Centrifuge modeling of seismic liquefaction of clean ottawa sand under low and high confining pressures for field conditions

Abstract

The current evaluation of liquefaction triggering is based on the liquefaction triggering charts, developed from Simplified Procedure. The most popular liquefaction triggering charts are based on field tests like Cone Penetration Test (CPT), Standard Penetration Test (SPT), and Shear Wave Velocity (Vs). However, there is a common problem for the SoP liquefaction evaluation charts: case histories, used to develop these charts, were under effective overburden pressure (σ’v0) less than 2 atm. To apply the liquefaction triggering charts to cases under high overburden pressures larger than 2 atm, overburden pressure correction factor, Kσ was proposed to correct the effect of overburden pressure on Cyclic Resistance Ratio (CRR). From then, there is a number of studies about the choices of Kσ values, however, the SoP Kσ curves were developed from undrained laboratory tests, rather than the field cases. To study the effects of overburden pressure on the liquefaction responses of idealized field cases, a series of centrifuge tests were conducted to simulate a saturated sand deposit with 5m thickness. These centrifuge models had two different relative densities (Dr = 45% and Dr = 80% to cover loose and dense sand deposits), two different overburden pressures (1 atm and 6 atm to cover low and high overburden pressures) and two different boundary conditions (single drainage (SD) and double drainage (DD)). Two novel centrifuge techniques were developed to obtain a designed level of overburden pressure with dry leadshot, and to realize double drainage with geocomposite materials. The objectives of the research were to study the responses of saturated sand deposits to the input motion under low and high overburden pressures and to determine overburden pressure correction factors, Kσ. The soil responses under low and high overburden pressures were compared in aspects of acceleration time histories, excess pore pressure histories during shaking and dissipation, pore pressure profiles, shear stress ratio, and shear strain time histories. There was also a comparison of soil responses with different boundary conditions (single and double drainage conditions) but the same relative density and the same overburden pressure. It was found that there was significantly more partial drainage under high overburden pressure of 6 atm than 1 atm in both the single drainage (SD) and double drainage (DD) conditions. To determine the overburden pressure correction factor, Kσ, based on centrifuge tests, both the stress-based and strain-based approaches were applied. Over the past 40 years, the most common approach used as the frameworks for assessing the liquefaction triggering potential at high overburden pressure is stress-based approach. The alternative approach: strain-based approach, is less common. All the Kσ values from centrifuge tests in the research of different relative densities (Dr = 45% and Dr = 80%), different boundary conditions (single drainage (SD) and double drainage (DD)) and different evaluation approaches (stress-based and strain-based), were consistently larger than 1.0, which is contradictory to the Kσ values from undrained laboratory tests, where Kσ < 1.0. This contradiction was caused by the significantly more partial drainage at high overburden pressure of 6 atm than at 1 atm. Further analysis was conducted to get the coefficient of consolidation, cv, during the dissipation period based on excess pore pressure and settlement records. It was found that the cv was 2 ~ 4 times larger for the sand models at 6 atm than at 1 atm in single drainage (SD) condition. And, cv was 3 ~ 5 times greater at 6 atm for double drainage (DD) conditions.