JP7346065B2 - Diagnostic method and device for low-speed rotating equipment - Google Patents

Description

The present invention relates to a diagnostic method and device for low-speed rotating equipment.

In equipment that operates continuously, such as petrochemical plants and power plants, it is often difficult to stop the equipment, so it is important to detect abnormalities as early as possible. If abnormalities in equipment can be detected early, it will be possible to lengthen the lead time until failure occurs based on the detection results.

One of the abnormalities in the above-mentioned equipment is flaws that occur on the rolling elements (balls, rollers), raceways and transfer surfaces of the inner and outer rings of bearings in rotating equipment. Generally, a vibration method is used to diagnose rolling bearings in rotating machines, and abnormalities are diagnosed based on vibration acceleration.

There are roughly two types of acceleration components of vibration acceleration used in such vibration methods.

(A) One is the acceleration component in the 1kHz to 10kHz range, mainly at several kHz, which captures the natural frequency of the bearing. However, the acceleration component in this region will not be excited unless the flaw in the bearing becomes large enough to generate a force that shakes the entire bearing. Therefore, it may not be suitable for early abnormality monitoring in continuously operating plants.

(B) The other is the acceleration component in the high frequency range of 10kHz to 30kHz. This is caused by contact between the bearing steels, and is a frequency component that is generated even by a small surface flaw because it is generated by the action of a small force. In other words, it can be said to be suitable for early abnormality monitoring of continuously operating plants.

However, since both of the above two types of vibration acceleration (A) and (B) capture the impact value caused by contact between the rolling elements and the flaws on the raceway surface, the signal level is determined by the peripheral speed of the rolling elements. depends on. In other words, when the circumferential speed of the rolling element decreases, the vibration acceleration level, which is the impact value, also decreases. Therefore, when the circumferential speed is low, such as when the rotational speed of the rolling element is several hundred rpm or less, the level decreases, and even vibration acceleration in a high frequency range is buried in the noise level and becomes difficult to detect.

On the other hand, the AE (acoustic emission) method in the 100k to 1MHz range is used for bearing diagnosis in the low rotational speed range. For example, when rolling fatigue flaking progresses, AE particularly occurs in the 100kHz to 500kHz range, and this can be used for diagnosis (see, for example, Patent Document 1). (AE is originally used to monitor crack growth in stationary machines such as spherical tanks, and is not affected by circumferential speed. In other words, AE is a signal with different properties from vibration acceleration. .)

JP2012-78288A

IEEJ Transactions D Vol.132No.4pp.501-509

However, the AE method used in low-speed bearing diagnosis requires a couplant to pass high-frequency sound between the equipment surface and the sensor surface, and the level changes due to the installation conditions. This may occur. Additionally, since the AE sensor must be installed in contact with the equipment surface, measurement may be difficult in high-temperature equipment or moving equipment.

Therefore, the present invention provides a low-speed rotating device that can suppress the influence of the installation state of the couplant between the equipment surface and the sensor surface, and that can perform non-contact measurements even in high-temperature equipment or moving equipment. The purpose is to provide a diagnostic method and a diagnostic device.

One aspect of the present invention is
detecting an air-propagated acoustic signal generated when the low-speed rotating equipment is in operation with a non-contact sensor that is not in contact with the low-speed rotating equipment;
A method for diagnosing a low-speed rotating device, including the step of determining whether or not an abnormality has occurred in the low-speed rotating device based on a signal detected by the non-contact sensor.

According to this diagnostic method, an airborne acoustic signal is detected by a non-contact sensor that is not in contact with the low-speed rotating equipment, so it is possible to detect and monitor abnormalities in a non-contact state.

In the above diagnostic method, the detecting step includes a step of non-contactly detecting an acoustic signal generated during rotation of a rotating body in a low-speed rotating device using a non-contact sensor;
The step of determining
extracting a signal in an inaudible range from the detected acoustic signal;
The method may also include a step of determining whether or not an abnormality has occurred in the low-speed rotating device based on fluctuations in the extracted acoustic signal.

In the above diagnostic method, the determining step includes performing envelope processing on the acoustic signal, amplifying the envelope processed signal, and calculating the frequency of occurrence by Fourier transforming the amplified signal. May contain.

In the above diagnostic method, the bandwidth of the acoustic signal may be limited to a frequency band of 20kHz to 100kHz.

In the above diagnostic method, the bandwidth of the acoustic signal may be limited to different values for one rotating body and another rotating body among the plurality of rotating bodies included in the diagnosis target.

