UC San Diego

Precast segmental construction of bridges can accelerate construction and minimize the cost of bridges in highly congested urban environments, environmentally sensitive regions, difficult to access ravines, and wide river crossings where medium to long repetitive spans are needed. Despite their proven benefits, the use of precast segmental bridges in seismic regions of the United States remains very limited. A main obstacle to their use is concern regarding the seismic response of segment joints. Recent research has shown that segment joints can undergo very large rotations that open up gaps in the superstructure, while maintaining its load carrying capacity, and with little, if any, damage. This dissertation investigates the seismic response of precast segmental bridges using detailed 2D non-linear time- history analyses and focuses on the behavior of segment-to -segment joints constructed using the balanced cantilever method. The joint model was calibrated using results available from experimental work on large scale sub- assemblages. Analytical models of full scale precast segmental bridges with geometries and characteristics, similar to the Otay River Bridge and the San Francisco- Oakland Bay Bridge Skyway in California, were also used in this study. A suite of twenty near field earthquake records was used to determine the median joint response as well as to quantify the effect of vertical motion on the joint response. The earthquake records were scaled using two different scaling methods and the effect on the structure response was studied. Four different pre- earthquake stress conditions were studied to determine if the effects of creep, shrinkage and temperature impact the seismic response of segment joints. In addition, a preliminary investigation into the possibility of removing a portion of the superstructure longitudinal post- tensioning and allowing non-linear elastic response of the segment joints during a significant seismic event was performed. Results indicated that vertical earthquake motions and the pre-earthquake stress-state can alter the response of segment joints. The results also showed that the method of record scaling does not significantly alter the median response, but may effect the variation in response. Furthermore, reducing the longitudinal post- tensioning near the piers may be possible provided that service load cases do not govern the design