Integrated vibe/ODM fusion and trending analysis for prognostic health management of engine bearing
A method for fault diagnosis of a bearing includes detecting, using an oil debris monitor (ODM) sensor, ODM data corresponding to an amount of debris flowing downstream from the bearing. The method also includes detecting, using a vibration sensor, vibration data corresponding to vibration of the bearing during use. The method also includes determining, by a controller, a vibration stage flag corresponding to a severity of damage of the bearing based on the vibration data. The method also includes determining, by the controller, a severity level of the damage of the bearing based on a combination of the vibration stage flag and the ODM data. The method also includes outputting, by an output device, the severity level.
1. A method for fault diagnosis of a bearing, comprising:
detecting, using an oil debris monitor (ODM) sensor, ODM data corresponding to an amount of debris flowing downstream from the bearing;
detecting, using a vibration sensor, vibration data corresponding to vibration of the bearing during use;
determining, by a controller, a vibration stage flag corresponding to a severity of damage of the bearing based on the vibration data;
determining, by the controller, a severity level of the damage of the bearing based on a combination of the vibration stage flag and the ODM data; and
outputting, by an output device, the severity level,
wherein:
determining the vibration stage flag based on the vibration data includes comparing the vibration data to a reference vibration data that was detected during an initial use of the bearing,
determining the vibration stage flag includes determining the vibration stage flag based on a time-domain analysis of the vibration data and a frequency-domain analysis of the vibration data, and
determining the vibration stage flag based on the frequency-domain analysis of the vibration data includes determining a frequency flag based on a maximum value of an analysis of at least two of a fundamental train frequency of the bearing, a ball pass outer race frequency, a ball pass inner race frequency, and a ball defect frequency.
2. The method of claim 1 , wherein determining the vibration stage flag based on the time-domain analysis of the vibration data includes determining the vibration stage flag based on root mean square (RMS), Kurtosis, and Skewness values of the vibration data.
3. The method of claim 1 , wherein determining the vibration stage flag further includes determining a time flag corresponding to the time-domain analysis of the vibration data and setting the vibration stage flag if either the time flag or the frequency flag is set.
4. The method of claim 1 , wherein determining the vibration stage flag includes computing a resonance frequency and a frequency band of the resonance frequency, bandpass filtering the vibration data to obtain a filtered signal, demodulating the filtered signal to obtain a demodulated signal, and performing time-domain and frequency-domain analyses of the demodulated signal.
5. The method of claim 1 , wherein determining the vibration stage flag based on the time-domain analysis of the vibration data includes normalizing the vibration data in a time domain, and wherein determining the vibration stage flag based on the frequency-domain analysis of the vibration data includes normalizing the vibration data in a frequency domain.
6. The method of claim 1 , wherein:
detecting the vibration data includes detecting vibration data from multiple vibration sensors; and
determining the vibration stage flag based on the vibration data includes: determining a vibration stage flag based on the vibration data from each of the vibration sensors, and determining the vibration stage flag to be a maximum value of all of the vibration stage flags determined from each of the vibration sensors.
7. The method of claim 1 , further comprising:
storing, in a memory, a decision table; and
determining, by the controller, a residual life of the bearing by comparing the combination of the vibration stage flag and the ODM data to the decision table.
8. A method for fault diagnosis of a bearing used in an engine, comprising:
detecting, using an oil debris monitor (ODM) sensor, ODM data corresponding to an amount of debris flowing downstream from the bearing;
detecting, using a vibration sensor, vibration data corresponding to vibration of the bearing during use;
determining, by a controller, a vibration stage flag corresponding to a severity of damage of the bearing based on a time-domain analysis of the vibration data and a frequency-domain analysis of the vibration data;
determining, by the controller, a severity level of the damage of the bearing based on a combination of the vibration stage flag and the ODM data; and
outputting, by an output device, the severity level,
wherein:
determining the vibration stage flag based on the vibration data includes comparing the vibration data to a reference vibration data that was detected during an initial use of the bearing, and
determining the vibration stage flag based on the frequency-domain analysis of the vibration data includes determining a frequency flag based on a maximum value of an analysis of at least two of a fundamental train frequency of the bearing, a ball pass outer race frequency, a ball pass inner race frequency, and a ball defect frequency.
9. The method of claim 8 , wherein determining the vibration stage flag based on the time-domain analysis of the vibration data includes determining the vibration stage flag based on root mean square (RMS), Kurtosis, and Skewness values of the vibration data.
10. The method of claim 8 , wherein determining the vibration stage flag further includes determining a time flag corresponding to the time-domain analysis of the vibration data, determining a frequency flag corresponding to the frequency-domain analysis, and setting the vibration stage flag if either the time flag or the frequency flag is set.
11. The method of claim 8 , wherein determining the vibration stage flag includes computing a resonance frequency and a frequency band of the resonance frequency, bandpass filtering the vibration data to obtain a filtered signal, demodulating the filtered signal to obtain a demodulated signal, and performing time-domain and frequency-domain analyses of the demodulated signal.
12. The method of claim 8 , wherein determining the vibration stage flag based on the time-domain analysis of the vibration data includes normalizing the vibration data in a time domain, and wherein determining the vibration stage flag based on the frequency-domain analysis of the vibration data includes normalizing the vibration data in a frequency domain.
13. The method of claim 8 , wherein:
detecting the vibration data includes detecting vibration data from multiple vibration sensors; and
determining the vibration stage flag based on the vibration data includes: determining a vibration stage flag based on the vibration data from each of the vibration sensors, and determining the vibration stage flag to be a maximum value of all of the vibration stage flags determined from each of the vibration sensors.
14. A system for fault diagnosis of bearings, comprising:
an oil debris monitor (ODM) sensor configured to detect ODM data corresponding to an amount of debris flowing downstream from a bearing;
a vibration sensor configured to detect vibration data corresponding to vibration of the bearing during use;
a controller configured to:
determine a vibration stage flag corresponding to a severity of damage of the bearing based on the vibration data including comparing the vibration data to a reference vibration data that was detected during an initial use of the bearing, and
determine a severity level of the damage of the bearing based on a combination of the vibration stage flag and the ODM data; and
an output device configured to output the severity level, wherein:
determining the vibration stage flag includes determining the vibration stage flag based on a time-domain analysis of the vibration data and a frequency-domain analysis of the vibration data, and
determining the vibration stage flag based on the frequency-domain analysis of the vibration data includes determining a frequency flag based on a maximum value of an analysis of at least two of a fundamental train frequency of the bearing, a ball pass outer race frequency, a ball pass inner race frequency, and a ball defect frequency.
15. The system of claim 14 , wherein the time-domain analysis of the vibration data is based on root mean square (RMS), Kurtosis, and Skewness values of the vibration data.