Effect of the random distribution of gravel particles on the mechanical behaviors of conglomerate

Oil and gas resources covered under gravel layer have had much attention. The mechanical behaviors and failure mechanism of conglomerate, main component of gravel layer, are still unclear for the complex distribution characteristics in gravel size, shape, and content. A numerical model considering gravel particles distributed randomly was established to study the effect of gravel distribution characteristics using a discrete element method. Results indicate that increasing the gravel size or content, or overlapping of the particles can reduce the number of independent gravel blocks, thus enhancing the heterogeneity of conglomerate and changing the failure types and crack development trend. The failure strength shows a parabolic trend with rising gravel content, while the elastic modulus increases linearly, causing the ability of specimen to resist deformation to improve. The mutual extrusion brings about the increase of failure strength rapidly and the reversal of peak strain and Poisson’s ratio when gravel content is over 70%. The cohesion and internal friction angle show a parabolic tendency, based on which the strength prediction model of conglomerate considering gravel content was established with high precision. This work provides a kind of numerical method for studying the physical and mechanical characteristics of the strongly heterogeneous rock considering the components.