|Abstract or Summary
- Infrastructures along the Oregon Coast are vulnerable to seismic events in the Cascadia Subduction Zone that could generate large tsunamis. Bridges along the coast are an important part of the transportation and lifeline system in the area. The major damages on these bridges due to earthquake and/or tsunami would result in traffic disruption along the coast which in turn could significantly damage the economic in the region. Most of these bridges, built between the 1950's and 1970's, were not designed to withstand such large seismic and tsunami loads; thus, they are at risk of being damaged. Since specific design codes for tsunami impact on bridge superstructures are currently not available in general, and in the State of Oregon in particular, this thesis develops and presents a guideline for estimating tsunami forces on bridge superstructures along the Oregon Coast for practical use in engineering design. These guidelines are generally expected to be applicable in other locations under similar situations.
In the development of the guideline, numerical models based on a finite-element code (LS-DYNA) were developed to analyze tsunami impact on full-scaled bridge superstructures of four selected bridges on the Oregon coast - Schooner Creek Bridge, Drift Creek Bridge, Millport Slough Bridge, and Siletz River Bridge. The numerical models were analyzed for a better understanding of the interaction between tsunamis and bridge superstructures and to calculate tsunami forces time-histories on the bridges. Two different types of bridge superstructure, deck-girder and box section, were developed in the case of the Schooner Creek Bridge to study the performance of both the cross-sectional configuration subjected to identical tsunami load conditions. The results showed that the tsunami forces on box section are significantly higher than the forces on deck-girder section; thus, the box section might not be appropriate to be used in a tsunami run-up zone. Moreover, numerical testing of bridge superstructures with rails and without rails under identical tsunami loads are performed to examine an effect of rails to tsunami forces. The results suggested that horizontal and vertical tsunami forces on bridge without rails is smaller than on bridge with rigid rails up to 20% and 15%, respectively. Furthermore, results obtained from the numerical models are incorporated into the mathematical formulations from the existing literature to develop a simplified method for estimating tsunami forces on bridge superstructures. Appropriate empirical coefficients for bridge superstructures under tsunami loads were evaluated in this study based on an average value of the scattering data from the numerical results. However, the guideline is intended to be used as a preliminary guidance for design only as it did not accounted for uncertainties, therefore, an
appropriate factor of safety must be included in the design. A previous analysis of tsunami forces on the Spencer Creek Bridge on the Oregon Coast, also performed by the research group in which the author is a member, is revisited to examine the applicability of the developed guideline. This thesis also presents computational performance studies and an optimal number of CPUs for running fluid-structure interaction (FSI) numerical models of bridge superstructures using LS-DYNA software on high-performance computing (HPC) systems.