Comparison between constant-volume and fully undrained cyclic simple shear tests using the strain-based energy concept

Abstract

The cyclic direct simple shear (CDSS) test has been widely used for dynamic analysis, including soil liquefaction. CDSS enables the replication of shear wave propagation and considers more representative stress conditions during earthquakes. Several studies have investigated the drawbacks of CDSS, including the prominent artificial pressure jump (R_u∗ jump) produced in the first few cycles during constant-volume tests, as demonstrated in the pioneer works of Prevost and Høeg (1976) [37]. However, the effect of this pressure jump on the overall CDSS results, including the number of cycles to produce liquefaction or cyclic mobility, has not yet been investigated experimentally. Therefore, this study investigated this aspect using a series of strain- and stress-controlled fully undrained triaxial simple shear (T_xSS) and constant-volume CDSS tests. Strain-controlled test results were used to establish relationships between dissipated energy and R_u to examine the difference between the fully undrained and constant-volume tests conducted on the same basis. The results show that the artificial R_u∗ leads to the overestimation of the R_u values and underestimation of the cyclic resistance. However, consistency between the CDSS and T_xSS results was achieved after correcting the R_u∗ jump error. The agreement between the experimental and numerical results confirms the efficiency of the stain-based energy concept to improve the cyclic CDSS test results.