A second-order particle Fokker-Planck-Master method for diatomic gas flows

TL;DR

This paper introduces a second-order accurate stochastic particle Fokker-Planck-Master method for diatomic gas flows, significantly improving computational efficiency and accuracy over traditional DSMC methods, especially in hypersonic flow simulations.

Contribution

It develops a unified stochastic particle Fokker-Planck-Master method with second-order accuracy for diatomic gases, enhancing efficiency and physical fidelity in rarefied gas flow simulations.

Findings

Achieves second-order accuracy in time and space for diatomic gas modeling.

Provides accurate solutions with coarser grids and larger time steps.

Achieves a 28-fold speed-up in hypersonic flow simulations around a cylinder.

Abstract

The direct simulation Monte Carlo (DSMC) method is widely used to describe rarefied gas flows. The DSMC method accounts for the transport and collisions of computational particles, resulting in higher computational costs in the continuum regime. The Fokker-Planck (FP) model approximates particle collisions as Brownian motion to reduce computational cost. Advanced FP models have been developed to enhance physical fidelity, ensuring the correct Prandtl number and the H-theorem. The FP model has further been extended to handle diatomic gases, such as the Fokker-Planck-Master (FPM) model. Alongside these developments in modeling, computational efficiency has also been improved by achieving second-order spatial and temporal accuracy, as demonstrated in the unified stochastic particle FP (USP-FP) method. However, these accuracy improvements have not yet been extended to diatomic gases, which are essential for engineering applications such as atmospheric reentry. This study proposes a unified stochastic particle Fokker-Planck-Master (USP-FPM) method for diatomic gases that achieves second-order accuracy in both time and space. Temporal accuracy is enhanced by reproducing second-order energy, viscous stress, and heat flux relaxations. Spatial accuracy is improved by employing a first-order polynomial reconstruction method. Three test cases are investigated: homogeneous relaxation, Poiseuille flow, and hypersonic flow around a cylinder. The results show that the USP-FPM method provides accurate solutions even with coarser cell sizes and larger time steps compared to the DSMC and FPM methods. In particular, for the hypersonic flow around a cylinder, the USP-FPM method achieves a speed-up factor of 28 compared to the DSMC method, while maintaining accuracy.