Meso-scale framework for modeling granular material using computed tomography

Numerical modeling of unconsolidated granular materials is comprised of multiple nonlinear phenomena. Accurately capturing these phenomena, including intergranular forces and grain deformation, depends on resolving contact regions several orders of magnitude smaller than the grain size. Here, we investigate a method for capturing the morphology of the individual particles using computed X-ray tomography, which allows for accurate characterization of the interaction between grains. Additionally, the ability of these numerical approaches to determine stress concentrations at grain contacts is important in order to capture catastrophic splitting of individual grains, which has been shown to play a key role in the plastic behavior of the granular material on the continuum level. Samples of Ottawa sand are numerically modeled under one-dimensional compression loadings in order to determine the effect of discretization approaches, such as mesh refinement, on the resulting stress concentrations at contact points between grains. The effects of grain coordination number and finite element type selection on these stress concentrations are also investigated.