Development of Performance Databases

The development of a multilevel, performance-based design methodology is the primary goal of the Pacific Earthquake Engineering Research Center (PEER). To implement such a methodology, it is necessary to accurately characterize structural behavior over a wide range of possible demands. Several efforts are under way to define and quantify limit states, and to develop analytical tools that reliably predict the point at which a structure will reach each state. The reliability of these tools needs to be evaluated by comparing predicted with observed behavior.

Although many experiments have been performed on reinforced concrete members and assemblages, the test results have never been systematically collected, organized, and made widely available. Two efforts initiated by the PEER Center are in progress to develop databases that catalog experimental test results in consistent formats. The PEER Column Performance Database provides test results for a large number of recent and older tests of reinforced concrete columns. The PEER Bridge Performance Database provides detailed performance information for bridge components and systems. Both databases are available over the Internet to the structural engineering community.

Peer Column Performance Database

The PEER Column Performance Database was developed to provide data with which to evaluate the reliability of column modeling and assessment strategies. The core of the database was originally assembled by researchers at the National Institute of Standards and Technology (Taylor and Stone 1993; Taylor et al. 1997). A PEER project at the University of Washington has expanded the number of tests in the database, increased the amount of information provided for each test, and made the information available on the World Wide Web (www.ce.washington.edu/~peera1).

The database currently describes tests of approximately 340 reinforced concrete columns. Of these columns, 188 had rectangular reinforcement, 135 had spiral or hoop reinforcement, 4 were spliced, and 11 were retrofitted. A schematic of the organization of the database and website is provided in figure 1.

Fig. 1. Organization of PEER Column Performance Database

For most of the tests, the database provides the column material properties, the column and test geometries, a plot and digital file of the force-deflection history, a classification of the observed damage, and notes on unusual characteristics, or the presence of P-D effects. For the spliced and retrofitted columns, the database provides only force-deflection data and overall dimensions for the columns.

The number of tests will continue to increase as more tests are performed and as more data from older tests are identified. The amount of information available for each test will also continue to expand. In particular, University of Washington researchers are now collecting damage information from the original test reports. For tests where such information is available, the database will soon include displacements at which the experimental investigator reported each type of damage.

PEER Bridge Performance Database

The University of California, San Diego (UCSD), has developed a PEER Bridge Performance Database that focuses on describing the structural behavior of concrete bridges as it pertains to performance response levels (www.structures.ucsd.edu/PEER/peer.html). The database addresses the response determination and parameterization of bridge components, subassemblages, and systems, in direct support of the development of multilevel performance design and evaluation procedures for bridges.

The proposed five-level performance evaluation approach (Hose et al. 1999), based on previous work on buildings (SEAOC 1996), consists of three stages of capacity evaluation that are described in detail in the background paper available on the Internet. The initial stage of evaluation, Bridge Damage Assessment, relates classifications of structural damage to socioeconomic definitions at five designated levels. The five levels of damage range from NO visible damage to LOCAL FAILURE/COLLAPSE of components or the entire system and are linked to corresponding damage and repair descriptions. The second stage in the capacity assessment approach for bridges, Bridge Performance Assessment, associates five performance levels ranging from CRACKING to STRENGTH DEGRADATION to qualitative and quantitative performance descriptions. Quantitative guidelines such as crack widths, crack angles, and regions of spalling are provided. The final stage of the capacity evaluation procedure consists of Bridge Performance/Design Parameters. The detailed cataloging and assessment of bridge components, subassemblages, and systems attempts to evaluate as many different parameters as possible such that the right combination of performance parameters can eventually be chosen for performance-based design. Parameters include concrete and steel strains, curvature ductility, plastic rotation, displacement ductility, drift, principal stress states, residual displacement index, normalized effective stiffness, and equivalent viscous damping ratio. The numerical determination of these parameters, which is performed at all five identified engineering limit states, requires instrumentation and/or analysis. This parameterization of bridge components, subassemblages, and systems is essential in the development of a consistent performance-based design methodology for bridges in seismic zones. Explicit instructions for calculating these parameters are provided on the Internet.

A standardized evaluation template is also available over the Internet such that a variety of test units can be analyzed and cataloged to develop a consistent performance-oriented design library. Instructions for each section of the template are provided during input, and the completed evaluation templates are submitted over the Internet. A key feature of the PEER Bridge Performance Database is correlating the five damage/performance levels to photographs, which is convenient as a postearthquake reconnaissance guide for the evaluation of structural safety. An example of this photo documentation is provided in figure 2 for one of the component tests conducted at UCSD as part of a PEER project to investigate the effects of subjecting a large velocity pulse on flexural bridge columns. This set of performance limit states related to photographs, together with the numerical quantification of performance parameters, forms an integral part of the bridge performance catalog.

 

Figure 2: Photo Documentation Example in PEER Bridge Performance Database



Test Setup


Level I


Level II


Level III


Level IV


Level V

Summary and Future Developments

The University of Washington and UCSD have collaborated in the development of their databases to ensure that the information provided by each is complementary. For example, the two databases will be using similar definitions of damage states. The purpose of the PEER Column Performance Database is to make extensive results of column tests available for the community to use. The PEER Bridge Performance Database correlates damage states to performance levels, with the ultimate goal of establishing a consistent multilevel performance-based design method that is based on quantifiable parameters rather than on the simplified physical models and empirical coefficients used in current bridge design practice. Although the two databases currently include information only on reinforced concrete columns and bridges, PEER anticipates that future phases will incorporate steel and composite construction that are more prevalent in the Midwest and East Coast.

For those interested in submitting data to either of these two databases, please visit the websites for more information.

Yael D. Hose
Graduate Research Assistant
Department of Structural Engineering
University of California, San Diego

Marc Eberhard
Associate Professor
Department of Civil Engineering
University of Washington

Frieder Seible
Professor and Chair
Department of Structural Engineering
University of California, San Diego

References

Hose, Y. D., P. Silva, and F. Seible. 1999. Performance library of concrete bridge components, sub-assemblages, and systems under simulated seismic loads. SSRP 99/08. La Jolla, Calif.: Structural Systems Research Program, University of California, San Diego. January.

SEAOC. 1996.Appendix B: Conceptual framework for performance-based seismic design. Vision 2000. 393–416. [Sacramento, Calif.:]: Structural Engineers Association of California.

Taylor, A. W., C. Kuo, K. Wellenius, and D. Chung. 1997. A summary of cyclic lateral-load tests on rectangular reinforced concrete columns. Report NISTIR 5984. Gaithersburg, Md.: National Institute of Standards and Technology.

Taylor, A. W., and W. C. Stone. 1993. A summary of cyclic lateral-load tests of spiral reinforced concrete columns. Report NISTIR 5285. Gaithersburg, Md.: National Institute of Standards and Technology.