The performance of bridges during large-magnitude earthquakes generated by the Cascadia Subduction Zone is an essential feature of the resilience of communities along the U.S. West Coast because bridges will be needed for pre-tsunami evacuation, emergency response, and economic recovery. This project developed a three-dimensional, nonlinear model of a typical bridge and subjected it to sets of simulated magnitude 9.0 earthquake ground motions to investigate the performance of bridges with a variety of structural characteristics.
As part of this study, typical characteristics of bridges along major highways in the Puget Sound region of Washington were identified from a database compiled by the Washington State Department of Transportation (WSDOT) and University of Washington engineers. On the basis of the properties that were common in the database, a three-dimensional, multi-degree-of-freedom model was developed to represent typical highway bridges in Washington state. The reference bridge model had three spans, a continuous superstructure, and an L-type (seat) abutment with bearing pads and internal transverse shear keys.
Using that model, the researchers conducted a series of parametric studies to evaluate the effects of six bridge locations, four site classes, abutment characteristics, and the properties of the columns at the intermediate supports. For each of the combinations of bridge abutment characteristics, column properties, location, site class, and ground motions, the researchers performed 300 analyses.
The researchers found that the calculated displacements of the bridges were consistently smaller in the longitudinal direction than in the transverse direction because of the high resistance provided by the backwall, abutment, and backfill soil. In the transverse direction, the likelihoods of flexural damage for the reference bridge were largest for locations on the Olympic Peninsula, reaching maximum values of 31 percent for concrete column spalling (Forks, Wash., Site Class D1) and 6.1 percent for longitudinal bar buckling (Port Angeles, Wash., Site Class D3). Neglecting the abutment resistance greatly increased the transverse displacements for all locations and site classes, with the largest increases occurring for long-period bridges in sedimentary basins. When the shear key resistance was neglected, and the column height was doubled for a bridge in Seattle, the maximum likelihoods of spalling and bar buckling increased from 1.53 percent to 32 percent and from 0.3 percent to 5.6 percent, respectively.
These findings provide WSDOT an opportunity to develop and prioritize cost- and time-efficient bridge retrofit plans. The results of this study suggested that, for typical bridges, WSDOT should prioritize retrofitting bridges where the abutments do not effectively constrain the column displacements of the bridge, either because the bridge has no shear key or because the superstructure is curved or discontinuous.
Report: WA-RD 908.2
Authors:
Kan-Jen Liu
Addie Lederman
Zachary Kortum
Marc O. Eberhard
Jeffrey W. Berman
Nasser A. Marafi
Brett Maurer
UW Department of Civil and Environmental Engineering
Sponsor: WSDOT
WSDOT Technical Monitor: Bijan Khaleghi
WSDOT Project Manager: Mustafa Mohamedali