Method and system for abnormality diagnosis of very low speed rotating machine

• US 8,534,128 B2
• Filed: 03/17/2008
• Issued: 09/17/2013
• Est. Priority Date: 03/26/2007
• Status: Expired due to Fees

• Current Assignee: Nippon Steel Corporation
• Original Assignee: Nippon Steel & Sumitomo Metal Corporation (Nippon Steel Corporation)
• Inventors: Murayama, Tsunemi
• Primary Examiner(s): Macchiarolo, Peter
• Assistant Examiner(s): Shah, Samir M
• Application Number:
  US12/593,195
• Publication Number:
  US 20100116060A1
• Time in Patent Office: 2,010 Days
• Field of Search: 735/93, 735/70, 735/79, 736/02, 736/46, 736/59, 736/60
• US Class Current: 73/593
• CPC Class Codes: F16C 19/527 related to vibration and noise
• CPC Class Codes: G01H 1/003 of rotating machines G01H1/10 takes precedence
• CPC Class Codes: G01M 13/04 Bearings
• CPC Class Codes: G01M 13/045 Acoustic or vibration analysis

Abstract

The present invention provides a high accuracy, inexpensive diagnosis method and system diagnosing an abnormality of a 10 rpm or less very low speed roller bearing using an acoustic signal, that is, detects an acoustic signal emitted from a roller bearing of a 10 rpm or less ultralow speed rotating machine using a sensor having a sound pickup performance of 10 kHz or less, uses frequency analysis to find the spectrum distribution, processes this to obtain the frequency analysis waveform, compares the spectral average value or effective value of a frequency of the natural frequency of the bearing or more of this with a reference value, and judges that the bearing is abnormal when this spectral average value or effective value exceeds the reference value.

Claims

1. A method for abnormality diagnosis of a roller bearing of a very low speed rotating machine comprising detecting an acoustic signal emitted from the roller bearing of the rotating machine accompanying a very low speed rotation of 10 rpm or less being diagnosed so as to diagnose any abnormality in the roller bearing,

said method for abnormality diagnosis of the roller bearing of the very low speed rotating machine comprising:

bringing a sensor having a sound pickup function into direct contact with said roller bearing to detect sound emitted from the roller bearing being detected as sound pressure, suitably amplifying it to obtain an acoustic signal, and finding a spectrum distribution of the same by frequency analysis;

finding a spectral average value or effective value of a frequency more than or equal to the natural frequency of the roller bearing and less than 10 kHz from a frequency analysis waveform obtained by said frequency analysis; and

comparing said spectral average value or effective value with a separately found reference value and judging the roller bearing to be abnormal when exceeding said reference value.

2. A method for abnormality diagnosis of a roller bearing of a very low speed rotating machine as set forth in claim 1, further comprising filtering out the natural frequencies of components of the rotating machine.

3. A method for abnormality diagnosis of a roller bearing of a very low speed rotating machine comprising detecting an acoustic signal emitted from the roller bearing of the rotating machine accompanying a very low speed rotation of 10 rpm or less being diagnosed so as to diagnose any abnormality in the roller bearing,

said method for abnormality diagnosis of the roller bearing of the very low speed rotating machine comprising:

bringing a sensor having a sound pickup function into direct contact with said roller bearing to detect sound emitted from the roller bearing being detected as sound pressure, suitably amplifying it to obtain an acoustic signal, and finding a spectrum distribution of the same by frequency analysis;

finding a spectral average value or effective value of a frequency more than or equal to the natural frequency of the roller bearing and less than 10 kHz from a frequency analysis waveform obtained by said frequency analysis;

comparing said spectral average value or effective value with a separately found reference value and judging the roller bearing to be abnormal when exceeding said reference value; and

outputting a signal obtained by amplifying said acoustic signal in an audible manner and screening the roller bearing for judgment of the abnormality.

4. A method for abnormality diagnosis of a roller bearing of a very low speed rotating machine as set forth in claim 3 wherein said acoustic signal is passed through a high frequency band pass filter of a frequency that is greater than or equal to the natural frequency of the roller bearing.

5. A system for abnormality diagnosis of a roller bearing of a very low speed rotating machine detecting an acoustic signal emitted from the roller bearing of the rotating machine being diagnosed along with a very low speed rotation of 10 rpm or less so as to diagnose an abnormality of the roller bearing,

said system for abnormality diagnosis of the roller bearing of the very low speed rotating machine comprising:

a sensor having a sound pickup function;

an amplifier amplifying an acoustic signal obtained by said sensor;

a frequency analysis part finding a frequency distribution of said acoustic signal;

a storage part of a natural frequency of the roller bearing;

a processing part finding a spectral average value or effective value of a frequency more than or equal to the natural frequency of the roller bearing and less than 10 kHz based on said frequency analysis;

a storage part storing a separately found reference value; and

an abnormality judgment part comparing said spectral average value or effective value with said reference value to judge if there is any abnormality,

wherein said sensor is configured to directly contact said roller bearing to thereby obtain said acoustic signal as sound pressure.

6. A system for abnormality diagnosis of a roller bearing of a very low speed rotating machine as set forth in claim 5, further comprising a filter filtering out the natural frequencies of components of the rotating machine.

7. A system for abnormality diagnosis of a roller bearing of a very low speed rotating machine detecting an acoustic signal emitted from the roller bearing of the rotating machine being diagnosed along with a very low speed rotation of 10 rpm or less so as to diagnose an abnormality of the roller bearing,

said system for abnormality diagnosis of the roller bearing of the very low speed rotating machine comprising:

a sensor having a sound pickup function;

an amplifier amplifying an acoustic signal obtained by said sensor;

a frequency analysis part finding a frequency distribution of said acoustic signal;

a storage part of a natural frequency of the roller bearing;

a processing part finding a spectral average value or effective value of a frequency more than or equal to the natural frequency of the roller bearing and less than 10 kHz based on said frequency analysis;

a storage part storing a separately found reference value;

an abnormality judgment part comparing said spectral average value or effective value with said reference value to judge if there is any abnormality; and

an output part outputting said amplified acoustic signal in an audible manner,

wherein said sensor is configured to directly contact said roller bearing to thereby obtain said acoustic signal as sound pressure.

8. A system for abnormality diagnosis of a roller bearing of a very low speed rotating machine as set forth in claim 7, further comprising being provided with a high frequency band pass filter extracting only a frequency band component of said amplified acoustic signal,

wherein the frequency band includes a frequency that is greater than or equal to the natural frequency of the roller bearing.