JP7619974B2 - Abnormality prediction diagnostic system, abnormality prediction diagnostic device and diagnostic method - Google Patents

Description

The present invention relates to an abnormality prediction diagnostic system for sliding parts used in valves, etc., an abnormality prediction diagnostic device, and a diagnostic method thereof.

A valve that has the function of closing the flow of a fluid stops the flow of the fluid by closing the flow path between the valve body and the valve seat using a drive unit, thereby preventing the fluid from leaking out of the valve.

For example, in an electric valve, the flow of fluid is stopped by the cooperation of elements consisting of a worm gear that changes the direction of the motor's rotational driving force by 90 degrees, a screw part of the valve stem that converts the rotational driving force into linear driving force, a gland packing part that prevents leakage of fluid from the surface of the linearly moving shaft, and a seat part that is the sliding part between the valve body and the valve seat, but sliding occurs in each of these elements. These sliding parts deteriorate due to friction, increasing the risk of leakage from the seat part and loss of driving force.
Furthermore, in an air-operated valve, deterioration occurs due to sliding of the piston seal portion instead of the worm gear.

The sliding parts mentioned above deteriorate depending on the number of times they are operated, so valves with deterioration due to sliding can be identified through disassembly and inspection, and valve malfunctions can be prevented by replacing the parts and carrying out maintenance. However, disassembly and inspection requires a lot of effort and costs.

Even at nuclear power plants, valves are disassembled and inspected during regular inspections to prevent accidents, but disassembly and inspection of large-diameter valves requires a crane, which entails a huge amount of labor and cost. Furthermore, disassembly and inspection of valves installed in systems with high levels of radiation can result in high radiation exposure.

For this reason, devices have been developed that diagnose valve operability by measuring the current value of the motor in the drive unit (Patent Document 1), devices that use acoustic sensors to diagnose valve stem damage and fatigue caused by fluid forces (Patent Document 2), and acoustic sensors that use information about solid-borne sound generated in at least one component of the vacuum valve as a measurement signal (Patent Document 3), reducing the labor hours and frequency of regular inspections.

JP 2006-083928 A Special Publication No. 2014-521045 Special Publication No. 2020-525720

In the above prior art, a system for detecting valve abnormalities or monitoring the condition using an acoustic sensor uses acoustic emissions caused by vibrations, cracks, and deformations due to the action of fluid in a flow control valve to detect deviations from the normal state and fatigue, thereby diagnosing valve failure. Therefore, while it is possible to diagnose the occurrence of structural defects in the flow control function and the valve, it does not completely diagnose the valve (the function of closing the fluid).

The objective of the present invention is to enable highly accurate diagnosis of signs of deterioration of the valve's dynamic function and its function of closing the fluid, without disassembling the valve.

In order to solve the above problems, the abnormality prediction diagnostic system of the present invention is provided with: a plurality of acoustic sensors attached to a valve, which detect the valve seat sliding sound of the sliding part between the valve body and the valve seat and the sliding sound of components other than the valve seat sliding sound; a signal processing unit that processes the detection signals of the acoustic sensors; an analysis unit that analyzes the detection signals of the plurality of acoustic sensors and the propagation rate of the sliding sound processed by the signal processing unit to extract the valve seat sliding sound from the detection signals of the acoustic sensors, determines the number of operations or operating time of the sliding part from the spectral intensity of the extracted valve seat sliding sound, and estimates the leakage amount, surface roughness, or friction coefficient of the sliding part; and a judgment unit that judges signs of deterioration, including whether the valve's function of closing the fluid and its dynamic function are normal or not, from the number of operations or operating time, leakage amount, surface roughness, or friction coefficient of the sliding part obtained by the analysis unit, and predicts the timing of maintenance, including disassembly of the valve.

According to the present invention, it is possible to diagnose signs of deterioration of a valve's dynamic function and its function of closing a fluid with high accuracy without disassembling the valve, thereby improving the accuracy of operation and maintenance of valves and plants and equipment in which the valves are installed.
In addition, even in valves that are not intended to shut off the flow of fluid, such as flow control valves, it is possible to diagnose the surface roughness of the valve body and valve seat, which are the flow restricting parts, thereby improving the accuracy of operation and maintenance of valves and plants and equipment in which the valves are installed.

1 is a diagram showing a cross section of an electric gate valve and the installation position of an acoustic sensor. FIG. 2 is a diagram showing a cross section of a manual globe valve and the installation position of an acoustic sensor. 1 is a diagram illustrating a configuration of an abnormality sign diagnosis device according to an embodiment. FIG. 2 is a diagram illustrating an example of a structure of an acoustic sensor. FIG. 4 is a diagram illustrating an example of a detection signal of an acoustic sensor. 13A and 13B are diagrams illustrating an example of a detection signal of an acoustic sensor during a closing operation. 13 is a diagram showing an example of a detection signal of an acoustic sensor during pressing. FIG. FIG. 11 is a flow diagram showing a method for extracting and dividing a waveform of a sliding sound. FIG. 11 is a diagram showing the correlation between the spectrum intensity of the sliding sound and the number of operations of the sliding part. FIG. 13 is a diagram showing the correlation between surface roughness and the number of operations. FIG. 13 is a diagram showing the correlation between the friction coefficient and the number of operations. FIG. 13 is a diagram showing the correlation between leakage amount and surface roughness. FIG. 1 is a flow chart showing a diagnostic procedure. FIG. 1 is a diagram showing the structure of a hearing probe. 1A and 1B are diagrams showing a method of mounting an acoustic sensor equipped with a heat insulating material.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.
The abnormality sign diagnosis device of the embodiment detects acoustic emissions (elastic waves) generated by the opening and closing operation of the valve as sliding sounds for each of the sliding parts, such as the worm gear, valve stem thread, sliding part between the valve stem and gland packing, and contact surface between the valve body and valve seat, with an acoustic sensor, and obtains the surface roughness, friction coefficient, and leakage amount of the sliding parts based on the detected sliding sounds, and diagnoses the condition of the sliding parts, including abnormality signs, maintenance timing, and defects. Because the sound of fluid flow becomes noise in the detected sliding sounds, the abnormality sign diagnosis device of the embodiment performs diagnosis while the operation of the equipment in which the motorized gate valve is installed is stopped.

In describing the abnormality sign diagnostic device according to the embodiment, first, the installation of an acoustic sensor in an electrically operated gate valve to be diagnosed will be described with reference to FIG.
FIG. 1 is a diagram showing a cross section of an electric gate valve (hereinafter, referred to as valve 1) and the installation position of an acoustic sensor 2.

When the valve 1 is opened or closed, sliding occurs at the contact surface (sliding part) between the worm gear 24 and the thread of the valve stem threaded part 25 of the valve 1, the sliding part between the valve stem 4 and the gland packing 28, and the sliding part between the valve body 29 and the valve seat 30. For this reason, acoustic sensors 2 are installed directly or via a listening rod 31 (also called a waveguide rod or wave guide) on the outer surface of the valve 1 near these sliding parts, which are the propagation paths for the direct waves of the sliding sound.

By the way, when the worm gear 24 rotates, the valve stem threaded portion 25 moves up and down, and the valve stem 4 integrated with the valve stem threaded portion 25 and the valve body 29 connected to the valve stem 4 also move up and down. That is, when the rotational motion of the worm gear 24 is converted into the up and down motion of the valve stem 4 by the valve stem threaded portion 25, sliding occurs between the worm gear 24 and the valve stem threaded portion 25. Sliding also occurs between the valve stem 4 and the gland packing 28, and between the valve body 29 and the valve seat 30.

In FIG. 1, the acoustic sensor 2 is installed via a sound bar 31 on the outer surface of the valve 1 at a position corresponding to the sliding of the valve stem 4 and the gland packing 28, and the acoustic sensor 2 is also installed directly at a position (the bottom of the valve 1) corresponding to the sliding of the valve body 29 and the valve seat 30.

