Investigating the Effect of Bidispersity and Number of Particles on Stress

One limitation of Computational Fluid Dynamics based investigations of industrial scale particle flow phenomena is the inability to properly describe the effects of particle-particle interactions. Discrete element methods (DEM) present an opportunity to study such interactions. In flows with higher solids loadings these particle-particle interactions give rise to microstructure formation. Significant progress has been made in the investigation of the nature, size, and composition of the observed microstructure. Recently DEM simulations have been used to investigate bidisperse systems with small numbers of particles. In this work we simulate bidisperse systems containing much larger system sizes with a computationally efficient DEM simulation. Specifically, we investigate the dependence of stress on the number of particles N with a bimodal distribution in a 3-D shear flow simulation. For the case of a coefficient of restitution of 0.6 and a solids volume fraction of 0.3, the ratio of the particle diameters, R, is varied from 1.0 to 3.0. As has been previously reported, there is an observed upward drift in stress with increasing numbers of particles when R equals 1.0. The upward drift occurs after the stresses appear to reach asymptotic behavior for small N. As R is increased, a similar upward drift in the stress is observed, however the effect is less pronounced because the overall magnitude of the stress decreases as R is increased. Previously reported clustering phenomena, observed in monodisperse systems, are also present in bidisperse systems. Also the effect of an increase in R on the initial onset of the increase in stress is investigated. Furthermore, the case of coefficient of restitution of 0.8 and solids volume fraction of 0.3 is studied in order to examine the effect of varying the coefficient of restitution. Finally, graphical visualizations of each simulation are used to illustrate the change in the system as N becomes large and to investigate number and size segregation effects as R increases.