JP2007192828A - Abnormality diagnostic device, rolling bearing system having this, and method of diagnosing abnormality - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an abnormality diagnostic device which can reduce time and cost concerning diagnostics while time and effort concerning resolving and assembling of the device are alleviating. <P>SOLUTION: The abnormality diagnostic device comprises at least one sensor out of a vibration sensor, an ultrasonic sensor, and an AE sensor fixed to a rotatable member or the above-mentioned rest member; a filtering processor 35 which extracts a specific frequency bandpass corresponding to one natural frequency out of the above-mentioned rotatable member, the above-mentioned rest member, and the above-mentioned sensor from the waveform of the detected signal; an envelope processor 37 which detects a waveform absolute value after filter processing; a frequency analyzer 38 which analyzes the waveform frequency transferred from the above-mentioned envelope processor 37; a comparing collation unit 39 which compares the frequency based on survey data frequency resulting from a damage on the above-mentioned rotatable member computed based on a rotation speed signal; and an abnormality determining unit 42 which specifies the existence of abnormalities or a locus of abnormalities based on a comparing result. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an abnormality diagnosing device for diagnosing an abnormality of a rotating part used in, for example, an axle, a gear box or a power generation windmill for a railway vehicle, and a rolling bearing device having the abnormality diagnosing device.

  Conventionally, after rotating parts such as railway vehicles and wind turbines for power generation are used for a certain period of time, the bearing device and other parts are inspected for defects such as damage and wear. This inspection is performed by periodically disassembling the entire apparatus, and an inspection person can visually detect damage and wear made on the rotating parts. And the main defects discovered in the inspection are indentation caused by foreign object biting, peeling due to rolling fatigue, other wear, etc. in the case of bearing devices, missing teeth or wear in the case of gears, etc. In the case of a wheel, there is wear such as a flat, and if there is unevenness or wear that is not found in a new product, it is replaced with a new product and reassembled in the apparatus.

Further, as a conventional rotating component, the bearing device 100 having the sensor module shown in FIG. 14 has a module hole 103 formed on the outer peripheral surface of the outer ring 102 of the rolling bearing 101, and a speed sensor, a temperature sensor, and an acceleration sensor are formed in the module hole 103. Is inserted and fixed. And the detection signal which each sensor in module 104 generated is transmitted to a remote processing unit in a locomotive which pulls a freight car and a passenger car in which rolling bearing 101 is installed through a communication channel.
As for the speed, by detecting the instantaneous speed of the journal based on the pulses generated by the rotating wheels, the speed is compared with the speed of other bearings operating under similar conditions. Save and record the full cycle history experienced by a solid.
Also, the temperature is compared with the temperature of other bearings operating under similar conditions by simple level detection.
In addition, for vibrations, a simple RMS measurement of the energy level over a predetermined time interval is made, the energy level is compared with the past energy level stored in the processing unit, and other operating under similar conditions. The energy levels of the bearings are compared (for example, see Patent Document 1).

  Further, in the abnormality detecting device 110 of the rolling bearing device shown in FIG. 15, a sensor mounting hole 113 is formed at the lower end portion of the outer ring 112 of the double row tapered roller bearing 111, and the rotational speed sensor 114 is formed in the sensor mounting hole 113. A sensor unit 117 having a temperature sensor 115 and an acceleration sensor 116 is inserted. (For example, refer to Patent Document 2).

  Further, the sensor-equipped rotating stationary member 120 shown in FIG. 16 has a sensor mounting hole 123 formed in the lower end portion of the outer ring 122 of the double row tapered roller bearing 121, and a sensor unit 126 having a rotation speed sensor 124 and a temperature sensor 125. Is inserted into a partition case 127 provided in the sensor mounting hole 123 (see, for example, Patent Document 3).

  In addition, the conventional bearing abnormality detection device 130 shown in FIG. 17 includes a pickup 132 that converts the mechanical vibration of the bearing 131 into an electrical vibration and outputs it, and an automatic gain control amplifier 133 that amplifies the output of the pickup 132. The gain control terminal of the automatic gain control amplifier 133 calculates the effective value of the output of the bandpass filter 134 and the bandpass filter 134 of 1 to 15 kHz that removes noise generated from the drive system and other mechanical systems from the output of the amplifier 133. The effective value calculator 135 to be supplied to the envelope, the envelope circuit 136 for inputting the output of the bandpass filter 134, the effective value calculator 137 for inputting the output of the envelope circuit 136, and the output of the effective value calculator 137 are input. And an alarm circuit 138 that issues an alarm by a lamp or contact output when the value exceeds a predetermined value (for example, See Patent Document 4.).

  18 has a microphone 142, an amplifier 143, an electronic device 144, a speaker 145, and a monitor 146, which are arranged in the vicinity of the rolling bearing 141. The conventional bearing abnormality diagnosis device 140 shown in FIG. Have a configuration. The electronic device 144 is an arithmetic processing device, and includes a transducer 147 as a converter, an HDD 148 as a recording unit, an abnormality diagnosis unit 149 as an arithmetic processing unit, and an analog conversion output unit 150 (for example, Patent Documents). 5).

