A signal processing approach for enhanced Acoustic Emission data analysis in high activity systems: Application to organic matrix composites. (March 2016)

• Record Type: Journal Article
• Authors: Kharrat, M.; Ramasso, E.; Placet, V.; Boubakar, M.L.
• Is Part Of: Mechanical systems and signal processing. Volume 70/71(2016)
• Page Start: 1038
• Page End: 1055
• Publication Date: 2016-03
• Subjects: Continuous AE signals -- CFRP -- Quasi-static tests -- Wavelet denoising -- Hit detection -- Pattern recognition; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2015.08.028
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: Structural elements made of Organic Matrix Composites (OMC) under complex loading may suffer from high Acoustic Emission (AE) activity caused by the emergence of different emission sources at high rates with high noise level, which finally engender continuous emissions. The detection of hits in this situation becomes a challenge particularly during fatigue tests. This work suggests an approach based on the Discrete Wavelet Transform (DWT) denoising applied on signal segments. A particular attention is paid to the adjustment of the denoising parameters based on pencil lead breaks and their influence on the quality of the denoised AE signals. The validation of the proposed approach is performed on a ring-shaped Carbon Fiber Reinforced Plastics (CFRP) under in-service-like conditions involving continuous emissions with superimposed damage-related transients. It is demonstrated that errors in hit detection are greatly reduced leading to a better identification of the natural damage scenario based on AE signals. Abstract : Highlights: The problem of continuous Acoustic Emission in CFRP composites was addressed. A high loading rate test under a high noise level has fathered continuous emission. The approach includes signal denoising, hit determination and feature extraction. It is managed to eliminate the hits saturation using signal segmentation. It allows a better identification of natural clusters using the GK algorithm.