JPS6251407B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a vibration diagnosing method and apparatus for a rotating machine, and more particularly to a vibration diagnosing method and apparatus for a rotating machine suitable for use in turbines, generators, etc. of thermal and nuclear power plants.

Recently, steam turbines and generator rotors in power plants have become larger as the capacity has increased, and vibrations are more likely to occur due to reduced shaft rigidity. In addition, relatively large-capacity power generation plants are now being used for load adjustment against peak loads, and as they are started and stopped frequently, there is a need to improve safety against these frequent starts and stops. There is. For this reason, the importance of vibration monitoring, which is an effective means of monitoring safety, has become even more important, and there is a demand for monitoring and diagnosing vibration phenomena and causes of vibration, not just monitoring the magnitude of vibration values. . Unbalanced vibration, which has a vibration component synchronized with the rotation speed, occurs most frequently in these rotating machines and is therefore an important vibration to be diagnosed.

Unbalanced vibrations can be reduced by balancing, and unbalances created during the manufacturing and assembly process of the rotor can be reduced at the beginning of operation. However, although the unbalanced vibration is small at the beginning of operation, unbalanced vibration may occur during operation due to bending of the rotor, contact with a stationary part (rubbing), damage to a member, etc. Among these, the occurrence of rotor bending is mainly caused by thermal imbalance occurring in the rotor due to changes in the load or current of the rotor system, and is often referred to as "thermal bending." It is.

Specific causes of thermal bending of the rotor include, for power generation rotors in power generation plants, unbalanced ventilation in field coils, interlayer shortening, or unbalanced growth in wedge stiffness, and for turbine rotors,
There are non-uniform thermal expansion coefficients of materials, differences in heat radiation due to surface rust or differences in paint thickness, and non-uniform heat transfer due to keys in shrink-fit rotors.

Although it is not easy to determine the specific cause of the vibration using only vibration data, if it can be determined that the unbalanced vibration is caused by thermal bending of the rotor, it is effective in dealing with the direct cause.

Conventionally, methods have been considered for determining the causes of various vibrations by recording vibration changes over time or changes in the rotating machine and comparing this with characteristic patterns for each cause.However, when it comes to diagnosing thermal bending of the rotor, No effective method has been found yet.

An object of the present invention is to provide a vibration diagnosing method and apparatus for a rotating machine that can automatically diagnose and display the occurrence of vibrations due to thermal bending of a rotor during operation.

In order to achieve this objective, the present invention analyzes the unbalanced amplitude synchronized with the rotation speed of vibration data within a certain period of time, sequentially determines the temporal changes in the unbalanced vibration vector, and determines the direction of the change and increase/decrease in vibration. The present invention is characterized in that it is confirmed that the trends of are almost constant, and a diagnosis is made that the cause of the unbalanced vibration is thermal bending of the rotor.

Hereinafter, the method of the present invention will be explained using the attached figures. Figure 1 is a vector diagram showing the amplitude and phase of the vibration component (unbalanced vibration component) synchronized with rotation when thermal bending vibration occurs, and as time passes, the vibration vectors x ₁ , x ₂ , ... x _N indicates a change from number 1 to N. What should be noted in this figure is that the lines connecting the tips of the individual vibration vectors (○) form a nearly straight line from 1 to N when viewed as a whole. In other words, the direction of the difference between the individual vibration vectors is approximately constant. This is considered to be a major feature of thermal bending vibration, and the present invention has been made with attention to this point.

FIG. 2 shows the procedure of the vibration diagnosis method according to the present invention, and this procedure will be explained below. First, the vibration of a rotor rotating at a constant speed is measured.
This vibration usually uses shaft vibration (journal vibration) of the bearing section. Since the measured vibration is an overall vibration and includes all frequency components, it is necessary to extract the component whose frequency is synchronized with the rotation. It is determined whether the vibration to be diagnosed is an unbalanced vibration or not, depending on whether the rotational synchronous vibration component is the majority. This can be done, for example, as follows.