In addition to the position of the outer surface of the valve 1, an acoustic sensor 2 is installed on the motor 22 or manual handle 23, which is the drive unit 5 of the valve 1, to determine the drive state of the drive unit 5 when the valve 1 is opened or closed. It is also used to reduce noise during detection by the other acoustic sensor 2. In FIG. 1, an acoustic sensor 2 is installed on both the motor 22 and the manual handle 23.

When the acoustic sensor 2 is directly installed in the vicinity of the sliding portion, a couplant 21 is interposed between the acoustic sensor 2 and the detection surface as an acoustic transmission medium.
When the acoustic sensor 2 is installed via the sensing rod 31 , a couplant 21 is interposed between the acoustic sensor 2 and the sensing rod 31 .
The material of the couplant 21 may be grease, wax, adhesive, lubricating oil, water, gel, or the like.

The acoustic sensor 2, or the acoustic sensor 2 and the sound probe 31, can be fixed by pressing them against the surface by hand or by using a fixing jig. This fixing jig can be attached to the detection surface by magnets, adhesives, welding, soldering, or by an arm or tripod placed on the ground. Since sound propagation is reduced by passing through an air layer, it is preferable to attach the acoustic sensor 2 to the same part.

The acoustic sensor 2 can be an acoustic emission sensor, a piezoelectric sensor, an ultrasonic sensor, or an acceleration sensor.

In FIG. 1, the installation position of the acoustic sensor 2 for detecting the sliding sound of the valve 1 has been described, but the acoustic sensor 2 can also be installed in a manual globe valve in a similar manner to detect sliding sounds.
FIG. 2 is a diagram showing a cross section of a manual globe valve and the installation position of an acoustic sensor 2.

In a manual globe valve, when the manual handle 23 is rotated in the closing direction to close the valve from an open state, the valve stem 4 descends, causing the valve body 29 and the valve seat 30 to slide at their sliding portions.
In the embodiment, the abnormality sign diagnosis device is provided with an acoustic sensor 2 on the handle portion and the side surface of the valve box,
The operation sound of the manual handle 23 (handle operation sound) and the sliding sound of the sliding portion between the valve body 29 and the valve seat 30 (valve seat sliding sound) are detected.

The sliding sounds of components other than the valve seat sliding sound are either the operation sound of the handle, or the sliding sound generated at the sliding part between the gland packing and the valve stem, the sliding sound generated at the sliding part of the threaded part of the valve stem, the sliding sound generated at the sliding part of the worm gear of an electric valve, or the sliding sound generated at the sliding part of the piston part of an air-operated valve, or a combination of several of these sliding sounds. Here (in the case of a manual globe valve), the operation sound of the manual handle 23 is also one of the sliding sounds of components other than the valve seat sliding sound.

In detail, the acoustic sensor of the manual handle 23 detects not only the handle operation sound but also the acoustic wave propagating from the sliding sound generated at the valve seat. Also, the acoustic sensor 2 on the side surface of the valve box detects not only the valve seat sliding sound but also the acoustic wave propagating from the handle operation sound generated at the manual handle 23.
The same applies to the valve 1 described in FIG.

The abnormality prediction diagnostic device of the embodiment extracts a specific sound source (e.g., a valve seat sliding sound) component from a detection signal containing a mixture of such multiple sound sources. The method is explained below, but it can also be applied to valves of other types and other operation methods other than motorized gate valves and manual globe valves.

The configuration of the abnormality sign diagnosis device 20 of the embodiment is described with reference to FIG. 3. In this specification, a system configured with the acoustic sensor 2 and the abnormality sign diagnosis device 20 is referred to as an abnormality sign diagnosis device system.

The abnormality sign diagnosis device 20 in FIG. 3 performs abnormality sign diagnosis by detecting the sliding sound of the worm gear 24 with the acoustic sensor 2a provided on the drive unit 5 of the valve 1 in FIG. 1, detecting the sliding sound of the sliding part between the valve stem 4 and the gland packing 28 with the acoustic sensor 2b, and detecting the sliding sound of the sliding part between the valve body 29 and the valve seat 30 with the acoustic sensor 2c. It goes without saying that the abnormality sign diagnosis device 20 is not limited to the acoustic sensors 2a, 2b, and 2c in FIG. 3, and can perform abnormality sign diagnosis using sliding sounds detected by acoustic sensors 2 installed in other locations.

The abnormality prediction diagnostic device 20 is composed of a signal processing unit 9 that processes the sliding sound signal detected by at least one acoustic sensor 2, an analysis unit 11 that determines the leakage amount value, surface roughness value, friction coefficient value, etc. of the sliding part based on the sliding sound of the sliding part acquired by the signal processing unit 9, and a judgment unit 16 that estimates the number of operations and operating time until leakage occurs or the flow rate tolerance is exceeded, estimates the partitioning capacity, and judges abnormalities based on the analysis results obtained by the analysis unit 11.

The signal processing unit 9 is a processing unit that processes each detection signal of the acoustic sensor 2 connected to the abnormality sign diagnosis device 20, and is composed of an amplifier unit 6 that amplifies the detection signal to a predetermined level, a high-pass filter 7 that cuts elastic waves, vibration components, noise, etc. in a specific frequency range, and an A/D converter 8 that digitally converts the detection signal of the acoustic sensor 2. The high-pass filter 7 may be a band-pass filter.
Moreover, the signal processing unit 9 may further include a low-pass filter, a band-pass filter, or an envelope detection circuit to enhance signal discrimination of elastic waves.

In detail, the amplifier unit 6 has amplifiers for the number of acoustic sensors 2, and the high-pass filter 7 has filters for the number of acoustic sensors 2 that filter the detection signals amplified by the amplifiers. The A/D converter 8 synchronizes and digitally converts the amplified and filtered detection signals of the multiple acoustic sensors 2.

The analysis unit 11 has a sliding sound waveform extraction/division unit 10 that extracts or separates the waveform of the sliding sound of a specific sliding part from the sliding sound detected by at least one acoustic sensor 2 processed by the signal processing unit 9, a spectrum analysis unit 15 that analyzes the waveform of the sliding sound in the frequency domain, and a leakage amount value acquisition unit 12, a surface roughness value acquisition unit 13, and a friction coefficient value acquisition unit 14 that determine a leakage amount value, a surface roughness value, and a friction coefficient value, respectively, from the time change in the spectral intensity of the sliding sound determined by the spectrum analysis unit 15.
The sliding sound waveform extraction/division unit 10, the leakage amount value acquisition unit 12, the surface roughness value acquisition unit 13, and the friction coefficient value acquisition unit 14 will be described in detail later.

The determination unit 16 has an abnormality determination unit 19 that determines abnormalities including an increase in friction loss based on the leakage amount, surface roughness, friction coefficient, and their change rates acquired by the leakage amount value acquisition unit 12, surface roughness value acquisition unit 13, and friction coefficient value acquisition unit 14, an operation count/operation time estimation unit 17 until leakage occurs/flow rate exceeds allowable value that estimates the operation count/operation time until leakage occurs/flow rate exceeds allowable value, and a partitioning capacity estimation unit 18 that estimates the partitioning capacity by combining with operating force information of the drive unit.

More specifically, the abnormality prediction diagnostic device 20 is composed of analog signal circuits such as an amplifier and an A/D converter, a microcomputer circuit that performs acoustic analysis and judgment using a built-in program, a display device that displays the analysis results and judgment results, etc.

