Report No.: CCEER-92-8

Title: An Evaluation of the Current Caltrans Seismic Restrainer Design Method

Authors: M. Saiidi, E. Maragakis, and S. Feng

Date: October 1992

A Report to the:

California Department of Transportation
National Science Foundation
Nevada Department of Transportation

Performing Organization:
Department of Civil Engineering/258
University of Nevada, Reno
Reno, NV 89557


The primary objective of this study was to develop an understanding of the implications of the current Caltrans hinge restrainer design procedure. Two aspects of the problem were studied. One was the effects of changing (a) the cross sectional area of restrainers and (b) the restrainer gap on the nonlinear response of a bridge with several hinges. The other was the sensitivity of the number of required restrainers to changes in some of the simplifying assumptions which are made in the current Caltrans restrainer design method.

Computer program NEABS-86 was used in the nonlinear analyses [3]. The focus of this part of the study was the relative displacements at the joints, restrainer forces, and restrainer stresses. Three earthquake records, the El Centro 1940, Eureka 1954, and Saratoga 1989. In addition to input earthquakes, the number of restrainers at each hinge, the restrainer gaps, and the hinge gaps were varied. It was found that when a restrainer gap of 0.75 in. is assumed, the number of restrainers does not affect the response significantly. It is recommended that the design should be based on cases with and without restrainer gaps to encompass all the critical forces, stresses, and displacements.

To study the effects of design assumptions on the required number of cables, several manual calculations of the example in the Caltrans restrainer design guidelines were carried out. The deviations from the method included the treatment of mass and stiffness of bridge segments as different hinges closed. Another variable was the simultaneous reduction in the restrainer gap and increase in the hinge gap. The results indicated that slight variation in some of the assumptions can change the number of restrainers significantly. A more streamlined design method that incorporates the nonlinear response of bridge components needs to be developed (Abstract by authors).