Evaluation of equivalent linearization methods of MDOF isolated structures in response spectrum analysis

In the seismic design of isolated structures, equivalent linearization of the hysteretic isolators is normally considered. There exist many equivalent linearization methods and evaluation of these methods has been widely studied. However, most of the previous studies concern a single degree of freedom system only, and the used analysis method is time history analysis. In such a case, the evaluation heavily depends on the ground motions selected. Note that the response spectrum analysis is a commonly used method for seismic design of structures. Therefore, it is more valuable to evaluate these equivalent linearization methods based on the response spectrum analysis. In this study, a multiple degree of freedom model is adopted due to the contribution of higher modes of the superstructure to the seismic responses, and the procedure of the response spectrum analysis in conjunction with the equivalent linearization method is presented in detail. Then, considering a series of parameters of isolators, a statistical study is performed to evaluate the accuracy of these equivalent linearization methods by comparing with the nonlinear time history analysis. Moreover, the effect of the damping reduction factor for the response spectrum on the accuracy of the equivalent linearization methods is discussed in terms of probability in a desirable error range. The analytical results show that the accuracy of the equivalent linearization methods is associated with the earthquake intensity. Meanwhile, for different response quantities the accuracy of each equivalent linearization method varies significantly. In addition, to estimate the seismic responses with a better accuracy in the response spectrum analysis, a suitable damping reduction factor should be considered.