Seismic Fragility Assessment of Highway Overpasses with Pier Walls in Low-to-Moderate Seismic Zones

Abstract

Published literature indicates that seismicity in Australia is going through an active phase and more seismic events may occur in the no-so-distant future. Besides, the short recorded history of Australia demonstrates that destructive earthquakes, such as the Newcastle earthquake of 1989, are probable even in metropolitan areas. Due to the growth in population and the economy, and also for the purpose of sustainable development, reliable seismic risk estimates are necessary for managing civil engineering infrastructure. In particular, this knowledge is essential for highway bridge structures with seismic deficiencies so as to enhance the public transportation networks’ resilience against earthquakes. A practical decision-making tool for long-term bridge performance prediction is the seismic fragility curve, which indicates the failure probability of a bridge with respect to a particular performance level at increasing intensity measures (IM). The objective of this study is to develop a methodology which can be utilised by bridge authorities and transport agencies for efficient estimation of seismic fragility of highway overpasses. A literature review has shown that while the Cloud analysis is an efficient method, some limitations adversely impact on its application for collapse fragility assessment and/or evaluation of non-collapse fragility of bridges with high structural nonlinearity. This is particularly true when comparing with the rigorous incremental dynamic analysis (IDA) method. To this end, a new method, namely the extended Cloud analysis (ECA), is proposed for the seismic fragility assessment. The ECA relies on the use of scaled Cloud analysis (SCA) data in combination with the original Cloud analysis (OCA) data. Different ground motion record (GMR) scaling approaches are also proposed and examined for this purpose. Special attention is given to the southeast Queensland region as the site of interest for GMR selection and, also, modelling of highway overpasses with pier wall configuration as a common class of bridges within this area. Nonlinear time-history analysis (NTHA) is performed for the developed bridge models subjected to selected GMRs, based on OCA, SCA, ECA and IDA fragility assessment methods. Correspondingly, probabilistic seismic demand analyses (PSDA) are carried out on the NTHA data to generate the relevant seismic fragility curves. This study reveals that the proposed ECA method provides a credible estimate of the collapse fragility of highway bridges, in close agreement with the IDA and is useful for conservative fragility assessment. In addition, the non-collapse seismic fragility curves generated by the ECA method for highway bridges with high structural nonlinearity are in remarkable agreement with the rigorous IDA method over both narrow and wide ranges of IMs. These valid seismic fragility results achieved by the ECA method reduce the NTHA required by the IDA method to one-seventh, thereby saving a notable amount of computational time. Subsequently, representative and reliable seismic fragility curves are developed for the highway bridges existing in southeast Queensland. This is achieved by using the ECA method and also by considering various sources of uncertainty, which allows for the long-term management of the local bridge maintenance, rehabilitation, and repair (MR&R) activities.