Search

The energetics and kinetics of interactions between microstructural point, line, and planar defects govern the most significant mechanical properties of structural materials, which is crucial in assessing, predicting, and controlling the material behavior in technological applications such as nuclear reactors and additive manufacturing (AM). Modeling microstructural evolutions under extreme conditions with high atomistic fidelity and at experimental time scales has been known as a longstanding challenge in material science. Under some cases, traditional atomistic modeling techniques like molecular dynamics (MD) enable the elucidation of molecular mechanisms of microstructural evolutions. Nevertheless, conventional MD could hardly go beyond nanoseconds and therefore simple extrapolation of MD results from short timescales could undermine the accuracy or even give a misleading prediction of materials performance at realistic environments. To overcome such challenges, advanced atomistic modeling techniques that can directly tackle the long timescale problems are greatly desired. In this thesis, we present a novel computational framework based on the concept of potential energy landscape (PEL), which enables the investigation of microstructural evolutions at long-time scales while fully retaining the atomistic details. To demonstrate its capabilities, we show the application of this framework to some key problems in material science. The following problems are addressed: (i) We investigate a local interaction between dislocation and a vacancy-type obstacle in BCC Fe at room temperature over a wide range of strain rates, from 108s-1 down to 103s-1. Very surprisingly, a non-monotonic correlation is found between the critical resolved shear stress (CRSS) of the system and applied strain rates. We demonstrate that the nSRS is due to the complex interplays between thermal activation and applied strain rate. (ii) We map the kinetic evolution of metastable symmetric tilt grain boundaries (GBs) in copper under non-equilibrium processing, which is significant in determining the strength and ductility of nanocrystalline materials. Our combined study, providing both atomistic simulations and novel experiments employing femtosecond laser-material interactions, demonstrates striking features of the energetics and kinetic pathways to achieve a multiplicity of grain boundary states. Specifically, it can be divided into an ageing regime and a rejuvenating regime over a broad energy—temperature parameter space. We further ascribe the origin of this ageing/rejuvenating behavior to the inherent energy imbalance along with elementary hopping processes in the system’s underlying PEL. (iii) We also investigate the non-equilibrium relaxations of metastable tilted GBs in copper under ultrafast thermal cycles, which is critical in assessing the structural materials’ properties during AM. We demonstrate that the structural evolution of metastable GBs is mainly driven by disorder and rough energy landscape rather than free volume. Most importantly, a universal scaling is observed between the GBs’ inherent structure energies and their structural transition temperatures during the rapid cooling stage. To further assess the applicability of the obtained scaling correlation, we also investigate a group of GBs in Ni. A similar correlation persists, which gives reason to expect the present study may apply to other materials as well. In summary, the PEL-based modeling framework is employed to investigate the microstructural evolution at long time scales, without invoking empirical assumptions or fitting parameters. This framework might shed light on the predictive design of advanced structural materials to be used in complex environments.