A Critical Evaluation of the Coefficient Method, Capacity Spectrum Method and Modal Pushover Analysis for Irregular Steel Buildings in Seismic Zones

Abstract

Classical design procedures are less advantageous than performance-based seismic design (PBSD) of buildings, which is included in existing standards such as ASCE 41-23 for new buildings or retrofitting. PBSD requires accurate assessment of building seismic responses. Such assessments can be done using either faster nonlinear static procedures (NSPs) or more time-intensive nonlinear time-history analyses (NTHAs). However, the reliability of NSPs can be questionable, as shown by previous research. Practitioners need to conduct further investigations to determine safety margins and the applicability scope of these methods. This is especially important for irregular buildings and near-fault zones. This problem is investigated in this paper by first using 1250 single-degree-of-freedom (SDOF) systems to evaluate the ASCE 41-23’s coefficient method and performing 25000 NTHAs for near- and far-fault records. Second, the responses obtained from two alternative approaches, the modal pushover analysis (MPA) and FEMA 440’s capacity spectrum method (CSM), are compared with NTHA responses for buildings with significant higher-mode effects. American standards are used to design 96 3D symmetric and asymmetric steel moment-resisting frame (MRF) buildings with different characteristics such as lateral, lateral-torsional, and torsional modes of vibration dominance as well as different stability conditions, which are considered in this paper. The MPA and CSM are compared with NTHAs in this paper. The results show that the ASCE 41-23’s coefficient method is unreliable for near-fault zones and that the MPA and CSM are unreliable for seismic evaluation of buildings with dominant lateral-torsional modes of vibration or significant P-Δ effects. The results also revealed that MPA is a conservative approach for seismic evaluation of torsionally dominant buildings while CSM is not.