JP2022024573A - Device and method of abnormality detection for rolling bearings - Google Patents

Abstract

To provide a device and a method of abnormality detection for rolling bearing capable of calculating the rotational speed of a rotary machine which is supported by rolling bearings and whose rotational speed changes.SOLUTION: An abnormality detection device comprises: a vibration detector 15 that detects vibration of rolling bearings 20 supporting a rotary machine 30 whose rotational speed changes; a data acquisition unit 11 that acquires a waveform in a time domain of the vibration detected by the vibration detector 15; a rotational speed calculation unit 12 that finds the envelope of the waveform in the time domain, converts the waveform of the envelope into intensity distribution data in a frequency domain, and finds the rotational speed of the rotary machine 30 using the intensity distribution data in the frequency domain; and an abnormality specifying unit 13 that specifies abnormal components of the rolling bearings 20 using the rotational speed found by the rotational speed calculation unit 12, and the characteristic frequency of the rolling bearings 20. The rotational speed calculation unit 12 finds, from the intensity distribution data in the frequency domain, vibration intensities at a plurality of frequencies proportional to the reference frequency in a predetermined range, and finds the rotational speed using the found intensities.SELECTED DRAWING: Figure 1

Description

The present invention relates to an abnormality detecting device and a method for detecting an abnormality in a rolling bearing, and more particularly to an apparatus and a method for detecting an abnormality in a rolling bearing by obtaining a rotation speed.

Anomalies in rolling bearings (eg, damage or defects) can be detected based on rotational speed and vibration. When a rolling bearing supports a rotating machine whose rotation speed changes, it is necessary to obtain the rotation speed of the rotating machine at the same time in order to analyze the measured vibration data when detecting an abnormality in the rolling bearing. However, in order to measure and collect the rotation speed of the rotating machine, it is necessary to increase the channel of the data logger by one to acquire the rotation speed information from the data logger, or to install the rotation speed sensor in the rotating machine. Therefore, there is a problem that the number of devices and elements used increases.

As a technique for solving such a problem, for example, there is a rolling bearing device described in Patent Document 1. The rolling bearing device described in Patent Document 1 is a device that calculates the number of rotations by using the vibration waveform detected by the rolling bearing, and is a bearing based on a detection signal from a vibration detector attached to a fixed wheel. It is equipped with a calculation unit that calculates the vibration and rotation speed of the main body. The calculation unit calculates the rotation speed of the bearing body based on the frequency component of the vibration obtained from the detection signal and the natural number of vibrations of the bearing body.

Japanese Unexamined Patent Publication No. 2017-133580

In the conventional technology such as the technology described in Patent Document 1, the measured vibration waveform is used for rotation without acquiring the rotation speed information from the data logger or installing the rotation speed sensor in the rotating machine. By calculating the number, the rotation speed of the rotating machine can be estimated. However, in the estimation of the rotation speed using the frequency analysis of the vibration waveform as in the technique described in Patent Document 1, in the case of a rotation axis with a small imbalance and a small vibration, or when the rotation speed is unique to a structure such as a housing. When the frequency is close to the vibration frequency, the vibration intensity at the frequency corresponding to the rotation speed is not always maximized, and the rotation speed may not be estimated accurately.

An object of the present invention is to provide an abnormality detection device for a rolling bearing and an abnormality detection method capable of calculating the rotation speed of a rotating machine supported by a rolling bearing and whose rotation speed changes.

The abnormality detection device according to the present invention is installed in a rolling bearing that supports a rotating machine whose rotation speed changes, and has a vibration detector that detects the vibration of the rolling bearing and a time region of the vibration detected by the vibration detector. The data acquisition unit for acquiring the waveform and the envelope of the waveform in the time region are obtained, and the waveform of the envelope in the time region is converted into the intensity distribution data in the frequency region by high-speed Fourier conversion processing, and the intensity distribution in the frequency region is obtained. The rolling bearing using a rotation speed calculation unit that obtains the rotation speed of the rotating machine using data, the rotation speed obtained by the rotation speed calculation unit, and a characteristic frequency indicating a factor of vibration of the rolling bearing. It is provided with an abnormality specifying part that identifies an abnormal part of the above. The rotation speed calculation unit obtains the vibration intensity at a plurality of frequencies proportional to the reference frequency, which is a frequency within a predetermined range, with respect to the intensity distribution data in the frequency domain, and uses the obtained intensity to obtain the rotation speed. Ask for.

