Motivated by recent advances in X-Ray Computed Tomography (XRCT) for analyzing phenomenon emanating from the grain-scale, we developed two grain-scale discrete simulation methods for modeling and characterization of real granular materials: the Granular Element Method (GEM) and coupled Level Set-Discrete Element Method (LS-DEM). The technologies underlying these methods are Non-Uniform Rational Basis-Splines (NURBS) and level set methods, which enable the representation of particle morphological features, namely, sphericity and angularity, to their fullest extent. For the first time, emergent properties such as shear strength can be captured entirely through particle geometry, without resorting to further artificial techniques for treating rolling resistance, and thus minimizing parameter calibration of classic DEM models. Through coarse-graining and homogenization techniques, these methods would allow for the extraction of plastic internal variables such as dilatancy and friction directly from the granular microstructure. Independent of image data, these tools can also be for used studying material behavior in what-if scenarios under loading conditions beyond the confines of experimental situations. We demonstrate the capabilities of GEM using realistic grain geometries, followed by a simple numerical example for the LS-DEM.
展开▼
