Developing a Fundamental Understanding of k-sigma for High Confining Stresses Under Field Conditions

Rensselaer Polytechnic Institute, Troy NY

Investigators

Abstract

The project addresses a scientific gap in our understanding of the fundamental behavior of liquefaction triggering and liquefaction resistance at high overburden pressure (deep ground elevations). Liquefaction triggering is an important issue which researchers and practitioners have been struggling with for decades. Liquefaction design and remediation decisions affecting large embankment dams and other impoundments are critical to public safety and they may involve great expense. The current State of Practice (SoP) of liquefaction triggering evaluation of saturated sand and other cohesionless soils uses mainly charts based on field penetration tests such as SPT (Standard Penetration Test), and CPT (Cone Penetration Test). Field earthquake liquefaction case histories used for calibration of these charts correspond to shallow ground elevations. For engineering projects involving earth dams, tailings dams and other impoundments, the SoP extrapolates the field experience from these charts using a correction coefficient K-sigma. The current SoP for the coefficient K-sigma can range between 0.45 and 0.85 for same type of soil, a range of values differing by a factor of almost two. Furthermore, preliminary findings by the PIs under an NSF EAGER research grant have shed new light on soil liquefaction earthquake response at deep ground elevations, throwing significant doubt on the current assumption. Centrifuge tests indicated that under the field conditions the values of coefficient K-sigma were above 1.0 and at least 50 percent higher than used in current practice. The results were also confirmed by numerical simulations. Such underestimation coefficient K-sigma could lead to unnecessary remediations, with the cost of remediation due to liquefaction hazard under individual dams sometimes reaching more than hundred million dollars. While the soil liquefaction response at shallow ground elevation (1 atm) in the centrifuge was comparable to that of undrained soil in the laboratory, it was very different at deep ground elevations (6 atm), with the sand being more partially drained during shaking at deep elevations, thus slowing down the pore pressure buildup and increasing the values of coefficient K-sigma. The results suggest that the current SoP of using undrained testing to extrapolate behavior from 1 atm to much higher overburden pressure - with no case histories validating the extrapolation, seem to underestimate significantly the liquefaction soil resistance deeper elevations. A comprehensive research plan using parallel programs of centrifuge and small-sample tests as well as 1D and 2D numerical simulations that go well beyond our limited centrifuge testing of one sand in the EAGER project. Two clean sands and one silty sand in both loose and dense condition - will be extensively tested at low and high confining pressures, with the centrifuge models providing a variety of drainage boundaries and the numerical simulations extending the centrifuge results through parametric studies and examination of the effect of the horizontal drainage under an earth dam. The research plan aims at providing practitioners and researchers with the proper design tools to evaluate triggering of soil liquefaction and liquefaction resistance in the field at high overburden pressure. Furthermore, the anticipated significant contribution of the experimental work and numerical analyses using computer programs Dmod2000, FLAC and/or OpenSees, will be made available to the engineering community. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.