  In the conventional bearing abnormality diagnosis device 160 shown in FIG. 19, the electrical signal waveform output from the sensor 161 is converted into a digital file by the analog / digital converter 162 and then sent to the waveform processing unit 163. Then, the waveform processing unit 163 performs envelope processing to obtain an envelope spectrum. Further, in the extraction process, the waveform processing unit 163 extracts the inner ring wound component, the outer ring wound component, and the rolling element wound component, which are specific frequency components of the bearing component, from the envelope spectrum using a predetermined formula. The calculation unit 164 performs a calculation process, the determination unit 165 performs a comparison process, the determination result is output from the output circuit 166, and is notified to the inspector via the speaker 167 and the monitor 168 (for example, Patent Document 6). reference.).

Further, as shown in FIG. 20, the abnormal vibration frequency generated in each member of the rolling bearing is obtained geometrically from the bearing rotational speed and the shape and size of the bearing parts, and various signal processing methods using this are known. (See, for example, Non-Patent Document 1).
JP-T-2001-500597 (pages 10-16, FIG. 1) JP 2002-295464 A (page 4-5, FIG. 1) JP 2002-242928 PR (page 4-5, Fig. 1) JP-A-2-205727 (page 2-3, FIG. 1) JP 2000-146762 A (page 4-6, FIG. 1) JP 2001-021453 (page 5-6, FIG. 1) Detection and prediction of abnormality in rolling bearings “Lubrication”, Vol. 23, No. 3 (1978) pp. 183 to 187

  However, in the method in which the entire mechanical device is disassembled and the person in charge inspects it visually, it takes a lot of time and cost to disassemble the entire device, and a lot of time may be required for reassembly. In particular, in the case of a wind turbine for power generation, it is often used offshore, and the number of wind turbines may be large. Therefore, at present, maintenance personnel often go to the site and inspect the rotating parts of individual wind turbines. In this case, it takes a lot of time and cost, and the efficiency of maintenance is sometimes poor.

  Further, since a large number of parts are visually inspected within a limited time, there is a possibility that a defect is overlooked. Further, there is an individual difference in the determination of the degree of defect, and parts may be replaced even if there is substantially no defect, which may be useless cost. Further, when reassembling, the inspection itself may cause a new cause of the defect of the component, such as making a dent on the rotating component that did not exist before the inspection.

  Further, depending on the mechanical device, even if an abnormality occurs in the rotating component, it may take time to replace the component, and thus the device cannot be stopped immediately. In this case, in order to confirm the extent of damage, use a vibration meter or the like to compare with the normal vibration value, or estimate the degree by hearing from a skilled worker, and determine the time to replace the parts. ing. However, since the vibration of the mechanical device is larger from the vibration source other than the rotating parts, it is difficult to identify the degree of damage. For this reason, efficient parts replacement may not be possible.

  In Patent Documents 1 to 6 and Non-Patent Document 1, there are known methods for diagnosing abnormalities in rotating parts from the outside using a temperature sensor or a vibration sensor. In these documents, for example, a vibration sensor is used. No measures have been taken against the processing burden when analyzing the detected signal to diagnose an abnormality, and the hardware of the processing circuit or the like may increase, or the load of software may increase.

  The present invention has been made to solve the above-mentioned problems, and diagnoses the presence / absence of abnormalities in the rotating parts of the apparatus, the identification of the parts, and the degree of damage while reducing the labor required for disassembling and assembling the apparatus. An object of the present invention is to provide an abnormality diagnosis device capable of reducing the time and cost required for diagnosis and a rolling bearing device having the same.

