Loading...

Chatter Detection in Machining Using Time-frequency Signal Analysis Methods

Khoshnazar, Haleh | 2014

787 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 46719 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Movahhedy, Mohammad Reza; Akbari, Javad
  7. Abstract:
  8. Regenerative chatter is a self-exited vibration that causes poor surface finish, tool wear and breakage and excessive noise. It is, therefore, essential to avoid chatter occurrence passively by selecting cutting conditions that guarantee the stability of the process, or detect and suppress chatter at its onset during the process. The latter has become more important due to the complex and nonlinear dynamics of metal removal and the rising interest in automated systems. Wavelet and Hilbert-Huang transforms are used to detect chatter in machining processes including milling and surface grinding, in this thesis. In the first part of the thesis, chatter detection in milling is studied. The radial cutting force is measured at different radial and axial depths of cut. The signal is, then, segmented and decomposed using wavelet transform. Instantaneous amplitudes of the intrinsic mode functions of the second-level wavelet detail is, then, calculated. It is observed that the mean value and standard deviation of the instantaneous amplitudes increased rapidly at the onset of chatter. The critical values of these two parameters for chatter detection are determined. The critical values are suitable for different radial depths of cut. Moreover, it is shown that using wavelet transform to de-noise the signal is essential. In the second part, the signal processing method used to detect chatter in milling is employed for chatter detection in surface grinding based on the grinding force. The intrinsic mode functions of the wavelet component with rich chatter information are obtained and the mean value and standard deviation of the instantaneous amplitudes are calculated. It is observed that these parameters are close to zero before chatter occurrence and increased after chatter. Therefore, the mean value and standard deviation of the instantaneous amplitudes can be used as chatter indices
  9. Keywords:
  10. Chatter Detection ; Milling ; Wavelet Transform ; Surface Grinding ; Hilbert-Huang Transform

 Digital Object List

 Bookmark

No TOC