Modeling confinement efficiency of reinforced concrete columns with rectilinear transverse steel using artificial neural networks

Artificial neural networks have attracted considerable attention and have shown promise for modeling complex nonlinear relationships. This paper explores the use of artificial neural networks in predicting the confinement efficiency of concentrically loaded reinforced concrete (RC) columns with rectilinear transverse steel. Fifty-five experimental test results were collected from the literature of square columns tested under concentric loading. A multilayer-functional-link neural network was used for training and testing the experimental data. A comparison study between the neural network model and four parametric models is also carried out. It was found that the neural network model could reasonably capture the underlying behavior of confined RC columns. Moreover, compared with parametric models, the neural network approach provides better results. The close correlation between experimental and calculated values shows that neural network-based modeling is a practical method for predicting the confinement efficiency of RC columns with transverse steel because it provided instantaneous result once it is properly trained and tested.