LIQUEFACTION-INDUCED FLOW FAILURE OF EMBANKMENTS AND RESIDUAL STRENGTH OF SILTY SANDS

Failure modes of embankments or slopes consisting of saturated cohesionless soils during earthquakes may be classified into two types, the one involving a limited amount of permanent deformation such as settlements or lateral building on the order of tens of centimeter to meters, and the other in which complete slumping or flow of materials takes place through distances of tens of meters. Associated with seismic risk evaluation of embankments, there are many cases where the occurrence of limited deformation is tolerated but the flow type of failure is never permissible. In such a case, a methodology becomes necessary for evaluating the safety of embankments against the flow type of failure, once it occurs. Post-earthquake stability analysis in which the conventional analysis is made by employing residual strength or steady-state strength of soils is used for this purpose. However, accurate evaluation of in-situ residual strength is a difficult task, particularly for cohesionless soils. Thus, it would appear useful if estimates of the residual strength can be made based on the back analysis of several embankments that have suffered the flow type failure during past earthquakes. In the present study, slope stability analyses were made for several earthquake-damaged embankments and tailings dams composed of silty sands or sandy silts to back-calculate the value of residual strength. The ground conditions at the failure sites were investigated by means of the Dutch cone penetration test or by the Swedish cone test. By comparing the values of the back-calculated residual strength with those of the cone resistance, empirical correlations between them are obtained which may be used in turn for the evaluation of residual strength of similar soils on other sites. (Author/TRRL)