The abnormality detection method according to the present invention includes a data acquisition step of acquiring a waveform in the time region of the vibration detected by the vibration detector that detects the vibration of the rolling bearing that supports the rotating machine whose rotation speed changes. , The anomaly detection device obtains the envelope of the waveform in the time region, converts the waveform of the envelope in the time region into the intensity distribution data in the frequency region by high-speed Fourier conversion processing, and uses the intensity distribution data in the frequency region. The rotation speed calculation step for obtaining the rotation speed of the rotary machine, the rotation speed obtained by the abnormality detection device in the rotation speed calculation step, and the characteristic frequency indicating the cause of the vibration of the rolling bearing are used. It has an abnormality identification process for identifying an abnormality part of a rolling bearing. In the rotation speed calculation step, the abnormality detecting device obtains the vibration intensity at a plurality of frequencies proportional to the reference frequency, which is a frequency within a predetermined range, with respect to the intensity distribution data in the frequency domain, and obtains the obtained intensity. The rotation speed is obtained by using.

According to the present invention, it is possible to provide an abnormality detection device for a rolling bearing and an abnormality detection method capable of calculating the rotation speed of a rotating machine supported by a rolling bearing and whose rotation speed changes.

It is a schematic diagram which shows the abnormality detection apparatus of a rolling bearing according to Example 1 of this invention. It is a schematic diagram which shows the structure of a rolling bearing. It is a schematic diagram of the cross section of the rolling bearing seen from the direction orthogonal to the axis of the rotating machine. It is a flowchart which shows the example of the process for identifying the abnormal part of a rolling bearing performed by the abnormality detection apparatus according to Example 1 of this invention. It is a flowchart which shows an example of the procedure of calculation of the rotation speed of a rotating machine performed by an abnormality detection device in step S5 of FIG. It is a flowchart which shows an example of the procedure of the process of performing the envelope process on the waveform of the time domain of vibration, and obtaining the envelope waveform, which is performed by the abnormality detection apparatus in step S3 of FIG. It is a flowchart which shows an example of the procedure of the process of identifying the abnormal part of a rolling bearing performed by the abnormality detection apparatus in step S6 of FIG. It is a flowchart which shows another example of the procedure of calculation of the rotation speed of a rotating machine performed by an abnormality detection device in step S5 of FIG. It is a flowchart which shows still another example of the procedure of calculation of the rotation speed of a rotating machine performed by an abnormality detection apparatus in step S5 of FIG. It is a figure which shows the example of the waveform of the time domain of the vibration of a rolling bearing. It is a figure which shows the example of the envelope waveform of the time domain of the vibration of a rolling bearing. It is a figure which shows the example of the frequency (intensity) analysis of the envelope waveform of the time domain of the vibration of a rolling bearing.

In the abnormality detection device and abnormality detection method for rolling bearings according to the present invention, the rotation speed of a rotating machine supported by a rolling bearing and whose rotation speed changes is calculated by using the vibration of the rolling bearing, and the calculated rotation speed is used. By identifying the abnormal parts of the rolling bearing, the abnormality of the rolling bearing is detected. In the rolling bearing abnormality detection device and the abnormality detection method according to the present invention, the rotation speed used for the abnormality detection of the rolling bearing is calculated as follows. That is, a reference frequency is set in the intensity distribution data in the frequency domain in which the waveform of the vibration of the rolling bearing is subjected to the envelope processing and then the high-speed Fourier conversion (hereinafter, also referred to as “FFT”) processing, and this reference frequency is set. The intensity of vibration at a plurality of frequencies proportional to is obtained, and the sum of the vibration intensities at the obtained multiple frequencies is obtained. Then, the sum of the vibration intensities is obtained for a plurality of reference frequencies, and the reference frequency at which the sum of the vibration intensities is maximized is calculated as the rotation speed of the rotating machine. The reference frequency at which the average of the vibration intensities at the obtained multiple frequencies is maximized can be calculated as the rotation speed of the rotating machine, and the squared average of the vibration intensities at the obtained multiple frequencies (root mean square). The reference frequency at which is the maximum can also be calculated as the rotation speed of the rotating machine.

Hereinafter, an abnormality detection device for rolling bearings and an abnormality detection method according to an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic view showing an abnormality detection device 10 for rolling bearings according to the first embodiment of the present invention. FIG. 1 also shows a rolling bearing 20 in which the abnormality detecting device 10 detects an abnormality, and a rotary machine 30 supported by the rolling bearing 20.

The abnormality detection device 10 includes a data acquisition unit 11, a rotation speed calculation unit 12, an abnormality identification unit 13, and an output unit 14, and is connected to the data logger 16. Further, the abnormality detection device 10 includes a vibration detector 15 installed in the rolling bearing 20 and a bandpass filter 18 (also referred to as “BPF”) for removing unnecessary vibration components.