The object of the present invention is achieved by the following configurations.
(1) An abnormality diagnosis device for diagnosing an abnormality of a rotating component that rotates relative to a stationary member,
At least one of a vibration sensor, an ultrasonic sensor, and an AE sensor fixed to the rotating component or the stationary member;
A filter processing unit that extracts a specific frequency band corresponding to a natural frequency of any one of the rotating component, the stationary member, and the sensor from a waveform of a signal detected by the sensor;
An envelope processing unit for detecting the absolute value of the filtered waveform transferred from the filter processing unit;
A frequency analysis unit for analyzing the frequency of the waveform transferred from the envelope processing unit;
A comparison / collation unit that compares the frequency based on the damage of the rotating part calculated based on the rotation speed signal and the frequency based on the actual measurement data;
Based on the comparison result in the comparison and collation unit, an abnormality determination unit that identifies the presence or absence of abnormality and a site of abnormality,
An abnormality diagnosis apparatus comprising:
(2) The filter processing unit extracts a plurality of the frequency bands corresponding to the plurality of natural frequencies, and the abnormality determination unit determines damage based on a comparison result in the comparison / collation unit in each frequency band. The abnormality diagnosis device according to (1), characterized in that the degree is diagnosed.
(3) An abnormality diagnosis device for diagnosing an abnormality of a rotating component that rotates relative to a stationary member,
At least one of a vibration sensor, an ultrasonic sensor, and an AE sensor fixed to the rotating component or the bearing housing;
A filter processing unit that extracts a plurality of specific frequency bands corresponding to a plurality of natural frequencies of any one of the rotating component, the bearing box, and the sensor from a waveform of a signal detected by the sensor;
An arithmetic processing unit that calculates an effective value or a crest factor of the waveform after filtering transferred from the filter processing unit;
A comparison unit that compares a calculation result in each frequency band calculated by the calculation processing unit with a normal value;
Based on the comparison result in the comparison unit, an abnormality determination unit for diagnosing the presence or absence of abnormality or the degree of damage;
An abnormality diagnosis apparatus comprising:
(4) The abnormality diagnosis apparatus according to any one of (1) to (3), further including a result output unit that displays a determination result in the abnormality determination unit.
(5) At least one abnormality detection sensor among the vibration sensor, the ultrasonic sensor, the AE sensor, and the temperature sensor is housed and fixed in the same casing in the load zone of the rotating component. The abnormality diagnosis device according to any one of 1) to (4).
(6) The rotating component is a rolling bearing, and the stationary member is a bearing box that supports the rolling bearing, and a flat portion is provided on a part of the outer peripheral surface of the bearing box on the load zone side, The abnormality diagnosis device according to any one of (1) to (5), wherein the sensor is fixed.
(7) The apparatus according to any one of (1) to (6), further including a headphone or a speaker that outputs a vibration signal detected by the vibration sensor and filtered by the filter processing unit as sound. Abnormality diagnosis device.
(8) A rolling bearing device for a railway vehicle comprising the abnormality diagnosis device according to any one of (1) to (7).
(9) A rolling bearing device for a speed reducer comprising the abnormality diagnosis device according to any one of (1) to (7).

  According to the abnormality diagnosis apparatus of the present invention, it is possible to simultaneously inspect the degree of defects and damage of a plurality of parts in actual operation without disassembling the apparatus in which the rotating parts are incorporated. Further, since the frequency band corresponding to the natural frequency of the machine part is extracted, it is possible to measure with high sensitivity and high S / N ratio. As a result, the inspection time and cost can be reduced, and the safety and reliability of the rotating parts of the machine or the apparatus can be improved.

Hereinafter, a rotating component abnormality diagnosis apparatus according to a first embodiment of the present invention will be described in detail with reference to the drawings.
FIG. 1 shows a rolling bearing device 10 for a railway vehicle to which an abnormality diagnosis device is applied. A rolling bearing device 10 for a railway vehicle includes a double-row tapered roller bearing 11 for a railway vehicle axle that is a rotating part, and a bearing box 12 that is a stationary member that constitutes a part of the railway vehicle carriage.

The double-row tapered roller bearing 11 includes a pair of inner rings 14 and 14 having inner ring raceways 15 and 15 inclined on the outer peripheral surface in a tapered outer surface shape, and a pair of outer ring raceway surfaces 17 inclined on the inner peripheral surface in a tapered inner surface shape. A single outer ring 16 having 17, and tapered rollers 18 that are a plurality of rolling elements arranged in multiple rows between the inner ring raceway surfaces 15, 15 of the inner rings 14, 14 and the outer ring raceway surfaces 17, 17 of the outer ring 16; Further, annular punching cages 19 and 19 for holding the tapered rollers 18 in a rollable manner and a pair of seal members 20 and 20 are provided.
The bearing box 12 includes a housing 21, a front lid 22 disposed on the front end side of the housing 21, and a rear lid 23 disposed on the rear end side.

  An inner ring spacer 24 is disposed between the inner rings 14 and 14. An axle 25 that is a rotating shaft is press-fitted into the inner rings 14, 14 and the inner ring spacer 24, and the outer ring 16 is fitted into the housing 21. The double row tapered roller bearing 11 is loaded with a radial load due to the weight of various members and an arbitrary axial load, and the upper portion of the outer ring 16 is in a load zone. Here, the load zone refers to a region where a load is applied to the rolling elements.

  One seal member 20 disposed on the front end side of the axle 25 is assembled between the outer end portion of the outer ring 16 and the front lid 22, and the other seal member 20 disposed on the rear end side is disposed on the outer ring. 16 is assembled between the outer end of 16 and the rear lid 23.

  The housing 21 constitutes a side frame of a railcar bogie, and is formed in an annular shape so as to cover the outer peripheral surface of the outer ring 16. A recess 26 is formed in the outer peripheral portion of the housing 21 at the central portion in the axial direction of the double row tapered roller bearing 11. The concave portion 26 is provided with a flat portion 27 at the bottom thereof, and an abnormality detection sensor 31 constituting a part of the abnormality diagnosis device is fixed to the flat portion 27.

