Experimental Study of the Performance Characteristics of Sandy Soil Debris Flow under the Effect of Artificial Rainfall

Abstract

Abstract The mechanisms and dynamic processes that drive debris flows are complex and require a multidisciplinary approach to study. A major component of debris flow is soil-water interaction, which is difficult to examine internally in a slope but can be studied using physical tests. In this paper, the results of a debris flow experiment in sandy soil are presented, with the aim of studying the performance of such soil when subjected to artificial rainfall. The method involves advanced data acquisition techniques, image measurement, and image analysis. The results show that the debris flow undergoes backward sliding failure in layers and blocks and has a large impact force. The failure can be divided into four stages: rainwater infiltration, start-up, main failure, and postfailure. The characteristic parameters, such as sand displacement, soil particle velocity, pore water pressure, and particle long-axis orientation, are analyzed. We show that the essential processes of sliding failure are hierarchical: sliding failure develops from a mesoscopic slope, followed by formation of a potential sliding surface in the sandy soil when subjected to rainfall, and ultimately a local sliding mass forms. Particle collision and separation can also be observed, along with expansion of the sliding mass until the debris flow occurs at the macroscopic scale. In practice, the complex soil-water interactions are hidden within the debris flow, which means that when monitoring and providing warnings of debris flows it will be necessary to examine global deformation, key local movement, and external deformation that occurs with changes in soil-water parameters for prediction purposes. This is over and above monitoring of conventional surface subsidence. This work provides an experimental method for studying other kinds of soil debris flow mechanisms, including key physical parameters, and may also be useful for developing engineering treatment and early warning systems.