Damage detection system, damage detection device, and damage detection method

Abstract

The objective is to provide a damage detection system, a damage detection device, and a damage detection method that can detect damage in a short amount of time. [Solution] The damage detection system of the embodiment comprises a plurality of sensors, a generation unit, and a damage detection unit. The plurality of sensors are arranged at predetermined intervals and detect elastic waves generated in the object to be inspected. The generation unit generates a spatiotemporal distribution representing the signal intensity of the elastic waves at different times for each elastic wave detected by each of the plurality of sensors, based on the elastic waves detected by each of the plurality of sensors. The damage detection unit detects damage to the object to be inspected based on the spatiotemporal distribution.

Description

Technical Field

[0001] Embodiments of the present invention relate to a damage detection system, a damage detection device, and a damage detection method.

Background Art

[0002] Cylindrical members such as plate-like members and pipes are damaged not only externally but also internally over time. Examples of internal damage include wall thickness reduction, peeling, and the like. However, internal damage cannot be determined from the appearance. Therefore, there is also a desire to detect internal damage in real time from the outside.

[0003] Conventionally, as a method for detecting internal damage, a technique has been proposed in which elastic waves are applied to an inspection object, and flaw detection is performed using the elastic waves that have passed through the damaged portion or the elastic waves reflected from the damaged portion. In Patent Document 1, an apparatus has been proposed that excites elastic waves in an inspection object and optically measures the shape of the surface elastic waves that propagate. In addition, in Non-Patent Document 1, a damage detection method focusing on guided waves that propagate through a structure has been proposed.

[0004] In the conventional method, in order to inspect the properties of guided waves over an area, a scanning method is used in which surface wave excitation and measurement are repeatedly performed while moving a point. Therefore, as the number of measurement points increases, the measurement time increases in proportion to the number of measurement points. In addition, since guided waves of different excitation trials are detected for each point, there is a possibility that the state of the inspection object changes at the start and end points of the measurement. In this case, it may become difficult to detect damage as the measurement time increases.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Non-Patent Documents

[0006] [Non-Patent Document 1] N. Hosoya, R. Umino, A. Kanda, I. Kajiwara, and A. Yoshinaga, “Lamb wave generation using nanosecond laser ablation to detect damage”, Journal of Vibration and Control, vol. 24, no. 24, pp. 5842-5853, Dec. 2018, doi: 10.1177 / 1077546316687904. [Overview of the project] [Problems that the invention aims to solve]

[0007] The problem that this invention aims to solve is to provide a damage detection system, a damage detection device, and a damage detection method that can detect damage in a short amount of time. [Means for solving the problem]

[0008] The damage detection system of this embodiment comprises a plurality of sensors, a generation unit, and a damage detection unit. The plurality of sensors are arranged at predetermined intervals and detect elastic waves generated in the object under inspection. The generation unit generates a spatiotemporal distribution representing the signal intensity of the elastic waves at different times for each elastic wave detected by each of the plurality of sensors, based on the elastic waves detected by each of the plurality of sensors. The damage detection unit detects damage to the object under inspection based on the spatiotemporal distribution. [Brief explanation of the drawing]

[0009] [Figure 1] Conceptual diagram of a Lamb wave. [Figure 2] A diagram showing an example of velocity dispersion characteristics calculation. [Figure 3] A diagram showing an example of the system configuration of the damage detection system 100 in the first embodiment. [Figure 4]Diagram showing a configuration example of the damage detection device 50 in the first embodiment. [Figure 5] Diagram showing an example of the spatio-temporal distribution generated by the generation unit 512 in the first embodiment. [Figure 6] Diagram showing an example of the spatio-temporal distribution interpolated by the interpolation unit in the first embodiment. [Figure 7] Flowchart showing the processing flow performed by the damage detection system in the first embodiment. [Figure 8] Diagram showing an example of a verification experiment. [Figure 9] Comparison diagram of the spatio-temporal distribution obtained from a normal test specimen and the spatio-temporal distribution obtained from a damaged test specimen. [Figure 10] Diagram showing an example of the system configuration of the damage detection system in the second embodiment. [Figure 11] Diagram showing an example of the arrangement of a plurality of sensors in the third embodiment.

