Generalized Langevin dynamics simulation of interacting particles

Abstract

We analyze the effect of including time-dependent memory functions in computer simulation of interacting Brownian particles. Numerical integration of generalized Langevin equations is carried out by an Allen-like algorithm. The inclusion of time-correlated random forces does not result in significative changes in the structure and self-diffusion coefficients, but the dynamic evolution of particles becomes closer to the real one.