Modeling and Computer Simulation: Molecular Dynamics and Kinetic Monte Carlo

Abstract

Recent years have witnessed tremendous advances in the realistic multiscale simulation of complex physical phenomena, such as irradiation and aging effects of materials, made possible by the enormous progress achieved in computational physics for calculating reliable, yet tractable interatomic potentials and the vast improvements in computational power and parallel computing. As a result, computational materials science is emerging as an important complement to theory and experiment to provide fundamental materials science insight. This article describes the atomistic modeling techniques of molecular dynamics (MD) and kinetic Monte Carlo (KMC), and an example of their application to radiation damage production and accumulation in metals. It is important to note at the outset that the primary objective of atomistic computer simulation should be obtaining physical insight into atomic-level processes. Classical molecular dynamics is a powerful method for obtaining insight about the dynamics of physical processes that occur on relatively short time scales. Current computational capability allows treatment of atomic systems containing as many as 10{sup 9} atoms for times on the order of 100 ns (10{sup -7}s). The main limitation of classical MD simulation is the relatively short times accessible. Kinetic Monte Carlo provides the ability to reach macroscopic times by modeling diffusional processes and time-scales rather than individual atomic vibrations. Coupling MD and KMC has developed into a powerful, multiscale tool for the simulation of radiation damage in metals.