Multi-objective optimization for predicting and enhancing machining performance in nanofluid- assisted turning using meta heuristic algorithms

Abstract

Abstract This study evaluates the performance of a multi-walled carbon nanotube (MWCNT)-based nanofluid as an alternative cutting fluid in the turning of EN31 steel. Experiments were conducted under dry, conventional coolant, and nanofluid-assisted conditions using a gravity-feed lubrication system with a flow rate of 1 L/h. The results indicate that the nanofluid improves machining performance, with reductions of 5–20% in cutting force, 11–37% in thrust force, 4–33% in feed force, and 2–34% in surface roughness compared to conventional cooling. Response Surface Methodology (RSM) with Central Composite Design was applied to develop predictive models for surface roughness and resultant cutting force. The models showed good statistical significance and prediction accuracy. Multi-objective optimization was performed to minimize surface roughness and cutting force while maximizing material removal rate. The optimal parameters were identified as 1500 rpm cutting speed, 0.10 mm/rev feed rate, and 1 mm depth of cut, resulting in surface roughness of approximately 0.47 μm, cutting force of 165 N, and material removal rate of 7.54 mm³/min. Comparative analysis of optimization algorithms showed that NSGA-II provided superior convergence and stability, achieving better performance than other methods. The findings confirm the effectiveness of nanofluid-assisted machining combined with intelligent optimization for improved machining outcomes.