2019 CSCE Annual Conference - Laval (Greater Montreal) Conference
Mr. Siyang Zhang, Concordia University
Many brittle materials, such as concrete, rock or other granular materials share a common characteristic; their intricate particle shapes. Due to their geometry, a high complexity of interaction arises between contacting particles, especially when they fragment from their original shape into smaller ones. Traditionally, it was common to use continuum methods, such as the finite element method (FEM), to reproduce the behaviour of these materials, even though these methods require complex constitutive models. In contrast to continuum methods, the Discrete Element Method (DEM), gained wide acceptance to model interacting solid bodies representing the behaviour of granular (spherical) and polyhedral (non-spherical) particle systems, with focus on micromechanics of soil particle interactions and displacements. One of the main shortcomings of DEM is that no information about a stress/displacement field within a particle is generated during a simulation. Uniting DEM and FEM, the Combined Finite-Discrete Element Method (FEM/DEM) presents a more comprehensive approach considering the material as both continua and discontinua. Since the number of particles used in any simulation is severely limited by the computing power available, it is customary to use simplified geometries representing a particle. The simplest being a sphere, and any particle is seldom represented by more than one or two hundred polygons. However, irrespective of the method used, the particle geometry has a potential to greatly influence the outcome of any simulation. The effect of surface detail at various lengths of scale can affect the location of particle contact points, which in turn govern the forces applied. This paper will summarize the findings of an investigation of the influence of changes in particle geometry (asperities) of polyhedral granular particles as simulated by the FEM/DEM. Both solution time and solution accuracy were considered, and the results will be critically reviewed and recommendations will be given for practical use in simulations.