Artificial Artificial Neural Network Modelling Approach to Predict the Effect of CFRP Composites on the Axial Strength of Rectangular and Square Columns

Abstract

The purpose of this study is to offer an analytical model based on artificial neural networks (ANN’s) that can predict axial strength of Corbon fibre reinforced polymers (CFRP)-confined concrete columns. The novel aspect of the suggested model is the way an analytical relationship is formulated without taking into account the conventional effectiveness parameter that is frequently present in models that are published in the literature. Using an experimental database from the literature, the ANNs were trained and evaluated. The database held various input parameters, such as the cross-sectional details, corner radius, initial compressive strength of concrete, longitudinal reinforcement ratio, Reinforcement tensile strength, , number of layers of CFRP, thickness of CFRP layers, and tensile e-modulus of CFRP, in addition to that one output parameter representing the final axial capacity of concrete Columns. It is advised to use the suggested model for rectangular columns that have continuous Unidirectional CFRP wrapping. The testing of model indicates the predictions' correctness, and an experimental vs theoretical comparison verifies their accuracy. The findings show that the suggested model is accurate and is suitable for the design of CFRP-confined concrete.