Issue |
Eur. Phys. J. AP
Volume 4, Number 2, November 1998
|
|
---|---|---|
Page(s) | 203 - 211 | |
DOI | https://doi.org/10.1051/epjap:1998261 | |
Published online | 15 November 1998 |
https://doi.org/10.1051/epjap:1998261
Study of monitoring load failure of an actuator system
CEGELY, Université Claude Bernard, 69622 Villeurbanne, France
Corresponding author: gael@leep.univ-lyon1.fr
Received:
22
July
1997
Revised:
29
May
1998
Accepted:
10
July
1998
Published online: 15 November 1998
This paper addresses the problem of an actuator's load monitoring through a particular fault pattern effects analysis. A load's failure, like a dip of torque superimposed on a constant or sinusoidal steady state load, has an effect on the motor's supply currents. In some cases the high stress and the toxicity of the operating environment of the actuator lead to the inaccessibility of the motor and load set and don't authorize the positioning of sensors. Consequently we use the motor current spectral analysis for the detection of the faults that occur in the load. To propose a global method, this paper presents an approach for the automation of the monitoring and diagnosis processes and studies the capabilities of an easy tool, the spectrogram, to detect the apparition of the fault during the work of the actuator and the signature obtained with different kind of faults and supplies. It also proposes a methodology to determine the parameters of the dip of torque fault. To conclude this study, we show the validity area of the method to other king of failure patterns.
PACS: 84.50.+d – Electric motors / 02.30.Nw – Fourier analysis / 07.05.Kf – Data analysis: algorithms and implementation; data management
© EDP Sciences, 1998
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.