A metamodeling based seismic life-cycle cost assessment framework for highway bridge structures

Abstract

This study presents a novel framework for seismic life-cycle cost assessment of highway bridges using parameterized fragility models rooted in modern statistical learning techniques. Unlike traditional fragility curves, often used for seismic loss assessment, the parameterized fragility models are multi-dimensional functions conditioned on ground motion intensity as well as critical field measurable bridge parameters. Such functions are particularly useful for prompt vulnerability estimation and life-cycle cost assessment of structurally dissimilar highway bridges across bridge portfolios in regions characterized by moderate to high seismicity. Consequently, the proposed framework can aid decision makers and stakeholders for efficient channeling of monetary resources for necessary seismic upgrade of highway bridges within a transportation network. As a case-study example, this study chooses the non-seismically designed multi-span simply supported concrete girders highway bridge class that constitutes a significant portion of the highway bridge inventory within Central and Southeastern US. The proposed parameterized fragility functions for this bridge are used along with seismic hazard data and probabilistic repair cost information for seismic loss assessment through a component-level approach. In addition to near-instant seismic life-cycle cost assessment, as demonstrated through several examples, this approach also helps avert potential double counting from multiple component damage while calculating system-level losses.