Identification of Deformation Mechanisms in Nanocrystalline B2 CuZr

Nanocrystalline B2 CuZr alloys are of significant interest due to their mechanical relevance in shape memory systems and metallic glass composites, where grain-scale deformation mechanisms critically influence macroscopic response. In this work, molecular dynamics simulations are combined with machine learning structural classification to investigate the tensile deformation behavior of nanocrystalline B2 CuZr with varying grain sizes. Uniaxial tensile tests reveal a clear grain-size dependence of the ultimate tensile stress, with larger average grain sizes exhibiting higher peak strength, while the flow stress shows a weaker sensitivity to grain size. Atomic scale analysis demonstrates that plastic deformation is dominated by grain boundary mediated mechanisms, including atomic rearrangements, atomic migration from grains to grain boundaries, progressive grain shrinkage, and void nucleation, rather than stress-induced martensitic transformation commonly observed in single-crystal B2 CuZr. Overall, the machine learning classifiers used here enabled the identification of distorted B2 environments, grain boundaries, defect structures, and void surfaces during deformation, providing a comprehensive description of the microstructure evolution. These results provide a general data-driven framework for analyzing complex deformation processes in crystalline materials.