Multi-Objective Optimization of the Nanocavities Diffusion in Irradiated Metals

Le résumé fourni par la source

Materials in fission reactors or fusion tokamaks are exposed to neutron irradiation, which creates defects in the microstructure. With time, depending on the temperature, defects diffuse and form, among others, nanocavities, altering the material performance. The goal of this work is to determine the diffusion properties of the nanocavities in tungsten. We combine (i) a systematic experimental study in irradiated samples annealed at different temperatures up to 1800 K (the created nanocavities diffuse, and their coalescence is studied by transmission electron microscopy); (ii) our object kinetic Monte Carlo model of the microstructure evolution fed by a large collection of atomistic data; and (iii) a multi-objective optimization method (using model inversion) to obtain the diffusion of nanocavities, input parameters of our model, from the comparison with the experimental observations. We simplify the multi-objective function, proposing a projection into the parameter space. Non-dominated solutions are revealed: two “valleys” of minima corresponding to the nanocavities density and size objectives, respectively, which delimit the Pareto optimal solution. These “valleys” are found to be the upper and lower uncertainties on the diffusion beyond the uncertainties on the experimental and simulated results. The nanocavity diffusion can be split in three domains: the mono vacancy and small vacancy clusters, for which atomistic models are affordable, the small nanocavities for which our approach is decisive, and the nanocavities larger than 1.5 nm for which the classical surface diffusion theory is valid.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.