A generic tool wear model and its application to force modeling and wear monitoring in high speed milling: Tool wear prediction data per tooth

The milling cutter is made of tungsten carbide, and the workpiece materials are Inconel 718 and AISI4340. The experiment was completed on a five axis precision machining center HSM600U. The force and vibration signals in the experiment were measured by the Kistler 9119AA2 three component dynamometer and the Kistler 8636C accelerometer. The sampling frequency is 50 KHz. In experiment 1, each tool used the same workpiece, the same conditions and the same tool to complete 315 cuts. Processing parameters of milling verification experiment are shown in Table 1.Fig. 12 shows the instantaneous theoretical milling force FMj per tooth, the instantaneous milling force Fj per tooth estimated by tool wear, and the tool flank wear VBj of each tooth in three groups of tests. It can be seen from the results that the two groups of change curves of FMj and Fj along with the tool flank wear VBj are very close, and the instantaneous milling force per tooth estimated by tool wear eliminates the interference of statistical error points during calculation, as shown in Fig. 12 (a)~(c). When the tool flank wear VBj of each tooth is small in the initial milling time, the milling force Fj estimated by tool wear has no significant difference on each tooth, and Fj increases slowly.With the increase of milling time, when the tool flank wear VBj of the three groups of experiments is 80~120 μ In the second group of experiments, Fj suddenly increases, as shown in Fig. 12 (d)~(f), and the milling force Fj increases by more than 5N. According to the curve of change of VBj of tool flank wear with milling time, the curve of change of VBj of milling force and subsequent tool flank wear suddenly increases in the stage of steady wear (II), and the tool flank wear increases slowly.Thereafter, the growth rate of the instantaneous milling force on the flank wear VBj is kept at a high value, and the difference between the milling forces on each tooth is more obvious due to the continuous accumulation of the tool runout effect. For example, the difference between the milling force F1 and the milling force F2 or F3 in the third group of experiments in the later milling time stage is greater than 5 N, as shown in Fig. 12 (g)~(i), which means that the increase rate of the milling force in the accelerated wear (III) stage is similar to the increase rate of the flank wear VBj. Therefore, the rear face of the first cutter tooth is more worn than the second and third cutter teeth. The paper proposes that the milling force Fj can be estimated to reflect the wear state of each tooth flank VBj, and the relationship between the milling force Fj and each tooth flank VBj can be determined by the mechanism analysis milling force model and the instantaneous milling force model of each tooth estimated by the tool wear.