Assessment of Dynamic Response of Cohesionless Soil Using Strain-Controlled and Stress-Controlled Cyclic Triaxial Tests

Abstract

Soils subjected to earthquake motions undergo random variations in stress, strain and frequency during the entire period of shaking. The spatial and temporal complexity of the strong motions is commonly addressed through simplistic strain-controlled or stress-controlled cyclic loading tests as a part of laboratory investigations. Such a methodology is utilized and reported in the present study for assessing the dynamic response of cohesionless sand obtained from River Brahmaputra, India. In order to assess the dynamic response and liquefaction potential of the cohesionless Brahmaputra sand, both stress-controlled and strain-controlled cyclic triaxial tests were performed on reconstituted cylindrical specimens prepared at different relative densities (Dr) ranging from 30 to 90%. The specimens were subjected to varying effective confining pressures (σ′c: 50–200 kPa), shear strain amplitudes (γ: 0.015–4.5%) and cyclic stress ratios (CSR: 0.05–0.3). For all the tests, the frequency of the applied harmonic loading was maintained at 1 Hz. The magnitude of excess pore-water pressure (PWP) generated during successive loading cycles of the strain-controlled tests was found to be considerably lower than that generated in a stress-controlled test. The strain-controlled test reveals a reduction in the development of excess PWP with the increase in confining pressure and relative density, thereby indicating a decrease in liquefaction potential with increasing confining pressure. The stress-controlled test highlighted that based on a particular CSR, an increase in confining pressure results in the requirement of higher deviatoric stress and smaller numbers of cycles are required to initiate liquefaction. Although apparently misleading, it is imperative that during earthquake, same deviatoric stress is applied on the entire soil deposit. Thus, for specimens at larger depths having higher confining pressure, the CSR value is smaller, and thereby successive deeper layers require more number of stress cycles to liquefy. Hence, the study revealed that both stress-controlled and strain-controlled tests can be successfully used to assess the dynamic properties and liquefaction potential of cohesionless specimens. Based on the findings, the developed cyclic shear stress ratio (CSR) = 0.5 and cyclic shear strain amplitude (γ) = 0.5% are considered as the limiting conditions to achieve the onset of liquefaction through strain-controlled and stress-controlled approaches, respectively.