An image segmentation-based method for quantifying the rock failure mechanism under true triaxial compression

Understanding the impact of geostress on the failure mechanism of surrounding rock is necessary for underground engineering applications. Here, we carried out a series of true triaxial tests on sandstone. The mechanisms of the areas on the fracture surface that appear in different colors are identified. Using the image segmentation method, the white shear fracture area on the grayscale image of the fracture surface is extracted. By calculating the proportion of the white shear fracture area, the rock failure parameter S for quantifying the rock failure mechanism is defined, and the influences of σ2 and σ3 on rock failure mechanisms are analyzed. Moreover, a criterion for determining the failure mechanism of rock under true triaxial stress is established. The results indicate that the shear fracture area that appears white is due to the adherence of rock debris and powder on the surface. The rock failure parameter S is appropriate to quantify the rock failure mechanism. A larger S indicates a higher proportion of shear fracture in the rock failure. The increase in σ3 has a significant promotional effect on the development of the shear fracture. An increase in σ2 can effectively promote the initiation of tensile fracture. According to the criterion, once the σ2 and σ3 of the stress environment are clarified, the failure mechanism of the rock can be predicted. This work contributes to an improved understanding of the failure mechanism of rock and a potential means by which to guide the design and construction of underground engineering.