Measuring and testing – Vibration – By mechanical waves
Reexamination Certificate
2002-03-27
2004-10-12
Williams, Hezron (Department: 2856)
Measuring and testing
Vibration
By mechanical waves
C073S593000, C073S660000
Reexamination Certificate
active
06802221
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to systems and methods for condition-based monitoring of machines. More specifically, the present invention pertains to the condition-based monitoring of bearing assemblies.
BACKGROUND OF THE INVENTION
In conducting a condition-based maintenance (CBM) program for machines, such as transportation machines of locomotives or other mobile assets, a single analyst using physical evaluation and a knowledge base or database can make a decision on the relative health of various components of the machine. One example of such components includes rotating components, e.g., a bearing.
The basis for determining the presence of bearing defects is the identification of amplitude peaks at certain frequencies in the vibration signature of the bearing.
Equations, known to those skilled in the art, relate the roller diameter, pitch diameter, number of rollers, and contact angle to “fault frequencies” associated with ball, race and cage damage. The presence of these fault frequencies in the vibration spectrum is indicative of bearing damage, and the relative amplitude of the measured parameters and the complexity of the features is a measure of the severity of the damage.
Certain methods of vibrational analysis of bearing machines may assume that various vibrational characteristics, such as the running speed and/or rotating speed are unknown. The frequencies of the bearing fault features are also assumed to be unknown. Moreover, the known geometric characteristics used to calculate fault frequencies are also assumed to be unknown. Accordingly, sample spectra are obtained to identify the various fault features. The data obtained from the sample spectra establish a baseline, or threshold parameters for the machine. However, in the operation of similar machine assets subject to a condition-based maintenance system, the generation of threshold requirements for each individual bearing assembly in the population may not be practical. Moreover, the identification of fault frequencies, running speeds or other operating parameters, may not effectively consider margin of errors that may otherwise be obtained from diagnostics of like bearing assemblies.
BRIEF DESCRIPTION OF THE INVENTION
The present invention is for a system and method for monitoring the condition of a bearing assembly by analyzing spectral data that is representative of the vibrational movement of the bearing assembly. The system generally includes a sensor, such as an accelerometer or vibration sensor, placed in proximity to the bearing assembly. The sensor generates a signal indicative of an amplitude and frequency of the vibrational movement of the bearing assembly. A processor, in communication with the sensor, receives the signals generated by the sensor, and generates spectral data representative of the vibrational movement of the bearing assembly with respect to the amplitude and frequency of the bearing vibrational movement.
A database, in communication with the processor, comprises data representative of a predetermined amplitude threshold for at least one bearing fault, and at least one predetermined frequency (also referred to as a “fault frequency”), wherein each frequency is characteristic of at least one bearing fault. These fault frequencies, and the amplitude threshold, are obtained from spectral data representative of the vibrational movement of a population of like bearing assemblies. In a preferred embodiment, an average overall amplitude is calculated and represents an average vibrational energy measure of the population of like bearing assemblies. The amplitude threshold may be empirically calculated as being contained within some standard deviation of the mean overall amplitude of the population of like bearing assemblies. In addition, the fault frequencies are estimated to appear within certain predetermined ranges of frequencies, based on data obtained from analysis of vibrational movement of the population of bearing assemblies.
Frequencies at which amplitudes exceed the predetermined minimum amplitude threshold and are within a predetermined frequency range for respective fault frequencies are compared to the amplitude threshold in order to evaluate the overall condition of the bearing assembly. The term “peak” may also be used as frequency, such that a peak appears on a spectrum and is characterized by a frequency and an amplitude measurement. In a preferred embodiment, the number of frequencies for each of the respective fault frequencies are counted if their amplitude exceeds the amplitude threshold, or a predetermined multiple thereof. A penalty, representative of a fault feature, is then added to the number of frequencies counted (“peak count”). A score that is representative of the overall condition of the bearing assembly is calculated by a summation of the individual penalties and peak counts. In this manner, a decision relative to the mechanical health of the bearing may be made by comparing features of a vibration frequency spectrum of the bearing to amplitude thresholds that have been determined empirically or statistically by analyzing similar bearing assemblies having similar mechanical and fault characteristics.
In addition, individual scores may be assigned to the different fault frequencies wherein each individual score may be indicative of a specific non-normal condition within the bearing. This capability to capture and evaluate individual scores is important data, which can be used as feedback to the total assembly design agency. Through the identification of such information, the selection and method of installation and manufacture of said bearings may be improved in cases where the quantities and time for production allow.
REFERENCES:
patent: 4129276 (1978-12-01), Svet
patent: 5109700 (1992-05-01), Hicho
patent: 5115671 (1992-05-01), Hicho
patent: 5140858 (1992-08-01), Nishimoto et al.
patent: 5381692 (1995-01-01), Winslow et al.
patent: 5446451 (1995-08-01), Grosskopf, Jr.
patent: 5922963 (1999-07-01), Piety et al.
patent: 6053047 (2000-04-01), Dister et al.
patent: 6321602 (2001-11-01), Ben-Romdhane
patent: 6499349 (2002-12-01), Aronsson
patent: 61294215 (1986-12-01), None
James E. Berry,How to Track Rolling Element Bearing Health with Vibration Signature Analysis,Sound and Vibration, Nov., 1991, pp. 24-35, USA.
Daniel Cecil M.
Graichen Catherine
Handler Larry R.
Hedeen Robert A.
Beusse Brownlee Wolter Mora & Maire P.A.
General Electric Company
Miller Rose M.
Rowold, Esquire Carl A.
Williams Hezron
LandOfFree
System and method for conditioned-based monitoring of a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for conditioned-based monitoring of a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for conditioned-based monitoring of a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3295835