Physics-based time-of-failure determination of rainfall-induced instability in lateritic soil slopes

Conventional time-of-failure estimated from slope surface displacement over time, ignores the crucial geotechnical and environmental causative factors that lead to slope instability. The instrumentation and monitoring are expensive, labour-intensive, and often not feasible for large number of hill slopes. This paper focuses on the physics-based determination of time-of-failure charts for laterite soil slopes prevalent in the Western Ghats of India, under rainfall infiltration. The finite element model was first validated by performing coupled flow deformation analysis of Kondavi soil cutting situated in the Ratnagiri district of Maharashtra, India. The soil samples were collected from the site for basic geotechnical characterisation in the laboratory. In addition, the soil water characteristics curve (SWCC) was determined using the filter paper method, and the unsaturated parameters were obtained using the van Genuchten model. Following the validation of numerical model with the failed Kondavi cutting, the factor of safety (FOS) and time-of-failure (TOF) were studied for varying rainfall intensity, and permeability of the soil. The FOS and TOF charts were then established for varying slope angles (30°, 45°, 60°), effective cohesion (10 kPa, 18 kPa, 25 kPa), effective friction angle (22°, 25°, 28°), water table depth (25 m, 28 m, 30 m, 35 m) and height of slope (15 m, 25 m, 35 m). Results indicate that if rainfall intensity is lower than soil permeability, TOF depends on both the intensity and duration of the rainfall. Higher rainfall intensity leads to a shorter time of failure, and vice versa. Conversely, when rainfall intensity exceeds soil permeability, TOF is determined by the duration of rainfall and the permeability of soil, as the rainfall infiltrates at the saturated permeability rate regardless of its intensity. It is also observed that friction is the dominant parameter for initial FOS, while cohesion is the dominant parameter for the TOF of a rainfall-induced landslide. Finally, charts are proposed that shall serve as a preliminary guide for the determination of the TOF for rainfall-induced instability in the lateritic soil slopes of India. The performance of the charts is further evaluated by comparing the observed and predicted TOF of two other failed laterite cuttings. The implications of these findings are profound, as the proposed TOF charts can be integrated into early warning systems, contributing towards improved disaster mitigation and preparedness with timely decision making for adequate management of landslide associated risks.