The aerospace industry would eventually replace chemical machining by mechanical machining which is more accurate, more predictable and more ecological. In fact, the discharges in the case of chemical machining contain especially carbon dioxide and solvents that are difficult to degrade in groundwater. The mechanical machining also avoids an important quantity of hazardous substances and provides better chips recycling. However, the control of mechanical machined parts quality goes through the prediction and the optimization of the metal cutting processes. The most attractive computational tool to predict and optimize metal cutting processes is the finite element modeling (FEM). The success and the reliability of any FEM depend strongly on the constitutive laws which describe the thermo-mechanical behavior of the machined materials. The most commonly used one is that of Johnson and Cook (JC) which combines the effect of strains, strain rates, and temperatures. The determination of the material constants of JC under high strains, strain rates, and temperatures during machining conditions has long been a major challenge but a necessity for those who apply finite element modeling techniques in machining processes at the chip formation scale.
This study aims at treating this subject in order to better understand the effect of the JC constitutive law on the prediction of cutting parameters (cutting forces, residual stresses, etc.) for aluminum alloys. In addition, in order to meet the interests of aerospace industry, three aluminum alloys (Al2024-T3, Al6061-T6 and Al7075-T6) commonly used in aircraft applications have been selected.
This research work is divided into three consecutive steps.
Firstly, a new approach to identify the material constants of JC for metal cutting is proposed. The approach is based on the inverse method (orthogonal machining tests) and the response surface methodology which allows generating a large number of cutting conditions within fixed ranges of cutting speed, feed rate, and rake angle. Based on this approach, the sensitivity of the material constants of JC to the rake angle for the three alloys was analysed. It was found that, for these three alloys, one set of the material constants obtained from the proposed approach predicts more accurate values of flow stresses as compared to those reported in the literature. Moreover, a 2D FEM investigation of the orthogonal cutting also showed a good agreement between the predicted cutting parameters (cutting forces and chip thickness) and experimental ones when using the material constants obtained by the proposed approach.
Secondly, a specific focus was put on the influence of the rake angle on the material constants of JC and hence on the predicted cutting parameters (cutting forces, chip morphology, and tool-chip contact length). To achieve this goal, different sets of JC constants obtained at different rake angles (-8°, -5°, 0°, +5°, and +8°) were used in conjunction with a 2D finite element model to simulate the machining behavior of Al2024-T3 alloy. It was found that the material constants set obtained with 0° rake angle gives overall more accurate predictions of the cutting parameters as compared to other studied sets.
Finally, the last step of this study is devoted to the prediction of induced residual stresses within the machined workpiece (Al2024-T3) and the temperature of the cutting tool(uncoated carbide). Three sets of JC based on the results obtained from the previous step with rake angles of -8°, 0°, and +8° were considered. Two finite element models were used; a 2D thermo-mechanical simulation to simulate chip formation and a 3D pure thermal analysis to obtain the temperature distribution. The results show that a better prediction of the residual stresses is obtained with JC at 0° while the other sets of JC at -8° and +8° tend to overestimate or underestimate the measured residual stresses, respectively. As far as the temperature of the cutting tool is concerned, the average values of the predicted températures of the cutting tool for each studied set of JC was considered in order to evaluate the best prediction. Based on these average values, the effect of the three sets of JC was not significant since the difference between the measured temperatures and the predicted average ones are less than 5.5% with the three cutting conditions.
Daoud, M. (Author),
Chatelain (Supervisor) &
Bouzid (Co-supervisor),
16 Nov 2016Student thesis: Doctoral thesis › Doctorate in Engineering: Engineering