SCANNING CURVE OF SOIL–WATER CHARACTERISTIC CURVE UTILIZING CAPACITANCE AND TENSIOMETER SENSORS

Abstract

Natural hazards such as landslides often occur above the groundwater table in the unsaturated soil zone. These partially saturated conditions arise from alternating events of evaporation and rainfall, with high-intensity rainfall potentially triggering slope failure by altering soil hydromechanical properties. While many researchers have explored slope instability using unsaturated soil mechanics and the drying soil–water characteristic curve (SWCC), this approach overlooks the hysteresis behavior of SWCC, limiting its ability to reflect actual field conditions. Some studies addressed this by incorporating the hysteresis behavior of the SWCC, including drying and wetting cycles, into slope stability assessments. However, limited research has utilized scanning curves in slope stability analysis despite their ability to capture intermediary paths within the hysteresis envelope and more accurately represent the natural state of soil under cyclic moisture changes. This study aims to determine the scanning curve of SWCC in real time using advanced tensiometers and capacitance moisture sensors within a controlled laboratory setting designed to simulate field conditions. Real-time monitoring of water content (WC) and soil suction (SS) in an engineered soil sample enabled the development of multiple drying and wetting SWCC cycles, including scanning curves. These experimental results were integrated into GeoStudio software for seepage and stability analyses. The findings demonstrate that the scanning curve, enclosing the hysteresis loop of SWCC, can be accurately determined through real-time monitoring of a single sample across repeated cycles. Numerical simulations incorporating scanning drying and wetting hydraulic properties, representing evaporation and rainfall flux boundary conditions, produced the most critical Factor of Safety (FoS). These findings indicate the impact of capturing actual soil behavior by determining scanning curves, preventing overestimation of the FoS, and thereby improving slope stability predictions.