The Effect of Vegetable Cutting Oils on Turning AISI 5140 Steel for Sustainable Machining

Abstract

This study presents a comprehensive machinability analysis of AISI 5140 steel widely used in industry by performing turning operations under dry and two different MQL (Minimum Quantity Lubrication) conditions. The analysis involves examining the machinability output parameters of cutting force and surface roughness at three different levels of feed rate and cutting speed, while maintaining a constant depth of cut. Physical experiments were conducted on a manual lathe, cutting forces were measured during the experiments, and surface roughness values were measured after the experiments were completed. The obtained data were first analyzed graphically and then subjected to regression, decision tree, heat map, and prediction table analyses, which represent machine learning models. The study concluded that low feed rates and high cutting speeds are required to achieve optimum cutting force. For surface roughness, low feed rate and high cutting speed values also yielded favorable results. The most advantageous performance was obtained in the MQL (Sunflower oil) environment. The accuracy of the prediction models was confirmed by achieving R² values higher than 70% for both output parameters. In conclusion, it was determined that integrating machine learning models into machinability studies can facilitate future manufacturing processes.