A comparative study of EMD and EEMD approaches for identifying chatter frequency in CNC turning

• Published in: European journal of mechanics, A, Solids A, Solids, 2019-01, Vol.73, p.381-393
• Main Authors: Shrivastava, Y., Singh, B.
• Format: Article
• Language: English
• Subjects: Chatter; Chatter frequency; Comparative studies; Decomposition; EEMD; EMD; Empirical analysis; Hilbert transformation; Numerical controls; Productivity; Signal processing; Surface finish; Time-frequency analysis; Tool life; Vibration
• ISSN: 0997-7538; 1873-7285
• DOI: 10.1016/j.euromechsol.2018.10.004

Tool chatter identification is one of the most important issues for today's researchers; as it adversely affects the tool life and surface finish which in turn reduces the overall productivity. In the past, researchers have proposed various techniques to compensate the effect of chatter. Signal processing is one such emerging technique that has proved to be quite efficient in exploring chatter. There are various approaches within this domain. Hilbert-Huang transform (HHT) is one of them. HHT is composed of empirical mode decomposition (EMD) and classical Hilbert transform (HT). EMD decomposes nearly any signal into a finite set of functions and a residue, whose Hilbert transform gives physical instantaneous frequency. However, there is a critical problem of mode mixing in EMD. To overcome this problem, a suitable EEMD approach has been adopted on experimentally acquired raw chatter signals in order to identify chatter frequency which has not been done by the previous researchers. In the present work, time-frequency analysis of experimentally acquired raw chatter signals have been done considering both EMD and EEMD techniques and thereafter their results have been compared to ascertain the suitability of each approach. It has been inferred from the present analysis that EEMD is quite apt in detecting chatter in turning operation with utmost accuracy. •The recorded chatter signals are non-stationary and non-linear in nature. Moreover, EMD/EEMD both are capable of processing such signals.•The non-contact type microphone is a suitable sensor for recoding the chatter signals.•EMD/EEMD is capable of eliminating the noise and delimit the influence of contaminations.•From the analysis, it is evident that the IMFs obtained are capable of reflecting the machining phenomenon.•The chatter frequency acquired at the given set of parameters is 208 Hz.