Graphical Representations of Experimental and ANN Predicted Data for Mechanical and Electrical Properties of AlSiC Composite Prepared by Stir Casting Method

Babalola, P.O. and Bolu, Christian and Inegbenebor, A.O. and Kilanko, O. O (2018) Graphical Representations of Experimental and ANN Predicted Data for Mechanical and Electrical Properties of AlSiC Composite Prepared by Stir Casting Method. In: IOP Conf. Series: Materials Science and Engineering.

PDF Download (692Kb)

Abstract

Artificial Neural network is a field of man-made intelligence that is able to undertake design prediction, mechanical property forecast, and process selection. In this paper, Aluminium Silicon Carbide composite was developed by reinforcing aluminium metal with silicon carbide powder using stir casting method. The produced aluminium matrix composites (AMC)were subjected to tensile, hardness and electrical tests to obtain tensile extension (mm), load (N), modulus (N/mm^2), yield strength (MPa), hardness (HV), ultimate tensile strength (MPa), tenacity at fracture (gf/tex), time at fracture (s), hardness (HV), conductivity(MΩ/m), and tensile stress (MPa) data. Artificial Neural Network (ANN) was then used to train, test, and validate the obtained experimental data and then predict new set of data. The experimental and ANN predicted data were represented using graphical illustrations. The results showed that ANN could be used to replace rigorous, costly and time consuming experimental exercise with minimal loss in accuracy.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	AlSiC composite, artificial neural network (ANN), mechanical properties, electrical properties, stir casting
Subjects:	T Technology > T Technology (General) T Technology > TJ Mechanical engineering and machinery
Divisions:	Faculty of Engineering, Science and Mathematics > School of Engineering Sciences
Depositing User:	Mrs Hannah Akinwumi
Date Deposited:	21 Sep 2018 11:52
Last Modified:	21 Sep 2018 11:52
URI:	http://eprints.covenantuniversity.edu.ng/id/eprint/11877

Actions (login required)

View Item