JPH0460545B2 - - Google Patents

Description

Detailed description of the invention [Technical field to which the invention pertains] The present invention relates to a rubbing detection method that uses acoustic vibration to detect the position of a rubbing phenomenon that occurs between a rotating part and a stationary part in a rotating machine. The present invention relates to a device, and particularly to a device configuration that can perform position detection without being affected by the waveform of acoustic vibrations associated with a rubbing phenomenon.

[Prior art and its problems]

If a rubbing phenomenon occurs during operation of a rotating machine such as a steam turbine or a generator, abnormal vibrations may occur, making it impossible to continue operation or leading to a serious accident. In particular, in thermal machines such as steam turbines, the distance between rotating and stationary members is originally set to be narrow, and structural members such as the casing are likely to undergo thermal deformation due to increased humidity.
There is a tendency for rubbing phenomenon to occur. For this reason, when a rubbing phenomenon occurs, driving operations such as lowering the rotational speed of the rotating parts are usually carried out to avoid the rubbing condition from continuing, but such measures are only temporary measures and should not be taken to eliminate the cause of the rubbing. Since the material itself still exists, the rubbing phenomenon often occurs again. In order to eliminate the cause of rubbing, it is necessary to clarify the location where the rubbing phenomenon occurs and change the assembly condition, such as widening the gap in the rotating machine.

The location of the rubbing phenomenon is usually detected by detecting the acoustic vibrations generated in the rotating machine due to rubbing (hereinafter this vibration may be referred to as "rubbing vibration"), and this rubbing vibration is synchronized with the rotation of the rotating member. There are characteristics that occur.

FIG. 4 is a configuration diagram of a conventional rubbing detection device that detects the position of rubbing occurrence. It is assembled. Reference numerals 21 and 22 denote first and second acoustic sensors made of piezoelectric elements formed to have a resonance frequency in a high frequency region.
The metal wire wave guides 2a and 2b are fixed to the other ends of the metal wire wave guides 2a and 11b, respectively. The wave guides 2a, 2b are provided to transmit the acoustic vibrations in the bearings 11a, 11b to the sensors 21, 22, and the sensors 21, 22 transmit signals 21 corresponding to the acoustic vibrations transmitted in this way.
It is configured to output signals a and 21b. 3
is a pulse generator that outputs a pulse signal 3a synchronized with the rotation of the rotating shaft 1a; in this case, the signal 3a is an electric pulse that is outputted every time the rotating shaft 1a rotates once. 41 and 42 are output signals 21a and 2
Amplifiers 51 and 52 amplify the signal 1b; bandpass filters 51 and 52 receive the amplifier output signals 4a and 4b and extract frequency components of the rubbing vibration from these input signals to improve the S/N ratio; 61 and 62 bandpass filters; This is a detector that performs envelope detection on the pass filter output signals 5a and 5b.

71 and 72 are input with the output signals 6a and 6b of the detectors 61 and 62, respectively, and the pulse signal 3a, and when the pulses constituting the signal 3a are input, a predetermined data collection time starts from the pulse input time. Each time waveform of the signals 6a and 6b is stored until T has elapsed, and then when the pulse forming the signal 3a is inputted, the same waveform storage operation is performed again and the waveform at this time is already stored. This waveform addition operation is performed a predetermined number of times N.
After repeating this, the waveform value of the cumulative time waveform of the (N+1) time waveforms formed as described above is divided by (N+1), and signals 7a and 7b having time waveforms corresponding to the division result are obtained. This is an arithmetic unit that outputs . Hereinafter, the above-described calculation operation will be referred to as a time average calculation, and the calculation units 71 and 72 that perform such a time average calculation may be referred to as an average value calculation unit. In this case, the data acquisition time T is normally set to correspond to several rotations of the rotating shaft 1a, and the arithmetic units 71, 72
is configured so that it does not respond to the pulse of signal 3a even if it is inputted once time T begins to elapse. Since the computing units 71 and 72 are configured as described above, the output signals 7a and 7
Both b are time waveforms from which white noise has been removed. 8 receives the signals 7a, 7b and the pulse signal 3a, and uses the arithmetic operation described below to control the rotating machine 1.
A position calculator 9 outputs a signal 8a corresponding to the rubbing occurrence position, and a display 9 receives the signal 8a and displays a display according to the rubbing occurrence position.

In FIG. 4, each part is configured as described above, so if rubbing occurs at position 100 between rotating shaft 1a and casing 1b, rubbing vibrations 100a and 100b are generated by this rubbing, Vibration 100a is caused by rotating shaft 1a, bearing oil film 1
3a, the bearing metal 11a, and the wave guide 2a to the sensor 21, and the vibration 100b is transmitted to the rotating shaft 1.
a, propagates to the sensor 22 via the bearing oil film 13b, the bearing metal 11b, and the wave guide 2b. Then, the output signal waveform of the main part in FIG. 1 becomes as shown in FIG. 5, for example. That is, the amplifier output signal 4 including the rubbing vibration component and the noise component
a, 4b are filters 51, 52, detectors 61, 6
2, the waveforms of the output signals 7a and 7b of the average value calculator are converted into the detector output signal 6.
The white noise seen in a and 6b is removed, resulting in a clean waveform. Then these signals 7a, 7b
Triangular pulse-shaped waveforms Q and R based on the rubbing vibrations 100a and 100b appear, respectively, and since these waveforms Q and R are caused by the rubbing vibration, their appearance is the pulse signal 3a.
is synchronized with. Then rubbing vibration 100
If the times when the waves a and 100b reach the acoustic sensors 21 and 22 are different, a time difference Δt occurs between the waveforms Q and R as shown in FIGS. 5 and 6. _t1 ,
As is clear from FIG. 6, _t2 is the time from the rise time of the signal 3a to the rise time of each of the waveforms Q and R, respectively.

