Probabilistic assessment of the rules used to combine peak floor responses of special concentrically braced frames under orthogonal seismic effects

Abstract

AbstractEstimating seismic demands and the associated damage to nonstructural components (NSCs) is a critical step in the performance‐based earthquake engineering (PBEE) framework. To facilitate this type of assessment, NSCs are classified as being acceleration, velocity, or displacement sensitive. When two‐dimensional structural models and response history analyses are used in the PBEE framework, heuristics are used to combine the orthogonal responses. Although there has been significant research to develop and evaluate combination rules for structural component responses, comparatively much less attention has been placed on NSC demands. FEMA P‐58 does provide a combination rule specifically for NSC seismic demands; however, the basis and accuracy of these provisions have never been examined. In this paper, the effectiveness of three popular combination rules used to estimate peak floor acceleration (PFA) and peak floor velocity (PFV) is investigated. Incremental dynamics analyses under unidirectional and bidirectional record‐sets are performed on six special concentrically braced frame buildings with different heights and torsional irregularity. The associated floor responses are used to examine the efficacy of the combination rules. Subsequently, a probabilistic approach is proposed to enhance the accuracy of combination rules. The performance of the enhanced combination rules is compared to that of a surrogate model that estimates PFA and PFV under orthogonal seismic effects. The results show that the performance of existing combination rules is slightly affected by building height and plan irregularity. Also, the combination rule suggested by FEMA P58 is found to overestimate the PFA and PFV demands by 10–15% compared to the demands derived from bidirectional response history analysis.