This work deals with the development of an intelligent inter-particle collision scheme using the Direct Simulation of Monte Carlo (DSMC) method. Conventional DSMC collision schemes cannot perceive collision distances by their own routines and, consequently, must rely on other techniques to provide a choice of collision pairs with a smaller inter-particle space. Here, we propose a modification of the Simplified Bernoulli Trials (SBT) scheme, called the Intelligent Simplified Bernoulli Trials (ISBT) scheme, which can create pseudo-circular subcells that reduce approximately 25-32% of the mean collision separation distance (MCS). The ISBT scheme tries to arrange the particle indexing and collision acceptance-rejection of the SBT scheme in a way that leads to the formation of virtual clusters. These inner-cell clusters then will cause the selection of the ‘near-neighbor’ pair, which leads to smaller mean collision separation distances. Different low- and high-speed test cases, e.g., lid-driven cavity flow, steady hypersonic flow over a two-dimensional cylinder, and Mach 15.6 nitrogen flow over a 25◦–55◦ axisymmetric biconic, are selected to assess the accuracy and efficiency of the ISBT collision scheme.

Keywords