Coupled hydro-mechanical analysis of two unstable unsaturated slopes subject to rainfall infiltration
Journal
Engineering Geology
Journal Volume
216
Pages
13-30
Date Issued
2017
Author(s)
Abstract
This paper presents a case study and numerical investigation of two unstable unsaturated slopes along the Taipei Maokong Gondola system. One slope collapsed because of torrential rainfall during Typhoon Jangmi in September 2008, and another nearby slope developed excessive deformation under consecutive wetting and drying cycles. Recorded rainfall, measured soil parameters, site geology, and slope geometry were used in coupled hydro-mechanical finite element analyses to investigate the failure and deformation mechanisms of these two slopes. The stress paths of these two slopes were also examined and are discussed in this paper. The numerical results demonstrated that the coupled hydro-mechanical analysis satisfactorily predicted the failure and deformation characteristics of the two unstable slopes. In the collapse case, failure occurred at a shallow depth when the surficial residual soil was saturated. The slope failure was attributed to a decrease in soil shear strength when the matric suction gradually decreased as rainfall progressed. In the excessive deformation case, the numerical results suggested that pore water pressure (PWP) varied at shallow depths under wetting and drying cycles, whereas positive PWP accumulated at the soil–rock interface and induced non-uniform lateral flows parallel to the interface. The accumulated positive PWP at the interface mobilized a considerable plastic deformation in the soil. Examination of the relationships between the slope factor of safety and the corresponding hydrological data (i.e., rainfall and soil PWP) revealed a positive correlation between the slope factor of safety and accumulated rainfall. The factor of safety predicted by the infinite slope equation significantly varied with the input of PWP. This study demonstrated that the infinite slope equation using the PWP measurements obtained from the lower half of the slope could effectively predict the slope factor of safety.
SDGs
Type
journal article