|x _N |≧αX _N …(1) Here, |x _N | is the amplitude of the rotation-synchronized vibration component at the vibration measurement time N, that is, at the start of diagnosis;
Similarly, X _N is the overall amplitude, and α is the unbalance determination coefficient. Unbalanced vibration occurs when formula (1) is satisfied. If the measured vibration is determined to be an unbalanced vibration, extract the unbalanced vibration component for the diagnosis time T from the present time from the past data recorded in advance, and calculate the change vector for each data one after another. Let's do it. Here, the change vector is the oldest unbalanced vibration vector x when a total of _N pieces of data are included in the diagnosis time T, with the current unbalanced vibration vector ₁ and each unbalanced vibration vector. This change vector can be expressed as follows.

y _o = x _o −x ₁ ...(2) where n=2, ...N.

Next, it is checked whether the phase change of the change vector y _o is constant. This is done using the following equation.

|θ _N −θ _o |=θ _B …(3) Here, θ _N is the phase angle of the latest change vector y _N , θ _o is the phase angle of the change vector y _o , and θ _B is the phase change judgment angle. , n takes a number from 2 to N-1. θ _N determines the direction of the vibration change over the entire interval of the diagnostic time T, and when it is compared with the phase angle θ _o of the change vector at each point in time, it is determined that when it falls within a certain range θ _B. Determine whether there is.

Next, the tendency of the amplitude change of the change vector is investigated in order to find out whether the vibration vector is increasing or decreasing over time. In other words, checking the direction using equation (3) alone can distinguish cases where the vibration vector does not change midway, for example, in Figure 3, the vector point after x _o remains constant in its vicinity. I can't. In the case of thermal bending, there is a gradual but continuous change in amplitude, and the above state is not thermal bending. This amplitude change is checked using the following equation.

|y _o+1 |−|y _o |＞0 …(4) Here, |y _o+1 |, |y _o | are the amplitudes of adjacent change vectors, and n is from 2 to N−1.
Take the number up to.

For all N data within diagnosis time T,
After checking the phase and amplitude of the change vector as described above, it is finally determined whether or not it is thermal bending. For all data, equation (3) and
If the relationship in equation (4) is satisfied, it is a perfect thermal bending vibration, but in reality, even if it is thermal bending, the relationship in the above equation may not be satisfied due to local variations in the data. It is practical to allow some margin for the number of data to be satisfied. That is, even if, for example, about 10% of the data does not satisfy the above conditions, it is determined that there is thermal bending.

Finally, the judgment results are displayed or output on a suitable display device such as a CRT (cathode tube) display tube or a line printer.

The vibration change rate of thermal bending in the above explanation is
Diagnosis time T is not fixed because it varies depending on the type of rotor or operating conditions in the power plant, but normally it should be selected to be about 10 to 60 minutes.

As described in detail above, by using the method according to the present invention, it is possible to reliably diagnose abnormal vibrations caused by thermal bending of the rotor, and this method is suitable for automating the diagnosis.

Next, an example of a vibration diagnosing device for a rotating machine that is most suitable for implementing the diagnosing method of the present invention will be described with reference to FIG.

In FIG. 4, reference numerals 11 to 13 indicate a multi-bearing rotor system consisting of a turbine/generator rotor to be subjected to vibration diagnosis, and are supported by bearings 41 to 46. The vibrations of the bearing detected by the vibration detectors 51 to 56 are amplified by the vibration meter 7 and input to the A/D converter 9 via the multipoint data input device 8. The multi-point data input device 8 is comprised of a sample-hold circuit for simultaneously capturing analog vibration signals from multiple points and a filter for removing unnecessary frequency components during signal capture.

The vibration signal digitized by the A/D converter 9 is input to a data processing device 10 . A data processing device 10 extracts a rotationally synchronous vibration component from input vibration data, and receives input from a waveform converter 11 and a clock 13 to process reference signals and time signals for vibration analysis. A pulse signal of one pulse per rotation detected by the pulse pickup 6 is inputted to the waveform converter 11, and this is converted into a sine wave to obtain a reference signal for vibration analysis of the data processing device.
On the other hand, the output of the pulse pickup 6 is simultaneously amplified and digitized by a tachometer 12 and input to a data processing device 10.

