Comparative Investigation on the Compression–Shear and Tension–Shear Behaviour of Sandstone at Different Shearing Rates

Abstract

Direct shear tests on rocks under compression–shear stress conditions have been widely conducted, whereas few have been performed under tension–shear stress conditions. However, rocks exhibit tension–shear failures in many scenarios, such as in the excavation disturbed zone in deep underground caverns and high slopes. A series of direct shear tests were performed with cuboid sandstone specimens under different normal tensile stresses (σn = − 3, − 2, and − 1 MPa) and compressive stresses (σn = 1, 3, and 5 MPa) at different shearing rates (v = 0.2, 1, 5, and 10 mm/min). The tension–shear tests were performed using an auxiliary device in combination with a compression–shear testing machine. The results showed that the fracture, shear stress–displacement curve, shear stiffness and shear strength were affected by both σn and v, and the differences in these mechanical behaviour between compression–shear and tension–shear cases were analysed in detail. The shear strength had a nonlinear relationship with both σn and v in the full region of tested normal stress (namely, the normal stress range from tension to low compression). The Hoek–Brown criterion $$(\tau = A(\sigma_{{\text{n}}} - \sigma_{{\text{t}}} )^{B} )$$ considering the shearing rate effect (the relationship between parameter A (B) and shearing rate v is represented by a natural logarithm function) was proposed as the optimal shearing rate-dependent strength criterion for sandstone in the tested normal stress range.