Mode for Carrying Out the Invention

[0010] Hereinafter, the damage detection system, damage detection device, and damage detection method according to the embodiments will be described with reference to the drawings. <00…… (Theoretical Background of Lamb Waves) First, before explaining the content of the embodiment, guided waves will be explained. Here, plate waves (Lamb waves), which are typical guided waves, will be explained. Consider the governing equation in elastodynamics. The displacement vector u is expressed as in the following equation (1) based on Helmholtz's theorem.

[0012]

Equation

[0013] The equation of motion is expressed in the form of the following equation (2) using material-specific constants (Lame constants λ, μ, density ρ).

[0014]

Equation

[0015] Based on equation (2) above, two independent wave equations can be derived as shown in equation (3) below.

[0016]

number

[0017] The propagation velocities in equation (3) above are given by equation (4) below, and each is a longitudinal wave (C L ), transverse wave (C T ) represents.

[0018]

number

[0019] Furthermore, in the case of thin waves, longitudinal and transverse waves are mutually converted at the boundary, and a composite wave satisfying predetermined phase conditions is observed as a traveling wave of a certain wave number, giving rise to a mode called a Lamb wave. Figure 1 is a conceptual diagram of a Lamb wave. When the plate thickness of the material shown in Figure 1 is set to infinite, it becomes a Rayleigh wave dispersion relation.

[0020] When the boundary conditions are set with thickness direction Z = ±h / 2, the frequency ω and wavenumber k satisfy the relationship shown in equation (5) below, which is known as the Rayleigh-Lamb equation.

[0021]

number

[0022] Lamb wave phase velocity V p , and group velocity V g This can be determined based on the velocity dispersion relation shown in equation (6) below.

[0023]

number

[0024] From this, we have frequency f and velocity V p ,V g The relationship between these can be determined. That is, the speed and wavelength of the Lamb wave can be determined based on material-specific values, shape parameters (in this case, thickness), and frequency. Figure 2 shows an example of calculating velocity dispersion characteristics. In Figure 2, an example of calculating the velocity dispersion characteristics with respect to frequency is shown for aluminum plates (plate thicknesses of 3 mm and 5 mm) as the material. In Figure 2, only the S0 mode, which is the lowest-order mode that vibrates symmetrically in the thickness direction of the plate, and the A0 mode, which is the lowest-order mode that vibrates asymmetrically with respect to the thickness direction of the plate, are shown. In Figure 2, the horizontal axis is frequency and the vertical axis is propagation speed.

[0025] As shown in Figure 2, it can be seen that the propagation speed changes as the thickness changes. Therefore, by observing the propagation process of elastic waves over time, it becomes possible to detect material thinning, foreign matter, etc. In this embodiment, based on the elastic waves generated in the object under inspection, time-series data of the propagation process of elastic waves is observed to detect damage in the object under inspection. The following describes the specific method with reference to an embodiment.

[0026] (First embodiment) Figure 3 shows an example of the system configuration of the damage detection system 100 in the first embodiment. The damage detection system 100 is a system for detecting damage to an object M to be inspected. The object M to be inspected is, for example, a plate-shaped member or a cylindrical member such as a pipe. The plate-shaped member may be a structure such as a bridge. The structure may be any structure that generates elastic waves due to internal deterioration such as the occurrence or propagation of cracks, or due to external impacts (e.g., rain, artificial rain). Note that bridges are not limited to structures built over rivers or valleys, but also include various structures built above the ground (e.g., elevated highway bridges).

[0027] Furthermore, damage in the damage detection system 100 may include, for example, thinning or delamination, but may also include cracks, cavities, and soil formation, as long as the damage occurs inside the structure. In addition, damage detection in the damage detection system 100 includes not only the presence or absence of damage, but also the identification of the size of the damage and the location where the damage occurred.

[0028] The damage detection system 100 comprises an impact application unit 10, a sensor 20, an amplifier 25, an ADC 30, a BPF 40, and a damage detection device 50. The sensor 20 and the amplifier 25, the amplifier 25 and the ADC 30, the ADC 30 and the BPF 40, and the BPF 40 and the damage detection device 50 are connected by wires. A noise filter circuit to remove noise may be provided between the sensor 20 and the ADC 30 as needed. The number of sensors 20, amplifiers 25, and BPF 40 are equal.