The multi-point vibration data analyzed by the data processing device 10 is stored in the storage device 14 together with the time and the number of rotations, while being input to the determining device 15. First determiner 1
In step 5, it is determined whether the vibration to be diagnosed is an unbalanced vibration based on whether the vibration of the input rotation synchronous component is above a certain level.
If the result is unbalanced motion, vector calculator 1
6, vector calculation based on the above-mentioned equation (2) is performed on the unbalanced vibration data of the diagnosis time specified in advance. At this time, the time T from the present time is stored in the storage device 14 as data of the diagnosis time T.
(For example, data of the rotation synchronization component up to T seconds ago) is used. Next, the comparator 17 checks the tendency of the phase change based on equation (3), and the comparator 18
The tendency of amplitude change based on equation (4) is checked by.

The second accompaniment determiner 19 receives the results of the comparators 17 and 18 and finally determines whether or not it is thermal bending vibration. As a condition for this judgment, in this example, the number of data for which the results of comparators 17 and 18 are positive, that is, the number of data that simultaneously satisfy equations (3) and (4), respectively, is 80% or more of the total number of data. I use that.

The result of the second determiner 19 is displayed on the display device 21 via the control device 20. Display contents can be expressed as graphs or text depending on the display device, and a CRT is an example of a display device that can be expressed in an intuitive and easy-to-understand manner. FIG. 5 shows a display example of the diagnostic results of the diagnostic device according to the present invention.
An example using CRT is shown. In the figure, the unbalanced vibration data used for diagnosis is displayed in a vector diagram along with comments on the diagnosis results. When the display ends, the above process is automatically repeated for the data at the next point in time.

In the above embodiment, the phase reference angle for investigating the direction of change of the unbalanced vibration vector is the direction of the difference between the target vibration vectors at the beginning and end, as shown in _FIG. The direction of the straight line connecting the tips of _N is used, but instead of this straight line, as shown in Figure 6, a straight line approximated by the method of least squares so that the variation in each data of x ₁ to x _N is minimized is used. b can also be used. At this time, the positions of the first and last data are 1 →
1', move from N to N'. Using the least squares method in this way reduces the influence of data variations,
The direction of vector change can be determined more reliably. Furthermore, the interval of vibration data x ₁ to x _N may be set to 1 minute, and the data every minute may be used for diagnosis, but if diagnostic data is obtained every few seconds during the 1 minute interval, , if the average value of these data is used every minute, the variation in the data will be reduced and the reliability of the data will be improved. Furthermore, if monthly data is used as the data to be diagnosed, rotor bending over time can also be diagnosed.

FIG. 7 is a diagram for explaining the principle of a modification of the vibration diagnosis method according to the present invention. In the same figure,
x ₀ is a vibration vector due to initial unbalance before the occurrence of unbalance due to thermal bending. x _t1 , x _{t2 .} . . x _tN is an unbalance vibration vector that changes with the passage of time t after the occurrence of bending. Therefore, the resultant vector z _tN of these two unbalanced vibration vectors is z _tN = x _t1 + x _tN (7), and this vibration is observed after thermal bending occurs, but the difference between the two resultant vectors is The direction of a certain change vector is constant for y ₁ , . . . y _N . That is, if the direction of the unbalance vector x _tN caused by thermal bending is constant, the direction of the change vector is constant. Therefore, if the position of occurrence of bending is estimated and the influence coefficient α for the unbalance position is known, then the unbalance W caused by bending is
_o can be calculated using the following formula.

W _o = y _o /α ...(8) Therefore, since y _o can be measured, by examining the temporal changes in the phase and amplitude of the unbalance vector found from equation (8), it is possible to diagnose whether thermal bending has occurred. can do. That is, there is no change in the phase angle of W _o ,
By checking whether only the amplitude |W _o | increases or decreases in a certain direction, it is possible to diagnose whether or not thermal bending has occurred.

As described above in detail, the method and apparatus for diagnosing vibrations of a rotating machine according to the present invention has the effect of automatically diagnosing and displaying the occurrence of vibrations due to thermal bending of the rotor during operation.

