An Artificial Neural Networks model for the prediction of the compressive strength of FRP-confined concrete circular columns

Nowadays, Fiber Reinforced Polymers are extensively applied in the field of civil engineering due to their advantageous proprieties such as high strength-to-weight ratio and high corrosion resistance in aggressive environments. It is well-known that the compressive strength of concrete significantly increases if a lateral confining pressure is provided. The present paper aims to present an analytical model, able to predict the strength of FRP-confined concrete, which is based on Artificial Neural Networks. The innovation of the proposed model consists of a formulation of an analytical relationship that does not consider the traditional effectiveness parameter commonly found in the models presented in the literature. An extensive experimental database was used to define the variables of the proposed equations. The proposed model is recommended for circular columns with continuous FRP wrapping. The validity of the predictions is indicated through a parametric study and the accuracy is tested by an experimental versus theoretical comparison. An additional comparison is shown by considering the theoretical predictions obtained from the proposed model and the outcomes of equations adopted by important international design codes. The results evidence that the proposed model is adapt for the design of FRP-confined concrete and guarantees an improved accuracy with respect the available competitors.