Automated identification and tracking of deformation twin structures in molecular dynamics simulations

Abstract

Deformation twinning significantly influences the microstructure, texture, and mechanical properties of metals, necessitating comprehensive studies of twin formation and interactions. While experimental methods excel at analyzing individual samples, they often lack the capability for temporal analysis of twinned structures. Molecular dynamics simulations offer a temporal dimension, yet the absence of suitable tools for automated crystal twin identification has been a significant limitation. In this article, we introduce a novel computational tool integrated into the visualization and analysis software OVITO. Our tool automates the identification of coherent twin boundaries, links related twin boundaries, validates twin structures through orientation analysis, and tracks twins over time, providing quantifiable data and enabling in-depth investigations. Validation on a copper single crystal under shear loading demonstrates successful tracking of various twins, revealing their genesis and growth over multiple timesteps. This innovative approach promises to advance the computational materials science domain by facilitating the study of deformation twinning, offering profound insights into the behavior and mechanical performance of materials.