| Seismic slope stability has been an important topic in geotechnical earthquake engineering for many years. Seismic stability of slopes and embankments is typically assessed by (1) conducting pseudostatic analyses, which will provide a stability parameter (e.g. factor of safety), and by (2) estimating permanent deformations likely to be induced by earthquake shaking. Among these, estimating permanent deformations has recently attracted much attention, and it is the primary topic of this thesis. A series of 1-g small-scale soil slope tests were performed on the U.C. Berkeley, Davis Hall shaking table to provide well-documented scale-model case history data including slope geometries, material properties, and slope response and performance. For each model slope, three types of physical tests were conducted: (1) Geophysical hammer-hit tests (to estimate shear wave velocity of the model soil), (2) Low-amplitude frequency-sweep tests (to acquire site responses of the model slope under a small-strain condition), and (3) Large-amplitude shaking tests (to acquire the slope responses and deformations data under strong shaking condition). The experimental data were then compared with calculations made by various analytical procedures for verification and calibration.; Direct use of the undrained shear strength of the model clay based on small vane shear tests led to some discrepancies. Factors that make the "representative" shear strengths of the model slopes differ from the values measured by the small vane shear tests are: (1) Shearing rate, (2) Progressive development of failure, (3) Number of significant pulses of the earthquake motion, and (4) Post-peak strength response. To understand the effect of the progressive development of failure on the clay's "representative" shear strength, the model clay was tested by a series of the large-scale direct shear tests (12 inches by 12 inches). Two vane shear tests with different scales (2.25 inches and 1 inch in diameter, respectively) were also performed using the same material. It was found that the peak strengths decreased with the increasing testing scales. Conversely, the failure displacement at peak strength increased with increasing testing scales (i.e. soil responses were more ductile with larger testing scale).; A procedure was developed to adjust the soil strength (measured by the vane shear tests) based on these factors. It was found that both conventional and advanced FE analyses provided better estimates using the adjusted "representative" shear strengths. (Abstract shortened by UMI.)... |
