Method and apparatus for analyzing rotating machines

• US 5,109,700 A
• Filed: 05/30/1991
• Issued: 05/05/1992
• Est. Priority Date: 07/13/1990
• Status: Expired due to Fees

• Current Assignee: Life Systems, Inc.
• Original Assignee: Life Systems, Inc.
• Inventors: Hicho, Michael D.
• Primary Examiner(s): Williams, Hezron E.
• Assistant Examiner(s): ARANA, LOUIS M
• Application Number:
  US07/707,441
• Time in Patent Office: 341 Days
• Field of Search: 73/660, 73/661, 73/593, 73/587, 73/602, 73/488
• US Class Current: 73/660
• CPC Class Codes: F04C 2270/125 Controlled or regulated
• CPC Class Codes: F04C 28/28 Safety arrangements; Monitoring
• CPC Class Codes: F16C 19/527 related to vibration and noise
• CPC Class Codes: F16C 2233/00 Monitoring condition, e.g. temperature, load, vibration
• CPC Class Codes: G01H 1/003 of rotating machines G01H1/10 takes precedence
• CPC Class Codes: G01M 13/045 Acoustic or vibration analysis
• CPC Class Codes: G01N 2291/2696 Wheels, Gears, Bearings

Abstract

A vibration transducer (22) is mounted to a rotating machine (20) for sensing vibration thereof. An output electrical signal from the vibration transducer is analyzed to generate a level display (18) of vibrational displacement per unit time, a speed display (22) indicative of rotational speed, and a bearing condition display (20) indicative of bearing condition, all displays derived directly from the vibration transducer signal. The electrical signal is transformed (36) into a frequency spectrum that has an amplitude for each of a plurality of narrow frequency ranges or bins. Each frequency bin has a characteristic center frequency and a predefined width or band of frequencies. A speed analysis program (38) identifies a set of at least first, second and third order related frequency bins, i.e. frequency bins whose center frequencies are an even multiple of each other, that have a significantly high amplitude and provides the lowest bin center frequency as a control signal to the speed display. A bearing condition analysis program (40) eliminates the bins that are integer multiples of the running speed and lower frequency, e.g. less than third order, identifies sets of bins with relatively large amplitudes that are integer multiples of each other, and selects the set of bins with the largest amplitude as being controlling of bearing condition. After elimination of frequencies not indicative of bearing condition, the bearing defect frequency remains. The amplitude of this frequency is displayed on bearing condition readout (22). The amplitude indicating the severity of the bearing defect.

Claims

1. A method for determining a condition of a rolling element bearing enclosed in a housing of a rotating machine, the method comprising:

sensing, from a vibration transducer input means in contact with said housing, mechanical vibrations of the rotating machine;

converting the sensed vibrations into an electrical signal;

converting the electrical signal into a frequency spectrum;

determining a running speed of the rotating machines from the frequency spectrum;

analyzing the frequency spectrum to identify frequency components characteristic of problems other than bearing defects;

eliminating the frequency components that characterize problems other than bearing defects;

determining an indicator of the rolling element bearing condition from a remaining at least one of the frequency components.

2. The method of claim 1 wherein the step of analyzing the frequency spectrum includes:

selecting larger frequency peaks of the frequency spectrum; and

,

refining a frequency resolution of the selected frequency peaks with data from adjacent frequency peaks.

3. The method of claim 1 wherein the step of eliminating the frequency components that characterize problems other than bearing defects includes:

removing low order frequency peaks from the frequency spectrum that are equal to or less than a multiple of the running speed;

removing the frequency peaks from the frequency spectrum that are integer multiples of the running speed; and

,

averaging corresponding remaining frequency peaks.

4. The method of claim 3 wherein the selecting step selects an averaged remaining frequency peak with the highest amplitude as the indication of the bearing condition.

5. An apparatus for determining an indicator of a condition of a rolling element bearing enclosed in a housing of a rotating machine, the apparatus comprising:

a vibration transducer means in contact with said housing means for sensing mechanical vibrations of the rotating machine and converting the sensed vibrations into an analog signal proportional to the sensed vibrations;

an analog-to-digital convertor means for converting the analog signal into a digital signal;

a means for converting the digital signal into a frequency spectrum;

an analyzing means for analyzing the frequency spectrum to identify frequency components characteristic of problems other than bearing defects;

a means for eliminating the frequency components related to problems other than bearing defects;

a means for selecting a remaining frequency component having a highest amplitude as the indicator of the rolling element bearing condition.

6. The apparatus as set forth in claim 5 further including a means affixed to the rotating machine for determining a speed of said rotating machine.

7. An apparatus for determining a vibration signal frequency component indicative of a condition of a rolling element bearing of a rotating machine, the apparatus comprising:

a vibration detector means for detecting and converting vibrations in the rotating machine into an electrical signal proportional to the vibrations;

a filtering means for filtering the electrical signal to obtain select frequencies;

an analog-to-digital convertor means to convert the electrical signal into a digital signal;

a means for converting the digital signal into a frequency spectrum;

a selecting means for selecting frequency components of the frequency spectrum characteristic of running speed and frequency components that are random, spurious and caused by other effects;

an eliminating means for eliminating the frequency components that are random, spurious and caused by other effects;

a determining means for determining running speed from the frequency components remaining after the step of eliminating the frequency components that are random, spurious and caused by other effects;

a means for analyzing the frequency spectrum to identify frequency components characteristic of problems other than bearing defects;

an eliminating means for eliminating the frequency components related to problems other than bearing defects;

a means for storing a plurality of remaining frequency components which are not eliminated by the eliminating means;

an averaging means for averaging corresponding stored frequency components to produce a plurality of averaged frequency components; and

,

a means for selecting an averaged frequency component having a highest amplitude as the frequency components indicative of the rolling element bearing condition.

