Multi-Objective Optimization of Three Different SMA-LRBs for Seismic Protection of a Benchmark Highway Bridge against Real and Synthetic Ground Motions

Abstract

Many researchers have taken advantage of adding shape memory alloy (SMA) wires to base isolators to control displacements and residual deformations. In the literature, different arrangements of SMA wires wrapped around the rubber bearings can be found, as examples, straight, cross and double-cross arrangements. SMA wires with various configurations and radii lead to the different characteristics of the isolator system and thus various shear hysteresis. Therefore, the aim of this study is to evaluate the performance of these three SMA wire’s configurations in the seismic retrofitting of a benchmark highway bridge by implementing them in the bridge’s existing lead rubber bearings (LRB). This system is referred to as SMA-LRB isolator. Firstly, because of the crucial influence of the wire’s radius, this parameter is determined using a multi-objective optimization algorithm (non-dominated sorting genetic algorithm (NSGA)-II). This algorithm simultaneously minimizes the deck acceleration and mid-span displacement. Secondly, the optimized SMA-LRBs are implemented in the highway bridge and nonlinear dynamic analysis is conducted. For the nonlinear response history analysis, two strong ground motion records are selected from the PEER database, by studying the site’s conditions. In addition, ten synthetic ground acceleration time histories are generated. The result illustrates that the double-cross SMA-LRB reduces the maximum and residual displacements more than two other devices; however, it causes the largest base shear force and deck acceleration. Besides, the cross-configuration results in the least displacement reduction and has the least shear force and acceleration. To find SMA-LRB with the best overall performance, a multi-objective decision-making method is utilized and the straight SMA-LRB is recognized as the most effective isolator.