Anchorage support is common in red-bed tunnels, and water–rock interactions at the grout–rock interface with seepage often lead to debonding failure. Meanwhile, the mechanical performance of grout–rock interface is closely related to the water content of the rock. This study examines the effect of water content on the shear behavior of grout–rock interface via slant shear tests, nanoindentation tests, and scanning electron microscopy. The macroscopic and microscopic mechanical characteristics of interfaces between grout and red-bed mudstone and sandstone are systematically studied under dry, natural, and saturated conditions. The results show that (1) mudstone exhibits a high-porosity “loosening zone” with numerous fractures near the grout–rock interface, reducing its bonding strength, but grout infiltration along these fractures enhances the frictional properties. Mudstone is more susceptible to degradation under dry and saturated conditions than in the natural state, with lower cohesion and a higher internal friction angle at the grout–rock interface due to greater grout infiltration. (2) The grout–sandstone interface is not tightly bonded, with obvious microscopic fractures, which the detached quartz particles can fill. Sandstone’s higher porosity allows shrinking under dry conditions and water filling under saturated conditions, resulting in higher cohesion but a lower internal friction angle at the grout–rock interface due to less grout infiltration. (3) The microscopic mechanical parameters of the interfacial transition zone of mudstone and sandstone exhibit monotonical increases and U-shaped variations, respectively, from the rock to the hardened grout. Compared to the natural state, the microscopic mechanical parameters of the interfacial transition zone in mudstone decrease following drying and saturation, while those of the interfacial transition zone in sandstone show an opposite trend. The microscopic hardness and cohesion of the grout–rock interface are positively correlated, and the cohesion can be predicted using a fitting equation based on the microscopic hardness measurements.
Shu et al. (Tue,) studied this question.