When inspecting the valve 1, the abnormality sign diagnostic device 20 stops the flow of the working fluid and drives the drive unit 5 of the valve 1 to move the valve stem 4 up and down or rotate it, thereby sliding the worm gear 24 of the valve 1, the sliding part between the valve stem 4 and the gland packing 28, and the sliding part between the valve body 29 and the valve seat 30. The abnormality sign diagnostic device 20 detects sliding sounds using acoustic sensors 2a, 2b, and 2c, and determines whether an abnormality has occurred, estimates the partitioning capacity, and estimates the number of operations and operating time until leakage occurs or the allowable flow rate is exceeded.

Here, the configuration of the acoustic sensor 2 will be described with reference to FIG.
4 shows a cross-sectional structure of the acoustic sensor 2 when it is installed on an installation surface 401. There are various types of known acoustic sensors 2, and any of them can be applied, but in the embodiment, the following type of acoustic sensor 2 is applied.

This type is a wideband and single-ended type. The wideband type has a flat band as a frequency characteristic, for example, a wide band from 100 kHz to 1 MHz. This acoustic sensor 2 includes a piezoelectric element 403, a receiving plate 404, a damper 405, a shield case 406, a cover 407, a connector 408, a signal cable 409, etc.

In addition, this type has a damper 405 provided above the piezoelectric element 403 to suppress resonance. The sensitivity of this acoustic sensor 2 is, for example, 40 to 55 dB. The standard for sensitivity is 0 dB = 1 V/m/s.

The heat resistance temperature of the acoustic sensor 2 is, for example, 80°C for a normal type and 200°C for a high-temperature type. The receiving plate 404, particularly its receiving surface, is fixed to the installation surface 401 via an acoustic coupler such as grease 402. There are two methods for fixing the acoustic sensor: directly fixing the acoustic sensor with adhesive, and using a sensor holder and fixing the sensor holder with adhesive, magnets, screws, etc. In the latter case, an acoustic coupler such as grease is applied between the acoustic sensor and the ground surface to ensure close contact so that air does not get in.

The signal cable 409 from the piezoelectric element 403 is connected to a connector 408, and is connected to an external signal cable through the connector 408.

Next, the details of the processing of the sliding sound waveform extraction/division unit 10 (see FIG. 3) will be described with reference to FIG.
FIG. 5 is a diagram showing an example of the change over time in the detection signals of the acoustic sensors 2a, 2b, and 2c (see FIG. 3).
The acoustic sensor 2a detects the sliding sound of the worm gear 24 (drive unit 5, see FIG. 1), the acoustic sensor 2b detects the sliding sound of the sliding part between the valve stem 4 and the gland packing 28 (see FIG. 1), and the acoustic sensor 2c synchronously detects the sliding sound of the sliding part between the valve body 29 and the valve seat 30 (see FIG. 1).

The signal waveforms of the sliding sound of the worm gear 24, the sliding sound of the sliding part between the valve stem 4 and the gland packing 28, and the sliding sound of the sliding part between the valve body 29 and the valve seat 30 each have their own unique characteristics. For example, in the case of the worm gear 24 detected by the acoustic sensor 2a, the sliding sound is generated by hammering, resulting in a sudden signal waveform that rises sharply and then decays. The sliding sound of the sliding part between the valve body 29 and the valve seat 30 when the valve is closed detected by the acoustic sensor 2c also has a sudden signal waveform caused by hammering. In addition, the sliding sound of the valve stem 4 and the gland packing 28 detected by the acoustic sensor 2b has a continuous signal waveform that is constantly generated when the valve is driven.

In addition, signals such as the sound of a driving motor or air, sliding noise signals that occur at regular intervals in gears and threaded parts, shock pulses caused by the wobbling of the valve disc just before closing, sliding noise signals that occur between the valve disc and the valve seat when closing, and other abnormal pulses caused by jamming, etc., are all unique patterns of generated sounds. Therefore, by classifying the patterns of sliding sounds, it is possible to identify the sliding parts.

The sliding sound propagates not only to the installed acoustic sensor 2 but also to other acoustic sensors 2. Therefore, the acoustic sensor 2 detects not only the target sliding sound but also other sliding sounds. For example, the acoustic sensor 2c detects not only the sliding sound of the sliding part between the valve body 29 and the valve seat 30, but also the sliding sound of the sliding part between the valve stem 4 and the gland packing 28 and the sliding sound of the worm gear 24. In other words, the detection signal of the acoustic sensor 2c contains the sliding sound of the sliding part between the valve stem 4 and the gland packing 28 and the sliding sound of the worm gear 24 superimposed thereon.

The sliding sound of the sliding parts of the valve body 29 and valve seat 30 detected by the acoustic sensor 2c is a sudden signal waveform, and the signal amplitude period is short. In contrast, the sliding sound of the sliding parts of the valve stem 4 and gland packing 28 and the sliding sound of the worm gear 24 detected by the acoustic sensor 2c are each signals with a predetermined delay time, and the signals are attenuated due to propagation. Note that Figure 5 is a diagram explaining the propagation signal of the acoustic sensor 2c, and does not explain all the propagation signals.

Therefore, in the signal detected by the acoustic sensor 2c, a certain period of time with a signal waveform of relatively large amplitude is identified as the sliding sound of the sliding part between the valve stem 4 and the gland packing 28, and the signal in other periods is eliminated as noise.

More specifically, the sliding noise of the sliding parts when the valve body 29 and the valve seat 30 are closed varies depending on the installation state of the valve.
When the valve stem is in an upright position, the valve disc is centered by its own weight, so the sliding that occurs when the valve disc seats occurs multiple times due to the behavior of the centering, and then the valve disc slides over the entire seat surface before seating and closing the flow path. For this reason, it is preferable to focus on the sliding sound caused by the final seating.

In addition, in a valve that is not in an upright position, the valve body and the valve seat are not aligned, and a sliding sound is generated due to a biased seating surface, and the first half of the acquired sliding sound is the sliding of the biased seating surface, and the latter half is the sliding of the entire seat part. Therefore, it is preferable to focus on the latter half of the sliding sound and diagnose the potential for leakage.
Next, another method of analyzing sliding sounds in the sliding sound waveform extraction/division unit 10 will be described.

Figure 6 shows the waveform (AE waveform) of the detection signal including the valve seat sliding sound detected by the acoustic sensor 2 installed in the valve box when the handle is rotated from the open state to the seated state during the closing operation of the manual globe valve in Figure 2.

The amplitude of the detection signal is large while the handle is being operated, and increases further when the valve seat is seated, after which the amplitude of the detection signal returns to the background level (background noise). From these characteristics, it is possible to determine the timing of seating and identify the detection signal when the valve disc and valve seat are sliding.

In Figure 6, two sudden AE waves are generated during steering operation, and whether this is a person sitting or not can be determined by whether the amplitude increases and then returns to the background level (background noise) or returns to the level it was at just before the increase. When the amplitude increases and then returns to the background level, as in the sudden AE wave generated at around 0.8 seconds on the time axis in Figure 6, this is when the person is sitting.

Figure 7 shows an AE waveform of the valve seat sliding sound generated between the valve disc and the valve seat, detected by the acoustic sensor 2 installed in the valve box when the manual handle 23 is rotated to a complete stop after the valve disc 29 seats on the valve seat 30 during the closing operation of the manual globe valve in Figure 2. The action of pushing the valve stem 4 until it comes to a complete stop after the valve disc 29 seats is hereafter referred to as "pushing in."

When the valve is pushed in, the valve body 29 and the valve seat 30 slide, and this sliding sound is mainly detected by the acoustic sensor 2 installed in the valve box. The waveform of this valve seat sliding sound has a continuous AE wave shape as shown in Figure 7.

When the sliding sound waveform extraction/division unit 10 detects a sudden AE wave, it determines that the detection signal is a sliding sound caused by the valve seating when the amplitude of the detection signal returns to the background level (background noise), and determines the period from the detection of the sudden AE wave to the return of the amplitude of the detection signal to the background level (background noise) as the push-in period, and the continuous AE wave during this period as a detection signal (AE signal) that includes the valve seat sliding sound.

