Numerical studies on stratified rock failure based on digital image processing technique at mesoscale

This paper investigates the failure behaviors of stratified rocks under uniaxial compression using a digital image processing (DIP) based finite difference method (FDM). The two-dimensional (2D) mesostructure of stratified rocks, represented as the internal spatial distribution of two main rock materials (marble and greenschist), is first identified with the DIP technique. And then the binaryzation image information is used to generate the finite difference grid. Finally, the failure behaviors of stratified rock samples are simulated by FDM considering the inho-mogeneity of rock materials. In the DIP, an image segmentation algorithm based on seeded region growing (SRG) is proposed, instead of the traditional threshold value method. And with the new proposed image segmentation algorithm, the mesostruc-ture of stratified rocks can be fully acquired. In the process of simulation, to sufficiently capture the inhomogeneity of stratified rocks, mechanical properties of each rock material are all characterized by the Weibull statistical manner. Several cases of stratified rocks with different homogeneity indices of rock materials have been discussed under two compression loading conditions considering the anisotropy of stratified rocks to some extent. Results from numerical simulations show that the inhomogeneity of stratified rocks arising from the presence of different rock materials has a great influence on the failure behaviors of stratified rocks and that the numerical failure behaviors of stratified rocks without confining stresses applied agree quite well with the general observations reported in the literature. Copyright © 2015 Tech Science Press.