Elliott Barnett | AIChE

Elliott Barnett

In dense granular flows of particles differing in size, small particles fall between large particles to segregate in lower portions of the flow thereby displacing large particles upward, a process known as “percolation.” Similarly, light particles rise above heavy particles as they flow due to “buoyancy.” The resulting segregation due to percolation or buoyancy depends on the differences between the particles and the flow conditions. We address granular segregation in two ways. The first approach is a continuum segregation model based on the advection-diffusion equation with a term added to account for particle segregation. This model can accurately predict mixing and segregation in both steady and transient flows for a variety of flow geometries and for a range of particle systems including multiple individual particle sizes, polydisperse (continuous) particle size distributions, mixtures of particles varying in both size and density, and non-spherical particles. The model can even be used to “design” combinations of particle size, density, and concentration that result in non-segregating bidisperse particle mixtures. The second approach is to characterize the particle level segregation force via discrete element method simulations, which provides the ability to predict whether an intruder particle will rise or sink. These single intruder results can be extended to cooperative segregation phenomena in particle mixtures. Our current challenge is to connect segregation parameters in the continuum segregation model to particle level forces as well as to extend the segregation model to particle systems with a wide range of particle sizes. Funded by The Dow Chemical Company, the Procter & Gamble Company, and the National Science Foundation (CMMI-1435065, CBET-1511450, CBET-1929265).