Estimation of FRP-To-Concrete Bond Strength of Carbon and Basalt Fabric Under Moisture Conditions: A Neural Network-Based Approach

The deterioration of reinforced concrete (RC) structures is the main issue throughout the globe. The researchers are persistently dedicated to resolving the issue of deterioration in RC structures. In the last few decades, fiber-reinforced polymer (FRP) has been continuously used to retrofit deteriorated structural members. The various advantages of FRP composites such as high-strength-to-low-weight, ease of use and transport, etc. increase the use of FRP composites in the retrofitting industry. The FRP-to-concrete bond strength is one of the main contributing parameters that decide the efficacy of the FRP composite with the concrete surface. Various analytical models are available to predict the FRP-to-concrete bond strength, but there is no analytical model that can estimate the FRP-to-concrete bond strength under moisture conditions. This study aims to evaluate the FRP-to-concrete bond strength under moisture conditions with an artificial neural network and analyse the durability of FRP composite under moisture conditions. The considered dependent parameters that affect the bond strength under moisture conditions were the width of the concrete block, bonded length of FRP composite, thickness of FRP composite, width of FRP composite, exposure type, temperature, relative humidity, duration, elastic modulus of FRP composite and compressive strength of concrete. The developed machine-learning model was accurate, fast, reliable, easy to use, and economical to predict the FRP-to-concrete bond strength under moisture conditions.