Comments on the determination of the undrained steady state strength of sandy soils

Seismically induced failures of earth structures generally involve relatively loose sandy soils. The failures can be divided into: (a) those caused by a loss of stability and (b) those in which limited, but large, deformations accumulate during the earthquake to constitute cause of "failure" for structures or lifelines. In both cases, the key soil parameter is the soil shear strength at large strains. The value of this strength determines whether instability could be triggered by the earthquake, and if instability is not possible, it determines the amount of soil displacement that will be accumulated. The strength that governs the two types of seismically induced failures discussed above is the undrained steady state strength (referred to as S-us). The laboratory test methods that are appropriate for determining steady state strength depend on the type of soil. The most important test characteristic is the ability to apply sufficiently large strains while maintaining specimen uniformity. The determination of in situ values of S-us must consider the heterogeneity of natural soil deposits and the changes in void ratio than even the most careful sampling procedure will cause in "undisturbed" soil samples. Empirical information based on field index tests (blowcounts, cone) is useful as a first approximation. However, in some cases these index tests do not properly represent field conditions.