Data-driven modeling and optimization for recast layer in electric discharge drilling of Inconel-718 material

Electric discharge drilling (EDD) is crucial in aerospace manufacturing due to its ability to create intricate geometries with high precision. Nonetheless, optimizing metallurgical features, such as minimizing the recast layer, remains challenging. This study addresses this issue by proposing a data-driven model to predict recast layer thickness during EDD of aerospace components. The study investigates the effects of discharge current, pulse on-time, pulse off-time, and flushing pressure on recast layer thickness in the EDD of Inconel-718 sheets. A total of 29 experiments were conducted based on response surface methodology. A quadratic regression model was developed to predict the recast layer thickness, revealing complex interactions between the EDD process parameters and the recast layer. The statistical model’s reliability is confirmed by satisfactory R 2 (88.2%) and adjusted R 2 (76.4%) values. Contour plots illustrate the interactive effects of process parameters, indicating optimal values for process parameters: discharge current: 7–7.5 A, pulse on-time: 7–8 µs, pulse off-time: 2–2.25 µs, and flushing pressure: 78–80 bar. The developed model aids in evaluating and optimizing recast layer thickness in EDD of Inconel-718, leading to improved performance and productivity in aerospace manufacturing.