In the above diagnostic method, between the step of acquiring an acoustic signal in a non-contact manner using a non-contact sensor and the step of limiting the bandwidth of the acoustic signal, an acoustic signal with a limited bandwidth is generated according to the rotation speed of the rotating body. It may include the step of extracting a signal of a predetermined time width from.

In the above diagnostic method, the step of determining may be performed on an acoustic signal acquired in a predetermined time width set according to the rotation speed of the rotating body.

In the above diagnostic method, when the first threshold value set in advance is compared with at least one of the maximum value of the detected acoustic signal and the maximum value of the spectrum of the acoustic signal, if the maximum value is larger; It may be determined that an abnormality has occurred.

In the above diagnostic method,
Comparing at least one of the maximum value of the detected acoustic signal and the maximum value of the frequency spectrum with a preset first threshold,
measuring the number of times at least one of the acoustic signal and the frequency spectrum exceeds a second preset threshold that is smaller than the first threshold;
The number of times at least one of the maximum value of the acoustic signal and the maximum value of the frequency spectrum exceeds a first threshold, and at least one of the acoustic signal and the frequency spectrum exceeds a preset second threshold When exceeds a predetermined number of times, first information regarding the abnormality may be generated.

In the above diagnostic method, at least one of the maximum value of the acoustic signal and the maximum value of the frequency spectrum exceeds the first threshold, and at least one of the acoustic signal and the frequency spectrum exceeds the preset second threshold. When the number of times the threshold value has been exceeded is equal to or less than a predetermined number of times, second information regarding the abnormality may be generated.

In the above diagnostic method, the acoustic signal may be detected using a pipe that propagates the acoustic signal and a non-contact sensor installed in the pipe.

In the above diagnostic method, an attenuation rate corresponding to the distance from the sliding part of the rotating body to the sensor is used, and abnormalities in low-speed rotating equipment are detected according to the level of the acoustic signal that takes this attenuation rate into account. It may also be used for diagnosis.

One aspect of the invention is a computing device having at least one memory that stores a program for executing the method as described above and at least one processor for executing the program.

1 is a schematic diagram showing an embodiment of the present invention, and shows an example of a configuration in which an abnormality in a low-speed rotating device including a shaft and a bearing can be detected using a non-contact sensor. FIG. 2 is a block diagram illustrating an example of a system that determines the presence or absence of an abnormality based on ultrasound signals. 12 is a flowchart illustrating an example of processing when determining the presence or absence of an abnormality based on an ultrasonic signal detected using a microphone placed in a pipe. It is a figure which shows an example of an abnormality cause identification flow. 2 is a table showing the relationship between the frequency generated when a rotating body rotates in a low-speed rotating device and damage to a bearing or the like. FIG. 3 is a diagram illustrating an example of a flow including a method for determining a time width for collecting data. It is a graph which shows the trend of the peak maximum value for every R rotation of a low-speed rolling roll. It is a graph which shows the trend of the event count for every R rotation of a low-speed rolling roll. It is a graph which shows the envelope spectrum of the ultrasonic sound of an abnormal bearing. This is a matrix for determination based on event counts and peak levels. It is a diagram showing an example of a sound collector with lines drawn at predetermined positions. FIG. 3 is a diagram illustrating laser irradiation from a pair of light sources disposed diagonally rearward on both sides of a sound collector in order to determine a measurement target distance. 12 is a view of the point where the laser beams transmitted through the vertical slit and the horizontal slit cross, viewed along the direction of the arrow in FIG. 11. FIG. FIG. 2 is a diagram schematically showing an example of a low-speed stirrer. This is a graph comparing the peak values of measurement points 1 to 4. 1 is a diagram illustrating an outline of a gear reducer in Example 1 of the present invention. FIG. 3 is a graph showing an envelope spectrum of ultrasonic sound obtained with normal device A. 7 is a graph showing an envelope spectrum of ultrasonic sound obtained in abnormal device B. FIG. It is a graph showing the envelope spectrum of AE obtained in normal device A. 7 is a graph showing an envelope spectrum of AE obtained in abnormal device B. FIG. FIG. 7 is sample data of (A) a spectrum other than when the bearing peeling progresses and (B) a spectrum when the bearing peeling progresses without filtering in Example 2 of the present invention. FIG. In a bearing accelerated life testing machine (65 rpm) for low-speed rotating equipment, (A) the peak value of the waveform after envelope processing of the signal detected by the conventional AE method, and (B) detection by the method described in this embodiment 3 is a graph showing a comparison between the peak value of the waveform of the acoustic signal and These are images taken (at a magnification of 50x) of a flawed portion of a bearing before (A) measurement and (B) after measurement.

