Incipient chatter fast and reliable detection method in high-speed milling process based on cumulative strategy

Area of Science:

• Manufacturing Engineering
• Mechanical Vibrations
• Signal Processing

Background:

• High-speed milling applications demand stringent time and reliability constraints for chatter detection.
• Existing chatter detection methods often struggle with heavy noise and require significant detection delays.
• Incipient chatter, or the early stages of unwanted vibrations, can significantly degrade product quality and tool life.

Purpose of the Study:

• To develop an incipient chatter detection method suitable for high dynamic applications like high-speed milling.
• To reduce detection delay and enhance reliability in the presence of significant environmental noise.
• To provide a statistically faster and more robust alternative to existing chatter detection techniques.

Main Methods:

• A multiple sampling per revolution (MSPR) technique was employed to gather information-rich data within short time windows.
• Data preprocessing involved Z-score normalization and mean-centering for chatter information consolidation.
• A modified adaptive cumulative chatter indicator was developed, adapting to noise variations and highlighting incipient chatter.
• A two-risk levels-based threshold was implemented to improve detection reliability by assessing multiple risk levels.

Main Results:

• The MSPR technique facilitated data acquisition with reduced detection delay.
• The modified adaptive cumulative chatter indicator demonstrated faster statistical decision-making compared to the adaptive cumulative log-likelihood ratio (ACLLR).
• The two-risk levels-based threshold enhanced detection reliability by overcoming single-threshold limitations.
• The proposed method was successfully applied to detect incipient chatter in a digital high-speed milling process.

Conclusions:

• The developed incipient chatter detection method effectively meets the time and reliability constraints of high dynamic applications.
• The combination of MSPR, adaptive indicator, and dual-risk thresholding offers a significant improvement over existing methods in noisy environments.
• This approach provides a reliable and efficient solution for early detection of chatter in high-speed milling operations.