Speaker
Description
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.