Real-time slope stability analysis of gentle expansive soil slope under rainfall infiltration condition

For gentle expansive soil slopes, rainfall-induced reduction in matric suction simultaneously mobilizes lateral swelling pressure (LSP) and diminishes shear strength, posing considerable threats to stability. Accurate prediction of failure timing remains challenging due to limitations in real-time suction monitoring, the lack of a robust LSP model for transitional suction states, and the absence of a shear strength model valid across the entire suction range (0–10⁶ kPa). To address these gaps, this study introduces a new LSP prediction model based on the matric suction equivalent concept and enhances the traditional Janbu method through integration with the Capillary-Adsorption Decoupled (CAD) shear strength model. This combined approach enables precise characterization of strength loss throughout the suction spectrum during rainfall. A comprehensive computational framework incorporating the improved Janbu method, CAD model, and LSP prediction was developed in MATLAB, combining GeoStudio-derived suction profiles with real-time tracking of critical slip surfaces and real-time stability assessment. Validation against published case studies confirms the model’s accuracy in predicting LSP evolution and slope stability, providing spatiotemporally explicit references for early warning systems in expansive soil regions.