Incidentally, in Figure 7, the sudden AE wave that occurs at about 0.3 seconds on the time axis is due to the valve seating, and the amplitude that follows until it returns to the background amplitude is due to the pressing period.

Next, we will explain a method for extracting only the valve seat sliding sound components from the detected continuous AE waves by analysis using the detection signal or signal waveform detected by the acoustic sensor 2 installed in two places, the valve box and the manual handle 23, during the closing operation of the manual globe valve in Figure 2.

<Step 1: Calculation of propagation rate α>
When the manual handle 23 is operated, the sliding sound of the manual handle 23 propagates through the parts where the metals are in close contact with each other, and reaches the acoustic sensor 2 in the valve box. Therefore, the acoustic sensor 2 in the valve box detects the sum of the sliding sound of the valve disc 29 and the valve seat 30 and the propagating sliding sound of the manual handle 23. With regard to the sliding sound of the valve disc 29 and the valve seat 30, it is necessary to remove the sliding sound of the manual handle 23 that propagates from the detection sound of the acoustic sensor 2.

Here, in step 1, when the propagation rate α, which indicates the propagation characteristics of the sliding sound, is introduced, the sliding sound of the manual handle 23 detected by the acoustic sensor 2 in the valve box section is the product of the detection result of the acoustic sensor 2 provided on the manual handle 23 and the propagation rate α. Therefore, the sliding sound of the valve body 29 and the valve seat 30 is obtained by subtracting the product of the detection result of the acoustic sensor 2 provided on the manual handle 23 and the propagation rate α from the sound detected by the acoustic sensor 2. Note that it is assumed that the acoustic sensor 2 provided on the manual handle 23 detects the sliding sound of the manual handle 23 without attenuation.

The propagation rate α is calculated as follows.
2, the only sliding sound before the valve disc 29 is seated on the valve seat 30 is the sliding sound of the manual handle 23. The signal strength of the AE signal detected by the acoustic sensor 2 of the manual handle 23 before the valve disc 29 is seated is x1, and the signal strength of the AE signal detected by the acoustic sensor 2 of the valve box is y1.
Propagation rate α = y1/x1 (Formula 1)
It is preferable to calculate the propagation factor α each time the sensor is attached, since the contact area of the sensor changes when the sensor is removed and reattached. Here, the signal strength of the AE signal or the sliding sound strength is, for example, an RMS value.

<Step 2: Calculating the valve seat sliding noise>
In step 1, the propagation rate α was defined as the ratio between the signal strength of the AE signal detected by the acoustic sensor 2 in the valve box section and the signal strength of the AE signal detected by the acoustic sensor 2 in the manual handle 23 when the sliding sound of the manual handle 23 propagates to the valve box section, but the propagation rate when the sliding sound of the valve body 29 and the valve seat 30 propagates to the manual handle 23 is also the same.

Regarding the continuous AE wave detected when the valve is pushed in, if the sliding sound intensity of the manual handle 23 is m, the sliding sound intensity of the valve disc 29 and the valve seat 30 is n, and the signal intensity of the continuous AE wave detected by the acoustic sensor 2 in the valve box is y2, then the acoustic sensor 2 in the valve box detects the sum of the propagated sounds of the valve sliding sound and the handle sliding sound,
y2 = n + α・m (Formula 2)
become.

Similarly, regarding the continuous AE wave detected when the manual handle 23 is pressed, the sliding sound intensity of the manual handle 23 is m, the sliding sound intensity of the valve body 29 and the valve seat 30 is n, and the signal intensity of the continuous AE wave detected by the acoustic sensor 2 of the manual handle 23 is x2. Since the acoustic sensor 2 of the manual handle 23 detects the sum of the propagated sounds of the handle sliding sound and the valve sliding sound,
x2 = m + α・n (Formula 3)
In addition, various sliding noises generated in the gland packing and threaded parts are considered to be included in the handle operation noise.

By solving the simultaneous equations of (Equation 2) and (Equation 3) above, the sliding sound m of the manual handle 23 and the sliding sound n of the valve disc 29 and the valve seat 30 can be obtained.
By the above two steps, it is possible to extract only the valve seat sliding sound components from the detected continuous AE waves by analysis using the detection signals of the acoustic sensors 2 installed in two locations, the valve box and the manual handle 23.

In this way, when it is desired to diagnose the sliding condition of the valve seat 30, not only is an acoustic sensor 2 installed near the valve seat to be diagnosed, but also near the manual handle 23, which is a component other than the valve seat, and by calculating the signals of the two acoustic sensors 2, it is possible to extract only the components of the sliding sound of the valve disc 29 and the valve seat 30 from the sensor signal, which contains a mixture of the sliding sound of the valve disc 29 and the valve seat 30 and the sliding sound of the manual handle 23. In other words, by using two acoustic sensors 2, it is possible to improve the accuracy of detecting the sliding sound of the valve disc 29 and the valve seat 30 compared to detecting only the vicinity of the valve seat with one acoustic sensor 2.

The above describes the case where there are two sound sources, but even if there are three or more sound sources, the sliding sound components of the area to be diagnosed can be extracted by using acoustic sensors 2 in a number equal to or greater than the number of sound sources, determining the propagation rate α of the sound source signal for each sensor, and solving the simultaneous equations.

Figure 8 is a process flow diagram of the sliding sound waveform extraction and division unit 10, which extracts the valve seat sliding sound by a method of extracting only the valve seat sliding sound components from the above-mentioned continuous AE wave through analysis.

In step S81, the background noise of the detection signal of each acoustic sensor 2 is obtained.

In step S82, the signal strength (RMS value) of each acoustic sensor 2 detection signal is calculated as an AE signal.

In step S83, it is determined whether the signal strength of the acoustic sensor 2 calculated in step S82 is greater than the value obtained by multiplying the background noise of the acoustic sensor 2 acquired in step S81 by a predetermined coefficient. In other words, it is determined whether the detection signal of the acoustic sensor 2 is a continuous AE wave, and if it is greater (Yes in S83), it is determined that the detection signal of the acoustic sensor 2 is a continuous AE wave and proceeds to step S84. If it is not greater (No in S83), it returns to step S82.

In step S84, the maximum signal strength (RMS value) of the continuous AE wave section of the detection signal of each acoustic sensor 2 is obtained.

In step S85, a predetermined sliding sound component is extracted from the detection signal of the acoustic sensor 2 by a method of analytically calculating the sliding sound of each sound source based on a simultaneous equation expressed by the sum of the products of the detection signals detected by each acoustic sensor 2 and the propagation rate of the sliding sound.

The signal strength of the extracted sliding sound is stored as needed for each sliding part in a database unit (not shown) of the abnormality sign diagnosis device 20 so that changes over time can be determined. Details will be described later, but when the signal strength of the valve sliding sound changes from the initial state, the abnormality sign diagnosis device 20 estimates the leakage amount, surface roughness, or friction coefficient of the sliding part between the valve body and the valve seat, determines valve deterioration, and predicts the time for maintenance.

As a result, the sliding sound waveform extraction and division unit 10 extracts the sliding sound of a specified sliding part, such as the sliding sound between the valve body 29 and the valve seat 30.

Hereinafter, the processing contents of the leakage amount value acquisition unit 12, the surface roughness value acquisition unit 13, and the friction coefficient value acquisition unit 14 in the analysis unit 11 of the abnormality prediction diagnosis device 20 of the embodiment will be described.

Figure 9 shows the correlation between the spectral intensity of the sliding noise and the number of times the sliding parts operate for the sliding noise of the worm gear 24, the sliding noise of the sliding part between the valve stem 4 and the gland packing 28, and the sliding noise of the sliding part between the valve body 29 and the valve seat 30.