Hereinafter, the configuration of the present invention will be described in detail based on an example of an embodiment shown in the drawings.

In one aspect of the present invention, it is determined whether an abnormality has occurred in a low-speed rotating device having a rotating body, and the low-speed rotating device is diagnosed. Low-speed rotating equipment refers to equipment whose rotating body typically has a rotational speed of 300 rpm or less or a dn value of 20000 [mm·rpm] or less. The diagnostic method and diagnostic device according to the present disclosure are suitable for rotating equipment with a rotation speed of 300 rpm or less or a DN value of 20000 [mm/rpm] or less, and are suitable for rotating equipment with a rotation speed of 100 rpm or less, such as a low-speed rotating extruder. It is even more suitable.

The rotating body of the low-speed rotating equipment that is subject to abnormality detection typically refers to the rotating shaft or the rolling bearing that supports the rotating shaft, and may also include gears, screws, etc. in low-speed rotating extruders. There is. In addition, specific examples of abnormalities include the development of flaking in rolling bearings, metal contact due to poor lubrication, contact between screw and barrel, occurrence of gear tooth surface damage, and metal peeling and metal contact due to damage. can. Gear metal peeling and metal contact include pitting, spalling, scoring, etc.

In this embodiment, a non-contact sensor that is not in contact with the low-speed rotating equipment detects an airborne acoustic signal generated when the low-speed rotating equipment is in operation, and based on the detected signal, an abnormality is detected in the low-speed rotating equipment. Determine whether or not this has occurred. This will be explained below with specific examples shown in the drawings (see FIG. 1, etc.). Note that the acoustic signal acquired in the present disclosure is an acoustic signal in a frequency band greater than 0 and 100 kHz or less, and may be referred to as an ultrasonic signal. Further, the acquired acoustic signal is typically an acoustic signal of 20 kHz or more and 100 kHz or less, which is an acoustic signal in a frequency band inaudible to the human body. Note that the acoustic signal used for abnormality detection may include a signal in a frequency band with a value larger than 100 kHz. Here, as the frequency band of the acoustic signal becomes larger, the propagation distance becomes shorter, so it becomes difficult to acquire the acoustic signal of the frequency band without contact. Therefore, it is preferable that the acoustic signal to be acquired has a maximum frequency of about 100 kHz.

FIG. 1 shows an example of a configuration in which an abnormality in a low-speed rotating device 10 including a shaft 11 and a bearing 12 can be detected using a non-contact sensor. The low-speed rotating device 10 is further provided with piping 14 (see FIG. 1).

The shaft 11 of this embodiment, which is an example of a rotating body in the low-speed rotating device 10, is supported by a bearing 12. The bearing 12 includes, for example, two ball bearings that support one end and the other end of the shaft 11, respectively. The bearing 12 of this embodiment is provided inside a casing (bearing box) 13.

The pipe 14 is provided with its end disposed in the casing 13. The piping 14 provided in this way collects abnormal noise generated in the bearing 12 within the casing 13, and makes it possible to measure the sound at a location away from the bearing 12. In addition, by using the piping 14 installed near the bearing 12, it is possible to prevent sound from attenuating due to diffusion and to accurately measure sound even at a location far from the shaft (rotating body) 11 and the bearing 12 (with the target equipment). This makes it possible to extend the measurement distance between the sensor and the sensor. The pipe 14 is preferably a pipe made of a material with high acoustic impedance (for example, a SUS pipe).

Further, in this embodiment, a manifold-shaped pipe 14 is used, which branches midway and has an end disposed in each of the two casings 13 . According to such piping 14, it becomes possible to simultaneously measure and monitor abnormal sounds generated in a plurality of bearings 13 using a single sensor (see FIG. 1).

The non-contact sensor is a sensor that detects an air-propagated acoustic signal (ultrasonic signal) that is generated when the low-speed rotating device 10 is in operation (that is, when the shaft 11 is rotating). In this embodiment, the microphone 21 is used as an example of a non-contact sensor. The microphone 21 is placed in the pipe 14 and is not in contact with the low-speed rotating device 10. For example, in this embodiment, the microphone 21 is arranged at the open end of the manifold-shaped pipe 14 on the opposite side to the bearing 12.

FIG. 2 shows an example of a device (monitoring device 20) that determines the presence or absence of an abnormality based on ultrasonic signals.