  The abnormality detection sensor 31 is a composite sensor in which a vibration sensor, a temperature sensor, an AE (acoustic emission) sensor, and an ultrasonic sensor are integrally housed and fixed in a housing. 1 includes a vibration sensor 32 and a temperature sensor 33. The abnormality detection sensor 31 shown in FIG. The vibration sensor 32 is a vibration measuring element such as a piezoelectric element, and detects peeling of the inner and outer ring raceway surfaces 15, 15, 17, 17 of the double row tapered roller bearing 11, missing gears, flat wear of wheels, and the like. Used for. The vibration sensor 32 may be an acceleration, speed, displacement type, or the like that can convert vibration into an electrical signal. When the vibration sensor 32 is attached to a mechanical device having a lot of noise, the use of an insulation type is more affected by noise. This is preferable because there is nothing.

  The temperature sensor 33 is a non-contact type temperature measuring element such as a thermistor temperature measuring element, a platinum resistance temperature detector, or a thermocouple. As the temperature sensor, when the ambient temperature exceeds a specified value, a temperature fuse that does not conduct when the bimetal contact is separated or the contact is melted may be used. In that case, when the temperature of the apparatus exceeds a specified value, the temperature abnormality is detected by blocking the conduction of the temperature fuse.

  Further, the abnormality detection sensor 31 is attached to the load area of the radial load of the bearing housing fitted to the non-rotating side race of the bearing. For this reason, for example, when the bearing raceway surface is damaged, the collision force generated when the rolling element passes through the damaged portion is larger in the load area than in the no-load area, and the bearing load area is more sensitive. Abnormal vibration can be detected.

  FIG. 2 shows a signal processing system diagram of the abnormality diagnosis apparatus using the vibration sensor 32 of the abnormality detection sensor 31. The vibration signal generated by the vibration sensor 32 is transferred to the filter unit 35 after amplification through the signal transmission means 34. The filter unit 35 is based on the natural frequency of any one of the double-row tapered roller bearing 11 that is a rotating part, the bearing box 12 that is a stationary member, and the abnormality detection sensor 31 that is stored in the natural frequency storage unit 36. Then, only a predetermined frequency band corresponding to the natural frequency is extracted from the vibration signal.

  This natural frequency is vibrated by a striking method using any one of the double-row tapered roller bearing 11, which is a rotating component, a gear and a wheel, a bearing box 12, which is a stationary member, and an abnormality detection sensor 31, as a measurement object. The vibration detector attached to the object to be measured or the sound generated by the impact can be easily obtained by frequency analysis. When the object to be measured is a double-row tapered roller bearing, a natural frequency due to any of the inner ring, outer ring, rolling element, cage, etc. is given. In general, there are a plurality of natural frequencies of mechanical parts, and the amplitude level at the natural frequencies is high, so the sensitivity of measurement is good.

  Thereafter, the envelope processing unit 37 performs absolute value detection processing for detecting the absolute value of the waveform for the predetermined frequency band extracted by the filter unit 35. Further, the frequency analysis unit 38 performs waveform frequency analysis processing, and the actual measurement data is transferred to the comparison and verification unit 39.

  On the other hand, in the theoretical frequency calculation unit 41, the calculated value data of the frequency component resulting from damage to the rotating parts such as bearing separation, gear loss, wheel flatness, etc., calculated based on the rotational speed information 40, is compared and compared 39. Forwarded to Then, the comparison / verification unit 39 compares the measured value data with the calculated value data, the abnormality determination unit 42 identifies the presence / absence of vibration abnormality and the abnormal region, and the result output unit 43 identifies the presence / absence of vibration abnormality. The part is output and an alarm can be output. Information transfer to the result output unit 43 is performed by wire or wireless.

The present embodiment focuses on the fact that an abnormality diagnosis of a rotating component can be performed by extracting and analyzing only a predetermined frequency band corresponding to the natural frequency from the vibration signal.
Specifically, the present applicant uses a double row tapered roller bearing having an outer diameter of 208 mm, an inner diameter of 130 mm, a width of 152 mm, and a number of rollers of 25, and vibrates the bearing box by a striking method while rotating the inner ring at 200 min ^−1. Thus, the measurement result of the natural frequency of the bearing box as shown in FIG. 3 was obtained. In addition, a radial load of 30 kN is acting on the double row tapered roller bearing, and the vibration detector was measured while being fixed at the bearing box load zone position. From this result, it can be seen that the waveform data shown in FIG. 3 substantially coincides with the result of the vibration frequency of the normal double-row tapered roller bearing shown in FIG. 4, and thereby, the predetermined frequency corresponding to the natural frequency. It was found that the presence or absence of abnormality can be diagnosed by extracting and analyzing only the frequency band.

  In this method, for example, based on the rotational speed information detected from an electric motor or the like and the design specifications of the rotating element parts, it is possible to easily perform frequency component calculation and comparison verification. In addition, the storage of the natural frequency or the processing of the vibration signal after amplification performs various data processing and calculation, and can be configured by, for example, a computer or a dedicated microchip. Further, the arithmetic processing may be performed after the detected signal is stored in a storage means such as a memory.

  Further, as a processing method for abnormality diagnosis based on the vibration signal performed by the comparison / verification unit 39, the following method may be used.

