Collapse risk of controlled rocking steel braced frames considering buckling and yielding of capacity-protected frame members

摘要

The response in the higher modes of controlled rocking steel braced frames (CRSBFs) significantly increases the frame member forces during earthquakes relative to those from a first-mode pushover analysis. Previous research has generally assumed that these large demands must be addressed either by designing frame members for the full elastic force demands or by mitigating the higher mode response by providing multiple nonlinear mechanisms. However, the minimum required design forces for structural elements have not been assessed through collapse fragility analysis. To address this need, this paper investigates the influence of member buckling and yielding on the collapse capacity of 3-storey, 6-storey and 12-storey buildings. For each building height, five frames are designed based on multiplying the estimated higher-mode forces by a different amplification factor, defined as ?HM, and adding these to the forces expected from a first-mode pushover analysis, with ?HM = 0,1.0, 1.5, 2.25, and 3.0. The collapse performance for each design is evaluated using multiple stripe analysis for conditionally selected ground motions considering both the first-mode and second-mode periods independently and using a model in which frame member buckling and yielding are included. Neglecting the higher-mode forces (i.e. using ?HM = 0) for design is only acceptable for the 3-storey and 6-storey buildings when a response modification factor of R = 8 is used. Increasing ?HM reduces the collapse risk, and designing for ?HM = 3.0 yields similar collapse probabilities compared to when frame member buckling and yielding are not modelled. Based on the results of this study, using ?HM = 1.0 is recommended for reliable collapse prevention during a 2%-in-50-year event, which represents a reduction of up to 50% relative to the elastic design forces that have been recommended by a variety of authors in previous studies.