System versus component response of a two-span reinforced concrete bridge system

A large-scale 20.5 m long asymmetric two-span reinforced concrete bridge was tested to failure using the shake table system at the University of Nevada Reno. Upon completion of testing, in depth analytical modeling was conducted to evaluate the accuracy of conventional methods in reproducing the bridge model response and to develop a model for further study. Utilizing the experimentally verified computer model, the system effect was investigated, comparing the system and response of individual bents as well as the response of several other bridge models. In comparing computational model of the shake table specimen and models of the individual bents with tributary mass, it was shown that for all of the columns in this study, there was generally not an increase in hysteretic energy or large displacement cycles from system response at given displacement demand. The response of the bents for each high amplitude test motion was also compared. It was shown that there were significant differences in the bent demands for a given excitation due to system effects. In addition to the shake table model, four bridge systems with a constant total lateral stiffness were used in a parametric study to determine the system effect. The symmetric and uniform versions of the bridge specimen were shown to be comparable in nonlinear performance to the bridge specimen for the same high amplitude demand. The failure progression of the bridge model and the analytical comparisons suggested that the reserve capacity from varied column heights could provide a beneficial substructure redundancy.