Traditional nonlinear static pushover analyses cannot fully address the multi-mode effects and response interaction that occur in a seismically excited structure. The recently developed modal pushover analysis (MPA) procedure improves on traditional pushover analyses by including multi-modal effects; however, its application is currently restricted to predicting the peak values of scalar response quantities. The peak responses computed using the MPA procedure could be used to form a rectangular demand envelope for any two specified response quantities. However, this rectangular envelope may be overly conservative as compared to the corresponding values computed by nonlinear time-history analyses. This paper describes a procedure, which uses the results of MPA analyses, to predict the envelope that bounds a vector of responses in a nonlinear structure. The accuracy of the proposed procedure is examined for selected pairs of responses by comparing the predicted envelope to the mean simulated response envelope obtained from an ensemble of nonlinear dynamic analyses. The response pairs considered include bi-directional interstory drifts and bending-moment-axial-load interaction in columns within a three-dimensional model of a three-story steel-moment-resisting-frame building. It is shown that the procedure has a level of accuracy that is appropriate for estimating the impact of seismic response interaction on the performance of a structure loaded into its nonlinear range.
展开▼