Impact of brace fracture modeling on seismic performance evaluation of braced frames

Abstract

Concentrically braced frame is prone to suffer brace fracture when subjected to design earthquake ground motions, and their seismic response evaluation is critically related to how brace fracture is modelled in the evaluation process. Seismic provisions are based on seismic analyses with brace fracture models developed with limited amount of test data. Recent studies found that these brace fracture models failed to predict fracture of the braces tested in structural laboratories over the last 40 years. These models may significantly overestimate, or underestimate tested brace fracture capacity. The conclusions of seismic performance evaluation studies strongly related to fracture models used in individual studies. The conclusions made in the first study in 1990 were the base of current seismic provisions. Would the conclusions be altered if different fracture models were used? This appears to be the most fundamental question for seismic design of CBF and surprisingly has not been addressed over more than four-decades long. This manuscript investigates the impact of brace fracture modeling on the output of seismic performance evaluation. Fracture ductility ratio was considered the most sensitive and uncertain modelling index. The manuscript concludes that seismic responses of CBF vary dramatically with small changes in brace fracture ductility assumptions, leading to unreliable evaluation of seismic performance. It is recommended that the seismic provisions for CBF based on single brace fracture models be revisited with more comprehensive fracture model that has been recently developed and shown superiority over all other brace fracture models used in current practice.