Investigating the effect of adding powders into dielectric in EDM machining of Inconel 718 Alloy and using an ANN model to predict the output parameters

One of the promising methods for improving output parameters in electrical discharge machining (EDM) is the powder mixed electrical discharge machining (PMEDM) process. In this process, the powder of a conductive or non-conductive material is added to the dielectric fluid. EDM is a very complex process which is influenced by many parameters. By adding powder to the dielectric, the complexity of process increases so that the determination of relations between input and output parameters becomes more difficult. In this paper, the effect of adding aluminum and silicon carbide powders to the dielectric on the output parameters of EDM process of Inconel 718 alloy is experimentally investigated. According to the results, both powders improve the EDM process performance. Moreover, an artificial neural network (ANN) model is developed for prediction of surface roughness (Ra) and material removal rate (MRR) and then the results are compared to a regression model. Results show that the accuracy of predicted Ra and MRR by ANN is higher than that of regression model, as the prediction errors are 5.22% and 9.16% for ANN and regression models, respectively.