This thesis presents results of ten centrifuge model tests of earthquake-induced lateral spreading in sand using a laminar box. The tests were conducted at a 50 g centrifugal acceleration field on the RPI 100g-ton geotechnical centrifuge. The centrifuge tests simulated a horizontal or sloping, 10 m thick layer of water-saturated coarse sand of infinite lateral extent on an impervious rigid base. The layer was subjected to lateral base shaking with a prototype peak acceleration ranging from 0.17 g to 0.46 g, frequency 1 to 2 Hz, and a duration of about 22 cycles in all cases. The slope angle simulated in the field ranged from 0{dollar}spcirc{dollar} to 10{dollar}spcirc{dollar}.; It was observed that the horizontal and sloping models behave differently. In the horizontal model tests, after liquefaction, the accelerations in the liquefied soil were dramatically reduced, with the upper layers of the stratum becoming isolated from further seismic excitation. In the sloping tests, no reduction in accelerations were observed after liquefaction; instead, large upslope acceleration spikes were recorded, which, as demonstrated by subsequent study, were due to dilative shear stress-strain response of the saturated soil.; In the sloping tests, at large cyclic strains of the order of 1% or 2%, the solid skeleton of the soil tries to dilate and induces an instantaneous reduction in pore pressure and a corresponding increase in soil shear strength. This dilative behavior typically shows up in the transducer raw data as upslope spikes in the acceleration records and simultaneous drops in the piezometric records. That is, the displacement in the downslope direction is arrested by dilatancy, which causes a sudden drop of pore pressure accompanied by deceleration which arrests the movement of the mass. This dilative response, which was observed to become stronger as the slope angle and the input acceleration increased and as the input frequency decreased, limited the downslope strain accumulation and thus the value of the final lateral ground displacement.; Extensive discussions and comparisons of the ten centrifuge test results are presented. These includes Table 10.3 which summarizes the observed effects of slope angle, input peak acceleration, and frequency, on the following measured parameters: pore pressures, thickness of liquefied soil, soil accelerations, lateral displacements, permanent shear strains, and surface settlements. It is found that in the sloping tests, the boundary between liquefied and nonliquefied soil has the maximum permanent shear strains and tends to act as a failure surface. The average shear stress-strain response of the soil during shaking is obtained for all model tests using system identification techniques, thus providing additional insights as well as observations on yield shear stresses and dilatant response.
展开▼
