Numerical study of the geometrical and hydraulic characteristics of 3D self-affine rough fractures during shear

This study presents a predictive model to estimate the permeability of 3D self-affine rough fractures during shear. Numerical simulations of shear-flow tests under different normal stresses were performed on a series of 3D rough fractures generated using the modified successive random additions (SRA) algorithm, and the evolutions of aperture and permeability were investigated. The results show that joint roughness coefficient (JRC) distribution of fracture profiles follows a Gauss function, where JRC is an extensively accepted parameter for characterizing fracture surface roughness in rock mechanics and rock engineering. As the shear displacement increases, aperture distribution becomes more anisotropic and more complex channels of fluid flow are formed. Permeability decreases with the increase in normal stress. For a smaller normal stress, the change of permeability is more significant because of the larger dilation inhibited by a larger normal stress. Due to the significant flow channels induced by shearing, the permeability in the direction perpendicular to the shear direction is approximately 1.12-10.98 times as large as that in the direction parallel to the shear direction. The validity of the proposed functions for permeability prediction is verified by comparisons with reported results in previous works. In practice, the permeability can be well calculated as a first-order estimation using the proposed model when the parameters such as JRC, shear displacement and normal stress are given.