Evaluation of Seismic Design Methods for Steel Multi-Tiered Special Concentrically Braced Frames

Steel Multi-Tiered Concentrically Braced Frames (MT-CBFs) are commonly used in North America as a lateral load-resisting system of tall single-story buildings. Multi-Tiered configurations are typically used when a single braced panel between the roof and ground levels is impractical in the case of tall building. Past studies show that MT-CBF columns designed in accordance with the 2010 US Seismic Provisions are prone to buckling due to a high axial compression force and in-plane bending demands induced in columns as a result of non-uniform distribution of brace inelastic deformations along the frame height. Special design provisions have been introduced in the current US Seismic Provisions to address flexural demands imposed on MT-CBF columns and prevent column instability. Nevertheless, the recent improvements lack full-scale experimental testing and comprehensive finite element simulations featuring current guidelines. In this study, the seismic design methods for Multi-Tiered Special Concentrically Braced Frames are evaluated using the nonlinear finite element method. First, a two-tiered SCBF was designed in accordance with the 2010 and 2016 requirements. A detailed finite element model of the frames was then created and a non-linear cyclic-pushover analysis was performed to evaluate the seismic performance of both frames. Special attention was paid to the validation of the design forces prescribed in the 2016 Seismic Provisions. Analyses results confirmed that the inelastic deformations in the frame designed using the 2010 requirements are not uniformly distributed but rather concentrated in one of the tiers; whereas, the current design method significantly reduces the concentration of inelastic deformations in a single tier and prevents column instability. Furthermore, it was found that column in-plane flexural demands are overestimated when designed in accordance with the current provisions. Further numerical simulations on a large number of frame configurations using the dynamic response history analysis procedure are required to verify this observation.

View content