[0029] Figure 3 shows a case with eight sensors 20, amplifiers 25, and BPF40, but it is sufficient to have three or more sensors 20, amplifiers 25, and BPF40. In the following explanation, we will assume that there are eight sensors 20, amplifiers 25, and BPF40. Multiple sensors 20-1 to 20-8 are installed at intervals based on wavelengths determined based on the material of the object to be inspected M and the shape of the object to be inspected M (e.g., thickness). Based on the explanation in Figure 2, the velocity and wavelength of elastic waves can be determined based on material-specific values ​​of the object to be inspected M, parameters of the shape of the object to be inspected M, and frequency.

[0030] Therefore, multiple sensors 20-1 to 20-8 are installed as a one-dimensional sensor array at predetermined intervals based on the wavelength determined as described above. For example, if the determined wavelength is wavelength λ and the axis direction of the object to be inspected is x, then the multiple sensors 20-1 to 20-8 are arranged at predetermined intervals along the x-axis such that the interval d between adjacent sensors 20 is less than half the wavelength λ (λ / 2), and the sum of the intervals in the x-axis direction (Σd) is greater than the wavelength λ. In the example shown in Figure 3, the interval between sensor 20-i (where i is an integer greater than or equal to 1) and sensor 20-(i+1) satisfies the relationship d < λ / 2. For example, d is between 5 mm and 1000 mm.

[0031] The impact-applying unit 10 applies an impact to the object M being inspected. This causes the impact-applying unit 10 to generate elastic waves in the object M being inspected. Applying an impact to the object M being inspected means, in other words, exciting elastic waves. The method by which the impact-applying unit 10 applies the impact (means for exciting elastic waves) may include, for example, the pencil lead break (PLB) method, steel ball impact, water droplet spraying, or pulse excitation using a piezoelectric element.

[0032] Furthermore, as the elastic waves generated at the object M under inspection, elastic waves (AE waves) that occur when the object M under inspection is performing its normal function can be used. For example, if the object M under inspection is a mechanical device, the vibrations that occur when the mechanical device is operating may be used as the elastic waves (AE waves). For example, if the object M under inspection is a pipe, the minute vibrations that occur when droplets flowing through the pipe collide with the pipe may be used as the elastic waves (AE waves). Note that if elastic waves (AE waves) that occur when the object M under inspection is performing its normal function are used as the elastic waves generated at the object M under inspection, the damage detection system 100 does not need to be equipped with an impact application unit 10.

[0033] Sensor 20 detects elastic waves generated in the object M being inspected. For example, sensor 20 detects elastic waves by contact or non-contact. When sensor 20 detects elastic waves by contact, sensor 20 is bonded to the surface of the object M being inspected for damage detection using an adhesive such as grease or couplant. In this case, multiple sensors 20-1 to 20-8 are installed in a line along the x-axis of the object M being inspected such that the distance between adjacent sensors 20 satisfies the relationship d < λ / 2. The distance between adjacent sensors 20 is, for example, the distance between the centers of the bonded surfaces of the sensors 20.

[0034] The surface of the object M on which the sensor 20 is installed may be the surface on which the elastic wave is excited, a different surface (for example, the surface opposite to the surface on which the elastic wave is excited), or a side surface of the object M. The same effect can be obtained regardless of which surface of the object M the sensor 20 is installed on. The sensor 20 has a piezoelectric element. The sensor 20 converts the detected elastic wave into an analog electrical signal (hereinafter referred to as "analog signal") and outputs it to the amplifier 25.

[0035] The sensor 20 uses, for example, a piezoelectric element with sensitivity in the range of 10 kHz to 1 MHz. The sensor 20 can be of any type, such as a piezoresistive type utilizing the piezoresistive effect, a capacitive type utilizing a change in capacitance, or a piezoelectric type utilizing the piezoelectric effect. The piezoelectric sensor 20 has the advantage of high flexibility in sensor installation and low power consumption because it does not require an external power supply. Furthermore, the piezoelectric sensor 20 is preferable because it can be excited by applying voltage, allowing for not only sensing but also calibration of the sensor itself. The sensor 20 is, for example, an AE sensor.

[0036] Note that sensor 20 is not limited to an AE sensor; other sensors may be used. For example, an acceleration sensor may be used instead of sensor 20. In this case as well, multiple acceleration sensors must be installed at the same intervals as sensor 20. The acceleration sensor detects elastic waves generated inside the object M being inspected. The acceleration sensor then converts the detected elastic waves into an analog signal by performing the same processing as sensor 20.

