DEM Simulations of Particle Clustering at High Solids Concentrations (Keynote)

One limitation of two-fluid computational fluid dynamics models for gas-particle flow is the inability of these models to properly describe the effects of particle-particle interactions in the case of large St at high solids concentrations. Discrete element methods (DEM) present an opportunity to study such interactions. In gas-particle flows with higher solids loadings, particle-particle interactions give rise to particle clustering and/or microstructure formation. In this paper we show how DEM can be used to determine not only the particle-phase stress (required for gas-particle two-fluid CFD models) in the case of particle clustering, but also how DEM can be used to determine the nature, size and composition of the observed microstructure. Specifically, we investigate, using a computationally efficient DEM simulation, the dependence of particle-phase stress in systems with high solids concentration on the solids coefficient of restitution (0.6 to 0.9) and solids volume fraction (0.05 to 0.3). We investigate both monodisperse and bidisperse particle systems. We show how a widening of the particle size distribution reduces the particle-phase stress and inhibits the formation of particle clusters. We also show in our bidisperse simulations how smaller particles (yet still inertia-dominated) are more evenly distributed throughout the system. Finally, we highlight the dependence of the particle-phase stress on the domain size of the simulation and emphasize that a sufficiently large domain size is necessary to accurately describe the clusters which form at higher particle concentrations.