The analysis unit 11 of the abnormality prediction diagnosis device 20 stores in advance the correlation between the measured spectral intensity of the sliding sound and the number of times the sliding part operates or the operating time for each sliding part. In more detail, the correlation between the spectral intensity of the sliding sound and the number of times the sliding part operates is stored in a database unit (not shown) of the abnormality prediction diagnosis device 20 for each sliding part: the worm gear 24, the sliding part between the valve stem 4 and the gland packing 28, and the sliding part between the valve body 29 and the valve seat 30.

Then, the analysis unit 11 determines the number of operations or the operating time taking into account the deterioration due to operation from the start of use of the sliding part when the sliding sound was detected and the deterioration due to fluid force, aging, etc., from the correlation between the spectral intensity of the sliding sound and the number of operations or the operating time of the sliding part shown in FIG. 9 based on the spectral intensity of the sliding sound detected by a specified acoustic sensor 2, which is calculated by the spectral analysis unit 15 that performs frequency analysis of the sliding sound. Here, it should be noted that the number of operations or the operating time determined from the spectral intensity of the detected sliding sound does not match the actual number of operations or the operating time, and that the deterioration corresponding to the determined number of operations or the operating time is detected.

FIG. 10A is a diagram showing the correlation between the surface roughness and the number of operations in each sliding portion.
The analysis unit 11 stores in advance in a database unit (not shown) of the abnormality sign diagnosis device 20 a correlation between the actually measured surface roughness and the number of operations or the operating time for each sliding part.

The surface roughness value acquiring unit 13 acquires the surface roughness value of the sliding portion that emits the sliding sound based on the sliding sound detected by the acoustic sensor 2 .
In detail, as described above, the surface roughness value acquiring unit 13 first obtains the number of operations or the operating time from the start of use of the sliding part based on the spectrum intensity of the sliding sound detected by the acoustic sensor 2 (FIG. 9). Then, from the correlation between the surface roughness and the number of operations or the operating time shown in FIG. 10A, the surface roughness value of the sliding part corresponding to the number of operations or the operating time is acquired.

FIG. 10B is a diagram showing the correlation between the friction coefficient and the number of operations or the operating time for each sliding portion.
The analysis unit 11 stores in advance in a database unit (not shown) of the abnormality sign diagnosis device 20 a correlation between the actually measured friction coefficient and the number of operations or the operating time for each sliding part.

The friction coefficient value acquisition unit 14 acquires the friction coefficient value of the sliding part that emits the sliding sound based on the sliding sound detected by the acoustic sensor 2 .
In detail, as described above, the friction coefficient value acquisition unit 14 first obtains the number of operations or the operating time from the start of use of the sliding part based on the spectrum intensity of the sliding sound detected by the acoustic sensor 2 (FIG. 9). Then, the friction coefficient value of the sliding part corresponding to the number of operations is acquired from the correlation between the friction coefficient and the number of operations or the operating time shown in FIG. 10B.

FIG. 10C is a diagram showing the correlation between the leakage amount and surface roughness in a sliding portion that seals a fluid.
The analysis unit 11 stores in advance the correlation between the actually measured leakage amount and surface roughness for each sliding portion in a database unit (not shown) of the abnormality sign diagnosis device 20 .

The leakage amount value acquisition unit 12 acquires the leakage amount from the sliding portion that emits the sliding sound based on the sliding sound detected by the acoustic sensor 2 .
In detail, the leakage amount value acquisition unit 12 first obtains the number of operations or the operating time since the start of use of the sliding part based on the spectrum intensity of the sliding sound detected by the acoustic sensor 2 as described above ( FIG. 9 ). Then, from the correlation between the surface roughness and the number of operations or the operating time shown in FIG. 10A , it acquires the surface roughness value of the sliding part corresponding to the number of operations or the operating time. After that, from the correlation between the leakage amount and the surface roughness shown in FIG. 10C , it acquires the leakage amount value corresponding to the acquired surface roughness value of the sliding part.

Next, the operation count/operation time estimation unit 17, partitioning capacity estimation unit 18, and abnormality determination unit 19 in the determination unit 16 will be described in detail. With this configuration, the abnormality sign diagnosis device 20 of the embodiment predicts the timing of maintenance, including disassembly of the valve 1, or determines signs of deterioration, including whether the function of closing the fluid, the dynamic function, and the flow rate adjustment function of the valve 1 are normal or not.

In the case of an ON-OFF valve, the operation count/operation time estimation unit 17 estimates the number of operations until leakage occurs based on the leakage amount acquired by the leakage amount value acquisition unit 12, the correlation between the leakage amount and surface roughness, and the correlation between surface roughness and the number of operations or the operation time.

In detail, when the leakage amount acquired by the leakage amount value acquisition unit 12 is "0", that is, when there is no leakage or the leakage amount is within the allowable range, the operation count/operation time estimation unit 17 determines the surface roughness of the sliding part at which the leakage amount becomes greater than "0" from the correlation between the leakage amount and surface roughness (Fig. 10C). Then, from the correlation between the surface roughness and the operation count or operation time (Fig. 10A), it determines the number of operations corresponding to the surface roughness of the sliding part at which the leakage amount becomes greater than "0". The operation count/operation time estimation unit 17 subtracts the number of operations at which the sliding sound was detected from the number of operations calculated above, and determines this as the number of operations until leakage occurs.

The operation count/operation time estimation unit 17 predicts the timing when the calculated number of operations until leakage occurs reaches a predetermined number of operations as the time for maintenance, including disassembly of the valve 1.

In addition, if the leakage amount acquired by the leakage amount value acquisition unit 12 is "0", the number of operations/operating time estimation unit 17 determines that the function of closing the valve fluid is "normal", and if the leakage amount is greater than "0", it determines that the function of closing the valve fluid is "not normal (abnormal)".
Furthermore, the number of operations/operating time until leakage occurs estimation unit 17 determines that there is a sign of deterioration in the valve's function of closing the fluid if the number of operations until leakage occurs is smaller than a predetermined number of operations.

Next, the case where the valve 1 is a flow rate control valve will be described.
In this case, the analysis unit 11 pre-stores the correlation between the measured surface roughness and operating time, and the correlation between the measured flow rate and surface roughness for each sliding part in a database unit (not shown) of the abnormality prediction diagnosis device 20.

The leakage amount value acquisition unit 12 then obtains the operation time from the start of use of the sliding part based on the spectrum intensity of the sliding sound detected by the acoustic sensor 2 ( FIG. 9 ). Then, from the correlation between the surface roughness and the operation time, it obtains the surface roughness value of the sliding part corresponding to the operation time. After that, from the correlation between the flow rate and the surface roughness, it obtains the flow rate value corresponding to the obtained surface roughness value of the sliding part.
In other words, the leakage amount value acquiring unit 12 functions as a flow rate value acquiring unit that determines the flow rate based on the sliding sound detected by the acoustic sensor 2 .

The operation count/operation time estimation unit 17 estimates the operation time until the allowable flow rate is exceeded based on the flow rate acquired by the leakage amount value acquisition unit 12 (flow rate value acquisition unit), the correlation between the flow rate and surface roughness, and the correlation between surface roughness and operation time.

In detail, when the flow rate acquired by the leakage amount value acquisition unit 12 (flow rate value acquisition unit) does not change, the operation count/operation time estimation unit 17 determines the surface roughness of the sliding part at which the flow rate increases from the correlation between the flow rate and surface roughness. Then, it determines the operation time corresponding to the surface roughness of the sliding part at which the flow rate increases from the correlation between the surface roughness and the operation time. The operation count/operation time estimation unit 17 subtracts the operation time at which the sliding sound was detected from the determined operation time to determine the operation time until leakage occurs.