The data acquisition unit 11 acquires the vibration waveform detected by the vibration detector 15. The rotation speed calculation unit 12 obtains the rotation speed of the rotating machine 30 from the waveform acquired by the data acquisition unit 11. The abnormality specifying unit 13 identifies an abnormal component of the rolling bearing 20 by using the rotation speed obtained by the rotation speed calculation unit 12. The output unit 14 can be configured with a display screen or an interface to an external device, and outputs an abnormality detection result.

The abnormality detection device 10 is a device that can be operated by a computer 17 such as a personal computer. The abnormality detection device 10 can also be configured by a computer 17.

The vibration detector 15 is composed of, for example, an acceleration sensor, and detects the vibration of the rolling bearing 20.

The data logger 16 is a storage device that stores the vibration data of the rolling bearing 20 detected by the vibration detector 15, that is, the waveform data of the vibration of the rolling bearing 20 in the time domain.

The data acquisition unit 11 of the abnormality detection device 10 acquires the waveform of the vibration of the rolling bearing 20 in the time domain from the data logger 16.

The rolling bearing 20 supports the rotary machine 30.

The rotary machine 30 includes a rotating shaft 31, and the shaft 31 is supported by a rolling bearing 20. The rotation speed of the rotary machine 30 (that is, the rotation speed of the shaft 31) changes during operation. Examples of the rotary machine 30 are a motor, a centrifugal compressor, and a pump.

FIG. 2 is a schematic view showing the configuration of the rolling bearing 20, and is a schematic view of the rolling bearing 20 as viewed from a direction parallel to the axis 31 of the rotary machine 30. The rolling bearing 20 includes an inner ring 21, an outer ring 22, a plurality of rolling elements 23, and a cage 24. The inner ring 21 is an annular member fixed to the shaft 31 of the rotating machine 30 and rotates together with the shaft 31. The outer ring 22 is an annular member fixed to the housing of the rolling bearing 20, and is arranged concentrically with the inner ring 21. The plurality of rolling elements 23 are members arranged in the space between the inner ring 21 and the outer ring 22, and rotate (revolve) in the same circumferential direction as the inner ring 21 while rotating. The cage 24 is a member that holds a plurality of rolling elements 23 so as to maintain relative positions in the circumferential direction with each other.

FIG. 2 also shows the diameter d of the rolling element 23 and the pitch circle diameter D (the diameter of the circle passing through the center of the rolling element 23).

FIG. 3 is a schematic cross-sectional view of the rolling bearing 20 as viewed from a direction orthogonal to the axis 31 of the rotary machine 30. FIG. 3 shows the contact angle α of the rolling bearing 20. The contact angle α is the direction 25 of the load applied between the rolling surface of the rolling element 23 (the surface of the inner ring 21 and the outer ring 22 in contact with the rolling element 23) and the rolling element 23, and the radial direction 26 (axis) of the rolling bearing 20. The angle between the direction perpendicular to 31).

FIG. 4 is a flowchart showing an example of a process for identifying an abnormal component of the rolling bearing 20 performed by the abnormality detecting device 10 according to the first embodiment of the present invention.

In step S1, the abnormality detection device 10 is started and the process is started.

In step S2, the data acquisition unit 11 of the abnormality detection device 10 acquires the waveform of the vibration of the rolling bearing 20 in the time domain from the data logger 16.

FIG. 10 is a diagram showing an example of the waveform in the time domain of the vibration of the rolling bearing 20 acquired in step S2. As shown in FIG. 10, in the rolling bearing 20, periodic vibration (change in acceleration) is generated in accordance with the rotation of the rotating machine 30 (rotation of the inner ring 21 of the rolling bearing 20).

In step S3, the rotation speed calculation unit 12 of the abnormality detection device 10 performs envelope processing on the vibration waveform acquired in step S2, and obtains the envelope of the waveform in the vibration time domain. That is, the abnormality detecting device 10 obtains a curve (envelope) that traces the contour of the waveform in the time domain of vibration.

FIG. 11 is a diagram showing an example of the envelope waveform in the time domain of the vibration of the rolling bearing 20 obtained in step S3. FIG. 11 shows the envelope of the waveform of the periodic vibration (change in acceleration) generated in the rolling bearing 20.

In step S4, the rotation speed calculation unit 12 of the abnormality detection device 10 performs a fast Fourier transform process (FFT process) on the time domain envelope waveform obtained in step S3, and converts the time domain envelope waveform into the frequency domain intensity. Convert to distribution data.

FIG. 12 is a diagram showing an example of frequency (intensity) analysis of the envelope waveform in the time domain of the vibration of the rolling bearing 20 obtained in step S4. As shown in FIG. 12, the frequency (intensity) analysis of the envelope waveform in the time domain of vibration shows that the vibration intensity is high (that is, a peak is generated) at some frequencies. ..

