Thermal Analysis of Micro-Channel Internal Cooling in Cutting Tools: A Machine Learning Approach

• Published in: Applied mechanics and materials Vol. 922; pp. 97 - 107
• Main Authors: Abid, Aman, Muaz, Muhammed, Alim, Md Kashif, Hussain, Shahid, Hamza, Syed Mohd, Arif, Sajjad
• Format: Journal Article
• Language: English
• Published: Zurich Trans Tech Publications Ltd 19.08.2024
• Subjects: Coolants; Cooling; Cooling effects; Cutting force; Cutting parameters; Cutting tools; Flood forecasting; Floods; Heat; Heat transfer; Inlet pressure; Machine learning; Metal cutting; Microchannels; Nickel base alloys; Notches; Regression models; Superalloys; Support vector machines; Surface properties; Surface roughness; Temperature; Thermal analysis; Turning (machining); Random Forest Model; Internal-Cooling Method; Support Vector Regression Model; Inconel 718; Machine Learning
• DOI: 10.4028/p-48gVXs
• ISSN: 1660-9336; 1662-7482

The use of coolants for cutting process in metal cutting operations is customary. Turning causes high cutting heat in nickel base super alloy Inconel 718. Nonetheless, it should be acknowledged that although flooding techniques are commonly used in the machining of super alloys, these flood cooling methods have extremely poor efficiencies. Another alternative to increase the cooling capabilities of fluids would be an internal-cooling approach that would enable to lower machining temperatures significantly. The heat dissipation ability in the tool is also greatly influenced by the micro-channel diameter of tool which further causes a significant effect on the coolant outlet velocity. A design of an internal-cooling single point cutting tool with micro channel structures for enhanced coolant heat transfer capability and reduced machining temperature is used for turning Inconel 718 under dry, flooded cooling and internal cooling to study the effects of cooling conditions on cutting force, cutting temperature and surface quality. A regression model is built using the Random Forest (RF) and Support Vector Regression (SVR) methods in machine learning framework. These models were then used to forecast input parameters, such as channel diameter and inlet pressure, which made it easier to obtain output data, such as pressure and maximum velocities at different notches. Eighty percent of the data in the dataset is used to train the model and with the remaining twenty percent set aside for evaluating the model's functionality. When comparing internal-cooling technology to traditional flood cooling, there are clear benefits including increased heat transfer efficiency, which leads to lower cutting temperatures, less cutting force, and better surface quality. More specifically, in the internal-cooling configuration, a direct relationship is shown between rising coolant inlet pressure and falling cutting force and temperature over time. Further highlighting the advantages of this cooling strategy is the relationship between increased intake pressure and decreased surface roughness.