How force chains in a steady surface granular flow may trigger a non-Bagnold behavior

We employ two-dimensional discrete element simulations to study how force chains and their network evolve in steady inclined surface granular flows to trigger the development of a non-Bagnold flow velocity profile. Circular disks with nonuniform sizes (10% variation) were used with a periodic boundary condition and a bumpy and hence no-slip base. Creeping to almost static flow regime were created by altering the flow height and inclination angles. Force network is extracted by the interaction forces between individual disks but various algorithms from the network science are modified and implemented to extract the needed force chain information. For example, the force community is identified by the highly correlated stress disks from the force network. Subsequently, unique force chains are identified by the network modularity to evaluate their size distribution and morphology and track their lifespans. Concurrently, we explore other force chain properties, including energy dissipation, degree of buckling, local dilation, and the non-affine deformation to understand the discovered force chain characteristics and the associated grain dynamics. In addition, we borrow the concept of information cycle to identify force cycles and explain the failure, reconstruction, and evolution progress of force chains and how they may trigger the development of a non-Bagnold flow.