JPH0152687B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a bearing damage diagnosing device for a rotating machine.

[Conventional technology]

Plain bearings, etc. traditionally used in rotating machinery,
The optimum diameter, width, and type are selected based on usage conditions such as rotor shape, weight, and transmitted torque. Taking a steam turbine generator as an example, in a large machine, the total amount
Since it weighs over 150 tons and rotates at high speed, it goes without saying that it is desirable that the bearings that support it are strong and always in a normal state, but there are risks such as burnout of the bearing metal, excessive or insufficient bearing load, etc. This directly causes excessive vibration and rubbing of the rotor, which may lead to excessive accidents such as rotor scattering accidents.

On the other hand, in recent large-capacity steam turbines,
Due to the decrease in shaft rigidity due to the increase in the size of rotating machines, unstable vibrations such as oil whips have become a problem. As a measure to prevent this oil whip, the pressure on the bearing surface is increasing. This inevitably leads to deterioration of bearing lubrication characteristics during low-speed operation, which often leads to bearing damage. In particular, in machines that perform turning operations (rotation speed approximately 2 rpm) for long periods of time, such as those seen in steam turbines, for a long period of one week or more, the lubrication state during low-speed operation must be controlled to prevent sliding damage that can lead to damage to the papillary metal. It is extremely important for preventive maintenance of machinery to detect or predict bearing abnormalities such as the metal wipe phenomenon (hereinafter abbreviated as metal wipe) at an early stage, and to avoid such abnormalities before they occur. However, as mentioned above, due to the high pressure on the bearing surface, complete control is becoming difficult, and metal damage accidents are occasionally seen.
Conventionally used methods for monitoring metal damage include bearing metal temperature and drain oil temperature measurement methods. These measurement methods will be explained below.

Figure 1 shows a simplified cross-sectional structure of the bearing. The metal temperature measurement mentioned above uses a thermocouple 3 inside the bearing back metal 2.
This is a means for monitoring the heat generated by metal damage, etc., by installing a thermocouple 3b inside the cover member 4, and monitoring the heat generated by metal damage, etc. through the exhaust oil. This is a means of monitoring metal damage, etc. by measuring the

Furthermore, methods for monitoring bearing abnormalities other than the above-mentioned temperature monitoring method include Japanese Patent Laid-Open No. 151237/1982 and Japanese Patent Laid-open No. 77189/1989.

All of these proposals are for detecting abnormalities from acoustic signals generated by bearing abnormalities.

[Problem to be solved by the invention]

In the metal temperature measurement mentioned above, thermocouple 3
Since a is installed on the bearing back plate 2 of the papitmetal 5, and is further away from the bearing sliding surface, an abnormality cannot be detected even if slight metal burnout occurs due to uneven contact or wipe, for example. Even if an abnormality occurs, it takes a certain amount of time for the thermocouple 3a to detect it. Similarly, when measuring the temperature of drained oil mentioned above, it takes a certain amount of time to detect an abnormality, so it is not only difficult to detect metal wipes that can lead to bearing damage, but also It is difficult to know its progress over time.

Furthermore, in JP-A-55-15137,
The abnormality is detected by counting all the wave numbers included in the envelope waveform of the detection signal generated by the bearing abnormality, and as is well known, it is determined that the abnormality is detected unless the threshold level H _L is exceeded. is impossible. Furthermore, there is no description of the concept of detection, that is, counting the envelope waveform itself as one pulse.

Therefore, it is not possible to detect low amplitude acoustic signals caused by the damage process of the metal wipe. In addition, since the above proposal uses a full wave number counting method, if the frequency of the acoustic signal generated due to a bearing abnormality changes, the number of detected pulse signals will vary greatly even if the acoustic signal has the same amplitude. I get tired. Depending on the type of bearing abnormality, the frequency of the generated acoustic signal may differ, and under such conditions, even if the bearing abnormality is of the same degree, there may be problems that can lead to misdiagnosis. There is.

JP-A-54-77189 does not detect the extent of bearing abnormality (damage process), but diagnoses the form of abnormality (detects the type of abnormality), and diagnoses the progress of bearing abnormality. It is not possible.

