Modelling and Multiobjective Optimization of Vibration Assisted Electrical Discharge Drilling Process

Abstract

The VA-EDD process has gained noteworthyconsideration in recent times due to its potential for enhancing material removal rates and improving surface finish in advanced machining applications. This study explores the synergistic integration of RSM and Multi-objective Optimization through PCA based GRA for the optimization of the VA-EDD process.The research involves the application of RSM to model and analyse the complex relationships between various process parameters and performance indicators in VA-EDD. Design of Experiments (DOE) is employed to systematically explore the parameter space, and the experimental data is used to develop predictive models for key process responses such as MRR and SR. The accuracy and reliability of the developed models are assessed through statistical analyses, providing insights into the optimal process parameter settings.The developed mathematical models for MRR and SR depict the experimental results with prediction errors of less than 9% for Ti-6Al-4V and for Al-TiB2, the developed mathematical models for MRR depict the experimental results with prediction errors of less than 13 % and for SR the developed mathematical models depict the experimental results with prediction errors of less than 8%.Further, Multi-objective Optimization techniques are employed to address the inherent trade-offs between conflicting objectives in the VA-EDD process. The optimization framework considers multiple objectives simultaneously, including maximizing MRR and minimizing surface roughness. To enhance the efficiency of the multi-objective optimization process and reduce the computational burden associated with a large number of variables, Principal Component Analysis (PCA) is applied. PCA is utilized to reduce the dimensionality of the input space by identifying the most significant process parameters that contribute to the overall variability of the system. The reduced set of principal components is then used for conducting multi-objective optimization, leading to a more streamlined and efficient