Controlling Preshear Relative Density in Triaxial Tests and Its Effects on Undrained Behavior of Sand

This paper introduces a new preparation technique, termed volume control (VC), which attempts to achieve a pre-determined, preshear relative density (D_rc) for granular soil samples used in triaxial tests. This technique provides more controllable density conditions on the samples in triaxial tests than conventional methods that apply a "target" preshear effective confining stress. Thus, the VC technique allows for a clearer examination of a granular soil's behavior by allowing the user to control the effects of preshear relative density on the results. In contrast, conventional preparation methods which set a predetermined confining stress cannot control the relative density prior to shearing. The new volume control technique was used to investigate the undrained behavior of a sand prepared at a given preshear relative density (D_rc), to study the stress-strain, pore pressure, and strength behavior. A series of monotonic undrained compression triaxial tests were performed on a sand obtained from a Cape Cod beach in Massachusetts (CCS). The VC technique was used to control the D_rc to 40, 70 or 95% with the final confining effective stresses for these samples ranging from 214 to 656 kPa. For validation and comparison, samples prepared by a stress control (SC) technique, the conventional method of performing triaxial tests on granular soils, was performed on samples with initial relative densities (D_ri), after the saturation phase of a triaxial test, in the range of 2.96 to 82.9%. These samples were consolidated isotropically to targeted effective confining stresses of 173, 345, and 690 kPa yielding D_rc in the range of 43 to 111%. Study results show that both sample preparation techniques are valid, however, the bulk moduli and measured friction angle results from VC technique tests were more consistent than those prepared under the SC technique.