Characterization of crystal defects during molecular dynamics simulations of mechanical deformation
PhD Thesis, Ruhr-Universität Bochum (2012)
Atomistic simulations such as molecular dynamics simulations are a well-established method to investigate the deformation mechanisms in crystalline solids during plastic deformation. Due to the progress in the available computational power over the last decades, the volumes that can be simulated within a reasonable amount of time by these methods have increased significantly, diminishing simulation artifacts introduced by small-scaling and enabling more realistic simulation setups. Nowadays a point is reached, where the bottleneck shifts from a lack of computational power towards a lack of proper methods to treat the enormous data produced with these simulations in such a way that the desired material properties become available. In the first part of this thesis, several new analysis methods for atomistic simulations are introduced that offer further insight in the ongoing mechanisms during plastic deformation. As plastic deformation is governed by the nucleation and evolution of dislocations, algorithms are developed that can derive dislocation networks directly from atomistic data and characterize each dislocation by its Burgers vector. For the study of deformation in shape memory alloys (SMAs) these methods are combined with algorithms to approximate phase boundaries, thus a link between plasticity caused by dislocations and pseudoelasticity by martensitic phase-transformation is established. In the second part, these methods are applied in nanoindentation simulations of two materials. On the one hand copper that deforms elastic-plastically by dislocation movement like most pure metals and on the other hand SMAs, in which pseudoelastic deformation, a diffusionless reversible austenitic-martensitic phase transformation takes place in addition to elastic-plastic deformation. Simulations of indentation in copper single crystals are performed to study the mechanisms of dislocation nucleation in advanced stages of plastic deformation and to investigate the distribution of geometrically necessary dislocation (GND) densities. Furthermore, the interaction of plastic deformation by dislocations and pseudoelasticity by phase-transformation is studied in simulations using a model potential for a shape memory alloy. In these simulations, dislocations are clearly identified as barriers for the transformation of martensite back to the austenitic phase.