Seismic response of isolated bridge superstructure to incoherent ground motions

This study investigates the seismic response of isolated bridge superstructure to incoherent ground motions. A ⅖th scale bridge model supported on elastomeric isolation bearings was tested on two independently controlled shake tables using harmonic and earthquake excitations. Ground motion incoherency due to wave passage effect was generated by phasing the harmonic motions applied at the ends of the bridge by 90° and 180°. The local site effect was modeled by modifying the table motions at each end of the bridge to represent, in turn, surface motions on rock, stiff soil and soft soil profiles respectively. Analytical studies were performed using Newmark's average acceleration method and the SAP 2000 computer program. Combinations of structure period and ground motion periods that give critical bearing displacements were calculated using 3-D spectra for incoherent motion. Newmark's method predicted the experimental results with good accuracy but SAP 2000 underestimated the response, possibly due to errors in the calculation of acceleration time histories from the input displacement time histories. It was observed that when the phase and amplitude differences between the support motions were small, the response to incoherent motion could be estimated by the average of the responses to coherent motions. However, response to incoherent motion can be greater than or smaller than that due to coherent motions, when applied in transverse direction and when there is large difference between support motions. Estimation of response using ii the displacement equation given in AASHTO Guide Specifications for Seismic Isolation Design was greatly improved when the one-second spectral acceleration was used, rather than the peak ground acceleration.