Prediction of performance parameters in Wire EDM of HcHcr steel using Artificial Neural Network

Abstract

Electrical discharge machining has been extensively used for cutting intricate contours or delicate cavities that would be difficult to produce with a conventional machining methods or tools. Wire EDM is in use for a long time for cutting punches and dies, shaped pockets and other complex shaped parts. Performance of the process is mainly depends on many parameters used during process. Machining input parameters provided by the machine tool builder cannot always meet the operator’s requirements. So, artificial neural network is introduced as an efficient approach to predict the values of performance parameters. In the present research, experimental investigations have been conducted to develop predictive models for the effect of input parameters on the responses such as Material Removal Rate, surface finish and kerf width. Material tested was HcHcr steel material. Molybdenum wires of diameters 0.18 mm were used for the WEDM machine. A feed forward back propagation artificial neural network (ANN) is used to model the influence of current, pulse-ON and pulse-OFF time on material removal rate, kerf width & surface roughness. Multilayer perception model has been constructed with feed forward back propagation algorithm using peak current, pulse-ON and pulse-OFF time as input parameters and MRR and surface roughness and kerf width as the output parameters. The predicted results based on the ANN model are found to be in very close agreement with the unexposed experimental data set. The modeling results confirm the feasibility of the ANN and its good correlation with the experimental results.