Enhanced Neural Evalution of Steel Fibre Reinforced Rubberized Concrete Beams

The present study investigates the structural behaviour of steel infused fiber-reinforced rubberised concrete beams, integrating analytical and experimental methods to improve their strength and sustainability. Novelty: Rubberised concrete, which contains recycled waste conveyor belt rubber, has positives for the environment but loses strength. To overcome these barriers, the usage of steel fibre reinforcement is investigated further, and experiments are carried out to evaluate its flexural strength of beams. The outcomes were fed into a model of an artificial neural network that used Scaled Conjugate Gradient, Levenberg-Marquardt and Bayesian Regularisation algorithms with back propagation. The input variables were the transverse section of the beam, span, strength of concrete, yield strength of steel, area of reinforcing steel, and the modulus of elasticity of concrete. The output variables were ultimate load, deflection at ultimate load. The test data required for ANFIS modeling were obtained from relevant literature. Various statistical measures, like Root Mean Square Error, R2 and MAPE have been obtained. Scatter plot have been drawn for validation purpose. The results show that the structural performance of rubberized concrete beams are greatly enhanced by the inclusion of steel fibers, This study emphasises the use of artificial neural network (ANN) models for performance, as well as the potential of steel fiber-reinforced rubberised concrete as a sustainable building material.