Measuring and testing – Vibration – By mechanical waves
Reexamination Certificate
2001-12-11
2004-01-27
Williams, Hezron (Department: 2856)
Measuring and testing
Vibration
By mechanical waves
C073S660000
Reexamination Certificate
active
06681634
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to rotating element bearings, and, more particularly, to methods and devices for identifying rolling element bearing defects.
BACKGROUND OF THE INVENTION
Prior Art
There are several processes used to identify a rolling element bearing defect. Some of the more common methods include (1) comparing a spectrum of vibration measurements with known bearing fault frequencies, (2) measuring and trending high frequency acoustic bearing noise, and (3) analyzing the modulation of an accelerometers natural frequency induced by bearing defects.
In accordance with the first known method, vibration measurements are displayed in either acceleration (g's) or velocity (inches/second). A spectrum of either measurement is generated and investigated to determine if any known bearing fault frequencies are present. Vibration measurements displayed in either acceleration or velocity have one major shortcoming. The known bearing fault frequency amplitude levels are extremely small compared to other rotating equipment vibration levels such as unbalance, misalignment, cavitation, and vane pass. The bearing fault frequency signals are often lost in more predominant vibration signatures.
Measuring and trending of high frequency acoustic noise coming from a bearing housing can also provide an indication of bearing defects. This is because the amount of high frequency acoustic bearing noise will increase as a bearing deteriorates, thereby indicating a deteriorating bearing condition. Measurement of high frequency acoustic noise is a very sensitive way to measure bearing faults. The drawback with this type of bearing detection measurement method is that there are other sources of high frequency acoustic noise found in centrifugal pumps. Pump cavitation, pump recirculation, dry running seals, rubbing laby seals, and pump-motor coupling interference can all be a source of high frequency acoustic noise.
Measuring an accelerometers natural frequency can also provide an indication of a bearing defect. Portable vibration equipment is employed to measure items such as, Spike Energy, HFD, and Peak View. Each of these approaches utilizes the concept whereby the impacts from bearing defects excite the natural frequency of the attached accelerometer. Digital signal processing monitors the excited accelerometer's natural frequency. That signal can be either displayed as an overall level or further analyzed. Further analysis of the excited natural frequency of the accelerometer involves filtering, enveloping and spectrum analysis to detect the presence of any of the known bearing defect frequencies. Bearing defect frequencies are obtained using the physical dimensions of the bearing and established equations.
As is well known in the art, early indications of bearing problems produce frequencies ranging from 250 to 350 kHz. As the wear on the bearing increases, the frequencies drop to around 20 to 60 kHz (1.2M to 3.6M CPM). It is well known in the art how to measure these frequencies, as well as the equations which solve for the frequencies that are involved. In later stages, the bearing defects began to ring at the natural frequencies of the bearing which occur in the range of 30 k to 120 k CPM. The wearing of bearings result in defects which can be expressed in terms of changes in frequency. One can then detect such frequencies as well as harmonics of such frequencies to provide data on bearing life.
There are many publications which describe this, including a publication provided by the Technical Associates of Charlotte, Inc., Copyright 1994 and showing equations as well as other data showing spectral and frequency responses relating to bearing defects. This publication has number R-0894-4 and is incorporated herein by reference.
Measuring the accelerometer's natural frequency has some appealing advantages. The natural frequency of a typical accelerometer is about 20-40 kHz. First, it uses the Accelerometer's Amplification Factor at its natural frequency as a built in amplifier of very low amplitude bearing defect frequencies. Second, filtering out low frequency signals (typically 5000 hertz and below) eliminates the traditional high amplitude pump/impeller generated frequencies such as 1×pump speed and vane pass frequency. Third, the enveloping process transposes a high frequency signal into a time wave form containing only low frequency signals that are easily detectable using standard vibration analysis digital signal processing. A drawback of this system is that it contains a lot of noise that results in a high spectrum noise base. This high-level noise base can easily mask or conceal the bearing defect frequencies.
SUMMARY OF THE INVENTION
This invention combines the advantages of measuring accelerometer signals with Time Synchronous Averaging (TSA) of the accelerometer signals to remove all the vibration frequencies that are not synchronized to the rotational speed of a rotating pump or other rotating equipment. Digital-signal processing is employed that further reduces the electrical noise thereby further reducing the spectrum noise base.
According to one inventive aspect, a method of identifying a rotating bearing defect includes the steps of: measuring an accelerometer signal generated at least in part by the rotating bearing to obtain a waveform signal; filtering the waveform signal using a bandpass filter to remove unwanted signal frequencies; enveloping the filtered waveform signal to obtain an enveloped low frequency time waveform; measuring the rotational speed of the rotating bearing; synchronizing the enveloped low frequency time waveform to the rotational speed of the rotating bearing and the repetitive phase relationship between trigger and bearing defect to obtain a time synchronized waveform; averaging the time synchronized waveform with at least one previously stored time synchronized waveform to obtain an average time synchronized waveform (TSA); spectrum analyzing the TSA; and identifying an amplitude of the TSA at rotating bearing defect frequencies and their multiples.
REFERENCES:
patent: 3699806 (1972-10-01), Weichbrodt
patent: 3733892 (1973-05-01), Rennick
patent: 4872337 (1989-10-01), Watts et al.
patent: 5663894 (1997-09-01), Seth et al.
patent: 6298725 (2001-10-01), Forrester
patent: 6526356 (2003-02-01), DiMaggio et al.
patent: 57179625 (1982-11-01), None
patent: 60076640 (1985-05-01), None
patent: 02222818 (1990-09-01), None
Hauenstein Kenneth L.
Henyan Oakley
Lorenc Jerome A.
Sabini Eugene P.
ITT Manufacturing Enterprises Inc.
Miller Rose M.
Williams Hezron
LandOfFree
Bearing defect detection using time synchronous averaging... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Bearing defect detection using time synchronous averaging..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Bearing defect detection using time synchronous averaging... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3253387