Highlights • The current study explored the behaviour of RC continuous deep beams when only the stress paths from the STM in ACI 318M-14 (ACI 2014) are reinforced as compressive and tensile members and the concrete from outside the STM is removed. In this case, the proposed specimens maintained the theoretical ultimate capacity from the STM in ACI 318M-14 while reducing the weight. Generally, the behaviour of the proposed specimens was similar to that of their reference specimens because the stress transfer paths from the loading points to the supporting points were those defined from the STM in ACI 318M-14. Accordingly, based on the experimental work presented in the current study, the following conclusions can be drawn: • The predominantly linear load-midspan deflection trend of the proposed specimens was similar to that of the conventional reference specimens. Until the development of the first cracks, all specimens had roughly equal initial stiffness and deflections, indicating that initial stiffness is dependent on the stress paths from the STM regardless of the concrete being omitted in the proposed specimens. The specimens exhibited lower stiffness and capacity under concentrated loading than under uniform loading. This can be attributed to the bottle-shaped strut that occurred under concentrated loading, whereas in the case of uniform loading, the strut was a fan type, which was not identically modelled by the proposed specimens. • In the proposed frame specimens, omitting the concrete that STM does not take into consideration led to transforming the type of strut from bottle or fan to prismatic one. Although the ACI 318 code, Table 23.4.3 deals with the two types of struts in a different way in terms of the strut coefficient βs, the experimental results did not record significant differences. • The proposed specimens exhibited approximately 7–26% higher loads than the ACI 318M-14 STM theoretical design load. This superiority ensured that they can be good alternatives for the reference specimens: compared to the reference specimens, the proposed specimens reduced the weight by approximately 22–40%.

Abstract This work proposed reinforced concrete two-span continuous deep beams by replacing the well-known conventional reinforcement by reinforcing struts and ties only and omitting the concrete where the struts and ties do not pass through. The strut and tie modelling recommended by ACI 318M-14 was adopted herein. The struts and ties were reinforced as compressive and tensile members, respectively. Twelve specimens, which were divided into four groups, were cast and tested. In every group, the first specimen was a conventionally reinforced reference beam, the second specimen was a specimen in which only the stress paths of the struts and ties were reinforced, and the third specimen was the proposed frame. The strut-tie angle, specimen width and loading type were taken into consideration as main parameters. Generally, the proposals gave less load capacity than the reference specimens, but they remained higher than the design load. In addition, the proposed specimens significantly reduced the weight.