The reliability of seismic evaluation depends mainly on the ability of the analytical model to capture the structural member response under earthquake excitations accurately. In concentrically braced frames (CBFs), the hysteretic response of braces dominates the seismic behavior of the structural system. The ability to predict the inelastic brace response and its fatigue life, therefore, is an essential constituent of a dependable seismic assessment approach. In fact, when the initial design requirements for ductile concentrically braced frames were developed, and first introduced in the seismic provisions, a fatigue life prediction, predominantly for square hollow structural section (HSS) braces, was considered in the seismic response analysis of CBFs. Ever since, particularly with the rapid evolution of the design practice, several seismic studies on CBFs proposing different brace fatigue fracture life predictions have striven to quantify the impact of brace fracture on the system performance accurately and more realistically. The intention has been to comprehend the member behavior of ductile CBFs as well as to address explicitly the important issues related to design. However, the variation in the predictions of brace fracture could lead to substantially different conclusions on the post-fracture response, and thus, different seismic design recommendations.

The research presented in this dissertation focused first on evaluating current brace fatigue life predictive models by comparing them with a comprehensive database of laboratory-tested steel braces and inspected their reliability by post-fracture seismic analyses. The study finds that existing ductility-based empirical models and strain-related fiber models were unable to predict fractures observed in laboratory tests. Accordingly, extensive research was undertaken to develop a practical, yet reliable, fatigue life model to predict the fracture life of square HSS accurately. In this research, a three-stage fatigue process concept was proposed, capable of predicting fracture initiation, fracture propagation, and partial or complete separation of bracings made of square HSS. The fracturing process concept is consistent with the fracture observed during experimental tests, and it provides a much-improved tool for numerical simulations of the post-fracture response of CBFs. Further, using the developed fatigue life prediction, the post-fracture seismic demands, for the first time in the literature, on column splices in CBFs was evaluated. This research finds that splices in ductile CBFs are vulnerable to undergo inelastic deformations when brace fracture occurs, as oppose to the intended design objective. The fundamental findings of this dissertation imply the need for mitigating the seismic risk of braces in existing CBFs along with more stringent requirements on the design and detailing of new braced frame buildings.