This CD-ROM consists of seven papers that were presented at a session sponsored by ACI Committee 341 at the ACI Fall Convention in St. Louis, MO, in November 2008. The papers focus on the most recent advancements in performance-based seismic design of reinforced concrete bridges, including analytical and experimental studies, and design and construction practices with relevant information related to reinforced concrete bridges based on performance-based design. Note: The individual papers are also available. Please click on the following link to view the papers available, or call 248.848.3800 to order. SP-271

This paper reviews the Direct Displacement-Based Design Method for seismic design of bridges and compares it to the design procedure implemented in the AASHTO Guide Specification for LRFD Seismic Bridge Design. It is shown with examples that DDBD has several advantages that could be put to use as an alternative design method to that proposed in the AASHTO Guide Specifications for LRFD Seismic Bridge Design.

This paper presents a collection of practical and readily implementable recommendations for the modeling of highway bridges and overpasses subjected to earthquake ground motions. The specifications were developed particularly for Ordinary Standard Bridges in California as defined according to the Caltrans Seismic Design Criteria. Bridge components that require special modeling considerations and nonlinear characterization are identified in this paper, establishing specific criteria for the level of sophistication required. To reduce possible errors that arise during modeling and analysis of bridge structures using a particular structural analysis program, a comparison between bridge models using SAP2000 and OpenSees analysis packages was carried out to assess sensitivities and characterize important modeling parameters. Comparisons were made between the two software packages using modal, pushover and nonlinear time history analyses. A total of six typical reinforced concrete bridges in California with box-girder superstructure and different geometries and cross sections were considered. Inconsistencies between the two analysis packages were found for peak displacements obtained through nonlinear time history analysis. Two methods of obtaining response estimate bias factors between the two programs are illustrated for the six bridges analyzed under three seismic hazard levels (50%-, 10%-, and 2%-in-50-year probabilities of exceedance).

Detailed data from 30 bridge column models, mostly tested on shake tables, were evaluated to determine the correlation between the apparent damage and internal and external seismic response parameters. The majority of the columns were of circular cross section. The columns were divided into five categories based on seismic design, shear demand, and ground motion type. Five distinct damage states were proposed based on the apparent damage. Six response parameters were also defined in terms of drift, frequency, strains, yielding, and ultimate displacements. The correlation between damage states and the response parameters were determined for each column category. The results demonstrate a general trend relating damage states to response parameters. Using the observed damage and the data presented in this article it was shown that an estimate of a non-dimensionalized push over curve may be made.

In the last two decades, seismic design of reinforced concrete bridges has shifted from a purely “life-prevention” design approach to a broader approach that also addresses “damage control” and “loss reduction” issues. This shift in philosophy requires the use of numerical tools that more accurately simulate the response of various bridge components. As importantly, a greater emphasis is placed on understanding the effects that these bridge components have on the overall seismic response of bridges. This paper presents the results of a series of nonlinear time history analyses of a RC bridge that was simulated under various modeling conditions using a finite element program called Opensees. These time history analyses were performed according to the following 16 modeling conditions: (a) two options for the nonlinear modeling of the columns, (b) two modeling conditions at the bridge columns’ foundation, and (c) four types of modeling conditions at the bridge abutments. Additionally, these modeling conditions were evaluated under two design earthquake levels that characterize the maximum considered earthquake and the frequent earthquake. To enforce the analysis, two more earthquake records were used: Inca-Peru and Northridge. Analytical results confirm that the various modeling options have significant influence on the seismic response of bridge systems, especially when nonlinear response of the abutment shear keys are included in the analyses. The different modeling runs were numerically evaluated and compared to each other in terms of the displacement ductility imposed on the columns. Within a performance-based design methodology, detailed results from these analyses are presented and discussed in further detail in the paper.