This research is an outcome of the National Science Foundation 07-506 program solicitation "George E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES) Research" competition. This project was awarded as a NEESR-payload. This project was led by the University of Vermont and included a subaward to Prof. Pedro deAlba of the University of New Hampshire.
It has long been observed that saturated sands subjected to shock or earthquake loading experience drastic loss of strength and behave as heavy fluids, gradually regaining strength as internal water pressures dissipate. As long as the liquefied state persists, the soil will flow down slopes, producing destructive landslides and large drag forces on obstacles such as piled foundations. Modeling this behavior for risk studies and engineering design, however, requires adequate measurements of how shearing strength is lost and is eventually recovered as internal water pressures build up and subsequently dissipate. There are currently no full-scale field measurements of these strength changes to guide development of such models; existing field case histories are limited to observing the final damage produced by the liquefaction process. Controlled laboratory measurements would be desirable, but the onset of liquefaction is accompanied by such large strains that soil samples in conventional laboratory tests become so drastically deformed that reliable strength measurements can no longer be made. As a first step in measuring the evolving behavior of liquefied sands, it was envisioned that the shear strength of liquefying sand can be measurable in-flight in a seismic geotechnical centrifuge model using a thin coupon (plate, about 25 millimeters by 75 millimeters by ~1.5 millimeters) pulled horizontally through the soil model, with its major dimensions parallel to the base of the model. The large strains and strain rates associated with liquefaction flow failures would thus be simulated by moving the coupon relative to the sand, through and after the shaking until the excess pore pressures dissipate. By measuring the drag force on the coupon, it was possible to observe the evolution of the soil shear strength as it decreases to a minimum (residual strength) and subsequently increases as pore pressures dissipate. The centrifuge results were compared to companion ring shear tests.
Equipment required to conduct the payload tests was designed and built by a group of undergraduate mechanical and electrical engineering students at the University of Vermont as their senior capstone design project, and calibrated before installation by a civil engineering undergraduate student. The UVM civil engineering undergraduate student and UNH graduate student conducted centrifuge tests at the University of Colorado at Boulder. The companion ring shear tests were conducted at UNH.
Data from this project is archived in the NEES data repository.
Publications from this work to date:
- Dewoolkar, M. M., Hargy, J., Anderson, I., de Alba, P., and Olson, S. M. (2016), "Residual and postliquefaction strength of a liquefiable sand," Journal of Geotechnical and Geoenvironmental Engineering, 142(2), 10.1061/(ASCE)GT.1943-5606.0001374.
- Anderson, I., Dewoolkar, M., Hargy, J., and de Alba, P. (2014), "Measurement of post-earthquake strength of liquefiable soils in centrifuge models," 8th International Conference Physical Modeling in Geotechnics, Perth, Australia.
- Anderson, I., Hargy, J., de Alba, P., and Dewoolkar, M. M. (2012), "Measurement of residual strength of liquefied soil in centrifuge models", GeoCongress 2012 Conference, Oakland, CA.
- Dewoolkar, M., Anderson, I., de Alba, P., and Hargy, J. (2010), Measurement of the Strength of Liquefied Soil in Physical Moels, a report submitted to the National Science Foundation
VTrans - Bridge Scour Project
Tropical Storm Irene had unprecedented impacts on transportation infrastructure in numerous regions of New England states and the state of New York. In Vermont alone, more than 300 bridges were damaged, which motivated this study Successfully mitigating scour and erosion related problems is dependent on our ability to reliably estimate scour potential, design effective scour prevention and countermeasures, design safe and economical roads and bridges accounting for erosion potential, and design reliable and economically feasible monitoring systems. These demand numerous research needs spanning from first-principle micro to system-level macro scales, which are currently underway in collaboration with Professors Donna Rizzo, Dryver Huston, Jeff Frolik and Yves Dubief. At the micro scale, understanding at the level of a soil particle is needed; i.e. how soil particles get lifted by flowing water, how the soil type affects its erosion potential, and what are the associated critical stresses and velocity of water especially under highly turbulent conditions during extreme events. First principle understanding of these mechanisms is underway by capturing concepts from the field of computational fluid dynamics (CFD) and general advancements in numerical simulations, computing power and visualization tools. An example of an intermediate scale study would be developing accurate and passive, yet low-cost sensing and monitoring technologies - a scour sensor is under development. An example of a system-level macro-scale study is developing a link between bridge scour and system-level geomorphic assessment of rivers and streams. Scour is affected by the geometries of the channel and the banks, flows, soils types and numerous other factors including human interference. Therefore, it is very difficult to develop a physics-based model that can reliably predict scour potential. On the other hand, modern numerical tools such as artificial neural networks can detect non-linear interrelations among variables, rank variable importance, and yield semi-empirical correlations that include rating schemes. About 1,370 stream miles (~2,500 stream reaches) in Vermont has been assessed with reasonable details for their geomorphic conditions including characteristics of the valley and floodplain corridor, the bankfull channel, riparian areas, evidence of flow modification and bed or planform changes; providing unique opportunities for linking these assessments to observed or predicted scour and erosion.
The geo-referenced data on damaged (and non-damaged) Vermont bridges from the 2011 Tropical Storm Irene can be downloaded from: ____________________________
Last modified February 10 2016 02:06 PM