In the monitoring device 20, an ultrasonic signal detected by a microphone 21 is amplified by a preamplifier 22, and then transmitted to an HPF (high pass filter) 23, where data in a predetermined frequency range is extracted, and then transmitted to a BPF (band pass filter). ) 24 (see Figure 2). The signal that has passed through the BPF 24 is transmitted to an A/D converter 27 and an envelope circuit 25, respectively. The waveform data subjected to envelope processing in the envelope circuit 25 is transmitted to the A/D converter 27 (see FIG. 2). A series of processes including the above may be realized by predetermined analog circuits, particularly for the preamplifier 22, HPF (high pass filter) 23, BPF (band pass filter) 24, A/D converter 27, and envelope circuit 25. Further, a series of processes including the above may be executed by an arithmetic processing device having a memory that stores programs for executing each processing method, and a processor for executing the programs. When a series of processes including the above are executed by an arithmetic processing unit, the ultrasonic signal detected by the microphone 21 is sent to the A/D converter, and the digitized ultrasonic signal is subjected to amplification processing, high-pass and band-pass processing. Each filter process and envelope process may be executed.

FIG. 3 shows an example of a process for determining the presence or absence of an abnormality based on an ultrasonic signal detected using the microphone 21 disposed in the pipe 14.

First, an ultrasonic signal (0 to 100 kHz) is detected by the microphone 21 (step SP1), the data in the measurement unit time T (s) is A/D converted (step SP2), and filtered to remove the inaudible range. After extracting the waveform signal (around 20kHz to 100kHz) (step SP3), perform envelope processing (step SP4), perform Fourier transform on the amplified signal (for example, the entire envelope waveform), and calculate the measurement unit time. After recording the peak value (maximum value) Ap (V) of the envelope waveform for each T (step SP5), the number A of times the envelope waveform value Av (V) measured in the measurement unit time T exceeds the first threshold value C is recorded. _E is recorded (step SP6), and the process moves to the "abnormality determination flow".

In addition, when calculating the occurrence frequency of the data (occurrence frequency data) after the Fourier transform of the envelope waveform, (i) the device automatically compares the relationship between the occurrence frequency and the damage and determines the location of the abnormality. (ii) The occurrence frequency is presented to the user, and the user compares the relationship between the occurrence frequency and the damage and determines whether there is an abnormality. .

In addition, although we explained here that the acoustic signal is filtered to be within the frequency band of the inaudible range (around 20 kHz to 100 kHz), there is a bandwidth of the ultrasonic acoustic signal that is the detection target. Needless to say, the width is not limited to a specific range. Furthermore, the acoustic signal may be limited within the frequency band of 20 kHz to 100 kHz as described above, but the meaning of "limited" in that case is that (i) the band (20- 100kHz), or (ii) after detecting the signal of 20-100kHz+α, extracting it with a bandpass filter (BPF24, etc.).

[Abnormality judgment flow]
In the "abnormality determination flow" of this embodiment, it is determined whether or not an abnormality has occurred in the low-speed rotating device 10 based on fluctuations in the extracted acoustic signal. Here, first, it is determined whether the peak value Ap exceeds a second threshold (peak threshold) P that is larger than the first threshold C (peak value Ap>threshold P) (step SP7). If it exceeds (Yes at step SP7), it is determined whether the number of measurements A _E exceeds the threshold E for the number of determinations (step SP8), and if it does not (No at step SP8), the peak value is noise. (Step SP9). On the other hand, if the number of measurements A _E exceeds the threshold value E for the number of determinations (Yes at step SP), the peak value is determined to be due to an abnormality (step SP10), and the process moves to the "abnormality cause identification flow". do.

Further, in step SP7 in the "abnormality determination flow", if the peak value Ap does not exceed a second threshold (peak threshold) P that is larger than the first threshold C, that is, if peak value Ap>threshold P does not hold (step No at SP7), it is determined whether the number of measurements A _E exceeds the threshold value E for the number of determinations (step SP11), and if it does not (No at step SP11), it is determined that the low-speed rotating device 10 is normal. (Step SP13). On the other hand, if the number of measurement _A ”.

FIG. 4 shows an abnormality cause identification flow, and FIG. 5 shows a relationship table between occurrence frequency and damage.

