Effects of spatially varying ground motions on bridge response

Abstract

Adjacent bridge structures can move relatively to each other in an earthquake, resulting in pounding or more severely, unseating. Pounding occurs when the relative closing movement is larger than the structural gap whereas girder unseating takes place due to the relative opening movements being larger than the seating length provided. Relative displacements arise from unequal fundamental frequencies of adjacent bridge structures, spatial variation of ground motions and different soil-structure interaction (SSI). The objective of this thesis is to investigate the influence of these factors, especially spatial variation of ground motions on the bridge response. To achieve the objective, a series of experiments were conducted on the bridge models made of polyvinylchloride (PVC) using either shake tables or inertial actuators. Ground motions of the soft soil, shallow soil and strong rock conditions based on the New Zealand design spectra were simulated. The ground motions of the soft soil condition were further classified as highly, intermediately and weakly correlated ground motions to account for coherency loss effect of excitation spatial variation. These experiments include testing a 1:125 scale bridge with three identical bridge segments to study the effect of spatially varying ground motions with pounding, testing the same model but with the footings on sand contained by rubber boxes to consider the effects of spatially varying ground motions and SSI with pounding, testing one of those bridge segments with movable abutments to investigate the effect of excitation spatial variation considering abutment excitation and pounding, testing a 1:125 scale bridge model with two identical bridge segments with artificial plastic hinges to study the effect of spatially varying ground motions on inelastic bridge response with pounding, and field testing a 1:22 scale bridge segment subjected to spatially varying ground motions to determine the minimum total gap of a modulus expansion joint required to avoid pounding. A total of 8660 tests were performed. Research found that spatial variation of ground motions can increase the relative displacement of adjacent bridge girders and pounding forces. Based on the experimental results, the New Zealand Transport Agency (NZTA) Bridge manual was reviewed. It was found that the current NZTA Bridge manual is unable to suggest sufficient minimum seating length to accommodate the measured relative opening displacements. A set of empirical equations is therefore proposed to calculate adequate minimum support seating lengths to prevent girder unseating.