(486b) Predicting Segregation in Flowing Granular Materials

As granular materials flow they can segregate due to particle size or density differences. The degree of segregation depends on the flow conditions and the differences between the particle species. Decades ago, Dr. Jerry Johanson was a pioneer in directly addressing practical segregation problems, particularly in hoppers and bins. However, only more recently has a theoretical approach been suggested to predict segregation based on a continuum advection-diffusion-segregation equation for the concentration of each particle species. The model requires knowledge of the flow kinematics, which can come from experiments, simple models, or DEM simulations. The challenge is in defining the appropriate form of the segregation term in the equation. Several approaches have been proposed based on various models for particle percolation. One of the most successful is a relation for the segregation velocity based on the local shear rate, particle concentrations, and a semi-empirical parameter obtained from discrete element method (DEM) simulations for bi-disperse systems of particles. This approach can be applied to many flow geometries, particles of different sizes, densities, or shapes, and bidisperse, multi-disperse, or poly-disperse particle systems, provided that the particle segregation properties can be characterized. Recent research suggests progress in extracting segregation parameters from simple measurements so that even complex segregating materials can be modeled. The challenge lies in connecting these approaches to practical flow geometries such as bins and hoppers to provide a fundamental basis for methodologies that can be adapted to systems that are pertinent to industrial particle handling and processing. Partially funded by NSF Grant No. CBET-1511450 and The Procter & Gamble Company.