Providing Models of the Compressive Strength of Square and Rectangular (S/R) Concrete Confined Using Genetic Programming

The accurate prediction of the compressive strength of FRP (Fiber-Reinforced Polymer)-confined is essential for structural engineers and designers. Several experimental studies have been conducted on concrete confined with FRP sheets. Different models in order to determine the compressive strength of FRP-confined concrete are provided in the previous researches. This study develops a practical model using genetic programming (GP) to reliably predict the compressive strength of FRP-confined concrete across various FRP types, enhancing its applicability for engineers. Firstly, a wide range of experimental data for square and rectangular (S/R) columns confined with a variety of FRP sheets has been collected (Including 463 specimens). 324 specimens (70 %) were used for modeling. For proposing models by using GP, the input and output variables were considered dimensionless. So input variables including b/h, r/b, r/h, r/tf, tf/h, Ff/ fco, and Ef/ fco and output is considered as fcc/fco. To present the model using GP, the three-transfer function set was selected. Finally, results compared with the existing models. The predictions of GP show satisfactory estimations, so that GP have averagely increased R2 approximately 9.91% rather than other models.