Furthermore, there is no description of detection of the metal wipe phenomenon that occurs only during turning operation of the rotating body. The above-mentioned phenomenon in which the form of the ultrasonic (acoustic emission) signal that occurs as the bearing is damaged during turning operation differs was first reported in June 1982.
Published on March 12th, Japan Society of Mechanical Engineers lecture proceedings and “lubrication”
(Volume 28, No. 12, 1983), this is a phenomenon that was first revealed through experiments conducted by the present inventors.

The purpose of the present invention is to detect bearing damage by detecting the ultrasonic signal generated by the metal wipe phenomenon, which not only makes it possible to immediately detect the occurrence of the metal wipe phenomenon, but also to diagnose its progress over time. The objective is to provide diagnostic equipment.

[Means to solve the problem]

The present invention includes an acoustic detection element that detects an ultrasonic signal generated by the metal wipe phenomenon of a sliding bearing during turning operation, an amplifier that amplifies the ultrasonic signal detected by the element, and a signal output from the amplifier. a detection circuit that detects the detection output, and a first comparison circuit that compares the detection output with a first reference value for determining the initial process of metal wipe, and determines and outputs the detection output that is larger than the first reference value. Then, the detection output is compared in magnitude with a second reference value (a value smaller than the first reference value) for determining the metal wipe damage process, and a detection output larger than the second reference value is determined. A second comparison circuit to output, a means for comparing the numbers of discrimination outputs of the first and second comparison circuits obtained for each measurement period, and a metal wipe in the measurement period when both numbers match. It is comprised of means for determining that the process is an initial process, and determining that it is a metal wipe damage process during a measurement period in which the number of determined outputs of the second comparison circuit is large after the two numbers match.

[Effect]

During the initial process of metal wiping, periodic ultrasonic signals are generated that are not necessarily synchronized with the rotational speed of the rotating machine. For example, the turning speed is 2
Hz, the number of ultrasonic signals generated during the initial process of metal wipe is several tens of Hz. The ultrasonic signal generated during the initial process of metal wiping has periodicity, and its amplitude is relatively large. On the other hand, in the process of damage to the sliding portion of the bearing due to the metal wipe, an ultrasonic signal with a small amplitude is generated between the periodic signals.

That is, the threshold level of the first comparison circuit described above is made large, and the threshold level of the second comparison circuit is made small. In the initial process of metal wiping, the numbers of the first and second pulses output from the first and second comparison circuits are the same, so the numbers of the first and second pulses match. . However, in the damage process caused by metal wipes, as mentioned above, ultrasonic signals with small amplitudes are generated.
For the number of pulses output from the first comparison circuit,
The number of pulses output from the second comparison circuit increases. Therefore, as a result of the comparison, unlike the initial process of metal wiping, the two do not match, and the number of pulses in the latter is large.

As described above, the comparison means compares the magnitude relationship between the numbers of pulses output from the first and second comparison circuits. Further, the comparison result is checked by a determining means to detect the initial stage of damage to the bearing due to the metal wipe to the stage of damage.

〔Example〕 Hereinafter, the present invention will be explained in detail using the drawings.

FIG. 2 shows a block diagram of a bearing damage diagnosis device that is a specific embodiment of the present invention. An acoustic detection element 22 (for example, one using a piezoelectric ceramic element) is installed by pressure bonding or gluing at a location where an ultrasonic signal generated by the papit metal of the bearing 21 supporting the journal 20 or the metal wipe on the outer surface of the bearing can be detected. The signal obtained by the acoustic detection element 22 is input to the preamplifier 23. Preamplifier 2
The signal amplified in step 3 is passed through a filter 24 to remove noise. Further, the signal from which noise has been removed is amplified by the main amplifier 25, and the signal is input to the turning determination circuit 26. The turning determination circuit 26 takes in the rotational speed v (rpm) of the journal and compares and determines whether it is turning or operating, and inputs the output signal from the main amplifier 25 to the detection circuit 27 only in the case of turning, and detects it. The signal detected by the circuit 27 is input to a signal amplitude discrimination circuit 28.

The signal amplitude discrimination circuit 28 is a circuit that discriminates the magnitude of the input signal amplitude to classify it into three stages: bearing normal, metal wipe initial stage, and papit metal damage stage, and outputs a corresponding signal. Two different types of comparison circuits 29a,
29b and a divider 30. In addition,
Although two types of comparison circuits with different comparison voltages are used in the signal amplitude discrimination circuit 28 described above, if more than two types of comparison circuits are used, it is possible to discriminate the amplitude with high accuracy. Finally, the bearing damage display device 31 displays the results described above.

