Detection of Chatter in Machining Processes by the Multiscale Maximum Approximate Entropy and Continuous Wavelet Transform
Chatter is a complex dynamic instability in machining processes and presents nonlinear and nonstationary behavior. Detection of this phenomenon before a catastrophic failure occurs has great importance in the industry today. This behavior demands online monitoring signal-processing techniques suitab...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-02-01
|
| Series: | Applied Mechanics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2673-3161/6/1/15 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Chatter is a complex dynamic instability in machining processes and presents nonlinear and nonstationary behavior. Detection of this phenomenon before a catastrophic failure occurs has great importance in the industry today. This behavior demands online monitoring signal-processing techniques suitable for facing these kinds of dynamics such as approximate entropy (AE) and wavelet transform. Moreover, AE is useful for dealing with noisy signals and requires a relatively small amount of observations. In this study, we propose an improved AE methodology, the multiscale maximum approximate entropy (MMAE), to detect chatter in milling processes. The maximum AE is achieved by the calculation of the parameter r proposed by Sheng and Chon. In the past, the calculation of this parameter was a drawback of the AE technique. The results show the effectiveness of this proposed technique in detecting clearly different gradual and drastic changes in chatter conditions. Moreover, a more known technique is presented: the time–frequency maps provided by continuous wavelet transform (CWT). The results also show the efficacy of this technique in detecting different levels of chatter. The results are corroborated by the machining piece observation of the chatter phenomenon. MMAE is also compared with sample entropy (SE) and the Hurst exponent obtained by the R/S analysis. At the end, a comparison analysis of the mentioned techniques is carried out, showing that they all have advantages and disadvantages. However, the disadvantages of MMAE and CWT can be solved, as mentioned in the comparison section. Thus, the conclusion is that MMAE and CWT techniques are optimal for the online monitoring of chatter in machining processes. |
|---|---|
| ISSN: | 2673-3161 |