2009 Portland GSA Annual Meeting (18-21 October 2009)

Paper No. 8
Presentation Time: 9:00 AM-6:00 PM

STABILITY OF UNSATURATED INFINITE-SLOPES UNDER TRANSIENT INFILTRATION CONDITIONS


SENER KAYA, Basak1, GODT, Jonathan W.2 and LU, Ning1, (1)Civil Engineering, Colorado School of Mines, 1500 Illinois St, Golden, CO 80401, (2)U.S. Geological Survey, Denver Federal Center, Box 25046, M.S. 966, Denver, CO 80225-0046, bsener@mines.edu

The reduction of hazard and risk from shallow landslides induced by rainfall requires understanding of the transient hydrological and mechanical conditions that lead to landslide occurrence. Rainfall infiltration into partially saturated hillside materials is a transient process that causes consequent time-dependent changes in hillslope moisture content, soil suction, and stability. We extend a framework for assessing the stability of partially saturated infinite slopes from steady to transient seepage conditions. This framework includes a generalized effective stress for variably saturated soils and considers the effects of soil weathering on porosity and shear strength as a function of depth.

We present results for hypothetical sandy and silty soils under transient seepage conditions to examine the influence of material properties on failure depth and timing. Time-dependent pore-pressure profiles were computed using a 1-D analytical solution. These profiles were then used to determine the timing and depth of a potential failure surface by means of an infinite-slope stability analysis extended to partially saturated soils using a generalized effective stress principle. For example, a 45° degree hillslope composed of coarse sand is destabilized over depths ranging between 0.1 m to 0.6 m below the ground surface after 36 to 48 hours of rainfall flux equivalent to the saturated hydraulic conductivity (1×10-5 m/s) of the material. In contrast, the same slope, composed of fine sand and subject to the same relative flux at the ground surface (5×10-6 m/s), is destabilized in less than 36 hours about 0.1 m to 1 m below the ground surface. Results illustrate the sensitivity of depth and timing of slope failure on soil hydrologic properties. We also compare model results with measured rainfall, soil suction, and soil moisture collected prior to shallow landslide occurrence from a steep coastal hillside near Seattle, WA. The comparison shows that the extended framework could be used to predict approximate depth and time of shallow landslides induced by rainfall.