Flexural strength prediction for concrete beams reinforced with FRP bars using gene expression programming

Abstract

FRP bars have been recently used as an alternative to the traditional steel bars in construction, especially in harsh environmental regions. Gene expression programming (GEP) is used in this research to develop a simplified model for predicting the flexural behavior of FRP reinforced concrete beams. An extensive and reliable database of 116 data points was collected from several experimental programs available in the literature to develop the GEP model. The model was established using six main parameters that predominantly control the flexural behaviour of the beams, which are beam width, concrete compressive strength, depth of the beam, FRP tensile reinforcement area, FRP modulus of elasticity, and FRP ultimate tensile strength. The accuracy of the model was then statistically evaluated, and a parametric study was conducted to check the sensitivity of the model to the input parameters. Finally, the performance of the GEP model was then compared to that obtained using the ACI-440-17 and the CSA S806-12 guidelines to further validate the model. The R-squared values of the three models are very high and are close to each other, indicating that the three models are accurate while the GEP model was the simplest one.