Abstract

Soil exhibits inherent spatial variability, creating a significant source of uncertainty in geotechnical assessments. This variability becomes particularly critical when evaluating the seismic performance of infrastructure such as multi-span highway bridges, since traditional methodologies in bridge design often oversimplify soil properties by assuming uniformity. This approach, however, may lead to considerable inaccuracies in determining structural response under seismic activity. The complexity of soil-structure interaction (SSI) in such multi-span structures further exacerbates the influence of soil spatial variability on the overall structural response to seismic events. Although numerous studies have explored the impact of spatial variation in ground motions on seismic performance, a noticeable gap exists in the literature addressing soil spatial variability in the SSI modeling and its impact in the seismic response of multi-span bridges. Accordingly, this research aims to address this gap by proposing a numerical framework that integrates the inherent spatial variability of soil in SSI modeling by means of random fields theory and 3D nonlinear dynamic finite element models into the seismic performance analysis of multi-span bridges. The findings from a case study reveals a significant influence of soil spatial variability on structural response, leading to discrepancies in vulnerability assessment between different bridge components and highlighting the importance of incorporating spatial variability in soil parameters into seismic assessments of bridges. Moreover, soil variability appeared to slightly impact system-level vulnerability. Although the main conclusions are developed from a case study and are applicable to bridges with similar characteristics and seismic demand, the proposed approach can readily be applied to other bridge configurations and seismic environments.