Nanocrystalline (NC) metals with averaged grain size smaller than 100 nm have shown promising mechanical properties such as higher hardness and toughness than conventional coarse-grained metals. Unlike conventional metals in which the deformation is controlled by dislocation activities, the microstructural evolution in NC metals is mainly dominated by grain rotation and stress-driven grain boundary motion (SDGBM) due to the high density of grain boundaries (GBs). SDGBM is thus among the most studied modes of microstructural evolution in NC materials with particular interests on their fundamental atomistic mechanisms. In the first part of this thesis, molecular dynamics simulations were used to investigate the influences of Triple Junctions (TJs) on SDGBM of symmetric tilt GBs in copper by considering a honeycomb NC model. TJs exhibited asymmetric pinning effects to the GB migration and the constraints by the TJs and neighboring grains led to remarkable non-linear GB motion in directions both parallel and normal to the applied shear. Based on these findings, a generalized model for SDGBM in NC Cu was proposed. In the second part, the interaction of SDGBM with crack, voids and precipitates was investigated. It was found that depending on the GB structure, material type and temperature, there is a competition between different atomistic mechanisms such as crack healing, recrystallization and GB decohesion. It is hoped that the findings of this work could clarify the micro-mechanisms of various experimental phenomena such as grain refinement in metals during severe plastic deformation, which can be used to design optimized route of making stabilized bulk NC metals. / February 2016

This thesis consists of a thermodynamically based kinetic model that more accurately predicts grain boundary mobility (GBM) over a large range of thermodynamic states including changes in temperature, pressure and shear stress. The form of the model was validated against calculated GBM values for Al bicrystals via molecular dynamics (MD) simulations. A total of 98,786 simulations were performed (164 different GBs, each with a minimum of 250 different thermodynamic states, and 2 different driving forces). Methodology for the computation of the GBM via MD simulations is provided. The model parameters are directly linked to extensive thermodynamic quantities and suggest potential mechanisms for GBM under combined thermal and triaxial loads. This thesis also discusses the influence of GB character on the thermodynamic mobility parameters. The resulting insights about GB character and thermodynamic state on GBM suggest an opportunity to achieve designed microstructures by controlling thermodynamic state during microstructure evolution.