In step S5, the rotation speed calculation unit 12 of the abnormality detecting device 10 calculates the rotation speed of the rotating machine 30 by using the frequency (intensity) analysis of the envelope waveform in the time domain obtained in step S4.

In step S6, the abnormality specifying unit 13 of the abnormality detecting device 10 identifies an abnormal component of the rolling bearing 20 by using the rotation speed of the rotating machine 30 calculated in step S5 and the characteristic frequency of the rolling bearing 20.

In step S7, the abnormality detection device 10 ends the abnormality detection of the rolling bearing 20 and stops.

The characteristic frequency of the rolling bearing 20 is a value determined by the geometrical dimensions of the rolling bearing 20 and the rotation speed (rotational frequency) of the rotating machine 30, and is a frequency indicating a factor of abnormal vibration of the rolling bearing 20. The characteristic frequencies are, for example, the revolution frequency of the rolling element 23 (rotational frequency of the cage 24) FTF, the rotation frequency BSF of the rolling element 23, the outer ring rolling element passing frequency BPFO, and the inner ring rolling element passing frequency BPFI. These characteristic frequencies can be calculated by the following formula.
Revolution frequency of rolling element 23 (rotation frequency of cage 24) FTF
FTF = 1/2 * (1-d / D * cosα) * fs (1)
Rotation frequency BSF of rolling element 23
BSF = D / (2d) * [1- (d / D * cosα) ² ] * fs (2)
Outer ring rolling element passing frequency (frequency at which the rolling element 23 passes through one point on the outer ring 22) BPFO
BPFO = z / 2 * (1-d / D * cosα) * fs (3)
Inner ring rolling element passing frequency (frequency at which the rolling element 23 passes through one point of the inner ring 21) BPFI
BPFI = z / 2 * (1 + d / D * cosα) * fs (4)
However, in the formulas (1) to (4),
d is the diameter (mm) of the rolling element 23,
D is the pitch circle diameter (mm)
α is the contact angle (radian) of the rolling bearing 20.
z is the number fs of the rolling elements 23 and is the rotation frequency (Hz) of the rotating machine 30.
Is.

When the rolling bearing 20 has an abnormality such as damage or a defect, a peak occurs in the characteristic frequency related to the abnormal portion and the frequency of its harmonics in the waveform of the vibration frequency region of the rolling bearing 20. That is, when there is an abnormality in the cage 24, a peak occurs in the frequency represented by the equation (1) and the frequency of its harmonics. If there is an abnormality in the rolling element 23, a peak occurs at twice the frequency of Eq. (2) and the frequency of its harmonics. When the outer ring 22 has an abnormality, a peak occurs in the frequency of the equation (3) and the frequency of its harmonics. When the inner ring 21 has an abnormality, a peak occurs in the frequency of the equation (4) and the frequency of its harmonics.

It should be noted that these peaks can often be confirmed to be weak even when there are almost no abnormalities such as damage or defects in the rolling bearing 20. Therefore, the frequency giving these peaks (that is, the characteristic frequency and the frequency of its harmonics) can be used for calculating the rotation speed of the rotating machine 30 as described later.

FIG. 5 is a flowchart showing an example of a procedure for calculating the rotation speed of the rotary machine 30 performed by the rotation speed calculation unit 12 of the abnormality detection device 10 in step S5 of FIG. In order to identify abnormal parts using the characteristic frequency of the rolling bearing 20, it is premised that the rotation speed (rotation frequency fs) of the rotary machine 30 is known as shown in the equations (1) to (4). be. In this embodiment, for the intensity distribution data in the frequency domain obtained in step S4, the reference frequency is set as a frequency within a predetermined range, and the vibration intensities at a plurality of frequencies proportional to the reference frequency are obtained and obtained. Find the sum of the vibration intensities at the frequency of. The sum of the vibration intensities is obtained for a plurality of reference frequencies, and the reference frequency at which the sum of the vibration intensities is maximized is obtained as the rotation speed of the rotating machine 30.

In step S501, the abnormality detection device 10 sets initial values for the variables Wmax, SPD, and ISP to be used. Wmax represents the maximum value of the sum of the vibration intensities at a plurality of frequencies proportional to the reference frequency. SPD represents the rotation speed of the rotating machine 30. isp is the value of the loop counter i when the sum of the vibration intensities at a plurality of frequencies proportional to the reference frequency is maximized. In this embodiment, as an example, Wmax = 0.00, SPD = 0, and ISP = 0 are set.

