(617g) Validation of the Direct Simulation Monte Carlo (DSMC) Method for Simulating Polydisperse Gas-Solid Flows

Flowing granular materials are ubiquitous in industry and nature, e.g. coffee, snow, coal, and Saturn’s rings, and exhibit complex phenomena not often observed in traditional gaseous or liquid fluids. A better understanding of granular flows can advance industries that use solid particles by providing insight towards the underlying mechanisms that give rise to different flow behaviors and ultimately the development of predictive modeling tools. Small scale experiments can probe the underlying physics of gas-solid systems, but extrapolating small-scale tests to large-scale industrial systems is unreliable. Performing experiments in industrial scale systems is costly and detailed non-intrusive measurements are difficult. Discrete element modeling (DEM) and other deterministic computational models have been used to simulate granular flows, but such methods are often restricted to relatively small-scale systems (~10⁷ particles). Continuum modeling approaches can simulate large-scale systems, but there can be challenges with accuracy and inclusion of particle-scale physics. This work assesses the feasibility of the direct simulation Monte Carlo method (DSMC), as an alternative modeling approach that is computationally efficient and accurate compared to discrete element simulations. The DSMC method uses representative particles and models particle collisions in a stochastic manner. The DSMC method has traditionally been used to simulate dilute gases but has been adapted to solve the granular kinetic theory equations (i.e., the revised Enskog equation). We have extended the direct simulation Monte Carlo method to simulate dense to dilute polydisperse gas-solid flows. To test the validity of the DSMC modeling approach for polydisperse flows, we evaluate its accuracy for relatively small scale systems via comparison to high fidelity discrete element simulations. The computational performance of DSMC is compared to discrete element simulations and we assess the suitability of DSMC for use in large-scale gas-solid systems.