(1) Method of Using Effective Value of Envelope Data as Reference Value In this method, a frequency component generated at the time of abnormality is obtained based on the equation of FIG. Then, an effective value of the envelope data is calculated, and a reference value for comparison is obtained from the effective value. Then, the frequency component equal to or higher than the reference value is calculated and compared with the frequency component generated at the time of abnormality. Hereinafter, description will be given with reference to FIG.

  First, the vibration of the bearing is detected through the vibration sensor 32 housed in the abnormality detection sensor 31 (step S101). The detected signal is amplified with a predetermined amplification factor, converted into a digital signal by an A / D converter (step S102), and filter processing for extracting only a predetermined frequency band is performed (step S103). Envelope processing is performed on the digital signal after the filter processing (step S104), and the frequency spectrum of the digital signal after the envelope processing is obtained (step S105).

Next, the effective value of the digital signal of the actual measurement data is calculated (step S106), and further, a reference value used for abnormality diagnosis is calculated based on the effective value (step S107). Here, the effective value is obtained as the root mean square of the frequency spectrum after the envelope processing. The reference value is calculated based on the following formula (1) or (2) based on the effective value.
(Reference value) = (effective value) + α (1)
(Reference value) = (effective value) × β (2)
α, β: Predetermined values that vary depending on the type of data

  On the other hand, based on the table shown in FIG. 20, the (theoretical) frequency generated due to the abnormality of the bearing is obtained (step S108), and the level of the abnormal frequency component of each member corresponding to the obtained frequency, that is, the inner ring scratch component Si (Zfi), outer ring wound component So (Zfc), rolling body wound component Sb (2fb) and cage component Sc (fc) are extracted (step S109) and compared with the reference value calculated in step S107. (Step S110). If all the component values are smaller than the reference value, it is determined that no abnormality has occurred in the bearing (step S112). If any of the components is greater than or equal to the reference value, It is determined that an abnormality has occurred (step S111).

(2) Method of Finding Peak of Spectrum and Comparing Peak Frequency and Abnormal Frequency In this method, a frequency component generated at the time of abnormality is obtained based on the equation of FIG. Then, it is checked whether or not a predetermined number or a peak equal to or higher than a reference value in the frequency spectrum obtained by the frequency analysis unit 38 corresponds to a frequency component in which an abnormality occurs.

(3) Method of using fundamental frequency and specific harmonics In this method, the primary value that is the fundamental frequency of the abnormal frequency component, the secondary value that has twice the fundamental frequency, and the frequency that is four times the fundamental frequency. Compare the peak frequency with the frequency generated at the time of abnormality for the quaternary value with, and if it is determined that there is an abnormality in at least two frequencies, it is finally determined that there is an abnormality. to decide.

(4) Method for estimating the magnitude of damage together with abnormality diagnosis In this method, the frequency spectrum after the envelope processing is used to confirm that the outer ring is damaged at a large peak frequency. By comparing the value with a reference level that is an average value of the entire frequency spectrum, the magnitude of damage in the outer ring causing the abnormality is estimated.

(5) A method in which the level difference from a harmonic component that is a natural number multiple of the fundamental frequency is used as a reference value. In this method, the fundamental frequency is 2, 3, 4, ... 2,3,4, ... with n times the frequency Count the number of the n-th level exceeding the reference value, and if the specified number exceeds the reference value, it is abnormal Is determined to have occurred. Specifically, counting is performed when the n-th order value is {(first-order level) − (n−1) · a} (dB) or more with respect to the first-order level. Here, a is an arbitrary value.

(6) Method using effective value for each frequency band In this method, abnormality diagnosis is performed using the effective value of the frequency band including the frequency caused by the abnormality, not the value of the peak level itself of the frequency caused by the abnormality. . Specifically, the effective value of the frequency band including the frequency caused by the abnormality is a root mean square or partial overall of the frequency band level. Here, the root mean square and the partial overall are obtained by a predetermined formula. Overall means the sum total of a specific designated section.

  According to the abnormality diagnosis apparatus of the first embodiment, paying attention to the fact that the natural frequency of any of the rotating parts, stationary members, and sensors substantially matches the vibration frequency of the normal rotating parts, the natural frequency is obtained by filtering. Only the frequency band corresponding to is extracted from the vibration signal, envelope processing and frequency analysis are performed, and abnormality diagnosis is performed. As a result, the frequency band to be diagnosed among the vibration signals can be specified, so that the influence of noise can be avoided, the hardware such as a signal processing circuit for detecting an abnormality can be made compact, and the software load can be reduced. . Further, analysis in a band with a high amplitude level is possible, and measurement with high sensitivity and high S / N ratio becomes possible.

