Abstract

ABSTRACT The purpose of this study is to verify the ability of a numerical method to predict earthquake-induced deformation of soil embankments. The method is dynamic response FE analysis that incorporates a cyclic elasto-plastic constitutive model for sand, a cyclic elasto-viscoplastic model for clay, and Biot’s two phase mixture theory. This method was applied to two sets of case records: 1) a series of dynamic centrifuge model tests of soil embankments resting on liquefiable sandy soils, and 2) a set of river embankments that were damaged and undamaged during the 1993 Hokkaido Nansei-oki earthquake. For the model tests, comparisons between measured and computed model responses were made. The results demonstrated that pore pressures within the foundation soil and vertical settlement of the embankment were qualitatively predicted. In addition, the mechanism of embankment settlement, even though the embankment foundation did not attain zero effective stress state, was explained. In the case of river embankments, the model parameters were estimated from the field and laboratory test data of foundation soils. A difference in the magnitudes of damage at two embankment sections was reproduced well, although it may not be quantitatively satisfactory. The results showed that the foundation sandy soils would have been liquefied. The analyses also indicated that a cohesive soil layer in-between sand layers that existed only at a non-damaged site prevented the liquefied soil beneath the embankment from spreading out lateraliy, and thus minimized the settlement of overlying soil.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call