Data-Driven Multi-Objective Optimization of Drilling Performance in Multi-Walled Carbon Nanotube-Reinforced Carbon...

Carbon fiber reinforced polymer (CFRP) composites are widely used in many engineering applications such as aerospace, automotive, and defense industries due to their superior properties such as high specific strength, stiffness, and corrosion resistance. However, these materials require drilling, especially during assembly processes. Damage mechanisms arising during this process, such as delamination, high thrust force, and torque, negatively affect structural integrity and production quality. This study proposes a data-driven, multi-objective optimization approach to solve problems encountered during drilling in multi-walled carbon nanotube (MWCNT)-reinforced CFRP nanocomposites. The study considers the MWCNT reinforcement ratio, cutting speed, and feed rate as process parameters and examines their effects on thrust force, torque, and delamination factor. Second-degree polynomial regression-based prediction models were created using the experimental data obtained, and these models were included in the multi-objective optimization process. During the optimization phase, thrust force and torque values were simultaneously minimized, while the delamination factor was kept below the statistically determined constraint of F_d ≤ 1.054. Pareto-optimal solution sets were obtained using NSGA-II and MOPSO meta-heuristic algorithms in the solution process. The results indicate that suitable combinations of drilling parameters can be identified through Pareto-based optimization, allowing significant reductions in thrust force and torque while maintaining the delamination factor below the specified limit. The study presents a reliable optimization approach for the more efficient machining of CFRP nanocomposites.