Moderate to large earthquakes typically trigger thousands of landslides, some of which significantly impact the built environment. In addition to co-seismic displacement, episodes of slow post-seismic movement have been noted in clay-rich slopes. Often these movements continue for days to weeks following the main shock. While analytical methods exist for predicting the potential for and displacement of inertially driven slope failures, no such procedures exist for identifying and predicting delayed and/or sustained displacement response in the post-seismic period.
This project investigated the role of pore pressure generation and subsequent dissipation on slow-moving landslides following earthquake shaking in normally consolidated clay slopes. Methods used to investigate this phenomenon focused on the testing of three geotechnical centrifuge models and complementary numerical modeling program.
Dynamic model response was grouped in two categories depending on whether or not permanent displacements occurred (i.e. non-destructive and destructive). For destructive motions displacement occurred along an initially localized slip surface, which then developed to a more diffuse (> 1 m) shear zone. This shear zone had the net effect of reflecting energy downward into the profile and preventing transmission of energy to the upper portions of the profile. This in turn concentrated residual pore pressure induced by cyclic strains to the lower one-half to one-third of the soil profile.
Post-seismic displacements in models were shown to be a result of transient states of lowered effective stress (i.e., factor of safety less than 1) induced by seismically generated excess pore pressure. These failures occur at shallow depth (< 2 m) as pore pressure generated deeper (> 5 m) in the profile migrated and dissipated at the slope surface. A power law viscosity model incorporated into the equation of momentum was shown to accurately predict displacements observed under assumed infinite slope conditions.
Primary factors contributing to this post-seismic response were geometry and subsurface conditions. Most sensitive to fluctuations in transient pore pressure (i.e. effective stress)were slopes of marginal stability (factor of safety less than 1.1). When a low permeability layer was located at the surface, post-seismic displacements were highest owing to impedance of excess pore pressures. Permeability contrasts within the soil profile had the cumulative effect of creating a zone of pore pressure accumulation and a preferred failure surface.
Cite this work
Researchers should cite this work as follows: