A NEW METHOD FOR ASSESSING POTENTIAL DISPLACEMENT OF PREEXISTING LANDSLIDES REACTIVATED BY EARTHQUAKE-INDUCED GROUND MOTION

Most large seismogenic landslides are reactivations of preexisting landslides with basal shear zones at residual strength. Although residual strength typically is constant for gravitational landslide reactivations, lab tests and inferences from coseismic landslide reactivations indicate that residual strength may vary significantly during rapid displacement. Still, the response of residual shear zones to seismic loading is largely unknown.

We used a ring shear apparatus to perform seismic loading and constant displacement rate tests on specimens from two landslides. Seismic tests resulted in finite to unbounded displacement. Dynamic shear resistance estimated from shear strength models developed from constant displacement rate test results correlated well with stresses observed during seismic tests, indicating that displacement rate and amount controlled failure characteristics. These findings reveal that methods typically used for estimating coseismic landslide displacement will produce inaccurate predictions unless they can be modified to account for shear strength variability; such modification is cumbersome when shear strength varies with cumulative displacement and intractable when shear strength varies with displacement rate.

We developed a new approach to estimate coseismic landslide displacement that accommodates variable shear strength models. This approach involves iteratively calculating through a time series of coseismic shear stresses (1) stress exceeding strength that varies with cumulative displacement, (2) displacement rate as a function of excess stress determined from (1), (3) incremental displacement resulting from (2), and (4) cumulative displacement from (3) and previous iterations. The approach produced results that compared favorably to observations made during seismic tests, indicating its utility for application to landslides. Our method can be readily incorporated into dynamic mathematical landslide models with varying degrees of complexity, thus displacement estimations may be improved by accounting for variable shear strength, site effects on seismic energy, and changes in landslide masses such as thickness, shape, etc.