Experimental Surface Modification, Prediction Modeling, and Multi-Objective Optimization of μ-EDM Process using GBR-NSGA-II for machining of Ti-6Al-4V+8%Ta

Although metal 3D printing offers a promising solution for producing customized Ti-alloy components. However, due to inherent process limitations, additional post-processing is required for the fabrication of micro-features. Therefore, precision machining techniques like micro-EDM are still required. Micro-EDM is an advanced machining technique better known for its precision while machining titanium alloys. Even though the μ-EDM provides superior performance, the optimization of machining parameters remains a significant challenge due to the complex interactions between electrical and thermal phenomena. Thus, the present research aimed at optimizing the parameters for μ-EDM machining of the SLM-printed Ti-6Al-4V + 8%Ta. A data-driven hybrid optimization framework combining GBR and the NSGA-II is developed to enhance the machinability of the alloy. The GBR models accurately predicted MRR, OC, and MH, achieving R 2 values of 0.987 (MRR), 0.999 (MH), and 0.999 (OC), with correspondingly low MAE and MSE values across training and testing datasets. Whereas the TWR has shown a low R 2 (>75%) value, which can be attributed to the inherent stochasticity. Additionally, the study is structured within a multi-objective optimization framework, emphasizing the maximization of MRR and MH while minimizing OC and TWR. After obtaining the possible number of optimal solutions from the surrogate model, the top five solutions are selected using TOPSIS. Lastly, confirmatory experiments validated the framework, showing MAPE deviations of less than ∼13.96% between predicted and experimental results. This integrated methodology effectively navigates the multifaceted nature of μ-EDM parameter optimization.