| Abstract Scope |
Deployment of new structural materials for nuclear reactors, such as novel austenitic stainless steels (SSs), requires materials qualification. Qualification of novel austenitic SS compositions requires understanding time-dependent mechanical properties, especially thermal and irradiation creep and radiation-induced segregation (RIS). Both properties depend on vacancy diffusion within the material, meaning accurate quantification of vacancy diffusion is crucial. However, though initial compositional heterogeneity and radiation-induced compositional segregation is common, the local compositional effects on vacancy diffusion have not been studied in austenitic SS. Vacancy diffusivity calculations require high accuracy vacancy migration energy barrier (VMEB) calculations. These in turn require the high computational expense of quantum mechanical calculations, such as density functional theory (DFT). Additionally, over 5x105 VMEB calculations are required to cover the vacancy’s local compositional space in austenitic SS. Therefore, we accelerate VMEB prediction by implementing a modified Gaussian process regression (MGPR) trained on a subset of DFT calculations. The MGPR-predicted VMEBs are more accurate and are produced over 300 times faster than those from four molecular dynamics embedded atom potentials. Employing the MGPR-predicted VMEBs in a kinetic Monte Carlo (KMC) algorithm, we predict vacancy diffusivity. Results agree with the available experimental trends reported within the austenitic stainless steel composition space. This methodology can be applied to any other ternary (or more complex) system for which local composition effects on vacancy diffusion are of interest, especially solid solution alloys used as fuel cladding or structural materials in nuclear power plants. |