Experimental probe on machining attributes of Al 7075-T6/SiC/crumb rubber/MoS2-based green hybrid composite using artificial neural network model

Abstract

Machinability of hybrid metal matrix composite is quite difficult because of an anisotropic nature, abrasive effect, and nonhomogeneous arrangement of two (or more) hard reinforcements, which are infused in an unreinforced base alloy. Therefore, this paper proposes an experimental probe on machining attributes of the novel green hybrid metal matrix composite which is prepared by an advanced vacuum-sealed bottom pouring stir casting technique with Al7075-T6 as a base matrix and SiC/crumb rubber/MoS2 as three distinct reinforcements. The variable input parameters such as SiC content (3.5 and 4.5 wt.%), crumb rubber content (0.3, 0.6, and 0.9 wt.%), MoS2 content (3.5, 4.5, and 5.5 wt.%), stirring speed (580, 600 and 620 rpm), stirring time (2, 5 and 8 minutes), and pouring temperature (670, 700, and 730°C) are selected for the fabrication of hybrid composites via Taguchi L18 mixed-level orthogonal array. For the machinability of fabricated composites in terms of material removing rate and surface roughness under dry conditions, a conventional turning operation on a lathe machine was opted for fixed input parameters such as spindle speed (200 rpm), depth of cut (1.0 mm), feed rate (0.25 mm/rev), and machining time (30 seconds). In contrast to an unreinforced base alloy, the material removing rate of the proposed green hybrid composite is slightly increased by 1.47%, whereas surface roughness is significantly decreased by 12.14%. Further, these experimental outcomes are validated with the model created using an artificial neural network, which is based on Levenberg–Marquardt algorithm. Besides, the predicted values obtained from the artificial neural network model lie in the close vicinity of the experimental outcomes, which in turn ensure the authentication of the developed artificial neural network model for the machined aluminum-based green hybrid composites.