In the signals 7a and 7b, waveforms Q and R appear as described above, so now the rubbing occurrence position 100
is located to the left of the Z-Z line by a distance of ΔL as shown in Figure 7, then ΔL is (1)
Expressed by the formula.

ΔL=(t ₂ - t ₁ )・V/2 ...(1) Here, V is the rubbing vibration 100 at axis 1a
At the propagation speeds a and 100b, the Z-Z line in FIG. 7 is a line passing through the midpoint between the bearing metals 11a and 11b.
In other words, if ΔL is known, the rubbing occurrence position 10
Since 0 is known, the position calculator 8 in FIG.
Detection of t ₁ and t ₂ and calculation of equation (1) are performed to output a signal 8a corresponding to ΔL, thereby making it possible to specify the rubbing occurrence position 100.
The rubbing detection device shown in FIG. 4 detects the position where rubbing occurs as described above, but in this case there are problems as described below. Figure 8 is a waveform diagram for explaining this problem. In the figure, the horizontal axis is the elapsed time, S is the time difference found from each intersection of the threshold value M indicated by a broken line in the diagram, and the waveforms Q and R. P is the waveform diagram for explaining this problem. This is the time difference obtained from each peak value of waveforms Q and R. As is clear from the figure, in condition 1, waveforms Q and R are similar, so S=P, but in conditions 2 and 3, waveforms Q and R are not similar, so S≠P. In the figure, for example, if S=P in condition 1 is equal to 1, then condition 2
Then, S=1.4, P=1.8, and in condition 3, S=0.5, P=
It is 0. The phenomenon that the waveforms Q and R are not similar is a phenomenon that is often observed in actual rubbing vibration detection, and one of the causes of this phenomenon is the attenuation of each vibration in both high and low frequency bands in the rubbing vibration with respect to the propagation distance. This may be because the rates are different.

Now, in FIG. 4, in order to calculate equation (1), the calculator 8 measures S or P shown in FIG. 8 as the time difference Δt between waveforms Q and R, and performs the calculation. ing. However, as can be seen from the above, S or P changes as the shapes of waveforms Q and R change even if the rubbing occurrence position is the same. Therefore, it can be seen that there is an error in determining ΔL based on the above calculation due to changes in the shapes of the waveforms Q and R. That is, the rubbing detection device shown in FIG. 4 has a problem in that there is a detection error due to a change in the shape of the waveforms Q and R.

[Purpose of the invention]

SUMMARY OF THE INVENTION An object of the present invention is to provide a rubbing detection device that can perform position detection without being affected by the waveform of rubbing vibration, by solving the problems of the conventional device as described above.

[Key points of the invention]

In order to achieve the above object, the present invention provides acoustic sensors on both ends of a rotating shaft of a rotating machine to detect acoustic vibrations of the rotating shaft, and performs signal processing on each output signal of both acoustic sensors to generate rubbing vibrations. The system calculates the peak value of , and detects the position where rubbing occurs based on the difference between the two peak values.By doing this, rubbing occurs without being affected by each waveform in both rubbing vibrations. This allows position detection.

[Embodiments of the invention]

FIG. 1 is a block diagram of an embodiment of the present invention, and the main points that differ from FIG. 4 are a peak detection stage 101,
102 and calculation means 103 are provided. The output signals 7a and 7b of the average value calculator are inputted to the peak detection means 101 and 102, respectively, and the pulse signal 3a is inputted to both detection means. Then, the detection means 101 detects the signal 7 shown in FIG. 5 in synchronization with the pulse signal 3a.
A first detection signal 101a is generated according to the peak level value X of the detected waveform by detecting the waveform Q from among the waveforms a.
The detection means 102 is configured to output
detects the waveform R from among the waveforms of the signal 7b shown in FIG. 5 in synchronization with the pulse signal 3a, and generates a second detection signal 1 according to the peak level value Y of the detected waveform.
02a. 106
consists of an amplifier 41, a filter 51, a detector 61, an average value calculator 71, and a peak detection means 101, and detects a first detection signal 10 corresponding to the peak value of the rubbing vibration in the output signal 21a of the acoustic sensor 21.
1a as described above, and 107 includes an amplifier 42 and a filter 52.
A second detection signal 102a is composed of a wave detector 62, an average value calculator 72, and a peak detection means 102, and outputs a second detection signal 102a corresponding to the peak value of the rubbing vibration in the output signal 21b of the acoustic sensor 22 as described above. It is a signal processing means. The calculation means 103 receives the detection signals 101a and 102a, and calculates the difference between the two signals, in other words, the signal 10 according to (X-Y).
3a as the third detection signal. Reference numeral 104 denotes a display means that performs calculations to be described later using the detection signal 103a and the output signal 105a of the storage means 105, and displays a display according to the rubbing occurrence position 100 described above.