[Abnormality cause identification flow]
After the above-mentioned "abnormality determination flow", the process moves to the "abnormality cause identification flow" in a predetermined case. That is, the abnormality determination flow as described above (acoustic measurement is performed (step SP1), data of the measurement unit time T (s) is A/D converted (step SP2), filter processing is performed to extract the inaudible waveform ( After step SP3), envelope processing (step SP4), and Fourier transformation of the amplified signal (for example, full envelope waveform) (step SP5), an abnormality has occurred in the low-speed rotating equipment 10. (Step SP10), or if it is determined that the low-speed rotating equipment 10 is in a state that requires attention (Step SP12), the "abnormality cause identification flow" is started (Step SP20), and the Fourier-transformed spectrum is It is determined whether a peak indicating the appearance of the peak appears (step SP21). If the appearance of the peak shown in FIG. 4 is recognized (Yes in step SP), the type of damage in the low-speed rotating equipment 10 is detected and the cause of the abnormality is identified (step SP22), and the abnormality cause identification flow is ended. (Step SP23).

[How to determine the time range for data collection]
FIG. 6 shows a flow including a method for determining the time width for collecting data. Note that steps in the processing flow related to determining the time width are indicated by numbers 31 and later (see FIG. 6).

After the start of measurement, the rotational speed N (rpm) of the shaft 11 is measured (step SP31), and a measurement unit time T(s) is automatically calculated from equation (1) (step SP32). However, R in Equation (1) is a "measurement rotation number" (an integral multiple of the rotation number) that is set in advance, and is a value that defines how many rotations are taken as one measurement unit.

After that, the peak value (maximum value) Ap (V) of the envelope waveform for each measurement unit time T and the number of times in which the time axis waveform exceeds a predetermined threshold (in this specification, the number exceeds such a predetermined threshold) are calculated. (The number of data obtained is called the "number of events") A _E (In this embodiment, the number of times the value Av (V) of the envelope time axis waveform measured in the measurement unit time T exceeds the first threshold C) ) is measured (step SP33). Then, the process moves to the previously described "abnormality determination flow" (see FIGS. 3 and 6).

As explained so far, when executing the process of the abnormality determination flow, the abnormality determination is performed on the acoustic signal acquired in a predetermined time width set according to the rotation speed of the shaft 11. It's fine. In other words, the step of extracting an acoustic signal of a predetermined time width may be performed at any time.

[Logic of abnormality determination]
FIG. 7 is a graph showing the trend of the peak maximum value for each R rotation (in other words, the transition for each measurement period of the peak value (maximum value) Ap (V) of the envelope waveform for each measurement unit time T), FIG. 8A is a graph showing a trend of event counts for each R rotation (in other words, a change in the number of envelope waveforms exceeding the first threshold value C for each measurement period for each measurement unit time T). Moreover, FIG. 9 is a matrix for determination based on the event count and the peak level (in other words, the peak value).

The logic of abnormality determination in the diagnostic method of this embodiment will be explained as follows. That is, the graph of the transition for each measurement cycle depicts the peak value and the number of events (see FIGS. 7 and 8A), and abnormality determination at each measurement point is performed based on a matrix (see FIG. 9). In a bearing where an abnormality has occurred, both the peak value and the number of events exceed the threshold (the peak value exceeds the threshold P and the number of events exceeds the threshold E), whereas when the peak occurs due to noise , the peak value exceeds the threshold P, but the number of events does not exceed the threshold E. Based on this, it is possible to discriminate whether the peak is due to an abnormality or a peak due to noise. Further, FIG. 8B shows an envelope spectrum used in the abnormality cause identification flow. In FIG. 8B, occurrence of an outer ring rolling element passing frequency and an inner ring rolling element passing frequency can be seen in the envelope spectrum, indicating that the cause of the abnormality has been successfully identified.

As explained above, in the diagnostic system of this embodiment or the diagnostic method using the system, the acoustic signal that propagates in the air in the pipe 14 is provided at a position where it does not come into contact with the low-speed rotating device 10 that is the monitoring target. It is detected by the microphone 21 in a non-contact manner. According to such a diagnostic method, abnormality detection and diagnosis can be performed even on low-speed rotating equipment 10 in high-temperature equipment, moving equipment, and the like.