Here, the above-mentioned double row tapered roller bearing (outer diameter 208 mm, inner diameter 130 mm, width 152 mm, number of rollers 25) having normal and scratches on the outer ring raceway surface is incorporated in a bearing box, and rotated at an inner ring rotational speed of 200 min ⁻¹ . Abnormalities were diagnosed by envelope analysis and frequency analysis of radial vibration of the bearing housing. FIG. 6 shows the results of envelope analysis and frequency analysis without filtering the vibrations of normal products and outer ring wound products, and FIG. 7 shows the results of filtering the vibrations of normal products and outer ring wound products in the 1 to 3 kHz band. The results of envelope analysis and frequency analysis are shown.

  In this example, a high-accuracy diagnosis was possible by using a reference value (rms + 3 dB) obtained by drawing a line in the figure as a reference for discriminating between normal and abnormal. That is, after extracting the frequency component exceeding this reference value, compared with the inner ring wound component Si (Zfi), the outer ring wound component So (Zfc), the rolling element wound component Sb (2fb) and the cage component Sc (fc), The presence / absence of abnormality and the part are specified from the degree of coincidence. In FIG. 6, no significant peak appears, but in FIG. 7, the outer ring wound product matches the frequency component caused by the wound. The component which appeared is appearing notably.

Thereby, about FIG.7 (b), it can determine with the outer ring | wheel of a bearing having a damage | wound. In addition, it can be seen that since the size of the wound of the outer ring wound product used in this test is small, the scratch component does not appear remarkably unless the envelope analysis and the frequency analysis are performed after the filter processing in the frequency band of 1 to 3 kHz. The reference value (rms + 3 dB) described above can be arbitrarily set / changed according to the operating state and application, but is preferably rms + 2 dB to rms + 6 dB in consideration of diagnostic accuracy.
Although the effective value is taken up here, an average value such as a moving average or a crest factor (= peak level / average value) may be used.

  Next, an abnormality diagnosis apparatus according to the second embodiment will be described with reference to FIG. The abnormality diagnosis apparatus of the second embodiment is for diagnosing the degree of damage in addition to the presence / absence of an abnormality and the identification of an abnormal part based on a vibration signal from a vibration sensor. The operation is almost the same as that of the first embodiment. For this reason, about the member which has the structure and effect | action similar to the member already demonstrated, description is simplified or abbreviate | omitted by attaching | subjecting the same code | symbol or an equivalent code | symbol in a figure.

  As described above, there are multiple natural frequencies of machine parts. When damage such as scratches occurs on the bearing raceway surface fitted in the bearing housing, impact vibration occurs due to the rolling elements passing over the scratches. The natural frequency of the bearing housing is excited. This natural frequency is transmitted in different ways depending on the size of the scratch and the rotational speed. That is, in the case of a large scratch, since the excitation source is large, a plurality of natural frequencies based on the structure of the bearing box and the like are excited. On the other hand, since the excitation source is small in the case of a small scratch, the natural frequency excited is smaller than in the case of a large scratch.

  For this reason, the natural frequency storage unit 36 stores the measurement result of the natural frequency of any one of the rotating component, stationary member, and sensor that is vibrated by the striking method, and the filter unit 44 is stored in the natural frequency storage unit. Based on the peaks and valleys of the waveform data of the stored measurement results, the vibration signal obtained by the vibration sensor 32 is divided into 2 to 7 specific frequency bands and extracted.

The envelope processing unit 37 performs absolute value detection processing for detecting the absolute value of the waveform for each frequency band extracted by the filter unit 44. Further, the frequency analysis unit 38 performs waveform frequency analysis processing, and the actual measurement data is transferred to the comparison and verification unit 39.
The comparison / verification unit 39 compares the actual measurement data with the calculated value data calculated by the theoretical frequency calculation unit 41 for each frequency band. The abnormality determination unit 45 identifies whether or not there is an abnormality in the double-row tapered roller bearing 11 and specifies an abnormal part, and confirms whether or not a damage component is generated in each frequency band for each component. Here, it is determined that the damage is large when the damage component occurs in the entire frequency band, and the damage is small when the damage component occurs only in the specific frequency band. Thereafter, the result output unit 43 outputs the presence / absence of vibration abnormality, the identification of the abnormal part, and the degree of damage.

  Here, in the double row tapered roller bearing having the natural frequency shown in FIG. 3, a bearing with a large scratch on the outer ring raceway surface and a small bearing are assembled and rotated in the bearing box, and the degree of damage in the second embodiment is diagnosed. It was. FIG. 9 shows the result of vibration frequency analysis of a bearing (hereinafter referred to as “outer ring small product”) whose outer ring raceway surface is greatly damaged, and FIG. 10 is a bearing whose outer ring raceway surface damage is small (hereinafter referred to as “outer ring small scratch”). The result of vibration frequency analysis is shown below. The size ratio of the damage magnitude is about 8.

  As shown in FIG. 3, there are a plurality of natural frequencies of the bearing housing in the frequency range of 10 kHz. Compared with the vibration frequency analysis results in FIGS. 9 and 10, this corresponds to the natural frequency of the bearing housing. It can be seen that the vibration level is high in the frequency band. Further, comparing FIG. 9 with FIG. 10, it is confirmed that the level of the natural frequency of the bearing box varies depending on the size of the scratch.

