Accuracy evaluation of the modal pushover analysis method in the prediction of seismic response of vertically irregular frames



The traditional pushover analysis procedure does not represent adequately, the effects of varying dynamic characteristics caused by structural yielding and the contributions of higher modes to the structural responses. The recently developed modal pushover analysis (MPA) procedure significantly improves the traditional pushover analysis by including the effects of a sufficient number of modes in the analysis.
The present study evaluates the accuracy of the modal pushover analysis in estimating the seismic demands of vertically irregular planar moment-resisting frames in comparison with the exact results from nonlinear time-history analysis (NLTHA). In this research, seven irregular 12-story frames representing three types of height-wise irregularities: mass irregularity, vertical geometric irregularity (i.e., setback) and difference in floor levels, and also one regular 12-story frame as a reference are analyzed due to an ensemble of seven ground acceleration records. The earthquake records are chosen such that a wide range of frequency contents, as well as the effects of both near and far fault distances, be taken into account in the analysis. Findings from this investigation indicate that in most of the irregular models considered, the method of MPA can reasonably estimate the values of critical parameters such as peak floor displacement and story-drift ratios. Furthermore, it is observed that the inclusion of the first two modes improves the accuracy of predictions relative to using only the fundamental mode in the analysis. Such degree of improvement in accuracy is not reached by including the third mode effects in the MPA procedure.