A low-variance particle collision scheme for variable weight DSMC simulations

Abstract

Numerical simulation of rarefied plasmas often makes use of Particle-in-Cell-Direct Simulation Monte Carlo (PIC-DSMC) codes [Birdsall, C. K., Particle-in-cell charged-particle simulations, plus Monte Carlo collisions with neutral atoms, PIC-MCC. Trans. Plasma Sci., 19(2), 1991], which simulate the motion of a large number of computational particles, each of which represents a large number of actual molecules, atoms, or electrons. Particle-particle and particle-surface collisions in such codes are performed in a stochastic manner: a single process is simulated for each collision, with the process being randomly chosen based on the probability of it taking place. In the present work a new collision and boundary condition simulation scheme, dubbed "event splitting" [Oblapenko, G. et al, Hedging direct simulation Monte Carlo bets via event splitting, J. Comput. Phys., 466, 2022], is investigated. It simulates multiple events simultaneously via splitting of the particles involved in the collision and has been shown to significantly reduce the level of stochastic noise in unsteady breakdown simulations. The scheme is applied to 1-D modelling of discharges in rarefied argon and helium plasmas, and compared to standard PIC-DSMC collision routines in terms of computational efficiency and numerical noise.