Investigating Triggers of Undrained Slope Instability Using Centrifuge Modeling

Abstract

Abstract This study investigated mechanisms by which undrained slope instability in contractive sand slopes can be triggered. Using centrifuge modeling, eight experiments were conducted on fine sand slopes, with the first prepared in a dilative state and the remainder in a contractive state. The tests examined potential triggers aimed at imposing undrained instability including a rising water table, sudden removal of toe confinement, rapid buttress and crest loading, and borehole-induced pore pressure spikes from borehole drilling with water. Seismic triggers were excluded. All models were tested at 30 g and subjected to a rising water table using either water or a viscous pore fluid. Spontaneous slope failure occurred only once the water table intersected the toe, at which point limit equilibrium slope stability analyses showed drained factors of safety of approximately unity. Triggers applied earlier did not cause global failure. Instrumentation captured total pore pressures, settlement, and total stresses. In all cases, failure was marked by an initial total stress redistribution, followed by small positive excess pore pressures localized along the developing failure surfaces and then by rapid slope deformation and a reduction in total stress and pore pressure. The undrained (or partially drained) responses were spatially limited to the shear zones and were not widespread within the saturated zone. Within the range of geometries, materials, saturation conditions, and triggers examined, undrained responses occurred as a consequence of drained instability once slope stability had become marginal. Triggers applied at higher drained factors of safety did not produce global failure.