Elevated Temperature Effects on FRP–Concrete Bond Behavior: A Comprehensive Review and Machine Learning-Based Bond Strength Prediction

Because of their improved properties, FRP composites are vastly used in the strengthening of aged concrete infrastructures. However, it has been observed that their performance is highly compromised when exposed to high temperatures, as expected during fire incidents, which critically affects FRP–concrete bond behavior, hence affecting the overall efficiency of the strengthening system. This paper critically presents the available literature concerning the degradation of bond strength between FRP systems with concrete substrates due to increased temperatures. Both analytical and numerical bond–slip models developed for the prediction of bond strength degradation under such conditions are reviewed. A generally confirmed fact is that exposure to high temperatures, especially those reaching glass transition temperature (Tg) for epoxy adhesives, leads to bond degradation. Therefore, cement mortar-bonded CFRP textiles display better performance in fire endurance. This present paper also utilizes machine learning algorithms for the prediction of bond strength under elevated temperatures based on an experimental database of 37 beams. The nonlinear relationships and variable interactions in the developed model provide a reliable method for the estimation of bond strength with reduced extensive experimental testing, where the critical role of temperature in bond behavior is identified. This paper emphasizes the use of advanced predictive models to ensure the durability and safety of FRP-strengthened concrete structures in thermally challenging environments.