)

Acknowledgments
This work was supported by the National R&D Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (Grant Nos. RS-2022-00155917 and NRF2021R1A2C2005654). This work was also supported by the R&D Program of the “KSTAR Experimental Collaboration and Fusion Plasma Research (EN2503)” trough the Korea Institute of Fusion Energy (KFE) funded by the Korean Ministry of Science and ICT.

Speaker: Seungmin Bong (KAIST)

465 4.115 Operation and Characterization of the Gas Puff Imaging System on the ADITYA-U Tokomak

Edge turbulence and cross-field transport in tokamaks remain critical unresolved challenges for the operation of future magnetic fusion devices such as ITER [1]. A variety of diagnostic techniques—including probes, passive imaging, and more recently Gas Puff Imaging (GPI)—have been employed to investigate these edge plasma phenomena [2, 3]. GPI enhances local visible emission through controlled neutral gas injection, enabling high spatial and temporal resolution measurements of edge turbulence, radial transport events, and blob-like structures in the scrape-off layer [4, 5].

A GPI system has been designed, developed, and installed for the first time on the ADITYA-U tokamak [6]. This work presents the characterization of the system, with emphasis on the dependence of GPI signal quality on the injected gas flux and the influence of gas puffing on plasma discharge parameters. A comparative study using hydrogen (the main discharge gas) and helium as the puff species is also performed. Differences in the spectral and statistical properties of the resulting GPI signals are analysed and discussed to assess the suitability of each gas for turbulence measurements on ADITYA-U.

References:

[1] W.M. Tang 1978 Nucl. Fusion 18 1089
[2] S.J. Zweben et al., Phys. Plasmas 9 (2002) 1981
[3] N. Offeddu, C. Wüthrich, et al., Rev. Sci. Instrum. 93, 123504 (2022).
[4] S. J. Zweben, D. P. Stotler, et al.. Phys. Plasmas 1 October 2017; 24 (10): 102509.
[5] Han, W., Pietersen, R.A., Villamor-Lora, R. et al. Sci Rep 12, 18142 (2022).
[6] Ruchi Varshney et al., To be submitted to peer peer-reviewed journal (2026)

Speaker: Ruchi Varshnedy (Institute for plasma research, india)

12:20 Lunch

Invited Talk: Afternoon session

466 I23 Filamentary transport in Quasi-Continuous Exhaust and detached X-Point Radiator regime simulations

Divertor detachment is mandatory for fusion reactors and must be reconciled with good confinement. While the H-mode remains attractive for its high confinement, heat and particle loads on plasma facing components must be mitigated, which result from ELM bursts and the narrow inter-ELM SOL width.

We present novel GRILLIX simulations of the Quasi-Continuous Exhaust (QCE) and detached X-Point Radiator (XPR) regimes, which reproduce and explain experimental measurements. Furthermore, dynamic regime transitions from L- to H-mode, EDA to QCE, and attached to detached divertor conditions are demonstrated.

In QCE, we find that the Quasi-Coherent Mode (QCM) which replaces ELMs is composed of a narrow kinetic ballooning mode (KBM) spectrum, destabilized linearly by high plasma shaping, and non-linearly by a reduction of the E×B shear by the Maxwell stress. At low density, this mode is coherent, and the SOL width remains narrow, consistent with EDA conditions. At high plasma density, additionally, the resistive X-point mode is destabilized and couples to the KBM. This results in decoherence of the QCM and production of plasma blobs in the near-SOL. The blobs increase SOL transport, increasing the heat channel width $λ_q$ by a factor 3.

In XPR conditions, we find fundamentally different mechanisms. The combination of turbulence and radiative condensation results in large fluctuations in the X-point region of up to 500% amplitude, as the plasma oscillates between ionizing and recombining conditions, and the spots move around. This has far reaching consequences for both detachment and transport. On one hand, the large and strongly anti-correlated density and temperature fluctuations locally reduce the ionization rate by a factor 2, and increase recombination by more than a factor 4, facilitating detachment. This is missed by mean-field simulations, e.g. with the SOLPS code, which evaluate reaction rates only from mean density and temperature. On the other hand, the radial transport coefficients increase by an order of magnitude, additionally to the formation a convective cell. This could explain the reduction of the pedestal by the XPR, and avoidance of ELMs.

Explaining these regimes and the dependence of their access on magnetic geometry and plasma composition bears promise for the design of fusion reactors.