[0037] Furthermore, a sensor capable of detecting elastic waves non-contact may be used as sensor 20. Other sensors capable of detecting elastic waves non-contact may include, for example, a microphone. A MEMS (Micro Electro Mechanical Systems) type piezoelectric sensor may also be used as sensor 20. A MEMS type piezoelectric sensor is a sensor formed from a cantilever-type piezoelectric thin film. MEMS type piezoelectric sensors offer significant advantages in terms of cost reduction and miniaturization when constructing an array. MEMS type piezoelectric sensors with a cantilever structure are preferable because they can improve sensitivity to minute vibrations.

[0038] The amplifier 25 amplifies the amplitude of the analog signal output from the sensor 20. The amplifier 25 outputs the amplified analog signal to the ADC 30.

[0039] The ADC30 receives the analog signal output from the amplifier 25 as input. The ADC30 is an analog-to-digital converter that performs analog-to-digital conversion on the input analog signal. In this way, the ADC30 converts the input analog signal into a digital signal. In this process, the ADC30 must perform sampling with a sampling period that is sufficiently short (for example, 20 times or more is recommended) relative to the propagation speed and phase speed of elastic waves. The ADC30 outputs the converted digital signal to the BPF40.

[0040] The BPF40 is a bandpass filter that extracts signals in a specific frequency band from the digital signal output from the ADC30. The frequency band extracted by the BPF40 is variable. The BPF40 can extract one or more signals in any frequency band. In the following explanation, as an example, the BPF40 will be described as extracting signals in a single frequency band. The BPF40 is one embodiment of the extraction unit.

[0041] The damage detection device 50 detects damage to the object M under inspection based on signals in a specific frequency band extracted by each BPF 40. The damage detection device 50 may include at least one of the amplifier 25, ADC 30, and BPF 40.

[0042] (Configuration of the damage detection device 50) Next, the specific configuration of the damage detection device 50 will be described. Figure 4 is a diagram showing an example of the configuration of the damage detection device 50 in the first embodiment. The damage detection device 50 comprises a control unit 51 and a storage unit 52. The control unit 51 controls the entire damage detection device 50. The control unit 51 is configured using a processor such as a CPU (Central Processing Unit) and memory. By executing a program, the control unit 51 functions as an acquisition unit 511, a generation unit 512, an interpolation unit 513, and a damage detection unit 514.

[0043] Some or all of the functional units of the acquisition unit 511, generation unit 512, interpolation unit 513, and damage detection unit 514 may be implemented by hardware such as an ASIC (Application Specific Integrated Circuit), PLD (Programmable Logic Device), or FPGA, or by the cooperation of software and hardware. The program may be recorded on a computer-readable recording medium. Computer-readable recording media include non-temporary storage media such as portable media like flexible disks, magneto-optical disks, ROMs (Read Only Memory), and CD-ROMs, and storage devices such as hard disks built into computer systems. The program may be transmitted via a telecommunications line.

[0044] Some of the functions of the acquisition unit 511, generation unit 512, interpolation unit 513, and damage detection unit 514 do not need to be pre-installed in the damage detection device 50, and may be realized by installing additional application programs in the damage detection device 50.

[0045] The acquisition unit 511 acquires signals in a specific frequency band extracted by each BPF 40. The acquisition unit 511 may store the signals in a specific frequency band acquired for each BPF 40 in the storage unit 52. If the BPF 40 extracts signals in multiple frequency bands, the BPF 40 may store them together in the storage unit 52 for each frequency band.

[0046] The generation unit 512 generates a spatiotemporal distribution based on signals of specific frequency bands extracted by each BPF 40 acquired by the acquisition unit 511. The spatiotemporal distribution is a distribution that shows the signal intensity for each detection time of the elastic wave detected by each sensor 20, with the installation position of each of the multiple sensors 20 on the horizontal axis and the detection time of the elastic wave on the vertical axis. In other words, the spatiotemporal distribution is a distribution that shows the signal intensity of the elastic wave at different times for each elastic wave detected by each of the multiple sensors 20. Here, the signal intensity of the elastic wave is the amplitude of the elastic wave. Information indicating the installation position of each of the multiple sensors 20 is input to the damage detection device 50 after the sensors 20 are installed.