Here, the operating principle of the bearing damage diagnosis device described above will be specifically explained.

FIG. 3 shows the operation of each circuit and its output signal waveform for ease of understanding when explaining the operating principle of the bearing damage diagnosis device. The ultrasonic signal output waveform (output of the main amplifier 25) shown in the figure is as follows:
This figure shows the changes over time from when the bearing is normal, through the initial process of metal wipe, to the process of papit metal damage. As shown in the figure, when the bearing is normal, there is only background noise, but in the initial process of metal wipe, a synchronous ultrasonic signal (referred to as the first metal wipe wave) is generated. The generation frequency of this ultrasonic signal is several tens of Hz, and one of its characteristics is that it is not necessarily synchronized with the rotation speed. The cause of the generation of the above-mentioned ultrasonic signal is due to slight metal contact between the journal and the papit metal due to lack of lubrication. Furthermore, in the papit metal damage process, the generation period of the above-mentioned ultrasonic signal is lengthened, and during that period, an ultrasonic signal (referred to as a second metal wipe wave) having a smaller amplitude than the above-mentioned ultrasonic signal is generated. This is due to deformation or peeling of papitometal.

Next, the ultrasonic signal is sent to the third detection circuit 27.
As shown in the figure, the signal amplitude determination circuit 28
is input. In the signal amplitude determination circuit 28, the comparison circuits 28a and 29b determine the respective comparison voltages Ea,
Eb (Ea>Eb) is compared and converted into a pulse signal as shown in the figure. Here, Ea is a first reference value set for the metal wipe initial process, and Eb is a second reference value set for the metal wipe damage process. The pulse signal is then input to a divider 30. In the divider 30, as shown in the figure, the calculation time (i.e. measurement time) (T ₁ -
(Tn), the calculation of (number of output pulses of comparison circuit 29a)/(number of output pulses of comparison circuit 29b) is performed. That is, the numbers of both pulses are compared for each measurement period. Figure 3 shows the calculation results. When the bearing is normal, no signal is output from the divider 30, but during the initial process of metal wiping, the comparison voltage is
Since the number of pulses compared at Ea and Eb is the same, the calculation result is 3/3=1. As the metal wipe progresses further and enters the papitmetal damage process, as mentioned above, a small amplitude ultrasonic signal is generated between the large amplitude ultrasonic signals, so the comparison voltages Ea and Eb are compared as shown in Figure 3. Since the number of pulses generated is different, the calculation result is 2/5, which is different from the initial process of metal wipe described above. Based on these operation results, the divider 30
Now, as shown in FIG. 3, different signals are output for each calculation time, and then the bearing damage display device 31 classifies and displays the signals in the three stages described above.

In the above explanation, each measurement interval Ti (i=1, 2,
3,...), the number of pulses is set to several, but in reality, it is several dozen or more. Furthermore, each period of each measurement period tends to become longer as the papit metal damage progresses. Therefore, the measurement interval itself needs to be long.

Note that although the calculation method described above is a digital method, even if the number of pulse signals outputted from the comparison circuits 29a and 29b is converted into analog and then the same calculation is performed in the divider 30, it is not a digital method. Needless to say, the same results can be obtained.

As explained above, according to the present invention, not only can the occurrence of bearing metal wipe be detected early, but also the extent of damage to the bearing can be known.
It can be used as an online monitor for bearing damage detection, and has the remarkable effect of being extremely effective industrially.

Next, other embodiments of the present invention will be described. Fourth
The bearing damage diagnosis monitor shown in the figure has the same function as the divider 30 described in FIG.
This is another example for determining the initial stage of metal wipe and the stage of papit metal damage. Here, in explaining FIG. 4, the operation of each circuit and its output waveform will be used for easy understanding.

First, the operation when the bearing progresses from a normal state to the initial stage of metal wipe will be explained. FIG. 5 shows the ultrasonic signal output waveform and the operation of each circuit.

