This study was part of a Washington State Department of Transportation (WSDOT) program to assess the vulnerability of highway bridges built before 1984. Researchers applied slowly-varying transverse loads to a three-span, reinforced concrete bridge, including the superstructure, piers, and abutments. The purpose of the tests was to measure the transverse stiffness of the bridge and to estimate each support's contribution to stiffness. The researchers also evaluated analytical models by comparing the calculated and observed responses.
The bridge was extremely stiff and strong. In two cycles to a load equal to 45 percent of the bridge's weight, the maximum bridge displacement was 0.15 inch. During these cycles damage was minimal. At a load equal to 65 percent of the bridge's weight, the pier displacement was 0.30 inch. After the bridge had been excavated, the stiffness decreased to 15 percent of its initial stiffness. The stiffness further decreased to 8 percent of the initial stiffness after the superstructure had been isolated from the abutments.
The University of Washington (UW), California Department of Transportation (CALTRANS) and WSDOT models underestimated the stiffness of the bridge in its initial state. The UW model probably overestimated the resistance of the polystyrene at the abutments and underestimated the stiffness of the soil at the wingwalls. The CALTRANS model was too flexible because it neglected the resistance of the bearing pads and polystyrene. The WSDOT model was too flexible because it neglected the resistance of the bearing pads and polystyrene, and underestimated the soil stiffness.
The researchers concluded that (1) the tests can serve as a valuable benchmark against which to evaluate proposed seismic-evaluation procedures and models, (2) bridges that are similar to the test bridge are not highly vulnerable to transverse motions, (3) complex soil modeling is not justified if soil test data are not available, and (4) nonlinear analysis was necessary to reproduce the details of the observed response.