A novel computational method was presented that made it possible to extract dislocation lines and their associated Burgers vectors directly from three-dimensional atomistic simulations. The on-the-fly dislocation detection algorithm was based upon a fully automated Burgers circuit analysis, which located dislocation cores and determined their Burgers vector. Using a subsequent vectorization step, the transition from the atomistic system to a discrete dislocation representation was achieved. Using a parallelized implementation of the algorithm, the dislocation analysis could 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 was 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 was able to identify individual segments as well as networks of perfect, partial and twinning dislocations. The dislocation density could be directly determined and even more sophisticated information was made accessible by the present dislocation analysis, including dislocation reactions and junctions, as well as stacking fault and twin boundary densities.

Dislocation Detection Algorithm for Atomistic Simulations. A.Stukowski, K.Albe: Modelling and Simulation in Materials Science and Engineering, 2010, 18[2], 025016