Pimentel, Joao; Universitat Bremen Fachbereich 01 Physik/Elektrotechnik; Germany
Pimentel, J.V.; University of Bremen; Germany
Klemm, R.; University of Bremen; Germany
Irretier, A.; University of Bremen; Germany
Dalgic, M.; University of Bremen; Germany
Zoch, H.-W.; University of Bremen; Germany
Krieger, K.-L.; University of Bremen; Germany
Session: Acoustic Emission 1
Time: 16:20 - 16:40
Acoustic emissions (AE) in a solid body can be used to monitor its structural health. Lightweight structures subjected to high and varying load, such as lightweight truck frames, are particularly susceptible to fatigue failures. Piezoelectric sensors can be used to continuously monitor structure-borne sound. However, signal-processing strategies are required to properly detect AE events relevant to crack formation. Previous work from the authors showed success in the detection of macro- and microfractures in steel samples of truck trailer structures during static tensile tests and dynamic fatigue tests. The physical processes that lead to the formation and propagation of cracks typically involve a sudden energy release, generating signals that are typically short-lived and span a relatively wide frequency range compared to the background noise. In these aspects, AE signals exhibit similarities to percussive instruments, although in rather different scales.
This work describes a new music-inspired approach to analyse vibroacoustic data from tensile and fatigue tests. The application of the method presented involves a short-time Fourier transform (STFT) applied sequentially to the sampled vibroacoustic signals and analysed both over time and in the frequency domain to detect such “percussive onsets”. A time registry of the AE events and their intensities is generated.
The method presented here was applied in the post-processing of measurement data to identify the emergence of cracks. Its success was verified by quantitative comparison with previously used detection methods and further examination of the tested samples (e.g. metallographic analysis), as described in the paper. The processing steps, relevant parameters and their influence on the results are discussed, as well as their physical significance in the acoustic emission process.