Effect of identification on seismic performance assessment of a tall building

Abstract

It is crucial to represent the realistic behavior of structures in order to estimate probability of failure for a prescribed damage level under earthquake actions. On the other hand, actual seismic responses of buildings may differ considerably from the simulated ones due to the assumptions made in Finite Element Models (FEM’s). Therefore, in recent years structural health monitoring techniques have been adopted to update the FEM’s based on modal parameters obtained from vibration-based identifications. In this study, a twenty-six story, core-wall tall building was instrumented with data acquisition system in order to examine feasible solutions to problems mentioned above. Modal parameters were obtained by Enhanced Frequency Domain Decomposition method and a detailed FEM of the structure is updated based on the identified modal characteristics in order to represent the actual dynamic characteristics of the building. Updating of a fine FEM allows eliminating the uncertainties in structural parameters such as the effects of soil springs and seismic gaps which are usually omitted in updating procedures. Along this line, nonlinear time history (NLTH) analyses were performed with the updated and the non-updated FEM with 5 % damping ratio. Additionally, NLTH analyses were carried out with updated FEM considering the identified (2 %) and 5 % damping ratios in order to observe the effect of damping ratio separately. Probability density functions in terms of interstory drift ratios were established in order to quantify the probability of failures under different ground motions. Results highlight that incorporation of system identification and model updating techniques into seismic performance assessment is crucial.