  Therefore, based on the measurement result of the natural frequency of the bearing box shown in FIG. 3 stored in the natural frequency storage unit 36, 0 to 3 kHz, 3 to 4.2 kHz, 4.2 from the vibration signal of FIGS. 9 and 10. The filter unit 44 extracted each frequency band of ˜6 kHz, 6-8 kHz, and 8-10 kHz, and envelope processing and frequency analysis were performed in each frequency band. FIGS. 11 and 12 are examples of the results of frequency analysis in the 0 to 3 kHz and 4.2 to 6 kHz bands, respectively. In the outer ring small product, the outer ring damage component is generated in both bands, while in the outer ring small chip product, the outer ring damage component is generated in the frequency band of 0 to 3 kHz, and the frequency band of 4.2 to 6 kHz. Then, the outer ring damage component is not seen.

  Table 1 summarizes whether or not the outer ring damage component is generated in the frequency analysis results in each frequency band. From this result, in the frequency band including each natural frequency of the bearing box, the outer ring damage component is generated in all frequency bands in the outer ring small product, and in the outer ring small scratch product in the frequency band of 0 to 3 kHz. Only the outer ring damage component has occurred. For this reason, it is confirmed that the damage is large when the outer ring damage component is generated in the entire frequency band, and the damage is small when the damage component is generated only in the specific frequency band.

  Therefore, according to the second embodiment, only a specific frequency band corresponding to the natural frequency of the bearing box is extracted, frequency analysis is performed in each frequency band, and presence / absence of frequency components due to damage for each frequency band By checking the level and the level, it is possible to diagnose the degree of defect and damage of each component in the actual operation state. As a result, it is possible to know the optimal replacement period of the parts and to perform efficient maintenance.

Next, an abnormality diagnosis apparatus according to the third embodiment of the present invention will be described. In the third embodiment, the signal processing system of the abnormality diagnosis device shown in FIG. 13 is applied to the rolling bearing device for a railway vehicle shown in FIG.
FIG. 13 is a signal processing system diagram using the vibration sensor 32 and the temperature sensor 33 of the abnormality detection sensor 31. The vibration signal generated by the vibration sensor 32 and the temperature signal generated by the temperature sensor 33 are sent to the calculation unit 50 via the signal carrying means 34. The calculation unit 50 allocates a vibration signal for each specific frequency band corresponding to a plurality of natural frequencies of any one of a rotating component, a stationary member, and a sensor, and the function of the filter unit 44 shown in FIG. Including. The vibration signal and the temperature signal allocated for each frequency band are input to the comparator 51.

  The comparator 51 compares the temperature data given by the calculation unit 50 with a preset temperature threshold value stored in the threshold value setting unit 52. At the same time, the vibration data calculated by the calculation unit 50 and the vibration threshold stored in the threshold setting unit 52 are compared for each frequency band. The abnormality determination unit 53 receives the comparison result of the comparator 51, and outputs a temperature abnormality determination signal to the result output unit 54 when the temperature data signal value exceeds the temperature threshold value. The result output unit 54 outputs a temperature abnormality alarm. Alarms are forwarded by wire or wireless and activated.

  When the vibration signal value exceeds the vibration threshold value, the abnormality determination unit 53 outputs a vibration abnormality determination signal to the result output unit 54. The result output unit 54 outputs a vibration abnormality alarm. Alarms are forwarded by wire or wireless and activated. At this time, the temperature threshold value and the vibration threshold value stored in the threshold value setting unit 52 and the temperature / vibration abnormality determination signal output from the abnormality determination unit 53 may use an effective value or a peak value within an arbitrary time.

  In this embodiment, the calculating part 50 calculates the vibration effective value or crest factor for every frequency band containing each natural frequency of the bearing box 12 mentioned above. The comparator 51 compares the calculated effective value in each frequency band with a vibration effective value of a normal bearing (hereinafter referred to as a normal product) or a normal value of the crest factor as a threshold value. The abnormality determination unit 53 determines the presence or absence of an abnormality or the degree of damage based on the comparison result. Specifically, when the ratio of the effective value or crest factor calculated in all frequency bands to the normal value is large, it is judged that the damage is large, and the effective value or crest factor calculated only in a specific frequency band is large. It is judged that the damage is small.

  Here, the vibration effective value (normal value) of a normal product in each frequency band corresponding to the natural frequency of the bearing housing 12 using the double row tapered roller bearing 11 and measurement conditions used in the first and second embodiments. Table 2 shows the ratio of the effective vibration value of the outer ring small product and the outer ring small product.

  As a result, in the effective vibration value in the frequency band including each natural frequency of the bearing box, the ratio of the effective value of the outer ring not larger than the normal product to the normal product is about 10 times or more in any frequency band. The outer ring small scratch product is about 5 times larger only in the 0 to 3 kHz band, and is almost the same as the normal product level in other frequency bands. For this reason, if the vibration effective value is large compared to the normal product in all frequency bands, the damage has progressed considerably. If only a specific frequency band has occurred, the damage may be in the initial stage. Recognize. Accordingly, it is possible to determine the degree of damage in addition to the presence / absence of abnormality of the rotating component.

