When an earth dam is subjected to earthquake loading, it is possible that certain zones within the embankment will liquefy, and their strength will be reduced to a low, residual, value until the excess pore water pressures causing liquefaction can dissipate. Embankments that are especially vulnerable in this regard are old hydraulic fills, and tailings embankments required by mining operations. Decisions regarding the safety of these embankments often hinge on assessing their post-earthquake stability assuming that certain zones within the dam have liquefied and their strength has been reduced to a small residual value. Current geotechnical engineering practice is to use residual strength values obtained from back-analysis of embankment failures. Unfortunately, there is a large scatter in these values. Recent experiments suggest that residual strength may depend on the velocity of motion of the sliding mass, i.e. liquefied sands behave as Bingham plastics, exhibiting a small threshold strength which increases with sliding velocity, and thus the residual strength at any moment during sliding depends on the particular geometry of the sliding mass. This may explain in part why there is so much scatter in back-analysis values. The proposed testing program will attempt to quantify, under controlled conditions, the extent to which the residual strength of liquefied granular materials is rate-dependent and the liquefied soil behaves as Bingham plastic. A ring shear device of new design is proposed for these tests. The materials to be tested represent a range of granular soils susceptible to liquefaction and flow. The test results will provide basic insight into the influence of such variables as gradation, fines content and plasticity of fines on rate-dependent residual strength values, applicable to the analysis of loose contractive materials in embankments and natural slopes subjected to earthquake loading. If, as expected, there is a significant variation of residual strength with strain rate in granular soils, there will be an important impact on analyses of earth embankments under earthquake loading, and on runout analyses of debris flows and tailings. In embankment dam stability analyses, the minimum threshold residual strength at very low strain rates would have to be employed; runout analyses would consider a range of strength values depending on the geometry and potential velocities of the sliding mass. These same factors will have to be considered in back-calculating residual strengths from field case histories.
Development of the proposed ring shear device will make available to the profession a new tool for studying the impact of other variables (e.g. partial drainage during sliding) on post-liquefaction behavior of natural sands under controlled conditions, as well as permitting the study of other materials such as tailings produced by different ore processing techniques. From the point of view of teaching and training, the design of the load control system for the ring shear device will be an excellent capstone design project for senior undergraduates in mechanical engineering, and will provide a beneficial interaction between the mechanical engineering students and at least one civil engineering graduate student; it is also hoped that one or more civil engineering undergraduates may be involved in the project through the university's very active undergraduate research opportunity program (UROP).
