A deterministic Gaussian-Mixtures Coulomb-collision algorithm for particle-in-cell

Coulomb-collision modules in PIC simulations are typically Monte-Carlo-based. Monte Carlo (MC) is attractive for its simplicity, efficiency in high dimensions, and conservation properties. However, it is noisy, of low temporal order (typically O(√△t), and has to resolve the collision frequency for accuracy.1 In this study, we explore a machine-learning- based, multiscale alternative to MC for PIC. The approach is based on the reconstruction of the particles’ velocity distribution function (VDF) using a Gaussian Mixtures Model (GMM) via the Maximum Likelihood Estimation principle.2,3 A key element of our algorithm is to decompose each Gaussian in the GMM into a convex linear combination of isotropic Maxwellians for which an exact set of evolution equations can be de- rived according to the Landau-Fokker-Planck collision operator.4 The proposed method is deterministic, free of instability, positivity-preserving, and strictly conservative, and is orders of magnitude faster than either MC or Eulerian Fokker-Planck solvers. We will illustrate the accuracy and performance of the proposed method with several examples of varying complexity.

1 Dimits, et. al., JCP, 228 p.4881 (2009)

2 Redner, Walker, SIAM Review, 26, (1984)

3 Chen, Chacon, Nguyen, JCP, 436 (2021)

4 Echim, Lemaire, Svendsen, Surv Geophys, 32, p. 1–70 (2011)