Countless hydraulic fill and tailings dams, levees, and other slopes contain loose soils with little or no internal cohesion. This state renders them vulnerable to failures under seismic events in which the soil becomes "liquefied." Although large flow failures of such earth structures are uncommon, the catastrophic consequences of such failures dramatically elevate the associated risk. Modeling this behavior requires reliable estimates of how the strength of a liquefying soil evolves to its minimum as internal water pressures are generated during an earthquake. Of equal importance is understanding how the soil subsequently regains its strength as water pressures dissipate once shaking ceases. To date, there are no full-scale field measurements of this strength evolution to guide model development; therefore, engineers must rely on strength estimates calculated after the fact from sparse, poorly documented flow failure case histories. The uncertainty in selecting design values of residual strength is exacerbated if the soil contains significant amount of fine particles, as each empirical method suggests a different approach to account for fines. This project will provide practicing engineers a better understanding of the large-strain behavior of liquefiable soils, improved estimates of residual shear strength, and new calibrated modeling tools that are required to make crucial decisions on high-risk, high-consequence projects involving tens or even hundreds of millions of dollars. The project's education and technology transfer plan includes training and mentoring graduate and undergraduate students from multiple disciplines (geotechnical, mechanical, and electrical engineering, as well as fluid dynamics).
This interdisciplinary project centers on developing a potentially transformative understanding of the shearing resistance of liquefiable soils (including its evolution during pore pressure generation, residual strength mobilization, and pore pressure dissipation) in an environment with greater realism than available in conventional laboratory tests. Specifically, novel centrifuge models will utilize a thin metal coupon (plate) pulled through a liquefiable soil before, during, and after shaking to directly measure the soil's residual strength and strength recovery. By embedding a pressure transducer in the coupon, it will be possible for the first time to measure pore pressures directly on the shearing surface in a liquefied soil. The large number of measurements will allow the researchers to explore the effects of fines content (soil compressibility), effective stress, and strain rate on residual strength. Cone penetration resistance and shear wave velocity will be measured in-situ in the centrifuge models, allowing the researchers to validate or improve empirical residual strength correlations. In turn, the physical measurements from the centrifuge testing program will enable the researchers to calibrate a multi-disciplinary, fluid dynamics-based stress-strain constitutive model for evaluating liquefaction problems such as flow slides, debris flows and lateral spreading around pile foundations. This project will utilize the NHERI Centrifuge facility at the University of California, Davis.
