Prediction of output parameters in wire electrical discharge machining of EN-31 steel by artificial neural networks

The objective of this work is to predict the effect of input parameters on output parameters using artificial neural network in wire electrical discharge machining. The input variables considered are peak current, pulse on time, flushing pressure of dielectric, pulse off time, wire tension and wire feed rate. The output variables taken into consideration are dimensional deviation and material removal rate. Workpiece used in the study is of EN-31 steel. Wire electrical discharge machining is a thermal cum electrical process and uses the electrical and thermal energy for cutting materials. Wire electrical discharge machining is utilised for cutting electrically conductive materials, which are difficult to machine with conventional machining methods. This process uses discrete electrical discharges between continuously travelling wire (tool) and the workpiece for cutting the workpiece. In this work, artificial neural networks are used for prediction of output parameters. Artificial neural networks have a highly connected set of nodes or processing elements that operate in parallel. Artificial neural networks can be trained using input and output data and can be used to predict data for new input values.