In step S502, the abnormality detection device 10 sets the range of the reference frequency fs (i). Assuming that the minimum value of the reference frequency fs (i) is f1 and the maximum value of the reference frequency fs (i) is fN, the range of the reference frequency fs (i) is f1 to fN. The range of the reference frequency fs (i) (minimum value f1 and maximum value fN) can be predetermined based on, for example, the range of the operating frequency of the rotary machine 30. The range of the reference frequency fs (i) is set in the abnormality detection device 10 by the operator of the abnormality detection device 10 inputting the minimum value f1 and the maximum value fN into the abnormality detection device 10.

In step S503, the abnormality detection device 10 sets the value of the loop counter i to 1.

In step S504, the abnormality detection device 10 sets f1 to the reference frequency fs (i).

In step S505, the abnormality detection device 10 sets the frequency step Δf that changes the reference frequency fs (i). The frequency step Δf can be arbitrarily set so that the reference frequency fs (i) changes within the range set in step S502. For example, the frequency step Δf is set in consideration of the accuracy of the obtained rotation speed and the time required for calculation. Can be done. The frequency step Δf is set in the abnormality detection device 10 by being input to the abnormality detection device 10 by the operator of the abnormality detection device 10.

The reference frequency fs (i) can also be determined by using the characteristic frequency of the rolling bearing 20 and the frequency of its harmonics. That is, if an abnormal portion of the rolling bearing 20 can be predicted, the reference frequency fs (i) may be set to a frequency (including its harmonics) determined from the characteristic frequency related to the abnormal portion. If the abnormal portion of the rolling bearing 20 cannot be predicted, the reference frequency fs (i) may be set to a frequency (including its harmonics) determined from all the characteristic frequencies of the rolling bearing 20.

In step S506, the abnormality detection device 10 obtains the vibration intensity at a plurality of frequencies proportional to the reference frequency fs (i) for the intensity distribution data in the frequency domain. At this time, the plurality of frequencies proportional to the reference frequency fs (i) are set to be the maximum frequencies in the range of Δf × proportional multiple. Further, in the case of a machine in which an abnormal (damaged) part can be specified in advance, a specific damage frequency (frequency determined from the characteristic frequency) may be used. In the case of a machine in which the abnormal (damaged) part cannot be identified in advance, all damaged frequencies (frequency determined from the characteristic frequency) may be used. These frequencies may include the high frequency. The vibration intensity can be obtained from, for example, the magnitude of the amplitude. The number of frequencies for which the vibration intensity is obtained can be determined, for example, based on the number of characteristic frequencies. In this embodiment, since the number of characteristic frequencies is four (equations (1) to (4)), the number of frequencies for which the vibration intensity is obtained is at least four, and even the nth harmonic is considered. If, the number can be 4n (the maximum value of n is, for example, 4).

In step S507, the abnormality detection device 10 obtains the sum W (i) of the vibration intensities at the plurality of frequencies obtained in step S506.

In step S508, the abnormality detection device 10 compares the sum W (i) and the magnitude of Wmax of the vibration intensities obtained in step S507. If the sum W (i) of the vibration intensities is larger than Wmax, the process proceeds to step S509, and if W (i) is Wmax or less, the process proceeds to step S510.

In step S509, the abnormality detection device 10 substitutes the value of W (i) for Wmax, sets i for ISP, and then proceeds to step S510.

In step S510, the abnormality detection device 10 adds 1 to the value of the loop counter i.

In step S511, the abnormality detection device 10 sets the reference frequency fs (i) to fs (i-1) + Δf. That is, the reference frequency fs (i) changes so as to increase by Δf.

In step S512, the abnormality detection device 10 determines whether the reference frequency fs (i) is equal to or less than the maximum value fN in the range of the reference frequency fs (i). When the reference frequency fs (i) is equal to or less than the maximum value fN, the abnormality detection device 10 repeats the processes from step S506 to step S511. Since the reference frequency fs (i) is changed by Δf in step S511, in step S506, the abnormality detection device 10 determines the vibration intensity at a plurality of frequencies proportional to the reference frequency fs (i) changed by Δf. You will be asked. If the reference frequency fs (i) is larger than the maximum value fN, the process proceeds to step S513.

In step S513, the abnormality detection device 10 determines whether the value of ISP is zero. If the value of ISP is zero, the processes from step S505 to step S512 are repeated. If the value of ISP is not zero, the process proceeds to step S514.

In step S514, the abnormality detection device 10 sets the reference frequency fs (isp) to the rotation speed SPD of the rotary machine 30. The rotation speed SPD of the rotary machine 30 is obtained by calculating in this way.

