Alternative designs of CBs 

Many alternative designs of CBs were proposed to increase the deformability and energy dissipation capacity of CBs, and particularly to suppress the shear failure. A review of proposed design methods including steel and composite CBs, as well as CBs with highperformance fiber-reinforced cement composites, is provided in the following sections.

Steel CBs and steel-concrete composite CBs 

To achieve ductile performance of CBs, steel and composite CBs may provide a viable alternative to reinforced concrete CBs. The steel CB would be efficient particularly where the use of deep reinforced concrete or composite CBs are not an option due to the height restrictions, or where a conventionally RC CB do not satisfy the required capacity and stiffness.

Based on appropriate behavior of steel link beams in eccentrically braced frame in ductility and energy absorption ability, Harries (1995) suggested steel CBs with their ends embedded in the walls. Therefore, three shear critical steel CBs and one flexure critical steel CB were tested under reversed cyclic loading. The test results indicated that steel CBs exhibited excellent energy dissipation and stable hysteretic response loops. It was found that the shear critical steel CBs exhibited a more ductile behavior and a better energy absorption feature compared to RC CBs. However, in shear critical specimen without stiffeners in embedded region, insufficient shear and local buckling resistance in the embedment region caused high concentration of compressive stress at the wall faces and inelastic deformation in which both shear yielding and web crippling occurred. The embedment length had great influence on the performance of steel CBs. The embedment length was calculated using the method proposed by Marcakis & Mitchell (1980) assuming a rigid body motion of the embedded steel section.

Teng et al. (1999) proposed concrete-filled rectangular tubes (CFRTs) as an alternative design of CBs with high ductility and energy absorbing capacity. Eight cantilever beam specimens consisting of two control rectangular hollow section tubes and six CFRTs were tested under static loading and reversed cyclic loading. The experimental results of four rectangular tubes, under cyclic loading indicated that the one without concrete infill had low ductility and rapid strength degradation and it failed by local buckling of both flanges. In contrast, the other CBs with concrete infill had higher ultimate strength and they failed through tensile cracks in flanges. However, steel-concrete slipping or formation of shear cracks due to concrete deterioration may cause strength and stiffness degradation. Therefore, the maximum load reached in a cyclic test was considerably lower than that in the corresponding static test. Overall, the CFRT CBs are suitable for seismic resistance if local buckling is minimized by using thick steel plate and slipping between the steel and concrete is reduced by using shear connectors.

Gong & Shahrooz (2001) carried out an experimental test on seven steel-concrete composite CB specimens to investigate the effects of different parameters including: effects of encasement, amount of web stiffener in the steel beam, presence or lack of face bearing plates at the wall-beam interface, level of shear force, and floor slab around the CB. It was observed that web buckling and flange instability could be prevented by encasement around steel CBs so that the use of web stiffeners is not required. In order to increase the beam ductility, the stiffness may also be increased with the use of reinforced encasement. However, the concrete encasement would generate a 25% extra strength and stiffness which would thereby lead to over coupling and hence larger forces in the walls. Consequently, the sequence of failure may change and be undesirable. Despite the effect of floor slab in increasing the stiffness at initial stage of loading, its contribution is lost after small deformations.

Although various alternative forms of CBs had been proposed, Lam et al. (2001) pointed out that none of them satisfied the demands on high deformability, good energy dissipation, easy construction and minimum disturbance to slab and wall detailing. Therefore, Lam et al. (2001) proposed steel composite CB in which shear studs are welded on the top and bottom of both sides of the plate in order to improve the horizontal shear transfer and bonding of steel plate and concrete. To evaluate the efficiency of the proposed method, Lam et al. (2005) studied the performance under reversed cyclic loading of three CBs with the span length-todepth ratio of 2.5 and identical dimensions. The reinforcement layout of one of the CBs was conventional whereas the others contained a vertically embedded steel plate without and with shear studs. The specimens were tested in vertical position under reversed cyclic loading. The experimental results indicated that embedded steel plates improve the shear strength and the stiffness of CBs. Adding shear studs enhanced the plate-reinforced concrete interaction and resulted in a good inelastic performance under large imposed shear deformation. An equation was also proposed to calculate the required strength for shear connection in the beam span to ensure the plate/RC composite action and to estimate the available plate/RC interface slips to mobilize shear studs.

Park & Yun (2005) conducted an experimental test of three steel CB specimens with three different failure modes (connection failure, shear critical failure and flexural critical failure) under reversed cyclic loading. The main test variables were the ratios of the CB strength to the connection strength. The test results revealed that steel CBs with shear yielding failure exhibited better characteristics of energy dissipation than the flexure critical CBs. Severe web buckling in the clear span of the steel CB led to its final rupture. Furthermore, an analytical study was carried out to develop a model for computing the embedment lengths of embedded steel sections taking into account the contribution of the auxiliary bars and the horizontal ties.

Su et al. (2006) investigated the importance of shear connectors on plate-reinforced concrete composite action (PRC) through experimental studies on five PRC CBs. The PRC CB specimens were in medium-length (l/h = 2.5) and short length (l/h = 1.17), containing a vertically embedded steel plate. One short beam was welded with expanded metal meshes on the plate surfaces and the others were welded with shear studs on the plates in the wall regions and/or the beam spans. The specimens were tested under reversed cyclic loading and the results indicated that the expanded metal meshes was not effective in the behavior of CB. The shear studs in the wall region rather than in the span, significantly improved the inelastic beam performance. The experimental results of three deep PRC CBs were compared to the obtained results by Kwan and Zhao (2002) for two conventionally and diagonally reinforced CBs (Su et al., 2009). It was found that the performance of PRC deep CBs with properly designed plate anchorage in the wall regions enhanced the shear capacity without causing the problem of steel congestion as in diagonally reinforced CB. However, poor anchorage of the plates caused brittle failure and prevented the development of full strength in a deep PRC CB. Furthermore, based on the experimental results and numerical studies (Su et al., 2008) on the behavior of PRC CBs, a design procedure was proposed by Su & Lam (2009).

Fortney et al. (2007) proposed a fuse steel CB based on the desirable behavior of CSWs in which the hinges should be propagated in the span of CBs and not into the wall connections. In the design of fuse CB, it is assumed that all inelastic deformations would be concentrated in the middle section (fuse section) of the beam. The fuse section was connected to the main sections of steel beams by slip critical bolted connections. To ensure that all inelastic deformations and damage were concentrated at the fuse, the main sections of CBs should have greater shear capacity than that of the fuse section. Experimental results of fuse steel CBs demonstrated good ductility and energy dissipation ability as steel CBs. Moreover, the damaged fuse could be replaced after a seismic event.

Motter et al. (2012) tested two large-scale composite CB specimens with different embedment length. Using the embedment models of Marcakis & Mitchell (1980) and Mattock & Gaafar (1982), the embedment length was calculated so that it represents a conservative design for the first CB and 75% of it was set for second one. The wall loads were 2.5 times greater for the second specimen than the first one, reaching nearly yielding at the embedment zone. The experimental results indicated that considerable outward slip of the steel section occurred due to lack of a specific prototype. Because of the difference of embedment length, the second beam specimen had less capacity and featured high pinching compared to the first one which showed less strength degradation and an enhanced ductility.