The operation count/operation time estimation unit 17 predicts the timing for maintenance, including disassembly of the valve 1, when the calculated operation time until the allowable flow rate is exceeded reaches a predetermined operation time.
In addition, if the flow rate acquired by the leakage amount value acquisition unit 12 (flow rate value acquisition unit) does not change, the operation count/operating time estimation unit 17 determines that the flow rate control function is "normal," and if the flow rate changes, it determines that the flow rate control function of the valve is "not normal (abnormal)."
Furthermore, the operation count/operating time estimation unit 17 determines that there is a sign of deterioration of the valve body and valve seat related to the flow rate adjustment function when the operating time until the calculated flow rate tolerance is exceeded is shorter than a predetermined operating time.

The shutoff capacity estimation unit 18 estimates the shutoff capacity for closing and opening the valve body of the shutoff valve by combining the operating force information of the drive unit.

In detail, the partitioning capacity estimating unit 18 measures the drive current of the drive unit 5 of the valve 1 in synchronization with the detection of the sliding sound, and obtains the rate of change of the drive force. Then, the partitioning capacity estimating unit 18 obtains the rate of change of the friction coefficient value of each sliding part obtained by the friction coefficient value obtaining unit 14. The partitioning capacity estimating unit 18 compares the rate of change of the drive force with the rate of change of the friction coefficient value of each sliding part, and when the rate of change of the drive force increases, predicts that the sliding part with the increased rate of change of the friction coefficient value is the part causing the increase in drive force.
The gate capacity estimation unit 18 obtains the number of operations required to reach the maximum driving force of the driving unit 5 from the correlation between the driving force and the number of operations, and predicts the number of operations at the limit of the valve gate capacity. The number of operations is calculated from the correlation between the spectrum intensity of the actually measured sliding sound and the number of operations of the sliding unit.

The abnormality determination unit 19 determines whether there is an abnormality, including an increase in friction loss, based on the friction coefficients and their rates of change acquired by the friction coefficient value acquisition unit 14, or whether there is an abnormality, including an increase in surface roughness loss, based on the surface roughnesses and their rates of change acquired by the surface roughness value acquisition unit 13.

In detail, the abnormality determination unit 19 acquires the friction coefficient value of each sliding part calculated from the sliding sound acquired by the friction coefficient value acquisition unit 14. Also, the surface roughness of each sliding part calculated from the sliding sound acquired by the surface roughness value acquisition unit 13 is acquired.
Then, the abnormality determination unit 19 calculates the amount of change in the friction coefficient and the amount of change in the surface roughness of each of the obtained sliding parts.

The abnormality determination unit 19 determines the number of operations at which the friction coefficient reaches the maximum allowable value from the correlation between the friction coefficient and the number of operations at each sliding part, and predicts the predetermined number of operations at which the friction coefficient reaches the maximum value as the timing for maintenance including disassembly of the valve 1. Then, the earliest timing for each sliding part is determined as the maintenance time. Furthermore, the sliding part with the largest change in the friction coefficient of each sliding part is determined as the sliding part predicted to be abnormal.

The abnormality determination unit 19 also determines the number of operations at which the surface roughness reaches the maximum allowable value from the correlation between the surface roughness of each sliding part and the number of operations, and predicts the predetermined number of operations at which the surface roughness reaches the maximum value as the timing for maintenance including disassembly of the valve 1. The earliest timing for each sliding part is then determined as the maintenance time. Furthermore, the sliding part with the largest change in surface roughness of each sliding part is determined as the sliding part predicted to be abnormal.

Furthermore, when the friction coefficient acquired by the friction coefficient value acquisition unit 14 is greater than the normal friction coefficient, the abnormality determination unit 19 determines whether the valve's function of closing the fluid, dynamic function, and flow rate adjustment function are normal or not according to the amount of change in the friction coefficient, and when the amount of change in the friction coefficient is greater than a predetermined value, the abnormality determination unit 19 determines that there is a sign of deterioration of the sliding part.

In addition, when the surface roughness acquired by the surface roughness value acquisition unit 13 is greater than the normal surface roughness, the abnormality determination unit 19 determines whether the valve's function of closing the fluid, dynamic function, and flow rate adjustment function are normal or not according to the amount of change in the surface roughness, and when the amount of change in the surface roughness is greater than a predetermined value, the abnormality determination unit 19 determines that there are signs of deterioration of the sliding parts.

When determining whether the valve's function of closing the fluid, its dynamic function, and its flow rate adjustment function are normal, or when determining signs of deterioration of the sliding parts, the abnormality determination unit 19 may make such a determination based on at least one of the friction coefficient acquired by the friction coefficient value acquisition unit 14 and the surface roughness acquired by the surface roughness value acquisition unit 13.

The correlation between the spectral intensity of the sliding sound and the number of operations or operating time of the sliding part, the correlation between the surface roughness of the moving part and the number of operations, the correlation between the friction coefficient and the number of operations or operating time, and the correlation between the amount of leakage and surface roughness, which are referred to in the analysis unit 11 and the determination unit 16 described above, differ depending on the sliding part. For this reason, they are recorded in a database in association with the valve type and the sliding part (sliding part name). Then, the analysis unit 11 and the determination unit 16 refer to the database depending on the sliding part of the sliding sound extracted based on the detection signal of the acoustic sensor 2.

If the database does not contain data on the sliding part from which the sliding sound was extracted, data on a valve of a model similar to the valve from which the sliding sound was detected is used based on the valve model. In this case, the valve model from which the data is reused may be associated in advance, a valve with similar specifications may be automatically selected, or the operator may be able to select from a list.

Next, the procedure for diagnosing a valve using the abnormality prediction diagnosis system of the embodiment will be explained using the flow diagram in Figure 11.

In step S111, acoustic sensors 2 are installed near the sliding part of the valve and near the drive part of the valve (such as the manual handle 23).

In step S112, the sensor installation locations and sensor model names of the two acoustic sensors 2 installed in step S111 are input to obtain the frequency characteristics (sensor F characteristics) of the acoustic sensors 2. The obtained installation locations (sliding part information), sensor model names, and frequency characteristics of the acoustic sensors 2 are stored in a database unit (not shown) of the abnormality sign diagnosis device 20 for each valve identified by the management number.

In step S113, the acoustic sensor 2 measures the background noise of the valve device.

In step S114, the valve is operated.

In step S115, the sensitivity of the acoustic sensor 2 is calibrated. This is done to eliminate the effect of changes in the contact area, since the contact area changes each time the acoustic sensor 2 is attached.

In step S116, the propagation rate α of the sliding sound is calculated based on the sliding sound detected by the two acoustic sensors 2 before the valve body 29 is seated on the valve seat 30. The calculated propagation rate α is stored in a database unit (not shown) of the abnormality sign diagnosis device 20 for each valve identified by the management number.

In step S117, the valve disc is pushed in after it has been seated, and the sensor signal (continuous AE wave) that returns to the background noise level is detected by acoustic sensor 2 as the sliding sound when pushed in, and the sliding time of the valve disc is obtained as the analysis interval from the detected sensor signal.

In step S118, the sliding sound of each sound source is analytically extracted from the sensor signal of the acoustic sensor 2 based on a simultaneous equation expressed by the sum of the products of the sensor signals of the analysis sections of each sound source detected by the two acoustic sensors 2 in step S117 and the propagation rate of the sliding sound.

In step S119, it is determined whether the signal strength of the valve sliding sound extracted in step S118 has changed from the signal strength of the valve sliding sound in the initial state of the valve, and if it has changed, the process proceeds to step S120. The signal strength of the valve sliding sound in the initial state is registered in the database for each valve.