[0047] The interpolation unit 513 interpolates the signal strengths between adjacent sensors 20 based on information indicating the installation position of each of the multiple sensors 20 and the signal strength of signals in a specific frequency band extracted by each BPF 40. As a result, the interpolation unit 513 generates a spatiotemporal distribution with interpolated signal strengths between adjacent sensors 20.

[0048] As described above, the spatiotemporal distribution generated by the generation unit 512 includes only data representing the signal intensity of elastic waves corresponding to the installation position of each of the multiple sensors 20. In this case, data corresponding to the positions between adjacent sensors 20 is missing. Therefore, the interpolation unit 513 estimates the signal intensity between adjacent sensors 20 based on information indicating the installation position of each of the multiple sensors 20 and the signal intensity of signals in a specific frequency band extracted by each BPF 40.

[0049] Using sensors 20-1 and 20-2 as an example, the interpolation unit 513 interpolates the signal intensity (amplitude) at a position between sensor 20-1 and sensor 20-2 based on the signal intensity (amplitude) of the elastic wave detected by sensor 20-1 and the signal intensity (amplitude) of the elastic wave detected by sensor 20-2. For example, at time t1, the interpolation unit 513 calculates the average value of the signal intensity (amplitude) of the elastic wave detected by sensor 20-1 and the signal intensity (amplitude) of the elastic wave detected by sensor 20-2 as an estimated value of the signal intensity (amplitude) at a position midway between sensor 20-1 and sensor 20-2 at time t1. Alternatively, the interpolation unit 513 may interpolate the values ​​between sensor 20-1 and sensor 20-2 using linear interpolation. Furthermore, the interpolation unit 513 may interpolate using polynomial approximation with the value from sensor 20-3. For example, spline interpolation can be used as the polynomial approximation. The interpolation unit 513 then assigns the calculated estimated signal intensity (amplitude) to the position between sensor 20-1 and sensor 20-2 at time t1 of the spatiotemporal distribution. The interpolation unit 513 performs this process for each time point to interpolate the signal intensity (amplitude) at the position between sensor 20-1 and sensor 20-2 at each time point.

[0050] Here, we have described the calculation of the estimated value at the position between sensor 20-1 and sensor 20-2, but the interpolation unit 513 performs the above process for all adjacent sensors. As a result, the interpolation unit 513 interpolates the spatiotemporal distribution generated by the generation unit 512. The interpolation unit 513 then generates the interpolated spatiotemporal distribution. When generating the interpolated spatiotemporal distribution, the interpolation unit 513 represents it in different ways depending on the signal intensity (amplitude) value. For example, the interpolation unit 513 generates the interpolated spatiotemporal distribution by representing the region where the signal intensity (amplitude) is positive in blue, the region where the signal intensity (amplitude) is negative in red, and the region where the signal intensity (amplitude) is zero in white. This is just one example; the interpolation unit 513 can represent it in different ways depending on the signal intensity (amplitude) value.

[0051] The damage detection unit 514 detects damage to the object under inspection M based on the interpolated spatiotemporal distribution generated by the interpolation unit 513. For example, the damage detection unit 514 detects damage to the object under inspection M based on the change in signal intensity, the velocity estimate, or the information indicating reflection shown in the interpolated spatiotemporal distribution. The method of damage detection performed by the damage detection unit 514 will be described later.

[0052] Figure 5 shows an example of a spatiotemporal distribution generated by the generation unit 512 in the first embodiment. As shown in Figure 5, the spatiotemporal distribution generated by the generation unit 512 includes only data at the locations where the multiple sensors 20 are installed (signal intensity of elastic waves detected by each sensor 20). For example, data d1 in Figure 5 includes data on the signal intensity of elastic waves detected by sensor 20-1. Similarly, data d2 to d8 in Figure 5 include data on the signal intensity of elastic waves detected by sensors 20-2 to 20-8, respectively. Thus, it can be seen that the spatiotemporal distribution generated by the generation unit 512 lacks data at the locations between adjacent sensors.

