Improved Component-Level Deterioration Modeling and Capacity Estimation for Seismic Fragility Assessment of Highway Bridges

Abstract

This paper presents a novel vector-based seismic vulnerability assessment methodology for deteriorating highway bridges by uniquely accounting for realistic deterioration of key structural components. The proposed framework offers notable enhancement over the state-of-the-art procedures that assume simplified and often unrealistic deterioration models of bridge components and promote inflexible unidimensional fragility curves for vulnerability assessment of aging bridge structures. Based on available data from past field investigation reports and laboratory experiments, this study proposes improved deterioration models that specifically encompass pitting corrosion of embedded reinforcing bars in columns under chloride attacks and necking failure of bearing anchor bolts. These deterioration models are incorporated within the finite element modeling of aging bridge components to develop multidimensional seismic demand estimates, capacity limit states, and parameterized seismic fragility functions using modern statistical learning algorithms. As a case study example, such vector-based fragility functions conditioned on a multitude of parameters are developed for the popular multi-span continuous steel girder highway bridge class of Central and Southeastern US. Twofold findings from the case -study reveal (1) the criticality of incorporating realistic deterioration modeling of critical bridge components for seismic vulnerability assessment, and (2) convenient utilization of parameterized seismic fragility functions by stakeholders and bridge engineers for prompt retrofit and rehabilitation decisions of aging highway bridges within the transportation infrastructure. Demonstrative examples reveal 25% and 10% overestimation of complete damage state median fragility for 75 year old bridge column and system, respectively, using the conventional modeling techniques when compared to the proposed improved deterioration models.