Hybrid ab initio-machine learning simulations of dislocations.

Dislocations are extended line defects which carry plastic deformation in crystalline materials. Understanding and optimizing dislocation behaviour by characterising dislocation interaction with point defects is a central topic in computational metallurgy. For this task, ab initio calculations, specifically density functional theory (DFT), are essential to capture dislocation core structures and complex bonding to impurity elements. However, the computational cost of DFT typically has cubic scaling with the number of atoms for metallic systems, which limits its direct applicability to the study of extended defects. In this seminar I will present how hybrid QM/MM methods in combination with modern machine learning force fields allow to study unfeasibly large systems with ab initio accuracy.