(723c) Segregation of Spherical and Non-Spherical Particles in DEM Simulations

 Most studies of segregation in flowing granular materials use spherical particles to evaluate the effects of particle size and density ratios, as well as particle surface roughness. However, in practical applications particles are generally non-spherical. Here we describe a discrete element method (DEM) simulation approach for spherical and non-spherical particles which are described mathematically as super-ellipsoids and, for contact resolution purposes, treated as potential particles in that the function that describes the particle also is used to determine if a point is on the interior, exterior, or surface of the particle. The contact resolution algorithm for non-spherical particles is more complex than for spherical particles. Screening methods limit the computationally expensive contact resolution algorithm to contacting particles, and the simulation is validated through comparison with previous results for “true geometry” models of cylindrical particles. We simulate bidisperse mixtures of non-spherical particles in quasi-2D bounded heaps and measure the segregation velocity as a function of local shear rate, species concentration, and shape difference. Shape-driven segregation velocities are compared to size-driven segregation velocities, which may eventually allow for shape differences for non-spherical particle mixtures to be mapped to equivalent size differences in spherical particle mixtures.

Partially funded by the Dow Chemical Company.