Assistant Professor, School of Civil and Environmental Engineering, Cornell University
Predicting Fracture in Metals with Atomistic Simulations
Friday, September 25, 1:00 – 2:00 pm
921 Benedum Hall
Many of the earliest atomistic simulations of deformation and fracture modeled the response of a several nanometer block of material under an applied load. While this
setup is analogous to elementary mechanical testing experiments, the loading rates, specimen dimensions, and resulting stresses differ by orders of magnitude. More recently, the aim of atomistic simulations has been more focused, investigating the behavior of an individual atomic scale process pertinent to macroscopic behavior. While the output of these focused simulations can provide great insight and constitutive input for coarser scale models, the problem of unrealistic strain rates and stresses remains. Most recently, simulations have begun to move away from modeling mechanical response directly and instead investigate the activation energy barriers of key deformation mechanisms.
In this presentation, I will address four substantial challenges to predictive atomic scale fracture simulation. Specifically, the challenges of a limited spatial and temporal domain, accurate interatomic interactions, and the vast complex interatomic configuration space will be addressed. In this talk I will highlight our group’s effort at using concurrent multi-scale modeling, transition state techniques, structural unit methodologies, high performance computing, and density functional theory to address these challenges.