Hybrid Grains: Adaptive Coupling of Discrete and Continuum Simulations of Granular Media

摘要

We propose a technique to simulate granular materials that exploits the dualstrengths of discrete and continuum treatments. Discrete element simulationsprovide unmatched levels of detail and generality, but prove excessivelycostly when applied to large scale systems. Continuum approaches are computationallytractable, but limited in applicability due to built-in modelingassumptions; e.g., models suitable for granular flows typically fail to captureclogging, bouncing and ballistic motion. In our hybrid approach, anoracle dynamically partitions the domain into continuum regions wheresafe, and discrete regions where necessary. The domains overlap alongtransition zones, where a Lagrangian dynamics mass-splitting coupling principleenforces agreement between the two simulation states. Enrichmentand homogenization operations allow the partitions to evolve over time.This approach accurately and efficiently simulates scenarios that previouslyrequired an entirely discrete treatment.