Preventing overpacking of compressed moderately dense granular gases

Extending the direct simulation Monte Carlo (DSMC) method to moderately dense granular flows has demanded recovering collisional pressures. There have been several reported DSMC variants that successfully recover the non-ideal equation of state but have largely been validated for systems with a homogeneous solids fraction field. Initially homogeneous systems that develop strong density gradients, such as the case of granular clustering, are seldom investigated due to the numerical phenomenon of particle packing past physical limits, or overpacking. In the presence of external gradients or internal clustering mechanisms, particles will compact into small volumes. We believe overpacking is a product of the unconditional acceptance of particle positions, and a proposed direct solution is modulating the flow of particles via a random walk. Termed as Energy DSMC (EDSMC), we show that integration of a modified Metropolis algorithm prevents overpacking to the Enskog Simulation Monte Carlo a la Bird.