Optimizing surface quality in PMEDM using SiC powder material by combined solution response surface methodology – Adaptive neuro fuzzy inference system

Abstract Improving the surface quality of products in the mold field is very necessary, because it directly affects their working ability in practice. Electrical discharge machining (EDM) is still a very popular technological solution in this field, and research on Powder-mixed EDM (PMEDM) is still a new research direction and it can overcome the limitations in EDM. In this study, the surface quality after PMEDM using silicon carbide (SiC) powder was studied, and AISID-3 tool steel material was used as the workpiece. The adjustable parameters in PMEDM were investigated, including current (I), gap voltage (Vg), cycle time (CT), duty factor (DF), and powder concentration (PC). To improve the accuracy of the surface roughness (SR) result model, the Response surface methodology (RSM) combined with the Adaptive neuro fuzzy inference system (ANFIS) was used to build the SR problem model. The research results showed that the influence of process parameters on SR was indicated by Analysis of Variance technique, and the calculation model of SR was determined with high accuracy. In addition, the surface layer of AISID-3 tool steel after PMEDM was significantly improved by using PMEDM with SiC powder. The ANFIS approach has inferred that the developed model with low error values of mean squared error, and the size of the micro cracks and craters formed on the machined surface is less in PMEDM, hardness was significantly improved with the increase in SiC PC.