The calculation of the rotation speed SPD utilizes the fact that the vibration intensity at a frequency other than the frequency proportional to the reference frequency fs (i) usually does not increase. The vibration of the rolling bearing 20 occurs at a frequency proportional to the rotation speed SPD of the rotating machine 30, and hardly occurs at frequencies other than this frequency, and even if it occurs, it is a small vibration (note that a multiple of the proportional vibration is vibration). Depends on the cause of). Therefore, when the reference frequency fs (i) coincides with the rotation speed SPD, the vibration intensity at a frequency proportional to the reference frequency fs (i) increases. However, since it is unclear which cause (in the case of bearings, the position of damage or defect) causes vibration, the sum of the vibration intensities W (i) at multiple frequencies that may cause vibration is obtained. , The reference frequency fs (isp) at which the sum W (i) is maximized is determined as the rotation speed SPD of the rotating machine 30.

FIG. 6 is a flowchart showing an example of a procedure of processing for obtaining an envelope waveform by performing an envelope process on a waveform in a vibration time domain, which is performed by the rotation speed calculation unit 12 of the abnormality detection device 10 in step S3 of FIG. be.

If there is an abnormality in the rolling bearing 20, a waveform including a component in which high-frequency vibration periodically fluctuates can be obtained as a waveform in the time domain of vibration. When the fast Fourier transform process (FFT process) is simply performed on the waveform in this time domain, only the spectrum of the high frequency component can be obtained, and the spectrum in which the high frequency vibration fluctuates periodically cannot be obtained. Therefore, in order to extract the periodicity of high-frequency vibration that fluctuates periodically, it is effective to perform envelope processing on the vibration waveform and then perform FFT processing.

In step S301, the abnormality detection device 10 performs FFT processing on the waveform (FIG. 10) in the time domain of vibration, and converts the vibration waveform data from the time domain to the frequency domain.

In step S302, the abnormality detection device 10 uses a bandpass filter 18 (also referred to as “BPF”) to remove unnecessary vibration components from the vibration waveform converted into the frequency domain. The rotation primary component and the low frequency vibration component become noise when extracting the waveform of the high frequency vibration. Therefore, the bandpass filter 18 is used to remove these unnecessary vibration components and take out the vibration components in the required frequency range. The frequency component taken out by the bandpass filter 18 can be determined based on the configuration of the rolling bearing 20 and the range of the operating frequency of the rotary machine 30. For example, since the component of the damaged rolling bearing 20 vibrates at the frequency of its natural vibration, the bandpass filter 18 is set so that the frequency component of this natural vibration can be taken out.

In step S303, the abnormality detection device 10 performs inverse high-speed Fourier conversion processing (IFFT processing) on the vibration waveform (waveform in the frequency domain) from which unnecessary vibration components have been removed by the bandpass filter 18, and the vibration waveform data. Is returned from the frequency domain to the time domain.

In step S304, the abnormality detection device 10 performs envelope processing on the vibration waveform returned to the time domain, and obtains an envelope waveform (FIG. 11) in the vibration time domain. The envelope processing is a process for obtaining a curve (envelope) that traces the contour of the waveform in the time domain of vibration. For the envelope processing, any existing method such as a method of applying a Hilbert transform to a waveform signal can be used.

By performing envelope processing on the vibration waveform as described above, it is possible to extract the periodicity of high-frequency vibration that fluctuates periodically. Therefore, the characteristic frequency used to identify the abnormality of the rolling bearing 20 is selected. Can be extracted.

FIG. 7 is a flowchart showing an example of a procedure for identifying an abnormal component of the rolling bearing 20 performed by the abnormality detecting device 10 in step S6 of FIG.

In step S601, the abnormality identification unit 13 of the abnormality detection device 10 determines whether the envelope waveform in the vibration frequency region obtained in step S4 of FIG. 4 contains an amplitude larger than a predetermined threshold value. investigate. If the intensity distribution data in the vibration frequency region contains an amplitude larger than this threshold value, it is considered that the rolling bearing 20 has an abnormality, and the process proceeds to step S602. If the intensity distribution data in the vibration frequency region does not include an amplitude larger than this threshold value, it is assumed that the rolling bearing 20 has no abnormality, and the process proceeds to step S603.

In step S602, the abnormality specifying unit 13 of the abnormality detecting device 10 identifies the abnormal component of the rolling bearing 20 by using the characteristic frequency. The characteristic frequency can be obtained from the equations (1) to (4) by setting the rotational speed SPD of the rotary machine 30 calculated in step S5 of FIG. 4 to the rotational frequency fs of the equations (1) to (4). .. The anomaly detection device 10 specifies a frequency that gives a peak of this amplitude with respect to an amplitude larger than the threshold value obtained in step S601. Then, the abnormality detecting device 10 examines which frequency corresponds to the characteristic frequency given by the equations (1) to (4) and the frequency of the harmonic thereof. In the abnormality detection device 10, the abnormal parts of the rolling bearing 20 are included in the cage 24, the rolling element 23, the outer ring 22, and the inner ring 21 from the characteristic frequency (including the harmonic of the characteristic frequency) corresponding to the frequency giving the peak. Identify which one.

