Modelling of spark erosion machining process for hybrid metal matrix composite by artificial neural network approach

Abstract

This paper discusses about predicting the machining parameters of Sink EDM drilling of Hybrid Metal Matrix Composite of Aluminum (LM6) strengthened with Boron Carbide (B4C) and Multi Wall Carbon Nano Tube (MWCNT), manufactured using stir casting technique. Three different specimens were prepared with different % wt Composition. The output process parameters like Metal Removal Rate (MRR), rate of tool wear, and surface finish of the dilled specimens were predicted by Artificial Neural Network (ANN) technique. To develop the model the input process parameters like pulse interval time (Toff), pulse duration time (Ton), current (I), voltage (V), MWCNT weight percentage and distance between electrode and work piece (G) were considered. From the experimental results it was found that after training the model, the developed Artificial Neural Network model was found to be very effective in predicting the output parameters with 98 % accurate with the experimental results better in terms of forming the better surface and stirring the reinforcement particles in matrix aluminum specimen.