Stress–strain pattern–based criterion to assess cyclic shear resistance of soil from laboratory element tests

Abstract

The cyclic shear response of soils is commonly examined using undrained (or constant-volume) laboratory element tests conducted using triaxial and direct simple shear (DSS) devices. The cyclic resistance ratio (CRR) from these tests is expressed in terms of the number of cycles of loading to reach unacceptable performance that is defined in terms of the attainment of a certain excess pore-water pressure and (or) strain level. While strain accumulation is generally commensurate with excess pore-water pressure, the definition of unacceptable performance in laboratory tests based purely on cyclic strain criteria is not robust. The shear stiffness is a more fundamental parameter in describing engineering performance than the excess pore-water pressure alone or shear strain alone; so far, no criterion has considered shear stiffness to determine CRR. Data from cyclic DSS tests indicate consistent differences inherent in the patterns between the stress–strain loops at initial and later stages of cyclic loading; instead of relatively “smooth” stress–strain loops in the initial parts of loading, nonsmooth changes in incremental stiffness showing “kinks” are notable in the stress–strain loops at large strains. The point of pattern change in a stress–strain loop provides a meaningful basis to determine the CRR (based on unacceptable performance) in cyclic shear tests.