Interpretation of the reactivation of slow-moving landslides based on ring shear tests and monitoring

Many landslides have been reactivated along the banks of the Three Gorges Reservoir (TGR) in China since 2003, many of which were slow-moving landslides. Normally, these landslides do not occur suddenly, but accelerations during short periods may occur, and they can still cause damage to buildings. The initiation of slow-moving landslides depends not only on the hydraulic characteristics of the sliding body, but also on the mechanical properties of the sliding zone soil. In this study, the Sifangbei landslide in the TGR was selected as a case study to analyse the residual strength of the sliding zone and the long-term monitoring data. The analysis of the long-term monitoring data indicated that the periodic landslide deformation reactivation was affected by seasonal rainfall and annual reservoir water-level fluctuations. Soil samples along the sliding zone collected from the front and middle of the landslide were tested using a ring shear test to study the influence of the water content and shear rate on the residual strength values. The results showed that an increase in water content can weaken the shear strength of sandy clay and clay. The sandy clay with fewer montmorillonite minerals and more sand particles had higher mechanical properties than the clay with more montmorillonite minerals and fewer sand particles. The damaging effect of the increased water content on the sandy clay was mostly reflected in the residual friction angle, while the damaging effect of the increased water content on the clay was mostly reflected in the residual cohesion. An increase in the shear rate had a positive effect on the shear strength of the sandy clay and clay. For the sandy clay with high particle friction, the shear mode changed from turbulent flow to slippage as large particles on the sliding surface broke into smaller particles when the shear rate reached a high speed (v ≥ 0.5 mm min −1). The shear strength under different scenarios revealed the mechanism of slow-moving landslide reactivation.