Parametric Probabilistic Seismic Performance Assessment Framework for Ordinary Standard Bridges

Abstract

This paper focuses on the assembly and implementation of a full-fledged parametric probabilistic seismic performance assessment framework for ordinary standard bridges (OSBs) in California. The framework stems from the performance-based earthquake engineering (PBEE) assessment methodology developed under the auspices of the Pacific Earthquake Engineering Research (PEER) Center. It involves a sequential execution of analytical steps to arrive at estimates of performance measures which, for example as considered in this study, are the mean return periods (MRPs) of exceedances for a selected set of limit-states (LSs). Improvements from state-of-the-art literature related to various stages of the PEER PBEE assessment framework are incorporated. This includes: (1) introduction of an improved intensity measure (IM), i.e., average spectral acceleration over a period range, for probabilistic seismic hazard analysis (PSHA), (2) conditional mean spectrum (CMS)-based site-specific risk-consistent ground motions selection for ensemble nonlinear time-history analyses involved in probabilistic seismic demand hazard analysis (PSDemHA), (3) introduction of material strain-based engineering demand parameters (EDPs), (4) identification of practical damage LSs, and (5) development of strain-based fragility functions required in probabilistic seismic damage hazard analysis (PSDamHA) for the considered LSs. Four distinct testbed OSBs are selected for the study. A two-dimensional design parameter space is defined in terms of typical primary design variables involved in seismic design of OSBs, i.e., the column diameter and the column longitudinal steel reinforcement ratio. Computational models of the as-designed bridges as well as their re-designs spawned by varying the primary design variables subject to practical constraints are assessed using the implemented framework. For each testbed OSB, and for each of the considered LSs, a smooth surface is fitted to the MRPs computed for all the re-designs of the bridge in the primary design parameter space. Topologies of these surfaces are explored. Feasible design domains in the two-dimensional design parameter space are identified. Safety of the as-designed version and feasibility domain for the re-designs of each testbed OSB are examined and discussed.