In addition, in this embodiment, the microphone 21 measures sound in the 20 kHz or higher (20 kHz to 1 MHz region, or 20 kHz to 100 kHz region) without contact, and calculates R times the rotation speed of the shaft 11 of the high-speed rotating device 10. Measurement unit time (formula (1)).In other words, calculate the time required for one rotation from the rotation speed N (rpm) of the rotating shaft, and calculate the time R times that time, that is, R rotations of the rotation speed. The time required is the unit measurement time T. By monitoring the maximum level of the peak occurring during that time and the number of events occurring, and distinguishing noise using a matrix, normality and abnormality can be evaluated. The diagnosis of low-speed rotating equipment 10 has become possible due to the discovery that sound in the frequency region has similar characteristics to AE.Specifically, as shown in FIGS. In the waveform B, it can be seen that the dominant frequency occurs at the same location with a high S/N ratio, and as shown in Figure 18, the peak value rising at the same time indicates a high S/N ratio. This shows that similar signals can be obtained for AE and acoustics for abnormalities in low-speed rotating equipment. On the other hand, when using the vibration method, it is not possible to diagnose with such a high S/N ratio.

In addition, with the conventional non-contact acoustic diagnostic method, when diagnosing low-speed rotating equipment, the dn value (expressed as the product of the inner diameter of the bearing and the rotational speed) is low and the S/N ratio is poor, so the envelope While it was not possible to identify the cause of the abnormality using FFT, this embodiment achieves a high S/N ratio similar to AE, and it is possible to identify the cause of the abnormality from the peak frequency. .

By the way, in low-speed rotating equipment of equipment that operates continuously, such as petrochemical plants and power plants, it is quite possible that an abnormality will occur only once every ten years. Even in such a case, according to the diagnostic method according to this embodiment, which achieves a high S/N ratio, the diagnostic method according to the present embodiment achieves a high S/N ratio. ) can be used to monitor the status.

Note that although the above-described embodiment is an example of a preferred implementation of the present invention, the present invention is not limited thereto, and various modifications can be made without departing from the gist of the present invention.

For example, in the above-described embodiment, an example was explained in which the rotating body (for example, the shaft 11) is single; In some cases, In that case, the bandwidth of the ultrasonic acoustic signal may be limited to different values for one of the rotating bodies and the other rotating body. In addition, for example, taking advantage of the fact that metal peeling sounds differ depending on the material and that poor lubrication sounds occur over a range of several kHz to 100 kHz, the bandwidth can be limited to 25 kHz to 35 kHz for each rotating body. You may.

Further, in the above diagnostic method, a step (step SP1) of acquiring an acoustic signal in a non-contact manner by a non-contact sensor (microphone 21 in the above embodiment), and a step (step SP1) of limiting the bandwidth of the ultrasonic acoustic signal ( Between step SP3), a step of extracting a signal with a predetermined time width from an acoustic signal with a limited bandwidth depending on the rotation speed of the rotating body (for example, the shaft 11) may be performed. This extraction step is executed, and the time width for acquiring acoustic signals is automatically adjusted so that the number of events (number of data exceeding the threshold) described above is the same (that is, standardized and does not vary). It's okay.

In addition, in the above-described diagnosis method, an attenuation rate is used that corresponds to the distance from the sliding part of the rotating body of the low-speed rotating device 10 to the non-contact sensor, and the attenuation rate is determined according to the level of the acoustic signal in consideration of the attenuation rate. An abnormality in the low-speed rotating device 10 may be diagnosed. For example, when installing the microphone 21 toward the shaft 11 and its sliding parts (bearings, sleeves, mechanical seals, etc.), the distance attenuation of the sound waves at the frequency at which the scratch occurs is calculated in advance, and calculated backwards from the measured distance. It is preferable to identify a location where the sound level is high. Although it is not possible to install a fixed sensor on the sliding part (itself), if a sensor such as the microphone 21 is used, it is possible to measure without contact, and the sound generated from the sliding part can be directly measured. be able to. At this time, if the location with the highest sound level is known, it will be possible to detect and measure with higher accuracy.

Further, in the case where a position with a high sound level is calculated and specified in advance as described above, the position may be indicated in advance to facilitate positioning of a non-contact sensor such as the microphone 21. Since the ultrasonic sound level depends on the measurement distance, it is effective if it is possible to easily confirm that the measurement distance is appropriate. For example, a line L may be drawn in advance at a predetermined position of the sound collector 28 for the microphone 21 and used as a mark for installing the microphone 21 at a predetermined position with a high sound level (see FIG. 10). . Alternatively, a laser beam is emitted from a pair of light sources 29 placed diagonally behind both sides of the sound collector 28 toward the vertical slit and horizontal slit provided in the sound collector 28, and if the crossed point 29x is at a predetermined position, the measurement is performed. The distance may be set to the target distance (see FIGS. 11 and 12).

