Statistical Modeling and Optimization of Process Parameters in Electro-Discharge Machining of Cobalt-Bonded Tungsten Carbide Composite (WC/6%Co)

摘要

In this paper, attempts have been made to model and optimize process parameters in Electro-Discharge Machining (EDM) of tungsten carbide-cobalt composite (Iso grade: K10) using cylindrical copper tool electrodes in planing machining mode based on statistical techniques. Four independent input parameters, viz., discharge current (A: Amp), pulse-on time (B: μs), duty cycle (C: %), and gap voltage (D: Volt) were selected to assess the EDM process performance in terms of material removal rate (MRR: mm3/min), tool wear rate (TWR: mm3/min), and average surface roughness (Ra: μm). Response surface methodology (RSM), employing a rotatable central composite design scheme, has been used to plan and analyze the experiments. For each process response, a suitable second order regression equation was obtained applying analysis of variance (ANOVA) and student t-test procedure to check modeling goodness of fit and select proper forms of influentially significant process variables (main, two-way interaction, and pure quadratic terms) within 90% of confidence interval (p-value≤0.1) It has been mainly revealed that all the responses are affected by the rate and extent of discharge energy but in a controversial manner. The MRR increases by selecting both higher discharge current and duty cycle which means providing greater amounts of discharge energy inside gap region. The TWR can be diminished applying longer pulse on-times with lower current intensities while smoother work surfaces are attainable with small pulse durations while allotting relatively higher levels to discharge currents to assure more effective discharges as well as better plasma flushing efficiency. Having established the process response models, a multi-objective optimization technique based on the use of desirability function (DF) concept has been applied to the response regression equations to simultaneously find a set of optimal input parameters yielding the highest accessible MRR along with the lowest possible TWR and Ra within the process inputs domain. The obtained predicted optimal results were also verified experimentally and the values of confirmation errors were computed, all found to be satisfactory, being less than 10%. The outcomes of present research prove the feasibility and effectiveness of adopted approach as it can provide a useful platform to model and multi-criteria optimize MRR, Ra, and TWR during EDMing WC/6%Co material.