AI and the World's Hard Problems · Nuclear fusion & energy · arXiv · 2026-10-20
Atomic-level study of dose-dependent lattice thermal conductivity degradation in irradiated tungsten
The authors used non-equilibrium molecular dynamics and machine learning to study how neutron-irradiation-induced defects reduce the lattice thermal conductivity of tungsten, a fusion plasma-facing material. Simulations covered doses up to 1.8 cdpa, and the ML analysis identified vacancy density as the dominant factor. They report a step-like degradation at a vacancy threshold of 0.05 at.%, with the maximum relative reduction capping at about 30 at.%.