[0053] Figure 6 shows an example of a spatiotemporal distribution interpolated by the interpolation unit 513 in the first embodiment. The interpolated spatiotemporal distribution shown in Figure 6 is the spatiotemporal distribution obtained when an elastic wave is excited near sensor 20-1 in the sensor arrangement shown in Figure 3. When an elastic wave is excited near sensor 20-1, the elastic wave first reaches sensor 20-1. Subsequently, the elastic wave reaches sensors 20-2, 20-3, 20-4, 20-5, 20-6, 20-7, and 20-8 in that order. Therefore, sensor 20-(i+1) detects the elastic wave later than sensor 20-i. As a result, the detection time of the elastic wave is shifted, as shown in the data d1 to d8 in Figure 6. Note that Figure 6 shows the spatiotemporal distribution obtained in a normal state where no damage has occurred inside the object M being inspected within the section where sensors 20-1 to 208 are arranged. In the spatiotemporal distribution shown in Figure 6, the horizontal axis represents the position of each sensor 20, with sensor 20-1, the closest to the elastic wave source, set to 0, and the vertical axis represents the time since the elastic wave was generated. The horizontal axis can be interpreted as the distance from sensor 20-1.

[0054] Since elastic waves are waves that vibrate and propagate to each sensor 20, their amplitude alternates between positive and negative values. Therefore, as shown in Figure 6, the spatiotemporal distribution shows alternating positive and negative signal intensity. When the object M under inspection is in a normal state, no disturbance occurs in the equiphase surface P. Here, disturbance means that a different phase occurs at a certain position within the same equiphase surface P, or that the signal level decreases at a certain position within the same equiphase surface P (for example, it is more than a threshold lower than the signal intensity at other positions).

[0055] On the other hand, if damage occurs inside the object M being inspected, it is expected that elastic waves will not propagate to some of the sensors 20 due to the damage, or even if they do propagate, the signal strength will be reduced. As a result, disturbances occur in the equiphase surface P. Therefore, the damage detection unit 514 detects that there is damage to the object M being inspected when disturbances occur in the signal strength shown by the spatiotemporal distribution. Furthermore, there is a high probability that some kind of damage has occurred at the location where the disturbances in signal strength occur. Therefore, the damage detection unit 514 detects the location where the disturbances in signal strength occur as the location of damage. In this way, the damage detection unit 514 detects damage to the object M being inspected based on changes in signal strength shown by the spatiotemporal distribution.

[0056] (Damage detection based on velocity estimates) The velocity of elastic waves can be estimated based on the slope of the equiphase surface P in the spatiotemporal distribution shown in Figure 6. As the velocity of elastic waves decreases, the slope of the equiphase surface P increases. That is, when the velocity of elastic waves is high, the slope of the equiphase surface P decreases (becomes nearly horizontal), and when the velocity of elastic waves is low, the slope of the equiphase surface P increases (becomes nearly vertical). Therefore, the damage detection unit 514 can estimate the velocity of elastic waves based on the slope of the equiphase surface P and detect damage to the object under inspection M based on the estimated velocity value.

[0057] Generally, the propagation speed of waves within a material is determined by the material itself. Furthermore, if damage occurs in part or throughout the interior of the object M that obstructs the propagation of elastic waves, the damage will slow down the propagation speed. Therefore, the damage detection unit 514 may detect damage to the object M by comparing the propagation speed (threshold) determined by the material of the object M with an estimated velocity value. For example, the damage detection unit 514 determines that there is no damage to the object M if the estimated velocity value is within the propagation speed range determined by the material of the object M. The damage detection unit 514 determines that there is damage to the object M if the estimated velocity value is outside the propagation speed range determined by the material of the object M. In this way, the damage detection unit 514 detects damage to the object M based on the estimated velocity value. It is also possible that an anomaly occurs where the propagation speed increases due to the adhesion of a substance with a faster propagation speed than the substance being inspected. In this case, the damage detection unit 514 can also determine that there is an abnormality in the object M if the estimated velocity is greater than or equal to the propagation velocity determined by the material of the object M being inspected.

[0058] (Damage detection based on information indicating reflection) If there are voids inside the object M being inspected, elastic waves may be reflected by these voids. In this case, the spatiotemporal distribution generated will exhibit reflection characteristics. Reflection characteristics refer to a state where the phase plane in the spatiotemporal distribution does not appear as a straight line, but rather as if it is refracted in the opposite direction. The damage detection unit 514 detects damage to the object M being inspected when reflection characteristics are observed in the spatiotemporal distribution. Furthermore, there is a high probability that some kind of damage has occurred at the location where the reflection characteristics were observed. Therefore, the damage detection unit 514 detects the location where the reflection characteristics were observed as the location of damage. In this way, the damage detection unit 514 detects damage to the object M being inspected based on the information indicating reflection shown in the spatiotemporal distribution.

