OPTIMIZATION OF THE PROCESS CONSTRAINTS IN SPARK EROSION MACHINING OF ALUMINIUM ALLOY AA 6061 HYBRID COMPOSITES USING ARTIFICIAL NEURAL NETWORK

Abstract

The foremost objective of this research work is to implement Artificial Neural Network (ANN), to improve spark erosion machining performance of aluminum alloy AA 6061 hybrid composites by controlling the process constraints, which is suitable for bio medical applications. Aluminum composites are mostly used to replace the conventional materials attributable to their less weight, notable wear and corrosion resistances. These composites are used in automotive, aerospace, electronics and bio medical applications. Machining of aluminium composites using conventional machining technique is one of the major challenges because of the presence of hard particles in aluminium matrix. Unconventional machining techniques have been preferred for machining aluminium composites to enhance better surface quality. In the present study the composite specimen was processed through stir casting and machining was carried out using spark erosion machining, by varying four process constraints with the application of design of experiments. ANN trained with multi-layer feed forward through the error back-propagation training algorithm, was used to model the network and predict the material removal rate (MRR) of the composite. The outcomes exposed that the projected values found from the ANN model were in good agreement with the investigational values and to study the machining characteristics of composites, the model could be effectively applied.