عنوان مقاله [English]
Reinforced concrete (RC) members with significant geometric discontinuities and complex stress distributions under loading require considerable analyses and usually complicated reinforcement detailing. Deep beams with large openings are one of the examples. Strut-and-tie models (STM) are a valuable tool for the design of these members. However, the actual stress fields in deep beams with openings are typically different compared with those predicated by STMs, as reported by many experimental investigations. Therefore, an optimization method on homogenization is used for finding optimal strut-and-tie models for RC elements. Previous truss optimization methods are strongly dependent on the element removal ratio and the FE mesh. Additionally, they require many algorithmic parameters to be finely tuned, and their convergence to optimal solution is uncertain. Full homogenization (FH) methods solve most of these problems because of their explicit formulas. Two examples from STM literature are used to illustrate the application of these methods.In the experimental part of the study, two reinforced concrete and two steel fiber reinforced concrete (SFRC) deep beams with opening were manufactured and monotonically loaded up to failure. The design of each test specimen was carried out using optimal truss models and the ACI 318-11 provisions for STM. The structural performance of the specimens is compared with that of some similar specimens in previous test series. All test specimens in this study carried loads greater than the factored design loads. Reinforced concrete specimens using optimization-derived truss models have efficient responses, in terms of ratio of ultimate load to steel weight.