3D DEM simulations of the variability of rock mechanical behaviour based on random rock microcracks

Abstract

In this study, to investigate the variability of mechanical parameters of rock, a three-dimensional (3D) discrete element model (DEM) with the consideration of random microcracks was proposed. The random microcracks follow the negative power-law distribution with upper and lower limits of microcrack size. The results showed that coefficient of variations (COVs) and average values of elastic modulus, Poisson's ratio, unconfined compression strength and tensile strength of Lac du Bonnet granite are well matched through the present model. The present simulations indicate that the large COVs were induced by the interconnected large microcracks. The parametric study clearly indicates that the random microcracks significantly affect COVs, while the effect of grain size variation can be ignored. The size of microcracks determines the variability of mechanical parameters, and the volume density of microcracks can significantly influence the COVs. The correlation between the fractal dimension of microcrack size distribution and COVs is not linear. The increasing variability of microcrack density can result in the increase of COVs. Finally, a calibration procedure was proposed to determine the present model's parameters.