A Bionic Inspired Dragonfly Algorithm for Parametric Optimization and Damage Investigation During Machining of Modified Polymer Nanocomposite

Polymer nanocomposite is commonly used to develop structural components of space, aircraft, biomedical, sensor, automobile, and battery sector applications. It remarkably substitutes the heavyweight metallic and nonmetallic engineering materials. The machining principles of polymer nanocomposites are intensely different and complex from traditional metals and alloys. The nonhomogeneity, abrasive, and anisotropic nature differs its machining aspect from conventional metallic materials. This investigation aims to execute the CNC drilling of modified nanocomposite using Graphene–carbon (G-C) @ epoxy matrix. The process constraints, namely, cutting speed (S), feed (F), and wt.% of graphene oxide (GO) vary up to three levels and are designed according to the response surface methodology (RSM) array. The nonlinear model is created to predict surface roughness (Ra) and delamination (Fd) on regression analysis. It has been found that the average error for Ra is 0.94% and for Fd it is 3.27%, which is acceptable in model predictions. The metaheuristics-based evolutionary Dragonfly algorithm (DA) evaluated the optimal parametric condition. The optimal setting prediction for the DA is observed as cutting speed (S)-37.68[Formula: see text]m/min, feed (F)-80[Formula: see text]mm/min, and wt.% of graphene oxide (GO)-1%. This algorithm demonstrates a higher application potential than the previous efforts in controlling Ra and Fd values. Both the drilling response values are found to be minimized when the cutting speed increases and the feed decreases. The best fitness value for the DA is 1.626 for surface roughness and 5.086 for delamination. This study agreed with the prediction model’s outcomes and the process parameters’ optimal condition. The defects generated during the sample drilling, such as fiber pull out, uncut/burr, and fiber breakage, were examined using FE-SEM analysis. The optimal findings of the DA module significantly controlled the damages during machining.