Flow mechanism and suction distribution in heteregenous residual soil slope under rainfall infiltration

Abstract

An imperative factor in rainfall induced slope failure is infiltration rate. Water that infiltrates into residual soil is predominantly controlled by two factors, i.e. rainfall intensity and saturated permeability of soil, which varies with depth as a result of weathering processes. Variation in permeability may either prevent or allow water to seep into deeper soil layers. Therefore, this study aimed to investigate the behaviour of suction distribution in a two-layered residual soil system consisting of Grade V and Grade VI residual soils with various saturated permeability functions using a laboratory physical slope model, in-situ or field work, and numerical modeling. The laboratory physical slope model was developed for the purpose of facilitating infiltration tests with three different permeability functions for each of the Grade V and Grade VI soils. A total of 42 infiltration tests were performed. The twolayered slope was then numerically simulated using SEEP/W GeoStudio software, which served to verify field data and determine the best modelling scheme that later be applied to signify the suction distribution behaviour of the residual soil slope model. Burrow holes present in the Grade VI soil layer caused the loss of the capillary barrier effect, which in turn allowed more rainfall to infiltrate into the soil layers. It was also found that when the ratio of permeability function between Grade V and Grade VI soils was high, an increase in the breakthrough time with corresponding decrease in the breakthrough matric suction occurred. From the seepage analysis, the numerical model incorporating burrow holes in Grade VI residual layer coupled with the effect of two sets of relict joints in Grade V yielded significant improvement in heterogeneous residual soil slope modelling. The findings of this study were then validated with previous findings using Prediction Accuracy (PA) analysis. It was established that burrow holes and two sets of relic joints conclusively improved the modelling of heterogeneous residual soil slope particularly at depths of 1.0 m and 1.5 m.