[0059] The damage detection unit 514 may also detect damage to the object M being inspected by identifying the following characteristics of the spatiotemporal distribution as disturbances: Damage to the object M can be detected if there are discontinuities in the equiphase surface in the direction in which the horizontal axis (position) and vertical axis (time) values ​​increase (upper right direction in Figure 6). Damage to the object M can be detected if the signal detected at each sensor installation position is discontinuous with the signal at an adjacent sensor installation position. Damage to the object M can be detected if there are locations where the propagation speed (mm / s) is 0. Damage to the object M can be detected if there are locations where the slope of the equiphase surface (s / mm) is smaller than a predetermined negative value (threshold).

[0060] (operation) Next, the operation performed by the damage detection system 100 will be described. Figure 7 is a flowchart showing the processing flow of the damage detection system 100 in the first embodiment. The impact application unit 10 applies an impact to the object to be inspected M near some of the sensors 20 among the multiple sensors 20 in response to user operation (step S101). From the viewpoint of ease of measurement, it is desirable to apply an impact to the object to be inspected M near sensor 20-1. For example, the location where the impact is applied is closer to sensor 20-1 than to sensor 20-2. For example, sensor 20-1 is located between the location where the impact is applied and sensor 20-2. Elastic waves are generated by the impact applied by the impact application unit 10.

[0061] Each sensor 20 detects the generated elastic wave. For example, sensor 20-i detects the elastic wave. Sensor 20-(i+1) detects the elastic wave a predetermined time α (= predetermined interval d / elastic wave velocity) after detection by sensor 20-i. Each sensor 20 converts the detected elastic wave into an analog signal and outputs it to the connected amplifier 25. Each amplifier 25 amplifies the signal strength of the analog signal output from the connected sensor 20 and outputs it to the ADC 30. The ADC 30 samples the analog signal output from each amplifier 25 and converts it into a digital signal (step S102). The ADC 30 outputs each converted digital signal to each BPF 40. For example, the ADC 30 converts the analog signal output from amplifier 25-i into a digital signal and outputs it to BPF 40-i.

[0062] Each BPF40 filters the digital signal output from the ADC30 (step S103). This allows each BPF40 to extract signals in a specific frequency band from the digital signal. The signals in the specific frequency band extracted by each BPF40 are input to the damage detection device 50. The generation unit 512 of the damage detection device 50 generates a spatiotemporal distribution using the signals in the specific frequency band extracted by each input BPF40 (step S104). The generation unit 512 outputs the generated spatiotemporal distribution to the interpolation unit 513. The interpolation unit 513 interpolates the spatiotemporal distribution output from the generation unit 512 (step S105). Specifically, the interpolation unit 513 interpolates the signal intensity at the positions between adjacent sensors. This allows the interpolation unit 513 to generate an interpolated spatiotemporal distribution. The interpolation unit 513 outputs the generated interpolated spatiotemporal distribution to the damage detection device 514.

[0063] The damage detection unit 514 detects damage to the object under inspection M based on the interpolated spatiotemporal distribution output from the interpolation unit 513 (step S106). The damage detection unit 514 can detect damage to the object under inspection M based on any of the above-mentioned changes in signal intensity, velocity estimates, or information indicating reflection. For example, when detecting damage to the object under inspection M based on changes in signal intensity, the damage detection unit 514 first extracts equiphase surfaces in the interpolated spatiotemporal distribution. The damage detection unit 514 determines whether or not there is a disturbance in any of the extracted equiphase surfaces. If there is no disturbance in any of the extracted equiphase surfaces, the damage detection unit 514 determines that there is no damage to the object under inspection M. On the other hand, if there is a disturbance in any of the extracted equiphase surfaces, the damage detection unit 514 determines that there is damage to the object under inspection M.

[0064] The damage detection unit 514 outputs a determination result (step S107). For example, the damage detection unit 514 may transmit the determination result to an external device, or it may display the determination result on a display device. The determination result output by the damage detection unit 514 may be information indicating the presence or absence of damage, information indicating the location of the damage, or an interpolated spatiotemporal distribution.

