Experimental Scheme to Monitor Eigenvalues of Vibration Signals in Cutting Tool Milling

Abstract

In order to better solve the problem of accuracy of tool wear status prediction, the extraction of feature values of tool wear information from the sensors is the basis for solving the problem. This chapter designs an experimental scheme to monitor the tool wear state by extracting the vibration signal of tool wear. Milling tool wear state recognition plays an important role in controlling the quality of milled parts and reducing machine tool downtime. However, the characteristics of milling process limit the accuracy and stability of tool condition monitoring employing vibration signals. A T-type cutting tool, a vibration sensor, an amplifier, a data acquisition card, and a computer make up the data acquisition and signal processing hardware. The vibration signal is statistically analysed in the time domain, and it is determined that the variance of the vibration signal caused by X-axis wear is positively connected with the level of tool wear. Moreover, the vibration signal is converted from time domain to frequency domain by Fourier transform, and the characteristic frequency bands of vibration signal are 2~4 kHz and 7~9 kHz in frequency domain.  The DB4 wavelet of Daubechies series wavelets is used as the wavelet packet base, and the DB4 wavelet packet base has features such as smoothness as well as matching wavelet fast algorithms Wavelet packet decomposition technology is used to extract the eigenvalues of vibration signals. In addition, the characteristics in time domain, frequency domain and frequency domain are also discussed.  It is further judged that the energy percentage of 2.5~3.75 kHz and 7.5~8.25 kHz is closely related to tool wear, so the energy percentage of the two characteristic frequency bands is selected as the characteristic value of tool wear monitoring.