Chatter detection in the turning process using acoustic wavelet scattergrams and transfer learning

Sharma, Sanjay and Gupta, Vijay Kumar and Rahman, Mustafizur and Saleh, Tanveer (2026) Chatter detection in the turning process using acoustic wavelet scattergrams and transfer learning. Measurement: Journal of the International Measurement Confederation, 265 (NA). pp. 1-10. ISSN 0263-2241 E-ISSN 1873-412X

	PDF (Journal) - Published Version Restricted to Repository staff only Download (8MB) | Request a copy
Preview	PDF (Scopus) - Supplemental Material Download (200kB) | Preview
Preview	PDF (WOS) - Supplemental Material Download (320kB) | Preview

Abstract

Chatter is an unwanted vibration in machining processes that degrades surface quality, reduces productivity, and shortens tool life. Reliable chatter detection is essential, particularly under varying machining conditions where traditional vibration-based approaches are often intrusive, complex, and require expert interpretations. This study introduces a novel time–frequency representation of acoustic signals using wavelet scattergrams for chatter detection in the turning process. These wavelet scattergrams generated through a wavelet scattering network (WSN) are used as an input to fine-tune a deep neural network (DNN) using a transfer learning framework. To ensure robust evaluation, experiments were conducted with five different overhang workpiece configurations, where the acoustic signals from the first four configurations were used for training and validation, and a completely new workpiece configuration was used to test the model performance. Among the evaluated models, ResNet50V2 achieved the highest classification accuracy of 99.2% while maintaining a low detection latency of about 62 ms, outperforming MobileNet and DenseNet121. The results demonstrate that the proposed method can reliably detect chatter using a non-intrusive microphone sensor and limited experimental data while maintaining robustness across varying machining conditions, making it suitable for accurate chatter detection in industrial applications.

Item Type:	Article (Journal)
Additional Information:	6808/127196
Uncontrolled Keywords:	Transfer learning, Chatter detection, Acoustic signal, Wavelet scattergram, Turning process
Subjects:	T Technology > T Technology (General)
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Engineering > Department of Mechatronics Engineering Kulliyyah of Engineering
Depositing User:	Dr. Tanveer Saleh CEng MIMechE
Date Deposited:	05 Feb 2026 10:04
Last Modified:	05 Feb 2026 12:47
Queue Number:	2026-01-Q1949
URI:	http://irep.iium.edu.my/id/eprint/127196

Actions (login required)

View Item