8. An apparatus for determining bearing condition comprising:

a vibration sensor for sensing vibration arising from a rolling element bearing and generating an electrical signal indicative thereof;

a running speed calculating means for calculating a running speed of the rolling element from the electrical signal including;

a transform means for transforming the electrical signal into a frequency spectrum including a plurality of frequency bins, each bin having a preselected frequency band, a center frequency, and a magnitude indicative of a strength of sensed frequency components within the corresponding frequency band,

a means for identifying a set of frequency bins in which the center frequencies are even multiples of each other and in which the collective amplitudes of the frequency bins exceed the amplitude of other sets of frequency bins with center frequencies that are even multiples of each other, and

a means for selecting the lowest center frequency of the identified set as the running speed; and

a bearing condition means for generating a bearing condition indicating signal from the electrical signal and the selected lowest center frequency.

9. The apparatus as set forth in claim 8 wherein the bearing condition means includes:

a means for removing running speed indicative frequency components from the electrical signal;

a means for calculating an average vibrational displacement per unit time from the electrical signal;

a means for calculating an average vibrational displacement per revolution from the average vibrational displacement per unit time and the running speed;

a means for comparing the frequency spectrum of the electrical signal with the running speed indicative frequency components removed with preselected frequency spectrums each indicative of a preselected corresponding bearing condition.

10. A method for determining an indicator of a condition of a rolling element bearing from a vibration signal obtained from a rotating machine, the method comprising:

a) converting vibrations in the rotating machine into an electrical signal proportional to the vibrations;

b) filtering the electrical signal;

c) converting the electrical signal into a digital signal;

d) converting the digital signal into a frequency spectrum;

e) selecting frequency peaks of the frequency spectrum characteristic of an actual running speed of the rotating machine, and frequency components that are random, spurious and caused by other effects;

f) eliminating the frequency components that are random, spurious and caused by other effects;

g) calculating a final running speed signal representing the actual running speed, from the frequency peaks remaining after the eliminating step;

h) analyzing the frequency spectrum to identify frequency peaks characteristic of problems other than bearing defects;

i) eliminating the frequency peaks related to problems other than bearing defects;

j) storing frequency peaks remaining after the step of eliminating problems other than bearing defects;

k) repeating at least steps (h), (i), and (j) to store a plurality of frequency peaks;

l) averaging corresponding stored frequency peaks; and

m) selecting an averaged frequency peak having a highest amplitude as the indicator of the rolling element bearing condition.

11. A method for determining an indicator of a condition of a rolling element bearing from a vibration signal obtained from the rotating machine, the method comprising:

converting vibrations in the rotating machine into an electrical signal proportional to the vibrations;

filtering the electrical signal;

converting the electrical signal into a frequency spectrum;

selecting (i) frequency peaks of the frequency spectrum which are characteristic of an actual running speed of the rotating machine and (ii) frequency peaks that are random, spurious and caused by other effects by;

selecting frequency peaks in the frequency spectrum to be designated as significant peaks, when the peaks are at least 0.005 in/sec, are at least as high as a next lower frequency peak, and are higher than a next higher frequency peak,

selecting such significant peaks

in a first band of frequencies 20% less than a nominal speed setting to 10% more than the nominal speed setting,

in a second band of frequencies 20% less than two times the nominal speed setting to 10% more than two times the nominal speed setting, and

in a third band of frequencies 20% less than three times the nominal speed setting to 10% more than three times the nominal speed setting;

eliminating frequency peaks that are random, spurious and caused by other effects;

determining a calculated final running speed corresponding to the actual running speed, from the frequency peaks remaining after the eliminating step;

analyzing the frequency spectrum to identify frequency peaks characteristic of problems other than bearing defects;

eliminating the frequency peaks related to problems other than bearing defects;

selecting a remaining frequency peak having a highest amplitude as the indicator of the rolling element bearing condition.

12. The method of claim 11 wherein the step of determining the final running speed includes:

summing energy levels of the significant peaks in the first band to obtain a first energy sum;

summing energy levels of the significant peaks in the second band to obtain a second energy sum;

summing energy levels of the significant peaks in the third band to obtain a third energy sum;

weighting each of the first, second, and third energy sums in accordance with proximity to a last previously determined final running speed;

selecting one of the first, second, and third weighted energy sums having the highest weighted energy sum value;

calculating an estimated running speed from the selected weighted energy sum;

comparing the calculated estimated running speed with at least one previously determined final running speed to see if they are within +/-20 RPM;

producing a non-detection signal when the comparing step finds a difference of more than +/-20 RPM;

identifying corresponding frequency peaks of greater than the third order;

comparing the calculated estimated running speed with a running speed corresponding to the greater than third order frequency peaks;

selecting one of the greater than third order frequency peak running speed and the estimated running speed, as a representation of the final running speed.

13. The method of claim 11 wherein the step of eliminating the frequency peaks related to problems other than bearing defects includes:

(a) eliminating significant peaks equal to or less than a preselected multiple of the final running speed;

(b) eliminating all integer multiples of the final running speed;

(c) looping through selected steps of the method to obtain multiple sets of peaks corresponding to the significant peaks remaining after steps (a) and (b);

(d) averaging the corresponding peaks.