Acoustic Emission Monitoring in Machining Processes

Acoustic emission monitoring exploits the elastic waves generated by rapid stress redistributions within a material during cutting, grinding or milling. These transient high-frequency signals contain rich information on micro-mechanical events such as crack initiation, chip formation and tool–workpiece interaction. Owing to their sensitivity to minute deformation processes, acoustic emissions offer superior temporal resolution compared to force or vibration measurements, enabling real-time assessment of tool condition and process stability. Modern implementations integrate advanced signal-processing techniques—such as wavelet packet transforms—and machine-learning algorithms to extract meaningful features from noisy data streams. This approach allows for the prediction of tool flank wear, onset of chatter and deterioration of surface integrity with high accuracy. The global drive towards intelligent manufacturing has elevated acoustic emission monitoring as a cornerstone of predictive maintenance and quality assurance, reducing unplanned downtime, optimising cutting parameters and minimising material waste. Despite challenges in sensor placement, bandwidth selection and feature thresholding, ongoing research is refining robust strategies for noise suppression and automated decision-making within the framework of Industry 4.0.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Acoustic Emission Monitoring in Machining Processes publication trend

The graph below shows the total number of articles in acoustic emission monitoring in machining processes across all publications each year (not limited to Nature Index journals).

Technical terms

Acoustic emission (AE): High-frequency elastic waves generated by rapid stress changes in a material under load.

Counts parameter: The number of times an AE signal’s amplitude exceeds a predefined threshold during a measurement interval.

Wavelet packet transform (WPT): A signal-decomposition technique that represents a signal in both time and frequency domains at multiple resolutions.

Flank wear: The width of material loss on the tool’s flank face used to quantify cutting-tool degradation.

Hidden Markov model (HMM): A statistical model that describes a system as a sequence of observable signals generated by underlying, unobservable states.

Random forest regression: An ensemble-learning method using multiple decision trees to predict continuous outcomes and reduce overfitting.

References

1. Identification of tool wear using acoustic emission signal and machine learning methods. Precision Engineering (2021).
2. A Novel Machine Learning-Based Methodology for Tool Wear Prediction Using Acoustic Emission Signals. Sensors (2021).
3. Analysis of Spindle AE Signals and Development of AE-Based Tool Wear Monitoring System in Micro-Milling. Journal of Manufacturing and Materials Processing (2022).