Investigating temperature variation of Al 6065 T6 during milling operation for aerospace applications using response surface methodology and support vector regression

Aluminium alloy finds increasing industrial applications due to its desirable properties. However, its low thermal conductivity often limits its application at elevated temperatures, thus the need to investigate the temperature variation during milling operation for aerospace applications. The response surface methodology (RSM) carried out in the Design-Expert 2022 software environment was used to investigate the temperature variation of Al 6065 T6 during milling operation and the designed experiment (DoE) comprises of three process parameters with a feasible range of values which produced 20 experimental trials. These include feed per tooth (0.1-0.25 mm/tooth), cutting speed (5-35 m/min) and axial depth of cut (0.10-0.35 mm). The milling operation was carried out on a 5-axis computer numeric control (CNC) milling machine (DMU80 monoBLOCK), and the response of the designed experiment (cutting temperature) was measured using the infrared video thermometer. The support vector regression (SVR) machine learning algorithm was also used to approximate the relationship between the process parameters (input variables) and a continuous target variable (cutting temperature) while reducing the prediction error. The statistical analysis of the physical experimentation results obtained produced a predictive model for estimating the magnitude of the cutting temperature. The range of process parameters that produced the least temperature (73.1 degrees C) are feed per tooth (0.25 mm/tooth), cutting speed (5 m/min) and depth of cut (0.10 mm). This study adds to the literature empirically and contributes to the understanding of temperature variation during the milling operation of Al 6065 T6. The results may promote the utilisation of Al 6065 T6 for component development for aerospace applications.