In step S120, the number of times the sliding part operates or the operating time is calculated based on the spectral intensity of the valve sliding sound detected in step S117, and the leakage amount, surface roughness, and friction coefficient of the sliding part corresponding to the calculated number of times the sliding part operates or the operating time are obtained to determine the deterioration of the valve.

In step S121, a diagnosis is performed to determine whether maintenance is required and whether there are signs of valve deterioration.

In the above embodiment of the abnormality sign diagnosis device 20, the sliding sounds of the multiple sliding parts are detected by the multiple acoustic sensors 2, and sliding sounds other than the target sliding sound detected by the acoustic sensors 2 are removed by the sliding sound waveform extraction/division unit 10 to perform deterioration sign diagnosis.

However, instead of this, the sliding sound waveform extraction and division unit 10 may analyze the pattern of the sliding sound detected by the acoustic sensor 2 and classify it for each sliding part. In this case, since the sliding sound propagates and attenuates, the acoustic sensor is placed near the sliding part to be diagnosed for signs of deterioration.

Specifically, in FIG. 1, an acoustic sensor 2 is provided to detect the sliding sound of the valve stem 4 and the gland packing 28, and the sliding sound is extracted by the sliding sound waveform extraction/division unit 10. At the same time, the acoustic sensor 2 detects the sliding sounds of the motor 22, which is the drive unit 5 of the valve 1, the worm gear 24 of the valve 1, the valve stem threaded portion 25, and the sliding surfaces of the valve body 29 and the valve seat 30 as propagation signals, and the sliding sounds are divided by the sliding sound waveform extraction/division unit 10 according to their waveform patterns.

However, when the motor 22 is running, it generates a continuous sound that may drown out other sliding sounds. For this reason, it is advisable to operate the valve 1 with the manual handle 23 to detect the sliding sound.

Furthermore, when the operation of the valve 1 starts, a pulse is generated due to hammering, and so that this is not confused with the sliding noise between the valve body 29 and the valve seat 30, a closing operation is preferable to an opening operation.
In this case, the following sliding sounds can be obtained as different patterns of sliding sounds: sliding sounds caused by initial hammering, sliding sounds of the gland packing 28 that occur constantly when driving, sliding sounds of the gear (not shown) of the manual handle 23 that occur at regular intervals, sliding sounds generated in the worm gear 24 and the valve stem threaded portion 25, impact sounds caused by the wobbling of the valve body 29 just before closing, sliding sounds generated in the seat portion of the valve body 29 and the valve seat 30 when closing, and other abnormal pulse sounds caused by jamming, etc.

The sliding sound waveform extraction/division unit 10 classifies the sliding sound according to the signal pattern of the sliding sound, so that the abnormality sign diagnosis device 20 can diagnose the condition of multiple sliding parts even with a single acoustic sensor 2.
The acoustic sensor 2 is preferably installed near a sliding location that should be diagnosed with the highest priority, or near the gland packing 28 where the amplitude of the sliding sound is small.

In the above embodiment, the valve 1 has been described as having a gate valve, but the type of valve is not limited to this, and similar diagnosis is also possible for globe valves, butterfly valves, and ball valves.
In a globe valve, sliding occurs at the seat, which is the sliding part between the valve body and the valve seat, like a gate valve, during opening and closing operations. Compared to a gate valve, the operating time and sliding distance are shorter, but a short sliding sound signal can be obtained after the collision sound is generated. This sliding sound signal makes it possible to diagnose the condition of the seat.
In addition, in butterfly valves and ball valves, the valve stem and valve body rotate to close the flow path, but sliding motion similar to that of the gate valve occurs, making diagnosis possible using sliding sound signals.

In addition, in the above embodiment, the driving method of the valve 1 is explained as an electric valve, but a similar diagnosis can also be performed on an air-operated valve. In an air-operated valve, sliding occurs due to the up and down movement of a piston, instead of the sliding of the driving part 5 of the electric valve. By acquiring this sliding sound with the acoustic sensor 2, a diagnosis similar to that of the electric valve can be performed.

Next, a case where the abnormality sign diagnostic device 20 according to the embodiment is applied to a nuclear power generation facility will be described.
There are various types of valves installed in nuclear power plants, such as high-temperature valves, valves through which fluid constantly passes, valves installed in locations where the ambient radiation level is high, valves installed in locations where noise levels are high during operation, etc. Therefore, when applying to nuclear power plants, it is preferable to carry out the test in a quiet state where the temperature is normal, such as during a periodic inspection, when the system is stopped, the radiation level is not high, and the test is carried out.

Many of the valves that become hot during operation of nuclear power generation facilities are covered with thermal insulation material 33, making it difficult to install acoustic sensor 2 near the sliding part of valve body 29 and valve seat 30 shown in Figure 1. Thermal insulation material 33 is usually designed to be removable, but in order to shorten the inspection period, it is desirable to perform the diagnosis with thermal insulation material 33 still attached.

Figure 12 shows the structure of a hearing rod 31 (also called a wave guide or wave guide rod), which is a jig for mounting the acoustic sensor 2 to a high-temperature part or a place with a small contact area, as an example of mounting the acoustic sensor 2. The acoustic sensor 2 is mounted to one end of the flat part of the hearing rod 31, and the other end is in contact with the object to be measured. The hearing rod 31 is used when the temperature of the object to be measured exceeds the heat resistance temperature of the sensor, or when the place where the sensor can be installed is small and it is difficult to directly install the sensor. When the temperature of the object to be measured is high, the hearing rod 31 can be used to install the acoustic sensor 2 at a location away from the heat source. When the temperature is high, the length of the hearing rod 31 is selected so that the temperature of the sensor part is below the heat resistance temperature of the sensor. As shown in Figure 1, a couplant 21 (grease, wax, adhesive, etc.) may also be used.

FIG. 13 is a diagram showing a method of mounting the acoustic sensor 2 for performing diagnosis while the heat insulating material 33 is still attached.
The heat insulating material 33 of the valve box 3 is provided with a cylindrical access passage that does not affect the heat insulating effect, and during diagnosis, a hearing probe 31 equipped with a cylindrical mounting jig 34 is inserted into the access passage. Here, it is important that the inner material such as cotton of the heat insulating material 33 does not come into contact with the hearing probe 31, and other structures may be used as long as this is achieved.

Since the listening probe 31 is designed so that it does not come into contact with anything other than the acoustic sensor 2 and the valve box 3, the signal generated from the sliding portion between the valve body 29 and the valve seat 30 is not dispersed within the listening probe 31, and no noise is generated within the listening probe 31, making it possible to perform more accurate diagnosis.
This will enable a reduction in the amount of material, labor, cost, and radiation exposure required for disassembling and inspecting valves at nuclear power generation facilities, as well as an improvement in the availability rate by shortening the inspection period.

The present invention is not limited to the above-mentioned examples, but includes various modified examples. The above examples have been described in detail to provide an easy-to-understand explanation of the present invention, and are not necessarily limited to those having all of the configurations described. It is also possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment.

