面向跨流域模拟的统一随机粒子方法和SPARTACUS 软件

In the study of gas flow across different regimes, traditional stochastic particle methods, such as the Direct Simulation Monte Carlo (DSMC) method, face limitations in efficiently simulating due to constraints imposed by the molecular mean free path and mean collision time on the computational spatiotemporal scales. The Unified Stochastic Particle (USP) method, a multiscale particle approach proposed in recent years, effectively overcomes these limitations inherent in the DSMC method by coupling the motion and collisions of molecules. The fundamental principles, algorithm implementation, and an open-source solver named SPARTACUS (SPARTA Combined with USP), developed by the authors′ team are presented. Derived from the open-source framework SPARTA (Stochastic PArallel Rarefied-gas Time-accurate Analyzer), SPARTACUS exhibits excellent versatility, capable of simulating internal and external flows around geometries of arbitrary dimensions. Additionally, the solver employs a parallel strategy based on the Message Passing Interface (MPI), showcasing potential scalability to support parallelism on the order of millions of cores. Several benchmark tests are provided, covering diverse shapes and spanning from low-speed to hypersonic conditions, with comparisons made against reference results from DSMC or experiments. The results demonstrate that SPARTACUS accurately simulates flows across different regimes, showcasing notable efficiency advantages compared to traditional particle methods.