Prediction of the Lateral Confinement Coefficient of The concrete Columns Confined by FRP using the Artificial Neural Network

Confining columns is one of the most commonly used methods of retrofitting structures. The confinement of concrete columns by FRP sheets is considered as one of the modern methods of retrofitting structures due to their properties. Tests conducted on a rounded-corner column by applying axial compression indicate that its behavior improves as well as the cross-section of column approaches the circular section. However, the behavior of columns confined with FRP or having transverse reinforcement has been studied by many researchers, yet study on wrapping columns by FRP when the transverse reinforcement does not satisfy the required confinement for the column has not been investigated by researchers. In this study, different models of the lateral confinement coefficient are presented. Also, by considering valid test results and the influence of different parameters such as length and width of column section, thickness of FRP sheet, the compressive strength of concrete, elasticity modulus of FRP and, radius of rounded-corner concrete a new model is presented using the Neural network in which the lateral confinement coefficient of confined columns can be predicted with high accuracy. Finally, the sensitivity analysis is carried out to evaluate the effect of each input parameter on the output parameter, and the results are presented.