Application of Fiber-Reinforced Polymer (FRP) sheets for external strengthening of concrete structures is a common practice nowadays. One of the important parameters that affect the performance of FRP strengthening is the bond strength between FRP sheets and concrete surface, which is more important at the joint locations. Various relations have been proposed for estimating this bond strength in which the effects of different parameters such as width of the structural member cross-section, compressive strength of concrete, modulus of elasticity of concrete, width of FRP sheet, and FRP modulus of elasticity, are considered. In some of the available relations, in addition to the above-mentioned parameters, the ultimate bond strength is predicted as a function of the local bond stress. Accordingly, a parameter called width factor is included in the relations. Unfortunately, in spite of relatively large number of existing predictive models, the accuracy of most of these models is not satisfying yet. Therefore, the main objective of the present study is to design an Artificial Neural Network (ANN) in order to provide highly accurate estimations for the bond strength between FRP sheets and concrete members. For this purpose, a database of experimental results collected from the previously published researches is used. Comparison of the derived ANN with some of the existing models proves its higher accuracy.

Keywords