The output waveform from the main amplifier 25 is as shown in the same figure, and then the output waveform is detected by the detection circuit 27 as shown in FIG. 5, and then applied to the comparison voltage Ea of the comparison circuit 29a. The signals are compared and converted into pulse signals as shown in the figure. Next, the pulse signal is transmitted to the inverter 50 shown in FIG.
a, and after being converted into the waveform shown in FIG. 5, the output waveform is input to the gate circuit 52. The gate circuit 52 opens the gate by the width of the pulse output from the inverter 50a, extracts the pulse output from the pulse oscillator 51 as shown in FIG. 5, and inputs the signal to the pulse division circuit 53. The pulse division circuit 53 divides the number of pulses from the pulse oscillator 51 output for each open gate of the gate circuit 52, and when the ratio is approximately 1, outputs a signal as shown in FIG. do. That is, the initial process of metal wipe is determined by checking the periodicity of the pulse signal output from the comparison circuit 29a. Next, the signal from the pulse divider 53 is input to the bearing damage display device 31 described above, and the bearing damage display device 31 displays whether the bearing is normal or the initial process of metal wiping.

Next, the operation shown in FIG. 4 when the state of the metal wipe progresses to the stage of damage to the papit metal will be explained. In order to make it easier to understand when explaining the method of detecting the papit metal damage process, the operation of each circuit and its output waveform are shown in FIG. The output waveform from the main amplifier 25 is as shown in the figure, and then the output waveform is output from the detection circuit 2.
7, the signal is detected as shown in FIG. 6, and further compared with the comparison voltage Eb of the comparison circuit 29b, and converted into a pulse signal as shown in FIG. Next, the pulse signal is inputted to an inverter 50b, converted into the waveform shown in the figure, and then inputted to a stable multivibrator (hereinafter abbreviated as M.M.) 55.
The waveform shown in the figure is obtained as the output waveform of M55.
Next, the output of the M.M55 is input to the AND circuit 56. The AND circuit 56 performs a logical product of the output of the comparator circuit 29b and the output of the M.M.55 to obtain a waveform as shown in FIG. By performing the above-described operations, only the small amplitude ultrasonic signals generated between the large amplitude ultrasonic signals, which is a characteristic of the papit metal damage process described above, can be extracted. Next, the signal output from the AND circuit 56 is input to the bearing damage display device 31, which displays that the bearing is in the process of papit metal damage.

In each of the above embodiments, the case where the number of channels is one is explained, but if the number of channels is increased, even higher precision can be obtained. Of course, the number of acoustic detection elements also increases accordingly.

〔Effect of the invention〕

According to the present invention, ultrasonic signals with different amplitude forms generated depending on the degree of damage to a sliding bearing due to the metal wipe phenomenon are separately detected, and
It is possible to evaluate. In addition, it is possible to detect the periodic ultrasonic signal generation characteristic of the initial process of bearing damage caused by metal wipes from the signal generation interval time. ) is also possible.

[Brief explanation of drawings]

Fig. 1 is a diagram showing a simplified cross-sectional structure of a conventional bearing and the installation locations of each thermocouple in the bearing metal temperature and drain oil temperature measuring method. Fig. 2 is a block diagram of the bearing damage diagnosis device of the present invention. Figure 3 is a waveform diagram for explaining the operation of the bearing damage diagnosis device of the present invention, Figure 4 is a block diagram of a bearing damage diagnosis monitor which is another embodiment of the present invention, and Figures 5 and 6 are waveform diagrams for explaining the operation of the bearing damage diagnosis device of the present invention. FIG. 3 is a waveform diagram for explaining the operation of the diagnostic monitor. 22...Acoustic detection element, 28...Signal amplitude discrimination circuit.

Claims (1)

[Scope of Claims] 1. In a bearing damage diagnosis device for diagnosing damage to sliding bearings used in rotating machinery, acoustic detection detects ultrasonic signals generated by the metal wipe phenomenon of sliding bearings during turning operation. an element, an amplifier that amplifies the ultrasonic signal detected by the element, a detection circuit that detects the signal output from the amplifier, and a magnitude comparison between the detection output and a first reference value for determining the initial process of metal wipe. a first comparison circuit that determines and outputs a detected output that is larger than the first reference value; value)
a second comparison circuit that compares the detected output with the second reference value and outputs the detection output that is larger than the second reference value; and the number of discrimination outputs of the first and second comparison circuits obtained in each measurement period. a means for comparing the sizes of each other, and determining that the metal wipe is in the initial stage during the measurement period when both numbers match, and after the numbers match and the number of pieces determined by the second comparison circuit becomes large; A bearing damage diagnostic device comprising: a means for determining that a metal wipe damage process occurs during a measurement period;