Design of Deep Beams Using Strut-and-Tie Models-Part II: Design Recommendations

The findings of the first part of this study showed that the use of current strut-and-tie modeling provisions of both ACI 318-05 and AASHTO Load and Resistance Factor Design (LRFD) can lead to an unconservative calculation of nominal capacities for some test specimens, with neither producing levels of safety that meet a 5% exclusion rule. This paper develops new provisions based on the lower bound to a database of 596 published test results. These new provisions incorporate the effects of strut inclination, concrete strength, geometry of a bottle-shaped strut, and shear reinforcement while increasing levels of conservatism relative to ACI 318 and AASHTO LRFD provisions. The equilibrium of the bottle-shaped strut was considered to determine the amount of required reinforcement. Struts were separated into two classes: those with sufficient reinforcement and those without sufficient reinforcement. Lower bounds were determined to the data compiled from tests of both sets of struts. When calculating the normal capacities of the specimens in the database with the new design provisions, 96% of the specimens had measured loads in excess of the nominal capacities. The new provisions satisfy the 5% exclusion rule.