1 Valve (electric gate valve)
2, 2a, 2b, 2c Acoustic sensor 6 Amplification unit (signal processing unit)
7 High-pass filter (signal processing section)
8 A/D converter (signal processing section)
9 Signal processing unit 10 Sliding sound waveform extraction and division unit (analysis unit)
11 Analysis section 12 Leakage amount value acquisition section (analysis section)
13 Surface roughness value acquisition unit (analysis unit)
14 Friction coefficient value acquisition unit (analysis unit)
15 Spectral analysis unit (analysis unit)
16 Determination unit 17 Operation count/operation time estimation unit (determination unit)
18 Partitioning capacity estimation unit (judgment unit)
19 Abnormality judgment section (judgment section)
20. Abnormality Prediction Diagnostics Device

Claims (14)

a plurality of acoustic sensors attached to the valve for detecting valve seat sliding noise of a sliding portion between the valve body and the valve seat and sliding noise of components other than the valve seat sliding noise;
A signal processing unit that processes a detection signal of the acoustic sensor;
An analysis unit that analyzes the detection signals of the acoustic sensors and the propagation rate of the sliding sound processed by the signal processing unit to extract the valve seat sliding sound from the detection signals of the acoustic sensors, calculates the number of operations or operating time of the sliding part from the spectrum intensity of the extracted valve seat sliding sound, and estimates the leakage amount, surface roughness, or friction coefficient of the sliding part;
a determination unit which determines whether the valve is in a normal state or not with respect to a function of closing the fluid and a dynamic function of the valve from the number of operations or operating time of the sliding part, the amount of leakage, the surface roughness, or the friction coefficient obtained by the analysis unit, and predicts the timing of maintenance including disassembly of the valve;
An abnormality sign diagnosis system comprising: 2. The abnormality sign diagnosis system according to claim 1,
The abnormality prediction diagnosis system is characterized in that the sliding sounds of components other than the valve seat sliding sound are either a handle operation sound, or a sliding sound generated at the sliding part between the gland packing and the valve stem, a sliding sound generated at the sliding part of the threaded part of the valve stem, a sliding sound generated at the sliding part of the worm gear of an electric valve, or a sliding sound generated at the sliding part of the piston part of an air-operated valve, or a combination of a plurality of sliding sounds. 2. The abnormality sign diagnosis system according to claim 1,
the analysis unit estimates a leakage amount, a surface roughness, or a friction coefficient of a sliding portion between the valve disc and the valve seat when a signal strength of the valve seat sliding sound extracted from the detection signal of the acoustic sensor changes from a signal strength of the valve seat sliding sound in an initial state,
The abnormality sign diagnosis system is characterized in that the judgment unit judges signs of deterioration and predicts the timing of maintenance. 4. The abnormality sign diagnosis system according to claim 3,
A database is provided which stores the signal intensity of the valve seat sliding sound extracted from the detection signal of the acoustic sensor so that the change over time can be understood, and stores, for each sliding part, the correlation between the spectrum intensity of the sliding sound and the number of operations of the sliding part, the correlation between the surface roughness of the sliding part and the number of operations or the operating time, the correlation between the friction coefficient of the sliding part and the number of operations or the operating time, and the correlation between the amount of leakage from the sliding part and the surface roughness,
when the signal strength of the valve seat sliding sound extracted from the detection signal of the acoustic sensor changes from the signal strength of the valve seat sliding sound in an initial state in the database, the analysis unit refers to the database to estimate the leakage amount, surface roughness, or friction coefficient of the sliding portion between the valve disc and the valve seat. 4. The abnormality sign diagnosis system according to claim 3,
When the sliding amount estimated by the analysis unit is within an allowable range, the determination unit determines the surface roughness of the sliding part corresponding to the leakage amount exceeding the allowable range from the correlation between the leakage amount and surface roughness of the sliding part, determines the number of operations corresponding to the determined surface roughness from the correlation between the surface roughness of the sliding part and the number of operations or the operating time, and subtracts the number of operations determined by the analysis unit from the valve seat sliding sound from the determined number of operations,
An abnormality prediction diagnostic system comprising: 5. The abnormality sign diagnosis system according to claim 4,
The abnormality sign diagnosis system is characterized in that, when data on the sliding part from which the valve seat sliding sound was extracted is not recorded in the database, the analysis unit refers to data on the sliding part of a valve of a type similar to the type of valve from which the signal was detected. 2. The abnormality sign diagnosis system according to claim 1,
The abnormality sign diagnosis system is characterized in that the analysis unit determines that the detection signal is the valve seat sliding noise when the amplitude of the detection signal returns to a background noise level after detecting a sudden AE wave. 2. The abnormality sign diagnosis system according to claim 1,
The abnormality sign diagnosis system is characterized in that the signal processing unit processes a detection signal from an acoustic sensor, which is any one of an acoustic emission sensor, a piezoelectric sensor, a vibration sensor, and an ultrasonic sensor. 2. The abnormality sign diagnosis system according to claim 1,
The signal processing unit processes a signal from an acoustic sensor that detects sliding sounds generated in the sliding part between the valve disc and valve seat of a valve, the sliding part between the gland packing and valve stem, the sliding part of the threaded part of the valve stem, the sliding part of the worm gear of an electric valve, or the sliding part of the piston part of an air-operated valve. 10. The abnormality sign diagnosis system according to claim 9,
The signal processing unit processes signals from acoustic sensors that are installed directly or via a waveguide rod at positions on the outer surface of the valve that serve as propagation paths for direct waves of the sliding sounds of the sliding parts of the valve body and valve seat, the sliding surface of the valve stem and gland packing, the sliding part of the threaded part of the valve stem, the sliding part of the worm gear of the electric valve, or the sliding part of the piston part of the air-operated valve. 10. The abnormality sign diagnosis system according to claim 9,
The acoustic sensor is installed by a built-in or removable installation jig, adhesive, soldering, or hand-holding;
The removable installation jig is installed directly on the outer surface of the valve by one or more of a magnet, a suction cup, a vice, and a bolt, or is installed by an arm extending from a tripod or a pole or other device installed on the ground. 10. The abnormality sign diagnosis system according to claim 9,
When the valve is covered with insulation or a wall, an openable and closable cylindrical access passage is provided in the insulation or wall, and a cylindrical mounting jig integrated with a waveguide rod is inserted into the access passage, and the waveguide rod is installed so that it does not come into contact with anything other than the acoustic sensor and the valve. a signal processing unit that processes detection signals from a plurality of acoustic sensors that are attached to the valve and detect valve seat sliding noise of the sliding portion between the valve body and the valve seat and sliding noise of components other than the valve seat sliding noise;
An analysis unit that analyzes the detection signals of the acoustic sensors and the propagation rate of the sliding sound processed by the signal processing unit to extract the valve seat sliding sound from the detection signals of the acoustic sensors, calculates the number of operations or operating time of the sliding part from the spectrum intensity of the extracted valve seat sliding sound, and estimates the leakage amount, surface roughness, or friction coefficient of the sliding part;
a determination unit which determines whether the valve is in a normal state or not with respect to a function of closing the fluid and a dynamic function of the valve from the number of operations or operating time of the sliding part, the amount of leakage, the surface roughness, or the friction coefficient obtained by the analysis unit, and predicts the timing of maintenance including disassembly of the valve;
An abnormality sign diagnosis device comprising: A diagnostic method for an abnormality sign diagnostic device that diagnoses a state of a sliding portion of a valve seat, comprising:
a step of detecting a valve seat sliding sound of a sliding portion between the valve body and the valve seat and a sliding sound of components other than the valve seat sliding sound by a plurality of acoustic sensors attached to the valve;
A step of analyzing the detection signal of the acoustic sensor and the propagation rate of the sliding sound to extract the valve seat sliding sound from the detection signal of the acoustic sensor;
A step of determining the number of operations or the operating time of the sliding part from the extracted spectrum intensity of the valve seat sliding sound based on a correlation between the spectrum intensity of the sliding sound and the number of operations of the sliding part;
a step of estimating the leakage amount, surface roughness, or friction coefficient of the sliding part from the number of operations or the operating time obtained in the step based on a correlation between the surface roughness of the sliding part and the number of operations or the operating time, a correlation between the friction coefficient of the sliding part and the number of operations or the operating time, or a correlation between the leakage amount of the sliding part and the surface roughness;
a step of determining whether the valve is functioning normally or not with respect to a function of closing the fluid and a dynamic function based on the number of operations or the operating time, the amount of leakage, the surface roughness, or the friction coefficient obtained in the step above;
and predicting when maintenance, including disassembly, of the valve is required.

Images