Analysis of RVE size based on three-dimensional fracture numerical network modelling and stochastic mathematics

Abstract

A new method to study representative volume element (RVE) size is developed based on three-dimensional (3D) fracture numerical network modelling and a stochastic mathematical model. A 3D fracture numerical network is established, and the network factors that affect RVE size, such as fracture density, diameter, strike, and dip angle, are collected. This paper utilizes these four factors to obtain the RVE size based on the contingency table method. It is concluded that the side length of the RVE is 2–8 times larger than the average trace length. Fracture frequency is chosen to validate the RVE size through 3D fracture numerical network modelling. RVE size calculated through fracture frequency is consistent with the contingency table method. It is proven that the method for calculating RVE size through the contingency table method is feasible and credible.