In step S603, the output unit 14 of the abnormality detection device 10 outputs a detection result as to whether or not there is an abnormality in the rolling bearing 20. When the rolling bearing 20 has an abnormality, the abnormality detecting device 10 can also output what the abnormal component identified in step S602 is. The abnormality detection device 10 can output the detection result and the identified abnormal component to the display screen of the abnormality detection device 10, or output to the computer 17, an external display device, a storage device, or a printing device.

In the flowchart shown in FIG. 5, in step S507, the abnormality detecting device 10 obtains the sum W (i) of the vibration intensities at the plurality of frequencies obtained in step S506, and in step S514, the sum W (i) is calculated. The maximum reference frequency fs (isp) is calculated as the rotation speed SPD of the rotating machine 30. The abnormality detecting device 10 according to the present embodiment uses the average W1 (i) of the vibration intensity and the squared average W2 (i) of the vibration intensity in addition to the sum W (i) of the vibration intensities at a plurality of frequencies. It is also possible to calculate the rotation speed SPD of the rotating machine 30.

As described above, when the reference frequency fs (i) coincides with the rotation speed SPD, the vibration intensity at a frequency proportional to the reference frequency fs (i) increases. Since the cause of the vibration is unknown, the average W1 (i) of the vibration intensity and the squared average W2 (i) of the vibration intensity at a plurality of frequencies that may cause the vibration are obtained, and the average W1 (i) is obtained. ) And the reference frequency fs (isp) at which the squared average W2 (i) is maximized can be obtained as the rotation speed SPD of the rotating machine 30.

FIG. 8 is a flowchart showing another example of the procedure for calculating the rotation speed of the rotary machine 30 performed by the rotation speed calculation unit 12 of the abnormality detection device 10 in step S5 of FIG. The flowchart shown in FIG. 8 is different from the flowchart shown in FIG. 5 in the processes from step S517 to step S519. In the flowchart shown in FIG. 8, the reference frequency fs (isp) at which the average W1 (i) of the vibration intensities at the plurality of frequencies obtained in step S506 is maximized is calculated as the rotation speed SPD.

In step S517, the abnormality detection device 10 obtains the average W1 (i) of the vibration intensities at the plurality of frequencies obtained in step S506.

In step S518, the abnormality detection device 10 compares the magnitudes of W1 (i) and Wmax of the vibration intensities obtained in step S517.

In step S509, the abnormality detection device 10 substitutes the value of W1 (i) for Wmax and sets i for ISP.

FIG. 9 is a flowchart showing still another example of the procedure for calculating the rotation speed of the rotary machine 30 performed by the rotation speed calculation unit 12 of the abnormality detection device 10 in step S5 of FIG. The flowchart shown in FIG. 9 is different from the flowchart shown in FIG. 5 in the processes from step S527 to step S529. In the flowchart shown in FIG. 9, the reference frequency fs (isp) at which the squared average W2 (i) of the vibration intensities at the plurality of frequencies obtained in step S506 is maximized is calculated as the rotation speed SPD.

In step S527, the abnormality detection device 10 obtains the squared average W2 (i) of the vibration intensities at the plurality of frequencies obtained in step S506.

In step S528, the abnormality detection device 10 compares the squared average W2 (i) of the vibration intensity obtained in step S527 with the magnitude of Wmax.

In step S529, the abnormality detection device 10 substitutes the value of W2 (i) for Wmax and sets i for ISP.

As described above, the abnormality detection device 10 according to the present embodiment can calculate the rotation speed of the rotating machine 30 whose rotation speed changes by using the vibration data of the rolling bearing 20. Therefore, the rotation speed of the rotation machine 30 can be accurately calculated by the abnormality detecting device 10 without acquiring the rotation speed information from the data logger 16 or installing the rotation speed sensor in the rotation machine 30. , Cost reduction and improvement of work efficiency and work accuracy can be achieved.

The present invention is not limited to the above embodiment, and various modifications are possible. For example, the above-described embodiment has been described in detail in order to explain the present invention in an easy-to-understand manner, and the present invention is not necessarily limited to the embodiment including all the described configurations. Further, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment. It is also possible to add the configuration of another embodiment to the configuration of one embodiment. Further, it is possible to delete a part of the configuration of each embodiment and add / replace another configuration.