Next, the operating principle of the apparatus shown in FIG. 1, which detects the position where rubbing occurs as described above, will be explained. Figure 2 is an explanatory diagram of the results of an experiment observed regarding the attenuation of the rubbing vibration propagating through the rotating shaft 1a shown in Figure 1, where the horizontal axis of the figure is the length of the rotating shaft 1a, and the vertical axis is shown in Figure 5. This is the peak level value of the rubbing vibration waveform obtained in the same manner as the rubbing vibration waveform Q or R in the signal 7a or 7b. As is clear from the figure, the characteristic line G indicating the experimental results is almost linear, and the absolute value K of the slope of this characteristic line G is K = 15/6 [dB/m] = 2.5 [dB/m] ]. When the rubbing vibration propagates along the rotating shaft 1a, it is attenuated as shown in FIG.
When the rubbing vibration occurs at various locations, the damping mode of the rubbing vibration is as shown in FIG. The vertical axis of FIG. 3 is the level of vibration according to the same definition as on the vertical axis of FIG. Then, X ₁ to _{X 5} are average value calculator 7
Rubbing vibration waveform Q included in the output signal 7a of No. 1
, and Y ₁ to _{Y 5} are the peak level values of the calculator 72.
This is the peak level value of the rubbing vibration waveform Q included in the output signal 7b. Note that A to E in FIG. 3 show cases where rubbing occurs at different positions as shown. Also, L in FIG. 3 is the distance between the bearing metals 11a and 11b like L shown in FIG. 7, and the Z-Z line has the same meaning as the Z-Z line in FIG. 7.

When rubbing occurs at different positions on the rotating shaft 1a, the peak level values X and Y of the rubbing vibration waveforms Q and R appearing in the signals 7a and 7b change as shown in Fig. 3, so in this case ΔL is
It is expressed by equation (2).

ΔL=(X-Y)/(2K)...(2) Therefore, by calculating equation (2), the position where rubbing occurs can be detected.
The storage means 105 in FIG.
The display means 104 is configured to output a.
calculates equation (2) using the signal 105a and the third detection signal 103a according to (X-Y), and calculates ΔL
The display is adapted to suit the situation. The storage means 105 is configured to be able to store different values of K, and in this case outputs a signal 105a corresponding to K=2.5 [dB/m].

In FIG. 1, since the rubbing occurrence position is detected as described above, it is clear that this position detection is not affected by the waveforms Q and R as in the conventional device. Furthermore, since position detection in this case is performed based on equation (2), it is also clear that the level of rubbing vibration at the rubbing occurrence position does not affect position detection. Note that in the present invention, the average value calculators 71 and 72 may perform the white noise removal operation in a manner different from that described above.

〔Effect of the invention〕

As described above, in the present invention, acoustic sensors are provided at both ends of the rotating shaft of a rotating machine to detect acoustic vibrations of the rotating shaft, and each output signal of both acoustic sensors is subjected to signal processing to detect rubbing vibration. Since the peak value of is determined and the rubbing occurrence position is detected based on the difference between both peak values, the present invention is capable of detecting the rubbing vibration at the rubbing occurrence position without being affected by each waveform in both rubbing vibrations. This has the effect that the rubbing occurrence position can be detected without being affected by the level.

[Brief explanation of the drawing]

Fig. 1 is a configuration diagram of an embodiment of the present invention, Fig. 2 is an explanatory diagram of experimental results, Fig. 3 A to E are explanatory diagrams of different rubbing vibration damping modes, and Fig. 4 is a conventional rubbing detection method. A configuration diagram of the device, FIG. 5 is an explanatory diagram of the waveforms of the main parts in FIG. 4, FIG. 6 is an explanatory diagram of the waveforms of the main parts in FIG. 5 over time, and FIG. 7 is an operating principle of the detection device of FIG. The explanatory diagram, FIG. 8, is a diagram illustrating the causes of detection errors in the detection device of FIG. 4. DESCRIPTION OF SYMBOLS 1... Rotating machine, 1a... Rotating shaft, 21... First acoustic sensor, 22... Second acoustic sensor, 101
a...first detection signal, 102a...second detection signal, 103...calculating means, 103a...third detection signal, 106...first signal processing means, 107...
Second signal processing means.

Claims (1)

[Claims] 1 first and second acoustic sensors that detect acoustic vibrations appearing on the rotating shaft at both ends of the rotating shaft in a rotating machine;
first and second signal processing means for respectively outputting first and second detection signals corresponding to the peak value of rubbing vibration in each output signal of the acoustic sensor; and a difference between the first detection signal and the second detection signal. a calculation means for outputting a corresponding third detection signal, and detects a position where the rubbing vibration occurs in the rotating machine based on the third detection signal.