| description abstract | Uneven distribution of mass, stiffness, and strength over the length of a bridge can produce rigid body rotation of the superstructure, or “twisting,” during an earthquake. Twisting response modes lead to the concentration or amplification of damage in the bridge substructure and, as such, are undesirable. AASHTO recommends a balanced stiffness approach to limiting bridge irregularities and provides possible modification techniques for bridges that do not satisfy the recommendations. However, there is a lack of research documenting the implementation details and benefits of these modification techniques on the performance of irregular bridges. This study investigated the effectiveness of three techniques to reduce the twisting response modes for geometrically irregular bridges during earthquakes: adjusting the effective heights of the bridge columns, modifying column end fixity conditions, and redistributing the superstructure mass profile. First, analytical equations were derived to estimate the values of parameters needed to achieve a balanced stiffness using each technique for any arbitrary bridge. Next, a finite-element model of an irregular, reinforced concrete, quarter-scale bridge previously tested on a shake table was validated against available experimental data. This model was used in conjunction with a metaheuristic optimization algorithm to determine the optimal values of the parameters such that the superstructure rotation was minimized for the experimental ground motions. Then, a parametric study was performed using a suite of hazard-consistent and spectrally matched ground motions to investigate the sensitivity of the optimization results to the number and spectral characteristics of the ground motions. The parametric study showed that reducing the effective heights of the columns by stiffening the lower portion of the columns was the most effective method to reduce the twisting response of the bridge and that three ground motions with different spectral shapes were the most computationally efficient subset when performing the optimization methodology. | |
