Artificial Neural Network Model for FRP Shear Contribution of RC Beams Strengthened with Externally Bonded FRP Composites

摘要

Fiber reinforced polymers (FRP) have been widely used in retrofitting or strengthening concrete structures in recent years. This is due to the various advantages of FRP composites, including high strength-weight ratio, high tensile modulus, superior corrosion resistance and ease of applying in strengthening applications. One of the most important applications of FRP is shear strengthening for reinforced concrete (RC) beams. In this paper, two different artificial neural network (ANN) models are proposed for predicting the FRP shear contribution of RC beams strengthened in shear with U-wrapping FRP sheets with/without additional anchorage system, respectively. To verify the accuracy of the ANN models, the predictions from five existing design guidelines are applied to comparison. It is found that the proposed ANN models can improve the accuracy of predicting the shear contribution for U-wrapping FRP configuration whether with anchorage or without one.