Every physical system is designed on microstructure-property relationships of materials for optimal performance, but the performance inevitably declines due to material degradation. Understanding a long-term microstructural evolution is important to ensure safe operation, and understanding defect generation in high-temperature or high-energy applications is invaluable as the material degradation process is rapid and the consequences can be fatal. Nevertheless, reliable identification and classification of lattice defects in atomistic simulations for polycrystals remain a long-standing challenge. The fundamental problem with conventional methods, such as the Wigner-Seitz cell method, is that point defects are identified not by actual lattice points but by initial atomic positions. Consequently, the defect analysis from existing methods is valid only when the initial atomic arrangement is the perfect lattice structure. In this study, we introduce two new defect analysis techniques based on the local atomic connectivity to classify and quantify point defects. Both methods capture the correct defect-production trend in collision-cascade simulations that is otherwise not captured by the existing methods. These scalable approaches provide robust, accurate defect classification for polycrystalline materials, which are inherently defective.
Shin et al. (Fri,) studied this question.