Seismic fragility updating of highway bridges using field instrumentation data

Seismic fragility assessment of deteriorating highway bridges using analytical methods often rely on empirical, semi-empirical or numerical models to predict the rate and nature of degradation. Consequently, the predicted structural vulnerabilities of bridge components or overall bridge system during seismic shaking are only as good as the adopted deterioration models. For the sake of simplicity and ease of computational modeling, these deterioration models are often far removed from observed manifestations of time-dependent aging. One such example is the nature of corrosion in reinforced concrete bridge components under chloride attacks. While this deterioration mechanism leads to the formation of pits along the length of the rebar, past literature often adopts the simplified uniform area loss model. This study proposes a probabilistic framework that assists in improved deterioration modeling of highway bridges by explicitly modeling pit formation and also provides the opportunity of updating the analytical models with field measurement data using Bayesian techniques. The framework and case-study results presented in this study are believed to render realistic seismic fragilities for highway bridges when located in moderate to high seismic zones.