In the present embodiment, vibration and temperature associated with the rotation state of the rotating component are detected simultaneously. Thus, when a bearing seizure abnormality occurs, it can be detected by temperature, and when an abnormality such as peeling of the bearing raceway surface, gear loss, or wheel flat wear occurs, it can be detected by vibration. Therefore, it is possible to simultaneously inspect the degree of defects and damage of a plurality of parts in an actual operation state without disassembling the apparatus in which the parts are incorporated.
Note that the signal processing system of this embodiment shown in FIG. 13 may be compared and collated by the signal processing of (1) to (6) of the first embodiment.

  The abnormality diagnosis apparatus according to the present invention is not limited to the above-described embodiment, and appropriate modifications and improvements can be made. In the present embodiment, the abnormality diagnosis device is applied to the rolling bearing device for a railway vehicle. However, the abnormality diagnosis device may be applied to a rolling bearing device for a reduction gear. In this embodiment, the natural frequency is stored in advance in the natural frequency storage unit. However, the measurement result obtained by the striking method may be directly taken into the filter unit.

  Further, in the above-described embodiment, the vibration signal detected by the vibration sensor and filtered by the filter processing unit is input to the headphones or the speaker, thereby performing an auditory determination in which the presence or absence of abnormality is confirmed by sound. Also good.

It is sectional drawing of the rolling bearing apparatus of 1st Embodiment of this invention. It is a signal processing system diagram of the abnormality diagnosis apparatus according to the first embodiment of the present invention. It is a figure which shows the measurement result of the natural frequency of a bearing box. It is a figure which shows the analysis result of the vibration frequency of a normal rolling bearing. It is a flowchart which shows the processing flow of the abnormality diagnosis in 1st Embodiment of this invention. It is a figure which shows the envelope analysis result without a filter process, (a) is an analysis result of a normal product, (b) is an analysis result of an outer ring | wound wound product. It is a figure which shows the envelope analysis result after 1-3 kHz filter processing, (a) is an analysis result of a normal product, (b) is an analysis result of an outer ring | wound wound product. It is a signal processing system diagram of the abnormality diagnosis device according to the second embodiment of the present invention. It is a figure which shows the analysis result of the vibration frequency of a bearing with a large damage of an outer ring. It is a figure which shows the analysis result of the vibration frequency of a bearing with little damage of an outer ring. It is a figure which shows the envelope analysis result of a 0-3 kHz frequency band, (a) is an analysis result of a bearing with a large damage of an outer ring | wheel, (b) is an analysis result of a bearing with a small damage of an outer ring | wheel. It is a figure which shows the envelope analysis result of a frequency band of 4.2-6kHz, (a) is an analysis result of a bearing with a large damage of an outer ring, (b) is an analysis result of a bearing with a small damage of an outer ring. It is a signal processing system diagram of the abnormality diagnosis device according to the third embodiment of the present invention. It is sectional drawing of the conventional bearing apparatus. It is sectional drawing of the other conventional bearing apparatus. It is sectional drawing of the conventional further another bearing apparatus. It is a block diagram of the conventional abnormality detection apparatus. It is a block diagram of another conventional abnormality diagnosis device. It is a block diagram of another conventional abnormality diagnosis device. It is a relational expression showing the relation between the defect of each member of the bearing and the abnormal vibration frequency generated in each member.

Explanation of symbols

10 Rolling bearing device 11 Double row tapered roller bearing (rotating parts)
12 Bearing box (stationary member)
14 Inner ring 16 Outer ring 18 Tapered roller (rolling element)
25 axle (rotating shaft)
27 Flat part 31 Abnormality detection sensor (sensor)
32 Vibration sensor 33 Temperature sensor 35, 44 Filter section (filter processing section)
36 natural frequency storage unit 37 envelope processing unit 38 frequency analysis unit 39 comparison verification unit 41 theoretical frequency calculation unit 42, 45, 53 abnormality determination unit 43, 54 result output unit 50 calculation unit 51 comparator (comparison unit)
52 Threshold setting unit

Claims (1)

An abnormality diagnosis device for diagnosing an abnormality of a rotating component that rotates relative to a stationary member,
At least one of a vibration sensor, an ultrasonic sensor, and an AE sensor fixed to the rotating component or the bearing housing;
A filter processing unit that extracts a plurality of specific frequency bands corresponding to a plurality of natural frequencies of any one of the rotating component, the bearing box, and the sensor from a waveform of a signal detected by the sensor;
An arithmetic processing unit that calculates an effective value or a crest factor of the waveform after filtering transferred from the filter processing unit;
A comparison unit that compares a calculation result in each frequency band calculated by the calculation processing unit with a normal value;
Based on the comparison result in the comparison unit, an abnormality determination unit for diagnosing the presence or absence of abnormality or the degree of damage;
An abnormality diagnosis apparatus comprising:

Images