Prediction of Concrete Cover Separation in Reinforced Concrete Beams Strengthened with FRP

Fiber-reinforced polymers (FRPs) are widely used to enhance the flexural capacity of existing concrete structures subjected to various loading actions. However, when FRPs are applied to structural elements as externally bonded flexural reinforcement, premature debonding of the FRP may occur in the form of concrete cover separation (CCS). This study explores the potential of gene expression programming (GEP) for the development of a new model for the prediction of the shear force required to induce FRP CCS failure. A comprehensive database consisting of FRP-strengthened RC beams that failed by CCS was compiled and comprised of 127 experimental data points. GEP was employed to develop an accurate prediction model for CCS failure, encapsulating all the influencing parameters into a single equation. A comparative study was also conducted to assess the performance of the new GEP model against several existing analytical models from the literature. The proposed GEP model demonstrated significantly higher correlations between predicted and actual CCS failure loads when compared with all available predictive models.