SEISMIC ANALYSIS OF BRIDGES WITH PILE FOUNDATIONS

Abstract Several types of models may be used for the seismic analysis of bridges with pile foundations. These range in complexity from detailed models where each pile is modeled down to its tip and depth-variable ground motions are input along the length of the pile through elastic or inelastic soil springs, to models where a complete foundation is modeled with a single set of impedance (stiffness and damping) matrices. Models intermediate in complexity may also be used, where the piles are modeled between the pile cap and the mudline and then replaced with individual impedance matrices below the mudline. In these individual pile models, it is possible to retain important nonlinear aspects of the foundation behavior, such as pile plunging and uplift, and hinging of the piles at the pile cap, with a model of reasonable size. Some practical aspects of the modeling are discussed, including the use of local coordinates, and the optimum model topology. These various approaches to the analysis of pile foundations are discussed in the context of two projects: the seismic retrofit of the San Diego–Coronado Bay Bridge and the seismic design of the new east segment of the San Francisco–Oakland Bay Bridge. ADINA was used for the analysis of both structures. For the San Diego–Coronado Bay Bridge individual pile models with nonlinear elements were used for the global analysis. For the San Francisco–Oakland Bay Bridge, detailed models were used to investigate the behavior of foundations with both vertical and battered piles. The different behavior of these alternative foundations is described.