Automatic Event Detection in Noisy Environment for Material Process Monitoring by Laser AE Method

Abstract

Abstract. Laser acoustic emission (AE) method is a unique in-situ and non-contact nondestructive evaluation (NDE) method. It has a capability to detect signals generated from crack generation and propagation, friction and other physical phenomena in materials even in high temperature environment. However, laser AE system has lower signal-to-noise ratio compared to the conventional AE system using PZT sensors, so it is difficult to apply this method in noisy environment. A novel AE measurement system to detect events in such difficult environments was developed. This system could continuously record all AE waveforms and enable unrestricted post-analyses. Noise reduction filters in frequency domain coupling with a new AE event extraction using multiple threshold values showed a good potential for AE signal processing. This system was successfully applied for crack monitoring of plasma spray deposition process of ceramic coating.

Empirical Modelling of Acoustic Emission Impulses

“…A morphological method for filtering unsteady noises was proposed in [7]; application of morphological operations to signal fragments, such as narrowing and expansion, allows filtering of pulsed noise of short duration. Filtering of unsteady impulse noise based on the distribution function of time intervals is considered in [8][9].A common drawback of the above filtering algorithms is the rather formal nature of the description of the signal and noise model. The filtering efficiency is considered on the basis of particular examples; there is no comparison of the shape of AE impulses with waveguide parameters.…”