Speaker: Wladimir Zholobenko (Max Planck Institute for Plasma Physics)

Oral: Afternoon session

467 O28 Tokamak boundary turbulence of detached plasmas with kinetic neutrals

Fusion reactors will operate with a strong scrape-off layer (SOL)/edge plasma-neutral interactions to benefit from the energy and momentum losses due to atomic reactions to protect the divertor walls, the so-called (semi) detached plasma. While access and basic properties of such regimes are well known, a complete and detailed understanding of the tokamak boundary of detached plasmas is absent. Progress has been made with mean-field simulations, however, this approach does not model self-consistently the cross-field transport, known to be dominated by turbulence and strongly dependent on the divertor regime as shown by experimental evidence. A first-principles understanding of the underlying physics of those plasmas is imperative to clearly define operational limits of future machines.
In this contribution, we discuss the results of the turbulence modeling of the TCV-X23 reference case, a lower single-null plasma at Btor=0.95T with an elongated outer divertor leg in attached and semi-detached conditions and portraying an extensive dataset including average and higher-order statistical moments of several observables. In particular, the filament dynamics and the plasma-neutral reactions are assessed by Gas Puffing Imaging (GPI) and spectroscopy, offering an ideal scenario for the validation of edge turbulence simulations. The modelling is carried out with SOLEDGE3X, a multispecies, electromagnetic, drift-reduced Braginskii code depicting the kinetic neutral dynamics via coupling with EIRENE, a feature recently available for turbulence simulations.
The reported simulations cover two different density levels in the reversed field direction and are rigorously validated against the experimental dataset containing average and fluctuations observables distributed across the entire tokamak boundary. Differently from the previous SOLEDGE3X simulations, where a simplified fluid neutral model including only the density equation was adopted, the current modelling includes all the atomic and molecular reaction terms for the momentum and energy equations in a kinetic description. The new physical model leads a divertor not fully detached in the high-density case, more in line with the experiments and demonstrating the improvement of code predictive capabilities. In high density conditions, a wider SOL and a larger heat flux spreading is found due to enhanced turbulent transport as consequence of the large collisionality. In particular, the dynamics of divertor-localized and upstream-connect filaments are investigated and new insights regarding the motion and associated transport are presented and confronted with GPI data of the divertor region.

Speaker: Diego Oliveira (CEA - IRFM)

468 O29 Accounting for cross-field transport and neutral recycling in a Lengyel-like detachment model

To manage erosion and protect against heat fluxes, tokamak power plants will need to be operated with detached divertors. To design future tokamaks around core-edge-integrated operating scenarios, we need fast, accurate detachment models. We developed a model to calculate the impurity concentration needed to detach the first $\lambda_q$ of the outer divertor. This model is based on the 0D analytical ‘Lengyel’ model, which forms the basis of other models such those by Goldston [1] and Reinke [2]. We extended the Lengyel model to include corrections for the cross-field transport in the divertor, power and momentum loss due to recycled neutrals, and turbulent broadening of $\lambda_q$ [3]. Our extended Lengyel model reproduces the empirical ‘Kallenbach’ scaling [4], which has been validated against detachment studies from several tokamaks including JET and ASDEX Upgrade [5]. In addition, a dedicated validation against a detached ASDEX Upgrade shot shows that our model can simultaneously predict the impurity concentration, divertor neutral pressure and separatrix temperature at detachment onset with reasonable accuracy [3]. The extended Lengyel model finds a transition from $c_z \propto n_{sep}^{-2}$ at low densities to about $c_z \propto n_{sep}^{-3}$ at high densities, encompassing the range of experimental scalings determined on ASDEX Upgrade and JET [6]. Applied predictively, the model finds that the ARC V3A tokamak ($P_{sep}$~120MW, $n_{sep}$~$10^{20}/m^3$) should access detachment with ~1% argon in the divertor. This is only slightly higher than on existing tokamaks due to the long divertor leg length and high absolute density — and significantly less than the ~6% predicted by the standard Lengyel model.

[1] R. J. Goldston et al 2017 Plasma Phys. Control. Fusion 59 055015
[2] M. L. Reinke 2017 Nucl. Fusion 57 034004
[3] T. Body, A. Kallenbach, T. Eich 2025 Nucl. Fusion 65 086002
[4] A Kallenbach et al 2016 Plasma Phys. Control. Fusion 58 045013
[5] S.S. Henderson et al 2024 Nucl. Fusion 64 066006
[6] S.S. Henderson et al 2021 Nucl. Mat. & Energy 28 101000

Speaker: Thomas Body (Commonwealth Fusion Systems)

Invited Talk: Afternoon session

469 I24 Near scrape-off layer decay lengths in D, T and DT plasmas on JET

In the near scrape-off layer (SOL), heat transport is dominated by parallel electron conduction, resulting in a radial heat flux decay length ($\lambda_q$) that is short relative to the machine size. The cross-field extent of the power entering the divertor region is set by $\lambda_q$, which strongly affects divertor performance, including the peak heat flux and access to detachment. In the conduction-limited regime, $\lambda_q$ is proportional to the near-SOL electron temperature decay length ($\lambda_T$). In addition, the near-SOL electron density decay length ($\lambda_n$) strongly impacts particle flux profiles at the divertor target. At high separatrix densities on ASDEX Upgrade, Sun et al. [1] reported broadening of both $\lambda_T$ (and hence $\lambda_q$) and $\lambda_n$ beyond the prediction of the attached $\lambda_q$ scaling [2]. These observations motivated Eich et al. [3] to propose new empirical scalings for $\lambda_T$, $\lambda_n$, and the electron pressure decay length ($\lambda_p$), based on a turbulence control parameter $\alpha_t$. Higher values of $\alpha_t$ correspond to increased cross-field transport driven by resistive-interchange turbulence, leading to profile broadening.

