| Autor: |
TSIAFIS, Ioannis, BOUZAKIS, K.-D., TSOLIS, Grigoris, XENOS, Thomas |
| Předmět: |
|
| Zdroj: |
Proceedings of the International Conference on Diagnosis & Prediction in Mechanical Systems Engineering; 2012, p260-266, 7p |
| Abstrakt: |
The scope of this work is to assess three spectral methods employed in journal bearing fault detection. To this end, two popular methods namely the Short Time Fourier Transform (STFM) and the Wavelet Transform and one innovative and very promising method, namely the Hilbert - Huang transform, were employed. Five experimentations were carried on employing five pairs of journal bearings. Vibration time series, measured by an accelerometer assembled on the base of the bearing, were obtained. All time series were processed by means of Fourier transform, wavelet transform and the Hilbert-Huang transform and the resulting spectra of each pair of bearings, sound and defective, were examined for possible differences. Both the Fourier Transform and the wavelet transform analysis did not reveal any differences between the spectra corresponding to sound and defective bearings. On the contrary, the analysis employing the Hilbert - Huang Transform revealed significant differences between the respective first five intrinsic mode functions (IMF) which reduced in magnitude as the order of the IMF increased i.e. as the spectral frequency decreased, whereas the Hilbert spectra obtained from the time series corresponding to sound and defective bearings strongly differed. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