[Brief explanation of the drawing]

FIG. 1 is a vector diagram showing the amplitude and phase of the unbalanced vibration component, FIG. 2 is a diagram showing the processing procedure of the vibration diagnosis method according to the present invention, FIG. 3 is a vector diagram showing the phase angle of the change vector, and FIG. FIG. 5 is an explanatory diagram showing an example of a diagnosis result, and FIGS. 6 and 7 are diagrams for explaining the principle of another embodiment of the present invention. It is a diagram. DESCRIPTION OF SYMBOLS 11, 12... Turbine rotor, 13... Generator rotor, 41, 46... Bearing, 51-56... Vibration detector, 6... Pulse detector, 7... Vibration meter, 8... Multi-point data input device, 9... A/ D converter, 10... data processing device, 15... first determiner, 16... vector calculator, 17, 18... comparator, 19... second determiner,
21...Display device.

Claims (1)

[Scope of Claims] 1. A method for diagnosing vibration properties due to rotor deformation of a rotating machine, which collects a plurality of pieces of vibration data within a certain period of time, and detects temporal changes in vibration vectors of unbalanced vibration components that are synchronized with rotor rotation. Find sequentially,
The difference between the first and last unbalanced vibration vectors is determined among the vibration data that targets whether the direction of change of the unbalanced vibration vector is constant, and based on the direction of this vector difference, the first unbalanced vibration vector and the By successively comparing the direction of the difference with the unbalanced vibration vector at subsequent points in time, the direction of change in the unbalanced vibration vector is determined, and the direction of change in the vibration vector and the tendency of amplitude change are each approximately constant. A method for diagnosing vibration of a rotating machine, characterized in that the cause of the vibration is diagnosed as thermal bending of a rotor. 2. Claim 1, characterized in that, as a reference direction for comparing the direction of change of unbalanced vibration vectors, the direction of a straight line based on least squares approximation is used for each unbalanced vibration vector of interest. Vibration diagnosis method for rotating machines described in . 3. To determine whether the amplitude change of the unbalanced vibration vector has a nearly constant tendency over time, use the initial unbalanced vibration data as a reference, and sequentially find the difference between this and the unbalanced vibration vector at each point in time, and check whether the amplitude is 2. The method for diagnosing vibrations of a rotating machine according to claim 1, wherein the diagnosis is carried out by comparing whether the vibrations increase or decrease over time. 4 Multiple pieces of vibration data within a certain period of time are
The vibration diagnosis method for a rotating machine according to claim 1, wherein the data is stored in a storage device. 5 Multiple pieces of vibration data within a certain period of time are
2. The vibration diagnosis method for a rotating machine according to claim 1, wherein the vibration diagnosis method is an average value of a plurality of vibration data obtained for each unit time obtained by dividing the certain period of time into the plurality of vibration data. 6 Multiple pieces of vibration data within a certain period of time are
Vibration data of several units or more on a monthly basis,
6. The vibration diagnosing method for a rotating machine according to claim 5, which enables diagnosis of rotor bending over time. 7. A device for diagnosing vibration properties due to rotor deformation of a rotating machine, which includes a vibration detector that detects vibrations of the rotating machine and facilitates A/D conversion of the signals of the vibration detector at many measurement points. a multi-point data input device, an A/D converter for A/D converting the output from the input device, a data processing device for extracting a rotation synchronous component from the A/D converted vibration signal; A waveform converter and a clock for providing a phase reference signal and a time signal to a data processing device, a rotation signal detector for providing a rotation signal to the waveform converter, and a rotation synchronous component input from the data processing device. A first determining device that determines whether the vibration is an unbalanced vibration, a calculation device that calculates an unbalanced vibration vector, a comparator that compares changes in the phase and amplitude of the vector difference of the unbalanced vibration, and a first determination device that determines whether the vibration is an unbalanced vibration. A vibration diagnostic device for a rotating machine, comprising: a second determiner that determines whether or not it is bending vibration; and a display device that displays the result of the second determiner.