Microstructural Transformations and Prediction Models of an Expansive Soil Subjected to Simulated Rainfall Conditions

Abstract

The pore size distributions of an expansive soil under different simulated rainfall conditions were studied using nuclear magnetic resonance (NMR) techniques. Four sets of treatments with different rainfall intensities (light, moderate, heavy, and rainstorm) and durations (0.5 d, 1 d, 2 d, and 3 d) were designed to analyze the effects of rainfall on the microstructure of the expansive soil. Results show that with increasing rainfall duration, micropores gradually form and develop, enlarge, and interconnect in the soil, eventually forming stable seepage channels. Under light and moderate rain conditions, the proportion of micropores gradually decreases, while the proportions of mesopores and macropores gradually increase. Under heavy and rainstorm conditions, the proportion of micropores sharply decreases and then stabilizes, while the proportions of mesopores and macropores increase. With increasing rainfall intensity, the dominant pore size and porosity both initially increase and then stabilize. A quantitative relationship model between pore size, porosity, and rainfall conditions is established and the fitting effect is good. This study shows that rainfall alters the microstructure of expansive soils, which stabilizes after dynamic equilibrium. This provides a theoretical basis for predicting and controlling the engineering behavior of expansive soils under different rainfall conditions.