Influence of modeling assumptions on the seismic response of multi-span simply supported steel girder bridges in moderate seismic zones

Steel girder bridges are one of the most common bridge types in the central and south-eastern United States (CSUS). An understanding of their seismic response and assessment of their seismic risk has become a focus of the earthquake engineering community due to the increased awareness of the seismic hazard in the region. Analytical assessment of this seismic risk requires an evaluation of typical seismic responses and more of an understanding of the modeling parameters that significantly affect those responses. A seismic evaluation of a typical configuration for a multi-span simply supported steel girder bridge is performed for an approximate hazard level of 2% in 50 years using a nonlinear three-dimensional (3-D) analytical model. The results show significant vulnerabilities in the reinforced concrete columns and in the steel fixed and expansion bearings. Although the longitudinal loading of the bridge results in much larger demands compared with the transverse loading, some components of the bridge may still have appreciable damage under the transverse loading case. An analytical design-of-experiments screening study shows that modeling parameters such as loading direction and damping ratio are the most important in determining seismic response. Fixed bearing stiffness, among others, also significantly affects the response and should be considered carefully.