Predicting shear capacity of Recycled Aggregate Concrete beams using Artificial Neural Network

This study investigates the application of an Artificial Neural Network (ANNs) utilizing a Multi-Layer Perceptron (MLP) architecture to predict the shear capacity of Recycled Aggregate Concrete (RAC) beams. The ANNs model was trained using the Levenberg-Marquardt algorithm with a comprehensive dataset comprising 232 experimental shear tests, reflecting a wide range of variables relevant to RAC beam performance. The model's predictions were compared to those derived from established design standards, including ACI 318-14 and Eurocode 2, to evaluate its performance. Various statistical criteria were employed to assess the model’s accuracy and reliability in predicting shear strength, including metrics that measure goodness of fit, error rates, and predictive consistency. This research highlights the growing potential of machine learning techniques in civil engineering, particularly for enhancing the precision of shear strength predictions for RAC beams. The findings suggest that the ANN model offers a valuable alternative to traditional prediction methods, with the potential to improve accuracy and address some of the limitations inherent in conventional design standards.