ABSTRACT

Discrete particle modeling is applied to build a numerical laboratory to study granular materials. It is expected that with basic information on the rock from conventional experiments, the numerical laboratory completes the data set and may establish how the mechanical and petrophysical data will vary given variations in e.g. porosity, mineralogy and texture. In this paper we present implementation of several new features into a discrete particle model (PFC) in order to establish such a numerical laboratory. These features are: (i) incorporation of complex particle shapes using clumps or clusters, (ii) coupling between grain skeleton deformation and pore fluid flow and (iii) implementation of elastic wave propagation. As application examples of this work, we present the results of biaxial tests with the numerical samples made of super particles. The formation of shear bands is observed at low confining pressures, and compaction bands are observed at high levels of confinement. We show the simulations of permeability alterations due to rock deformation and micro damage. We also look at stress dependence of wave velocities in uncemented and cemented granular media, including development of stress-induced elastic anisotropy.