10 ... Abnormality detection device, 11 ... Data acquisition unit, 12 ... Rotation speed calculation unit, 13 ... Abnormality identification unit, 14 ... Output unit, 15 ... Vibration detector, 16 ... Data logger, 17 ... Computer, 18 ... Band path filter , 20 ... rolling bearing, 21 ... inner ring, 22 ... outer ring, 23 ... rolling element, 24 ... cage, 25 ... load direction, 26 ... radial direction, 30 ... rotating machine, 31 ... shaft.

Claims (10)

A vibration detector installed on a rolling bearing that supports a rotating machine whose rotation speed changes, and a vibration detector that detects the vibration of the rolling bearing.
A data acquisition unit that acquires a waveform in the time domain of the vibration detected by the vibration detector, and
The envelope of the waveform in the time domain is obtained, the waveform of the envelope in the time domain is converted into intensity distribution data in the frequency domain by a fast Fourier transform process, and the rotation of the rotating machine is performed using the intensity distribution data in the frequency domain. The rotation speed calculation unit that obtains the speed, and
Using the rotational speed obtained by the rotational speed calculation unit and the characteristic frequency indicating the cause of vibration of the rolling bearing, an abnormality specifying unit for identifying an abnormal component of the rolling bearing, and an abnormality specifying unit.
Equipped with
The rotation speed calculation unit obtains the vibration intensity at a plurality of frequencies proportional to the reference frequency, which is a frequency within a predetermined range, with respect to the intensity distribution data in the frequency domain, and uses the obtained intensity to obtain the rotation speed. Ask,
Anomaly detection device for rolling bearings. The rotation speed calculation unit changes the reference frequency, obtains the intensity of vibration at a plurality of frequencies proportional to the changed reference frequency, and obtains the rotation speed using the obtained intensity.
The abnormality detection device for rolling bearings according to claim 1. The rotation speed calculation unit obtains the sum of the vibration intensities at a plurality of frequencies proportional to the reference frequency, and obtains the reference frequency at which the sum is maximum as the rotation speed.
The abnormality detection device for rolling bearings according to claim 2. The rotation speed calculation unit obtains an average of vibration intensities at a plurality of frequencies proportional to the reference frequency, and obtains the reference frequency at which the average is maximum as the rotation speed.
The abnormality detection device for rolling bearings according to claim 2. The rotation speed calculation unit obtains the squared average of the vibration intensities at a plurality of frequencies proportional to the reference frequency, and obtains the reference frequency at which the squared average is maximum as the rotation speed.
The abnormality detection device for rolling bearings according to claim 2. The data acquisition process of acquiring the waveform in the time domain of the vibration detected by the vibration detector that detects the vibration of the rolling bearing that supports the rotating machine whose rotation speed changes.
The anomaly detection device obtains the envelope of the waveform in the time domain, converts the waveform of the envelope in the time domain into the intensity distribution data in the frequency domain by the fast Fourier transform process, and uses the intensity distribution data in the frequency domain. The rotation speed calculation step for obtaining the rotation speed of the rotary machine, and
An abnormality specifying step of identifying an abnormal component of the rolling bearing by the abnormality detecting device using the rotational speed obtained in the rotational speed calculation step and a characteristic frequency indicating a factor of vibration of the rolling bearing.
Have,
In the rotation speed calculation step, the abnormality detecting device obtains the vibration intensity at a plurality of frequencies proportional to the reference frequency, which is a frequency within a predetermined range, with respect to the intensity distribution data in the frequency domain, and obtains the obtained intensity. The rotation speed is obtained by using.
A method for detecting anomalies in rolling bearings, which is characterized by the fact that. In the rotation speed calculation step, the abnormality detecting device changes the reference frequency, obtains the intensity of vibration at a plurality of frequencies proportional to the changed reference frequency, and uses the obtained intensity to determine the rotation speed. Ask,
The method for detecting an abnormality in a rolling bearing according to claim 6. In the rotation speed calculation step, the abnormality detecting device obtains the sum of the vibration intensities at a plurality of frequencies proportional to the reference frequency, and obtains the reference frequency at which the sum is maximum as the rotation speed.
The method for detecting an abnormality in a rolling bearing according to claim 7. In the rotation speed calculation step, the abnormality detecting device obtains the average of the vibration intensities at a plurality of frequencies proportional to the reference frequency, and obtains the reference frequency at which the average is maximum as the rotation speed.
The method for detecting an abnormality in a rolling bearing according to claim 7. In the rotation speed calculation step, the abnormality detecting device obtains the squared average of the vibration intensities at a plurality of frequencies proportional to the reference frequency, and obtains the reference frequency at which the squared average is maximum as the rotation speed.
The method for detecting an abnormality in a rolling bearing according to claim 7.

Images