An integrated artificial neural network and Taguchi approach to optimize the squeeze cast process parameters of AA6061/Al2O3/SiC/Gr hybrid composites prepared by novel encapsulation feeding technique. (December 2020)

• Record Type: Journal Article
• Authors: Natrayan, L.; Senthil Kumar, M.
• Is Part Of: Materials today communications. Volume 25(2020)
• Publication Date: 2020-12
• Subjects: AMC Aluminium metal matrix composite -- ANN Artficial neural network -- HMMC hybrid metal matrix composite -- SP squeeze pressure -- MT Melt temperature -- DT Die temperature -- HT Holding time -- OA Orthogonal array -- HV Hardness -- TS Tensile strength -- SEM Scanning electron microscope -- EDS Energy dispersive X-ray spectroscopy; Hybrid metal matrix composites -- Squeeze casting -- Optimization -- Taguchi -- ANN; Materials science -- Periodicals; 620.11
• DOI: 10.1016/j.mtcomm.2020.101586
• Languages: English
• ISSNs: 2352-4928

Abstract

Graphical abstract: Highlights: A novel feeding technique used to fabricate AA6061/Al2 O3 /SiC/Gr hybrid metal matrix composites by squeeze casting method. Taguchi L16 orthogonal array designed the experiment. Uniform distribution and high mixing ratio was achieved via encapsulated feeding technique. ANOVA exposed melt temperature showed a significant contribution in improving the mechanical properties of hybrid composites. ANN predicted the results with 95 % accuracy. Abstract: Aluminum matrix composites known for their superior mechanical properties finds its use as liners in engine cylinders, discs, drum brakes, and pistons in automotive applications. This paper investigates the influence of squeeze casting process parameters on AA6061/Al2 O3 /SiC/Gr hybrid metal matrix composite using the encapsulated feeding technique. Four levels of factors selected for the L16 orthogonal array to optimize the process parameters were squeeze pressure (60, 80, 100, 120 MPa), melt temperature (700, 750, 800, 850 °C), die temperature (100, 150, 200, 250 °C) and pressure holding time (5, 10, 15, 20 s). Hardness and tensile strength were measured for the designed experiments. Scanning electron microscope with energy-dispersive X-ray spectroscopy identified surface morphologies and elemental analysis. Optimized results were predicted using the artificial neural network. The melt temperature and squeeze pressure exhibited a significant contribution in controlling the mechanical behaviour of the hybrid composites. Taguchi analysis suggested that SP3, MT2, DT4 and HT2 casting conditions presented the optimal process parameter level that showed the maximum hardness of 131 HV and the tensile strength of 329 MPa. Scanning electron microscopy and energy-dispersive X-ray spectroscopy showed uniform distribution of reinforcement in the encapsulation process compared with the regular feeding technique. ANN predicted the hardness and tensile strength with 95 % accuracy. Compared with the regression model and experimental data, the ANN prediction was more accurate. The defined hybrid metal matrix composite stands out as the substitute for AA6061 alloy that meets the demands of the modern automotive industry in engine cylinder liner applications.