Speaker
Description
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.