Numerical study of the cyclic behavior of a proposed all-steel brace with reduced fuse length

Concentrically braced frames are among the prevalent seismic force-resisting systems used in the construction of steel structures. This type of system provides a suitable level of stiffness for structures under low and intermediate seismic oscillations. However, under strong motions, it has noticeable deficiencies such as stiffness loss under compressive force, the unacceptable difference between the tensile and compressive strength of the brace, low energy-dissipating capacity, and overall poor cyclic behavior. To overcome these deficiencies, the idea of the Buckling Restrained Brace (BRB) was proposed a few decades ago.  Since the invention of BRB, extensive studies have been carried out to optimize the new brace system. These studies have resulted in the emergence of different generations of buckling restrained braces. In the first generation of BRB, a concrete-filled sheath had been used around the inner core of the brace. To upgrade that heavy brace, the researchers developed an all-steel brace system that was considerably lighter in weight, faster to build, and easier to inspect its yielded core after an earthquake. Later on, the idea of reducing the length of the core, as well as the sheath, was proposed which led to an even lighter brace, while keeping all the major advantages of the traditional BRB. In this paper, twelve all-steel BRB samples, based on a reduced fuse length, have been investigated numerically. Each brace sample is composed of three boxes, which include the main box, the outer sheath, and the inner box. The outer sheath and the inner box are used to prevent the local buckling of the core in the fuse zone. The outer sheath and the inner box are connected to the brace core at one end only. In this study, the cross-sectional area of the brace core in the fuse zone was considered to be less than half the total cross-sectional area of the original brace section. The samples were loaded by the quasi-static cyclic loading protocol of AISC. The numerical analysis showed that the proposed brace withstood an axial strain level of around 4%. The numerical modeling of the proposed brace was verified by the data reported for an earlier experiment that had been carried out in the laboratory of the Housing and Urban Development Research Center (BHRC). In the numerical study, the effect of influential parameters of the proposed brace on its cyclic behavior was investigated. These parameters included the ratio of the fuse length to the total brace length, the gap between the core and the inner/outer boxes, the inner/outer box thickness, and the friction coefficient between the core and the contact surfaces of the boxes. Using the hysteretic curves of the brace, obtained from the numerical analyses, the ductility parameters, and the amount of dissipating energy were evaluated. The results showed that the obtained amount of the relative lateral displacement of the proposed brace is acceptable according to the code regulations. Moreover, the cumulative inelastic deformation of the proposed brace surpasses the minimum requirement of the code for the predefined loading protocol. The studied samples were stable and had relatively symmetric cyclic behavior in the compression and tension zones. The study showed that the proposed bracing is suitable for the rehabilitation of buildings.