Sheared Granular Material: Single and Multi-Particle Experiments and Physics-Based Cellular Automata Modeling

Granular flows continue to be a complex problem in nature and industry where solid particles exhibit solid, liquid, and gaseous behavior in a manner which is often difficult to predict both locally and globally. In solids processing applications such as pharmaceutical production, food processing, and coal processing, the ability to understand and accurately predict granular flows under shear can improve process efficiency and product effectiveness. This ability requires both the execution of detailed granular flow experiments and the development of predictive models. Equally important is the study of interactions at the single particle level, as the aggregation of these interactions defines global flow behavior. Ultimately, in terms of modeling, the ability to accurately simulate flows in a fast and efficient manner is necessary in these industrial settings where large particle counts are present and accurate design decisions are required. Towards this end, this work develops experimentally-validated modeling frameworks for the prediction of granular shear flows through a three-phase approach consisting of: (1) single particle interaction property investigations, (2) multi-particle granular flow experiments, and (3) physics-based cellular automata modeling. In phase 1, detailed experiments are performed on the coefficient of restitution and coefficient of friction for single particles interacting with flat substrates. Multi-particle shear flow experiments in phase 2 examine the interaction between rough surfaces and granular flows when grain materials, and hence interaction properties, are varied. Consequently, the single particle investigations are used to interpret these results. In terms of modeling, lattice-based cellular automata (CA) presents itself as a possible high-speed supplement to the physically rigorous but computationally demanding discrete element method (DEM), which is the "gold standard" in granular flow modeling. As such, three CA frameworks are developed in phase 3, with phase 2 experiments used for validation. These include a two-dimensional model which considers friction and particle spin during collision processing, a three-dimensional model which provides the ability to handle flows beyond solely the kinetic regime, and a multiphase model which combines computational fluid dynamics with CA to model flows comprised of particles immersed in a fluid.