Author : Santosh Tamang and M.Chandrasekaran

Metal matrix composites (MMC) are widely used for producing components in automotive, aerospace and bio-medical industries because of their improved properties in comparison with non reinforced alloys. These materials are known as difficult-to-machine materials because of hard abrasive particles were used for reinforcement. Among these, the aluminium based silicon carbide particulate (Al-SiCp) reinforced MMC have received more commercial attention due to their high performance. In this work, machinability study and predictive model development was carried out for machining Al-SiCp MMC using polycrystalline diamond (PCD) tool. Turning experiments based on full factorial design (33), a total of 27 machining trials are carried out to study the effect of turning parameters viz., spindle speed (N), feed (f) and depth of cut (d) on the responses such as surface roughness (Ra) as product quality and material removal rate (MRR) as productivity improvement in the machining process. Multi response predictive modeling has been developed using artificial neural network (ANN). The ANN architecture having 3-6-2 is found to be optimum average percentage error of 4.46% for surface roughness and 7.26 % for material removal rate. The predictive model exhibit close correlation with the experimental result as confirmed by the validation test. The methodology found to be effective tool and can be developed with minimum effort.

Keywords: MMC, Turning, ANN, Ra, MRR.

Article published in International Conference on Advances in Mechanical Sciences 2014, Special Issue-2 (Feb 2014)