Modeling and optimization of multi-hole rotary electrical discharge drilling of die steel

Abstract

This research proposed a mathematical model focused on the artificial neural network (ANN) and the response surface methodology for gas-assisted electrical discharge drilling (GAEDD) of die-steel. The GAEDD was performed with multi-hole rotary electrode in this work. Orthogonal central composite rotable design (CCRD) method employed for conduction of experiments. An overall of thirty two experiments were executed at five levels. In this exploration, pulse duration, discharge current, duty cycle, tool rotation, and discharge gas pressure has been selected as input parameters, while the material removal rate (MRR) and the surface roughness (SR) were selected as process output. Analysis of variance was implemented to instigate and to know the competency of the established models of MRR and SR. The fit outline supports that the statistical model is proper and the lack of fit is irrelevant. Mean square error and root mean square error obtained through response surface methodology was utilized for the modeling and anticipating potential of models developed through artificial neural network (ANN). The exploratory and forecast estimates of MRR and SR of the GAEDD, acquired by RSM and ANN were observed to be as per one another. In addition, the ANN technique has shown that, when contrasted with the RSM, the response is even more appropriate. A genetic algorithm has been used to extract optimum parameters by optimizing the GAEDD process to maximize MRR and minimize SR. The optimal value of MRR and SR obtained through genetic algorithm was found to be 22.34 mg/min. and the SR at 3.80 µm respectively.