Reliability of reinforced concrete bridge systems subjected to pulse-type ground motions

Seismic records obtained near the causative fault, especially those containing large impulsive motions, have been considered to produce severe structural response in both buildings and bridges. This level of structural response has been inadequately anticipated through past design and analysis methodologies, frequently resulting in severe damage. As such, it was necessary to determine the source of the inadequacy in design methodologies. The inadequacy could result from inadequate PGA levels being used, design response spectra and design motions which do not capture the frequency content of near-field motions, or inadequate detailing and modeling practices for reinforced-concrete construction.; In order to validate current modeling methods for application to impulsive ground motions, experimental testing was conducted to determine the influence of large-amplitude early drift cycles on column response. The experimental studies revealed that limited changes occur in force-displacement behavior for ductile-detailed column samples as a result of the addition of a large-amplitude displacement cycle early in the test protocol, although the distribution of plastification at the base of the columns was impacted somewhat, resulting in a substantially reduced plastic hinge length. For columns detailed to produce brittle-shear and limited-ductility failures, the results were substantially different. Increased scatter of response characteristics—failure mechanism, drift and load capacity, and onset of performance states—was observed with the addition of early large-amplitude displacement cycles to the testing protocols. However, where plastic hinging was observed—even if ultimate failure occurred by brittle fracture—the overall force-displacement response did not appear to be substantially affected. This indicated that the inadequacy must lie in the modeling assumptions for ground motion models.; Based on the validation of current structural modeling methods, nonlinear numerical SDOF models of ductile bridge columns were constructed and subjected to two suites of earthquake records: one containing large-amplitude pulses, the other without. These ground motion suites were scaled based on peak ground acceleration to generate eight sub-suites of ground motions. The results of the simulations demonstrated that near-field motions generate distinct differences in response probabilities from those observed for far-field motions. Brittle column response was omitted from the analytical modeling as these columns are now quite rare in California, and the results of the experimental testing suggest that structural modeling methods are adequate. Furthermore, numerical modeling of ductile columns was adequate to determine if changes in ground motion representation to include near-field impulsive phenomena influenced model response.; It was the conclusion of this research program that substantial consideration must be given to the influence of near-field impulsive phenomena on the response of even ductile-detailed bridges in order to avoid unanticipated failures. Additionally, higher levels of PGA must be considered for structures which are believed to lie in the near-field of hazardous faults in order to increase the overall reliability of bridge structures.