The aim of this contribution is to systematically investigate the dependence of $\lambda_T$, $\lambda_n$, and $\lambda_p$ on $\alpha_t$ in ELMy conditions on JET, including the effects of divertor configuration and isotope fuelling (D, T, and DT) without external impurities. Most empirical $\lambda_q$ scalings are based on experiments with deuterium plasmas, introducing uncertainty when extrapolating to deuterium–tritium (DT) operation. Both isotope [4] and divertor configuration [5] can affect plasma evolution under identical engineering parameters.

The decay lengths were strongly correlated with $\alpha_t$ and showed no significant isotope dependence within $\alpha_t < 0.35$. Vertical–vertical (inner–outer) divertor configurations exhibited broader profiles than vertical–horizontal configurations at the same $\alpha_t$, indicating that target recycling dynamics are not fully captured by the $\alpha_t$ framework. Trends in $\lambda_T$ and $\lambda_p$ showed reasonable agreement with the ASDEX Upgrade empirical scalings [3], providing a valuable inter-machine comparison that validates the machine-size dependence inherent in $\alpha_t$. Larger scatter in the $\lambda_n$ measurements indicates that particle and heat exhaust are governed by distinct mechanisms. For a given divertor configuration, $\alpha_t$ provides a simple parameter to characterise near-SOL broadening, offering guidance for reactor power-exhaust design.

[1] H.J. Sun et al., Plasma Phys. Control. Fusion 57, 125011 (2015).
[2] T. Eich et al., Nucl. Fusion 53, 093031 (2013).
[3] T. Eich et al., Nucl. Fusion 60, 056016 (2020).
[4] C.F. Maggi, Nucl. Fusion (2025).
[5] E. Joffrin et al., Nucl. Fusion 57, 086025 (2017).

Speaker: Peter John Ryan (UKAEA)

Oral: Afternoon session

470 O30 Pedestal origin, SOL transport and extrapolation of small-ELM energy flux in ITER and SPARC

Experimental analysis and simulations with the BOUT++ code[1] show that small edge-localized modes (ELMs) in reactor-relevant high-density regimes originate in a region close to the separatrix and only marginally perturb the pedestal structure. The measured divertor peak parallel energy flux for a database of small ELMs in DIII-D and ASDEX Upgrade can be reproduced, within 40% accuracy on average, if an ad-hoc modification of the peak parallel ELM energy flux model[2] is applied to account for the small ELMs pedestal birth location. This allows for first-order extrapolation of small ELM divertor parallel energy fluxes to ITER and SPARC, resulting in values that satisfy the nominal melting threshold of Tungsten monoblocks of 12 MJ/m2[3].

An inverse dependency is observed between the experimental ELM energy flux to the divertor and the separatrix turbulence control parameter, αt[4]. As αt increases from ∼0.2 to ∼0.8, transitioning from a type-I ELM to a small-ELM regime, the divertor peak energy flux decreases by a factor of five. The scrape-off-layer (SOL) transport associated with small ELMs is found to be related to divertor conditions, where higher divertor collisionality correlates with stronger upstream radial heat and particle fluxes. While large αt at the separatrix leads to substantial mitigation of intra-ELM divertor heat loads, the accompanying enhancement of first-wall fluxes might present challenges for first-wall integrity and/or impurity sources in future machines.

The findings reported in this study, both via modeling and experiments, constitute a step forward toward the assessment of high edge-collisionality scenarios as viable plasma regime for the operation of near-future fusion machines. The extrapolations to SPARC and ITER constitute first quantitative projections that can support future PMI-focused analyses aimed to inform reactors design.

[1] T. Y. Xia and X. Q. Xu, Nucl. Fusion, 55, 113030 (2015)
[2] T. Eich et al., Nucl. Mater. Energy 12 (2017) 84-90
[3] J. Gunn et al., Nucl. Fusion 57 (2017) 046025
[4] B. D. Scott, Phys. Plasmas 12 (2005) 062314

---

Work supported by US DOE under DE-FC02-04ER54698, DE-FG02- 07ER54917 and DE-AC52-07NA27344.

Speaker: Renato Perillo (UCSD)

15:20 Closing