Optimization of mechanical properties of Al7150/Si3N4/C composites using artificial neural network

Abstract

The study aims to investigate and predict the effect of reinforcements such as silicon nitride (Si3N4) and graphene (C) in aluminum 7150 matrix. Al7150/Si3N4/C hybrid composite is fabricated by a stir casting technique and subsequently T6 heat treated for applications such as body stringers, spar chords, seat tracks, and stringers of wing surfaces of aircraft. A feedforward propagation multilayer neural network was developed for modeling and prediction of hardness, tensile strength, and tensile elongation. The results show that the addition of fillers and T6 heat treatment enhances the mechanical properties of the Al7150/Si3N4/C composite. The artificial neural network (ANN) model suggested for Al7150 composites demonstrates beneficial results when compared to experimental measurements. The prediction model, which has a mean absolute percentage error of 0.64%, 0.3%, and 2.49% for hardness, tensile strength, and tensile elongation can accurately predict the effect of reinforcement contents and T6 heat treatment on mechanical properties of Al7150/Si3N4/C composites.