Strength Loss and Softening of Sensitive Clay Slopes.

摘要

Strength loss and strain softening/localization are important issues for sensitive clay slopes. There is a question that if the ground moves a fraction of a meter due to earthquake shaking, sensitive clays would be just as easily remolded and trigger a catastrophic failure. Two possible reasons why we expect different relationships between deformation and strength loss are: seismic (dynamic) failure mechanisms are different from gravity driven (static) failure mechanisms, and rate dependent shear strength hinders localization of deformations into distinct shear planes.;In this dissertation, a literature review is performed to understand prior research on the behavior of sensitive clay slopes subjected to static and dynamic loading. Then, as a part of effort to simulate sensitive clay behavior with simplicity and accuracy, two constitutive models---rate-independent and rate-dependent sensitive clay models are developed. As an experimental approach, a construction technique to create sensitive clay in laboratory is also developed using a weak cement mix. In terms of repeatability and workability, physical modeling for artificially sensitive clay is valuable. The effects of sensitivity and earthquake intensity on the instability and deformation of slopes were studied using 12 centrifuge tests for cement treated clay slopes. Additional laboratory tests were performed to find material properties and static/dynamic behavior of artificially sensitive clays.;Based on physical modeling and laboratory tests, shear band thickness is governed by sensitivity, strain rate, and test conditions. Regarding the behavior of sensitive clay slopes, two different counterbalancing effects exist together. One is the effect of strain softening and initial stiffness. Strain softening of sensitive clay tends to make soil deformation bigger, but higher stiffness of earlier stage of loading for sensitive clay hinders the amount of deformation. Therefore, the importance of strain level and corresponding mobilized strength should be highlighted when comparing the performance of sensitive and insensitive clay slopes. Only after remolding at large strains might we expect insensitive clays to perform better than weakly cemented sensitive clays. Sensitive clay slopes can suffer catastrophic failure if they are deformed enough to destroy their structure.;The other counterbalancing effect exists between strain rate and strain softening. Strain rate effect may be an important parameter to affect how fast strain is localized. Faster strain rate (such as dynamic loading in centrifuge test) leads to higher strength. Then, the thickness of shear band tends to increase in terms of strain rate because larger shear bands develop smaller strain rates and smaller strengths. Hence, the rate effect hinders strain localization. However, because of strain softening, shear bands tend to localize as strain increases. Therefore, two competing effects exist together and there might be a rate-dependent optimal shear band thickness.;Thus, it is crucial to account for the level of shear strain of concern. Depending on the shear strain level, mobilized strength can be different. Accordingly, mobilized strength is affected by the rate of softening and the rate of strain. Counteracting effects summarized above can explain different aspects of static and dynamic as well as laboratory and field scale shearing behavior of sensitive clay slopes.