Evaluation of deformations of earth structures due to earthquakes

Conditions leading to post earthquake-induced deformations of an embankment due to liquefaction were investigated by a physical and numerical modeling program. A set of centrifuge experiments were performed in the National Geotechnical Centrifuge (NGC) and Schaevitz Centrifuge (SC) at the University of California at Davis. A Nevada Sand embankment model overlaid/confined by a clayey Yolo Loam layer was liquefied by an injection system at the base of the embankment model at NGC. The dynamic compaction of liquefiable soil was simulated by the volumetric increase of pore fluid through the water injection system in these tests. Models of the NGC model were also prepared and statically and dynamically tested at the SC facility to compare the deformation patterns between injection- and dynamic-induced liquefactions.; The dynamic and static centrifuge models were numerically modeled. In the process a new constitutive model was developed that was utilized in modeling of the injection-induced liquefaction process. The numerical models showed that the void ratio was redistributed during both injection- and shaking-induced liquefaction. The shearing deformations of Nevada Sand were, experimentally and numerically, observed to cease whenever the shaking, water injection, and/or in general the void ratio redistribution within the sand was stopped. The numerical model of injection-induced liquefaction also highlighted that the injection rate has a significant effect on the location of failure surface. A faster rate of injection may induce additional water accumulation/dilation at the base of Nevada Sand layer and induce a deep-seated failure. The injection rate reduction will push the failure towards the interface of less permeable cap and Nevada Sand layer.; The concept of residual strength is generally inconsistent with the concept of void ratio redistribution. However, the comparison between the mobilized shear strengths of the centrifuge models and the expected residual shear strengths was made based on the widely use of residual shear strength in practical approaches (for evaluation of dynamic and post earthquake stabilities) to highlight that the blindly use of residual shear strength without consideration of potential void ratio redistribution may not be a conservative assumption and may mislead the stability evaluations, especially for dilative sands.