Including Friction and Spin Rules in the Lattice-Based Cellular Automata Approach to Model Granular Flows

Granular flows have been proposed as an alternative lubrication mechanism to conventional liquid lubricants in sliding contacts due to their ability to carry loads and accommodate surface velocities. Their load carrying capacity has been demonstrated in the experiments of Yu and Tichy [1]. Alternate lubrication techniques are becoming necessary due to the failure of conventional liquid lubricants in extreme temperature environments, and their promotion of stiction in micro-/nanoscale environments. Yet, understanding granular behavior has been difficult due to its non-linear and multiphase behavior. Cellular Automata (CA) has been shown to be a viable first order approach to modeling some complex aspects of granular flow. Previous work by the authors successfully modeled granular shear with a CA model [2]. Additional work combined CA computational efficiency with particle dynamics to effectively model collision events. This work builds upon and modifies the prior CA modeling approaches by adding friction modeling and spin of particles. This modification maintains the computational efficiency of CA, while increasing accuracy of the predicted granular flow properties, such as speed, solid fraction, and granular temperature. The current work compares the CA model with friction and spin physics relations to the authors’ prior CA model which neglected friction. Both CA models are also evaluated against experimental data to quantify the benefits of including friction and spin in the CA modeling approach for granular flows.