It goes without saying that the low-speed rotating equipment 10 that is the target of diagnosis based on the method described above is not limited to a specific equipment. For example, in the above description, reference was made to low-speed rotating equipment in petrochemical plants and power plants, but these are merely examples, and it is naturally possible to target other low-speed rotating equipment 10 such as low-speed stirrers.

FIG. 13 schematically shows an example of a low speed stirrer (designated 100). 101 is a sliding part, 102 is a bearing part, 103 is a sliding part, and 104 is a bearing/mechanical part. In addition, in FIG. 14, when the sliding part 101 in a normal device is taken as a measurement point (measurement point 1), when the bearing part 102 in a normal equipment is taken as a measurement point (measurement point 2), the sliding part 103 in an abnormal equipment A graph comparing the respective peak values when the measurement point (measurement point 3) is taken as the measurement point (measurement point 3) and the measurement point (measurement point 4) when the bearing/mechanical part 104 in the abnormal equipment is taken as the measurement point is shown as a reference diagram. As shown in FIG. 14, the peak value is higher in the abnormal sliding portion 103 (measurement point 3) and the mechanical seal portion 104 that extends between the bearings than in the sliding portion 101 (measurement point 1) and the normal bearing portion 102 (measurement point 2). It can be seen that (measurement point 4) is higher. In particular, it can be seen that the highest point is measurement point 3, and the sliding part (rotating part) on the upper part of the bearing is the highest position. An open inspection revealed that the cause of the abnormality was wear of the bearing sleeve.

[Specific example of abnormality detection in envelope spectrum]
A test was conducted to detect gear abnormalities by applying the above-described diagnostic method to a gear reducer (indicated by reference numeral 200 in FIG. 15), which is a type of low-speed rotating equipment 10. Example 1 will be described below (see FIGS. 15 and 16).

In the gear reducer 200 shown in FIG. 15, reference numeral 201 represents a coupling, 202 and 204 represent bearings, and 203 represents a gear. The rotational speed of the output shaft of the device to be measured in this example is low, 100 rpm or less, and it is known that pitching occurs in the gear on the output shaft side. Under these conditions, an ultrasonic signal caused by an abnormality in the gear on the output shaft side was detected using a point indicated by symbol MP in FIG. 15 as a measurement point. The gear mesh frequency (=shaft rotation frequency x number of teeth) was 77.5Hz.

The levels of envelope spectra obtained in this test are shown in FIGS. 16A, 16B, 16C and 16D. In contrast to the normal equipment, where no specific frequency increase occurred, in the abnormal equipment, a clear peak level increase was observed at the gear meshing frequency for both AE and ultrasonic sound.

[Limitation of frequency band of bearing flaking progress sound]
The sound of flaking progress in the bearing was measured, and sample data was obtained by limiting the frequency band of the ultrasonic signal. This will be described below as Example 2 (see FIG. 17).

FIG. 17 shows sample data of (A) a spectrum other than when the bearing peeling progresses and (B) a spectrum when the bearing peeling progresses without filtering. In addition, in FIG. 17, different bearings are used as sample data.

Comparative example 1

[Conventional monitoring and evaluation methods]
Figure 18 shows the peak value of the waveform after envelope processing of the signal detected by the conventional AE method in a bearing accelerated life tester (65 rpm) for low-speed rotating equipment, and the acoustic signal detected by the method described in this embodiment. The peak value of the waveform is shown. Further, FIG. 19 shows images taken before and after measurement of the flaw portion that occurred in the bearing.

Comparing the peak values in the graph of FIG. 18 between (A) the conventional AE method and (B) the present embodiment, it was found that the peaks occurred at almost the same time (Fig. 18, especially the part surrounded by a rectangular frame). From FIG. 18, it is recognized that the peak values of the AE waveform and the ultrasonic acoustic waveform rise at almost the same time. FIG. 19 is a photograph of the bearing before and after the test in FIG. 18, in which flaking originating from minute cracks is observed. From this, it was found that the ultrasonic acoustic waveform, like the AE waveform, captures a rise in peak value at the moment when bearing flaking is progressing. Further, from FIG. 18, the envelope peak values of AE and ultrasound show an upward trend almost simultaneously. The results of the open inspection after stopping the test shown in FIG. 19 showed that flaking occurred. From this, it can be said that, like AE, ultrasonic waves are able to detect the occurrence and progress of flaking at a rotation speed of 65 rpm.

