Predicting the brittle response of reinforced concrete (RC) members is a complex challenge. Different industries and tools provide varying accuracy and analysis times, and advanced finite-element (FE) tools such as Abaqus, Ansys and Diana require high computational costs and expertise. To overcome these issues without extra computational cost, a method combining an artificial neural network (ANN) and finite-element analysis (FEA) method is proposed. The proposed method was designed for the analysis and design of both new and existing RC structures, including multi-span beams. In this study, two experimental control model beams (CM-0 and CM-180) and four new FE models (models with half-diameter stirrups (HDN-0 and HDN-180) and models with double spacing of stirrups (DSN-0 and DSN-180)) were examined, where 0 and 180 represent the values of axial loads (in kN). The analysis assessed the impact of critical design parameters, specifically the transverse reinforcement ratio, on the load-carrying capacity of multi-span beams, particularly in brittle conditions. The results showed that the ANN–FEA model closely aligned with the experimental values and Abaqus results for the control models. For the other four models, both the ANN–FEA and Abaqus yielded similar results, while SAP2000 displayed uniform values regardless of the stirrup arrangements.
Ahmad et al. (Tue,) studied this question.