Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Mechanical measurement system
Reexamination Certificate
1998-07-15
2003-01-14
Hilten, John S. (Department: 2863)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Mechanical measurement system
C702S042000
Reexamination Certificate
active
06507790
ABSTRACT:
FIELD OF THE INVENTION
This invention generally relates to the field of monitoring a sound source for determining its operating condition, particularly relates to the fields of machinery condition monitoring, acoustics, and digital signal processing, and specifically relates to the real-time digital filtering of an acoustical signal to obtain its power spectrum, and the comparison of the power spectrum with a previously determined baseline spectrum as a means of detecting developing machinery faults.
BACKGROUND OF THE INVENTION
There has been commercial activity in the field of machinery condition monitoring for at least 25 years, almost all of it based on either periodic or continuous measurement of machine vibration. Acoustic monitoring has rarely been used for detecting machinery faults, even though sound and vibration are closely related. A rotating or reciprocating machine, for example, produces dynamic forces (forces which are rapidly changing functions of time) which cause various parts of the machine to vibrate. These vibrations also cause sound to be radiated from the machine. The relationship between the dynamic forces acting within a machine and the sound radiating from the machine is complex. The vibration spectrum (the displacement amplitude as a function of frequency) depends on the measurement location on the machine, as well as the orientation of the vibration transducer with respect to the axis of rotation of the machine. The sound spectrum (the acoustic power as a function of frequency) depends on the orientation of the microphone with respect to the machine, the directional characteristics of the microphone, and the acoustical characteristics of the surrounding objects and structures. The point of origin of a vibration component may not be an efficient radiator of sound; nevertheless, it is still possible to hear this vibration component if it is transmitted to another part of the machine which is mechanically resonant at that frequency. The design of the machine, and especially the damping characteristics of the materials used, greatly affects the intensity and spectral distribution of the radiated sound.
A machine which is in good condition and functioning properly will have a certain vibration spectrum, which in turn will generate a certain sound spectrum, that is, an acoustic signature which can be used as a reference or baseline. In general, the vibration spectrum and the sound spectrum are not the same; in fact, they may be quite different. But, if the condition of the machine deteriorates, or if there is a sudden failure, the vibration spectrum, and therefore the sound spectrum, will change. The deteriorating machine condition can be detected by continuously monitoring the sound coming from the machine, computing the power spectrum, and comparing the power spectrum to the baseline spectrum stored in memory. If the real-time power spectrum deviates from the baseline spectrum by more than a predetermined amount, an alarm can be activated, along with automatic shutdown of the monitored machine, if desired.
There are many types of machine faults that could be detected by such a monitor; for example, rotating imbalance, reciprocating imbalance, misaligned or bent shafts, damaged rolling element bearings, damaged journal bearings, damaged or worn gears, broken drive belts or chains, mechanical looseness, jamming, overloading, friction, windage, impacts, explosions, and escaping air, water, or steam. An acoustic monitor could also provide protection for non-rotating equipment such as boilers, electrical transformers, and flow processes.
The art and science of vibration-based machinery condition monitoring is highly developed, and there are many commercially available products for measuring vibration, and for collecting, storing, analyzing, and displaying vibration data. In recent years, there has been a significant increase in activity in this field because of the widespread availability of digital signal processing (DSP) hardware such as DSP microcomputers. These are high speed single-chip computers which incorporate a high degree of operational parallelism and which are designed to implement computationally intense DSP algorithms such as the fast Fourier transform (FFT), widely used to compute the power spectrum of a vibration signal. Vibration analysis techniques have been developed to detect and diagnose specific machine faults, using commercially available hardware and software tools. The usual approach is to measure vibration with an accelerometer which is in direct contact with the machine being monitored or studied, and then process the resulting signal with an instrument known as a dynamic signal analyzer (DSA). This equipment is expensive, the placement and orientation of accelerometers on the machine can be critical, and skilled personnel are required to operate the DSA and correctly interpret the resulting vibration spectra.
With regard to the early detection of machinery problems, in many cases the first indication of trouble is the sound that a machine makes. In fact, it may be argued that acoustic monitoring (by human observers) is the oldest form of machinery condition monitoring in existence. Experienced machine operators or plant maintenance personnel can often recognize that a machine is in distress because they are familiar with what the machine sounds like when it is operating normally. An acoustic monitor could, in effect, replace human observers in situations where machinery is operating in remote, inaccessible, or hazardous locations, or any other situation where machinery requires continuous monitoring. The acoustic monitor according to the teachings of the present invention is intended to be affordable, dependable, easy-to-use, and easy-to-install and has, as its purpose, machinery protection rather than machinery fault diagnosis or testing. Once the user has been alerted to the fact that machinery is in distress, more sophisticated equipment can be used to diagnose the specific problem. It is not necessary to continuously monitor the machinery with costly vibration-based instrumentation.
DESCRIPTION OF THE PRIOR ART
At the present time, there is only one commercially available product known to be capable of continuous acoustic monitoring of industrial processes and equipment: the Model 261 Sound Level Detector/Controller, manufactured by Quest Electronics of Oconomowoc, Wis. This product is essentially a sound level measuring instrument with an output relay having an adjustable threshold calibrated in decibels (dB). It measures the root-mean-square (RMS) sound pressure level (SPL) sensed by a microphone and actuates a relay if the threshold setting is exceeded. Other than providing the A and c frequency weighting commonly used for sound level measurements, this product does not perform any type of filtering or spectral analysis. It is a broadband instrument which simply measures the combined effect of all the frequency components of a signal. Primarily intended for industrial hygiene purposes (noise control and warning), it can also be used to provide an alarm signal or automatically shut down a machine if the sound pressure level exceeds the threshold setting.
SUMMARY OF THE INVENTION
An acoustic monitor according to the teachings of the present invention is a self-contained system which detects faults in the operating condition by continuously analyzing the sound produced by the sound source being monitored and comparing the resulting power spectrum to a previously recorded “acoustic signature” used as a baseline. In the preferred form, the acoustic monitor performs real-time {fraction (1/12)}th octave digital bandpass filtering over an eight octave range (midband frequencies of 33.108 Hertz to 8,000 Hertz) and computes the acoustic power output, in decibels, of each of the resulting 96 bandpass filters. Fractional octave bandpass filtering produces a constant percentage bandwidth analysis, that is, the bandwidth of each bandpass filter is a constant percentage of its midband frequency. In the case of a {fraction (1/12)}th octav
Hilten John S.
Horton, Inc.
Kinney & Lange , P.A.
Lau Tung S
LandOfFree
Acoustic monitor does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Acoustic monitor, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Acoustic monitor will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3034982