Abstract:
In this study, the accuracy level of geometric scaling factor (herein, lambda = 45) selected for the shaking table test model represents the actual site conditions as discussed by comparing the results of both experimental data and numerical analysis obtained from the soil-structure oscillation and free-field response under seismic loads. The difficulty to be overcome in this experimental campaign was to investigate the limited soil size considered for this complex geodynamic problem due to the small capacity of the shaking table having payload of 2.5 kN and the geometry of testing platform of 1 m x 1 m, respectively. The superstructure used in the analysis of the soil-structure interaction is specified as six story building structure with height of 30 m resting on a silty sand soil with a shear wave velocity of 536 m/s. Within the frame of this work, the underlying soil is handled as a homogeneous single layer on the top of rigid bedrock. The diameter of the ground specimen and the bedrock depth were considered to be 26.23 m and 10 m, respectively. Special cylindrical flexible testing equipment was designed to hold the excavated soil volume in the container without deforming the experimental setup. In order to examine the seismic responses at the control points of the scaled test models under the nonlinear supporting effects, 2-D finite element analyses of the soil-structure system have been executed by using the PLAXIS simulation program. It is obviously seen that the soil layer parameters have significantly increased the free-field site response compared with the bedrock response under different seismic excitations. Based on the comparison of calculated results with experimental recordings, it is suitable to mention that small capacity shaking table experiments with appropriate scaled test models provides an alternative means of laboratory investigations even for such complex soil-structure coupling problems.
