Synergistic modeling for precision in electrochemical discharge mechanism: a triad of 3D finite element, ANN, and regression approaches

Electrochemical discharge machining (ECDM) using electrolyte as a primary source emerged as advanced hybrid micromachining process for micromachining advanced materials like glass, ceramics, carbon fiber composites, etc. This process involves multiple electrochemical process parameters involving electrodes, electrolytes, working energy, and the setup. The complex interactions between the numbers of parameters significantly give rise to stochastic nature of the process. Thus, reliable modeling techniques are required to explore its behavior. In this paper, 3D finite element-based thermal model is developed followed by the artificial neural network (ANN) and multiple regression models to analyze the material removal rate (MRR) and the radial overcut (ROC) during micro-drilling. The temperature variations produced during simulation studies are utilized for MRR prediction. The predicted FEM results and model predictions are verified with experimental results. Results demonstrate a notable agreement between predicted and experimental outcomes. MRR improves by 28.5% when the concentration raises from 25 to 30 wt.%, and by 30.4% when the voltage is increased from 45 to 50 V. Overall, the analyses affirm that the developed models can effectively anticipate process responses, with ANN model outperforming the multiple regression model in providing realistic values for MRR and ROC within the specified range of parameters.