Data processing: measuring – calibrating – or testing – Measurement system – Performance or efficiency evaluation
Reexamination Certificate
1999-03-13
2004-02-17
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system
Performance or efficiency evaluation
C702S056000, C073S660000
Reexamination Certificate
active
06694285
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to monitoring systems for monitoring the operation of a rotating machinery.
2. Related Art
In the ever-increasing competition in the industrial field, industrial equipment, such as rotating machinery, must operate at or near full capacity and sustain such operation for long periods of time. With this type of demand placed on such equipment, periodic maintenance to avoid a catastrophic failure becomes important. Of course, periodic preventative maintenance requires that the equipment be taken off-line for service, thereby potentially resulting in unnecessary down time. Maintenance engineers have been challenged to establish proper time intervals for scheduled preventative maintenance in order to reduce such unnecessary down time.
Alternatively, some maintenance engineers have concluded that the equipment should operate until catastrophic failure. This stems from the fact that, in some instances, it may be better to operate equipment until it fails than to accept the maintenance and the resulting penalty costs of shutting down the equipment prematurely. Also in lieu of scheduled maintenance, some defects may be found by a trained operator. Because such detection is subject to human interpretation, pass/fail criteria may vary between operators and also from day to day with the same operator. Other defects may not be detected at all.
Attempts have been made to automatically monitor such equipment for defects through the use of a sensing element disposed within the equipment itself or through the use of a hand-held device which is periodically attached to one or more discrete locations on the machine being monitored. More sophisticated monitoring systems are permanently installed and carry out essentially continuous monitoring of a machine-mounted transducer along with computer-based analysis of all monitored data.
Most automatic monitoring systems typically sense vibration or temperature. Vibration is produced by the moving parts in the rotating machinery due to causes such as unbalance, misalignment of shafts, worn out bearings, broken gear teeth or foreign particles lodged within the machine. Excessive levels of vibration indicate malfunction of the machine, which can result in machine failure. The temperature of a bearing, for example, can also be monitored to detect the occurrence of over-heating. In some instances, the oil level in the machine may be monitored, automatically through the use of a float system or manually through the use of a dipstick or a sight glass, so that the likelihood of defects or malfunction of the device due to low oil level may be reduced. Other automatic means to detect oil level include beam techniques that measure time of flight or frequency modulation of an ultrasonic, microwave or light/laser beam. Electrical methods have also been employed that detect changes in current, voltage, capacitance or inductance of the liquid to determine the fluid level.
Vibration analysis methods in simple monitoring systems generally average all frequencies occurring in a given device, typically using a root-mean-square analysis, and compare the root-mean-square vibration energy to a threshold. These methods are not capable of detecting discrete component malfunctions because the detrimental vibration of that component may be dampened by the averaging. Other vibration analysis methods in more sophisticated monitoring systems utilize techniques which envelope discrete frequencies known to correspond to certain frequencies of a given component in the device such as, for example, a bearing. These enveloped frequencies are compared to threshold values to determine whether the component is experiencing detrimental vibration levels. A disadvantage with these more sophisticated systems is that, if a detrimental vibration of certain frequency is occurring within the device, but the monitoring system is not programmed to analyze that particular frequency because it has not been enveloped, the defect may go undetected.
In addition, in conventional monitoring systems, the thresholds level are set prior to placing the device to be monitored in service. As a result, the pre-set threshold levels are necessarily a compromise and therefore cannot take into account the operating environment surrounding the site the of the device, such as vibrations induced by adjacent machinery. A particular disadvantage with such pre-set thresholds is that the monitoring system may trigger an inordinate amount of false alarms, thereby reducing user confidence in the system.
Conventional gear boxes are generally designed to include a margin of safety such that the gear box is operated within acceptable limits. Typical margins of safety are one and a half to four indicating that the stated capacity of the gear box may be much less than the actual maximum capacity. Generally, it is thought that high margins of safety are provided so that in the event that the gear box is operated at a higher demand, the gear box will not inadvertently fail.
Mechanical devices having higher margins of safety generally are bulkier and heavier than a corresponding mechanical device having a smaller margin of safety. To minimize the cost and size of these mechanical devices without jeopardizing the margin of safety, mechanical engineers have redesigned these mechanical devices by substituting higher performance materials. However, this too is limited in terms of cost and practicality of the substitute materials.
SUMMARY OF THE INVENTION
One feature of the invention is a monitoring method and device which is capable of determining the overall fitness or condition of the device being monitored while simultaneously detecting unknown or unrecognized vibration frequencies.
In one embodiment, a vibration signal is sensed to produce a vibration signal in the time domain and a Fourier transform on the vibration signal in the time domain is performed to produce a vibration signal in a frequency domain. The vibration signal in the time domain may comprise a debiased vibration signal and a debiased-rectified vibration signal. The vibration signal in the frequency domain is then divided into a plurality of bands. The average spectral energy or the spectral peak in each band may be calculated and compared to one or more thresholds. Values representing the thresholds that have been crossed are assigned, accumulated and compared to one or more thresholds.
In another embodiment, a Fourier transform on the vibration signal in the time domain is performed to produce a vibration signal in a frequency domain. The spectral component at zero frequency may be calculated and compared to one or more thresholds. Values representing the thresholds that have been crossed are assigned, accumulated and compared to one or more thresholds.
In another embodiment, an overall energy or the peak vibration of the vibration signal in the time domain is determined and compared to a threshold.
In another aspect, a method for setting threshold values corresponding to the environment surrounding the device to reduce the incidence of false alarms is performed. In one embodiment, the method includes the steps of sensing an operating condition during a learning period and calculating the mean or the standard deviation of the sensed condition obtained during the learning period. If both are calculated, then they may be added together.
In another embodiment, a plurality of the sensed conditions may be accumulated during the learning period. The learning period may be predetermined period of time or may be an input value. In another embodiment, a plurality of threshold values for a single operating condition may be set.
According to another aspect, controlling false alarms while maximizing detection of true alarms in a device monitoring system may also be performed. In one embodiment, a method includes the steps of sensing an operation condition of the device; comparing the sensed operating condition to first thresholds; assigning a value indicative of one of the first threshold values cross
Chachich Alan C.
Choe Howard C.
Woodroffe Jaime A.
Barbee Manuel L
Hoff Marc S.
Textron System Corporation
Wolf Greenfield & Sacks
LandOfFree
Method and apparatus for monitoring rotating machinery does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for monitoring rotating machinery, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for monitoring rotating machinery will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3314440