Dislocation detection algorithm for atomistic simulations

Abstract

We present a novel computational method that makes it possible to directly extract dislocation lines and their associated Burgers vectors from three-dimensional atomistic simulations. The on-the-fly dislocation detection algorithm is based on a fully automated Burgers circuit analysis, which locates dislocation cores and determines their Burgers vector. Through a subsequent vectorization step, the transition from the atomistic system to a discrete dislocation representation is achieved. Using a parallelized implementation of the algorithm, the dislocation analysis can be efficiently performed on the fly within a molecular dynamics simulation. This enables the visualization and investigation of dislocation processes occurring on sub-picosecond time scales, whose observation is otherwise impeded by the presence of other crystal defects or simply by the huge amount of data produced by large-scale atomistic simulations. The presented method is able to identify individual segments as well as networks of perfect, partial and twinning dislocations. The dislocation density can be directly determined and even more sophisticated information is made accessible by our dislocation analysis, including dislocation reactions and junctions, as well as stacking fault and twin boundary densities.