DESCRIPTION OF SYMBOLS 10...Low speed rotating equipment, 11...Shaft (rotating body), 12...Bearing, 13...Casing, 14...Piping, 20...Monitoring device, 21...Microphone (non-contact sensor), 23...HPF (high pass filter), 24... BPF (band pass filter), 25...envelope circuit, 27...A/D converter, 28...sound collector, 29...light source, 29x...point where laser beams cross, 100...low speed stirrer (an example of low speed rotating equipment) , 101...Sliding part, 102...Bearing part, 103...Sliding part, 104...Bearing/mechanical part, 200...Gear reducer, 201...Coupling, 202...Bearing, 203...Gear, 204...Bearing, MP... Measuring point

Claims (12)

a step of detecting an air-propagated acoustic signal generated during operation of a low-speed rotating device with a rotation speed of 300 rpm or less or a DN value of 20000 [mm/rpm] or less with a non-contact sensor that is not in contact with the low-speed rotating device;
a step of determining whether or not an abnormality has occurred in the low-speed rotating equipment based on the signal detected by the non-contact sensor;
In the step of determining,
Comparing at least one of the maximum value of the detected acoustic signal and the maximum value of the frequency spectrum of the acoustic signal with a preset first threshold,
measuring the number of times at least one of the acoustic signal and the frequency spectrum exceeds a second preset threshold that is smaller than the first threshold;
At least one of the maximum value of the acoustic signal and the maximum value of the frequency spectrum exceeds the first threshold, and at least one of the acoustic signal and the frequency spectrum exceeds a preset second threshold. determining that an abnormality has occurred in the low-speed rotating device when the number of times the threshold value has been exceeded exceeds a predetermined number;
Diagnosis method for low-speed rotating equipment. The detecting step includes a step of non-contactly detecting an acoustic signal generated during rotation of a rotating body in the low-speed rotating device by the non-contact sensor,
The step of determining
extracting a signal in an inaudible range from the detected acoustic signal;
determining whether an abnormality has occurred in the low-speed rotating equipment based on fluctuations in the extracted acoustic signal;
The method for diagnosing low-speed rotating equipment according to claim 1. The step of determining
enveloping the acoustic signal;
amplifying the envelope processed signal;
A step of Fourier transforming the amplified signal to calculate the generated frequency;
The method for diagnosing low-speed rotating equipment according to claim 2. The method for diagnosing low-speed rotating equipment according to any one of claims 1 to 3, wherein the bandwidth of the acoustic signal is limited to a frequency band of 20 kHz to 100 kHz. According to any one of claims 1 to 4, the bandwidth of the acoustic signal is limited to different values between one rotating body and another rotating body among a plurality of rotating bodies included in the low-speed rotating device. Diagnosis method for low-speed rotating equipment described. Between the step of contactlessly acquiring the acoustic signal by the non-contact sensor and the step of limiting the bandwidth of the acoustic signal, the bandwidth is limited according to the rotation speed of a rotating body in the low-speed rotating device. the step of extracting a signal with a predetermined time width from the acoustic signal obtained by
The method for diagnosing low-speed rotating equipment according to claim 4 or 5. The step of determining is performed on the acoustic signal acquired in a predetermined time width that is set according to the rotation speed of a rotating body in the low-speed rotating device .
The method for diagnosing a low-speed rotating device according to any one of claims 1 to 5. A preset first threshold value is compared with at least one of the maximum value of the detected acoustic signal and the maximum value of the frequency spectrum of the acoustic signal, and if the maximum value is larger, an abnormality occurs. The method for diagnosing a low-speed rotating device according to any one of claims 1 to 7, wherein the method determines that At least one of the maximum value of the acoustic signal and the maximum value of the frequency spectrum exceeds the first threshold, and at least one of the acoustic signal and the frequency spectrum exceeds a preset second threshold. When the number of times the threshold is exceeded is less than or equal to a predetermined number, second information regarding the abnormality is generated. The method for diagnosing low-speed rotating equipment according to claim 1. Any one of claims 1 to 9, wherein the acoustic signal is detected using a pipe that propagates an acoustic signal generated in a bearing of a rotating body in the low-speed rotating device , and the non-contact sensor installed in the pipe. Diagnosis method for low-speed rotating equipment described in section. diagnosing an abnormality in the low-speed rotating device according to a level of the acoustic signal that takes the attenuation factor into consideration, using an attenuation rate that corresponds to a distance from a portion of the low-speed rotating device on which a rotating body slides to the sensor; The method for diagnosing a low-speed rotating device according to any one of claims 1 to 10. A diagnostic device for low-speed rotating equipment, comprising an arithmetic unit having at least one memory storing a program for executing the method according to claim 1 and at least one processor for executing the program. .

Images