An artificial neural network model for estimating the shear contribution of RC beams strengthened by externally bonded FRP

This paper provide an artificial neural network model for prediction the shear contribution of FRP in reinforced concrete (RC) beams strengthened in shear with externally bonded FRP. Although there are some models and equations for estimating the contribution of FRP, these models in some cases have significant error in calculation of FRP contribution. One of reasons for these errors is neglecting the effect of shear span (a) to effective depth of beam (d) ratio in FRP performance. In this model mechanical and dimensional properties of RC beams strengthened and strengthening materials and also the shear span to effective depth of beam ratio (a/d) are taken as input parameters and the shear contribution of FRP is the target of the network. After a comprehensive review in existing literature, 96 strengthened RC beams which all of them have FRP rupture failure mode were selected which 92 of them were used for training, validation and testing the network and four of them were used for controlling the generalization of network. Finally the outputs of model have compared with ACI 440.2R, fib 14 and CIDAR guidelines and the result indicated that the ANN model is more accurate than the existing guideline equations based on experimental result.