An efficient approach for optimum shape design of steel shear panel dampers under cyclic loading

Abstract

The low-cycle fatigue performance of shear panel damper (SPD) highly depends on the geometry of its shape and the criterion considered for its design. The main contribution of the current study is to find the optimum shape of the SPD subjected to cyclic loading by considering two different objective functions. The maximum equivalent plastic strain and the ratio of energy dissipation through plastic deformation to the maximum equivalent plastic strain are selected as the first and second objective functions, respectively. Since the optimization procedure requires high computational efforts, a hybrid computational approach is used to perform two paramount phases of estimating the inelastic responses of the SPD and solving the optimization problem. In the first phase, as an alternative for the time-consuming finite element analysis of the SPD, a weighted-support vector machine model is developed to predict the inelastic responses of the SPDs during the optimization process. In the second phase, the optimum shape of the SPD is found by using the whale optimization algorithm (WOA). The results indicate that both design criteria lead to the optimum-shaped SPDs with a significant improvement in their low cycle fatigue performance in comparing with the initial rectangular shape while a slight reduction in their energy dissipation capacity. Moreover, the second design criterion is slightly better in the performance improvement of the optimum-shaped SPDs compared with the first one. In addition, the weighted-based SVM approach can accurately predict the inelastic responses of the SPDs under cyclic loading, and its combination with WOA results in finding the optimum solutions quickly.