Seismic intensity measures for risk assessment of bridges

Abstract

In performance-based seismic assessment, structural response is characterised using fragility functions based on a seismic intensity measure (IM). IMs are typically related to the characteristics of ground shaking and structural dynamic properties, with the spectral acceleration at the first and dominant mode of vibration, Sa(T1), being a popular choice for buildings. In bridge structures, where no single dominant mode typically exists for bridges with some degree of irregularity, the use of Sa(T1) may be inefficient (i.e. large dispersion) due to multi-modal transverse response. To avoid having to choose a single bridge mode when using Sa(T1) and to appease the needs of bridge portfolio assessment, peak ground acceleration (PGA) can often be the IM used for bridge fragility functions in some countries. This study examines the efficient assessment of simple bridge structures characteristic of the European context by exploring different IMs based on Sa(T), peak ground velocity (PGV) or a recent candidate average spectral acceleration, AvgSa. Several case study bridges are evaluated via multiple stripe analysis with hazard-consistent ground motion records. The results indicate that PGA and PGV are indeed inefficient IMs compared to other IMs of similar complexity, especially at serviceability limit states, for the bridge structures examined. Also, a relatively casual record selection strategy is seen to not be suitable for risk assessment of bridges and can result in notable differences in risk. In contrast, AvgSa, which is an IM based on a simple combination of Sa(T) values across a range of periods, showed very good predictive power and robustness in terms of its risk estimates across all ranges of structural response. This was observed for the structure-specific IMs in addition to the group IMs used for assessing multiple structures with the same ground motion records. This study has thus shown these AvgSa-based IMs to be an appealing choice to consider for further examination in future fragility function and risk model development for bridge structures.