Developing Cross-scale Machine Learning Interatomic Potential for Fe-Cr-Ni Stainless Steel

The overarching goal of this project is to provide foundational, holistic understanding on hydrogen-metal interaction to shed light on the mechanical properties of stainless steel at elevated temperatures in hydrogen environment. The approach is to first develop cross-scale machine learning interatomic potential and then simulate mechanical behavior using molecular dynamics. Key crystal defects considered include H interstitials, vacancies, dislocations, stacking faults, grain boundaries, surfaces, and precipitates. The Moment Tensor Potentials platform is adopted for this work since it demonstrates a fine balance between model accuracy and computational efficiency. The potential is well trained based on large amount of high-fidelity density functional theory calculations of the constituent systems and compositions. The validation is carried out by comparing various important properties including short range order, coefficient of thermal expansion, elastic properties, stacking fault energy, grain boundary energy, and surface energy with experiments and density functional theory calculations. This work illustrates an efficient and reliable approach for atomistic simulation of high temperature hydrogen attack of stainless steel.