[0065] (Example of experiment) Next, we will explain the results of a demonstration experiment based on the above configuration. In the demonstration experiment, eight sensors 20-1 to 20-8 were installed in a straight line at predetermined intervals d (25 mm intervals) on the first surface of a 5 mm thick aluminum plate (object to be inspected M), and a simulated foreign object O was brought into contact with the second surface, which is the opposite surface of the first surface of the aluminum plate (object to be inspected M). The simulated foreign object O was placed between sensor 20-4 and sensor 20-5 (for example, between 75 mm and 100 mm) as shown in Figure 8. Then, elastic waves were generated by bringing a steel ball (impact-applying part 10) into contact with the aluminum plate (object to be inspected M).

[0066] Figure 9(A) shows the spatiotemporal distribution obtained from a normal test specimen, and Figure 9(B) shows the spatiotemporal distribution obtained from an aluminum plate (inspection target M) in contact with a simulated foreign object. A normal test specimen is an inspection target M used for testing that has no internal damage. As shown by circle inf1 in Figure 9(B), it can be seen that disturbance in elastic wave propagation due to the simulated foreign object O is detected at a position of 75mm to 100mm. The damage detection unit 514 detects damage based on this disturbance in elastic wave propagation.

[0067] The damage detection system 100 configured as described above comprises a plurality of sensors 20 arranged at predetermined intervals to detect elastic waves generated in the object to be inspected M, a generation unit 512 that generates a spatiotemporal distribution based on the elastic waves detected by each of the plurality of sensors 20, and a damage detection unit 514 that detects damage to the object to be inspected M based on the spatiotemporal distribution.

[0068] This allows for the detection of damage within the range where multiple sensors 20 are installed with a single elastic wave excitation. Therefore, as in conventional methods, there is no need to repeatedly perform surface wave excitation and measurement while moving points in order to inspect the properties of the guided wave over a surface. Furthermore, since there is no need to detect guided waves from different excitation trials at each point, damage can be detected on the inspected object M, which is in the same state at the start and end of the measurement. As a result, damage can be detected in a short amount of time.

[0069] Furthermore, the damage detection unit 514 detects damage to the object M being inspected based on either the change in signal intensity shown in the spatiotemporal distribution, the velocity estimate, or information indicating reflection. In this way, the damage detection system 100 detects damage to the object M being inspected using information shown in the spatiotemporal distribution obtained based on the elastic waves detected by each sensor 20. Therefore, it is possible to easily identify where the damage has occurred. As a result, it becomes possible to identify not only whether there is damage, but also the location where the damage has occurred in a short amount of time.

[0070] Furthermore, the damage detection system 100 includes an interpolation unit 513 that generates a spatiotemporal distribution by interpolating the signal strengths between adjacent sensors based on information indicating the installation locations of multiple sensors 20 and the signal strength of elastic waves detected by each of the multiple sensors 20. This makes it possible to detect damage even in locations where no sensors 20 are installed. Consequently, the number of sensors 20 required for damage detection can be reduced. As a result, the overall cost of the damage detection system 100 can also be reduced.

[0071] (Second embodiment) In the first embodiment, a plate-shaped member was used as an example of the object to be inspected. In the second embodiment, a cylindrical member such as a pipe was used as an example of the object to be inspected. In particular, in the second embodiment, thinning of the pipe wall was used as an example of damage to the object to be inspected.

[0072] Figure 10 shows an example of the system configuration of the damage detection system 100a in the second embodiment. In Figure 10, the amplifier 25, ADC 30, BPF 40, and damage detection device 50 are omitted from the description, but the damage detection system 100a also includes the amplifier 25, ADC 30, BPF 40, and damage detection device 50. In Figure 10, it is the same as the damage detection system 100 except that the object M to be inspected on which the multiple sensors 20 are installed is different, and that it does not have an impact application unit 10. The differences from the first embodiment will be explained below.

[0073] In the elbow section of a pipe, which is the object of inspection M as shown in Figure 10, damage called wall thinning may occur, where the inner wall is eroded by the impact of droplets D passing through the pipe. Figure 10 shows an example where wall thinning has occurred in region R1 of the pipe. In the damage detection system 100a of this embodiment, such wall thinning can also be detected based on elastic waves obtained by multiple sensors 20 placed on the outside of the pipe. In the damage detection system 100a, multiple sensors 20-1 to 20-5 are arranged in a one-dimensional array along the longitudinal direction of the elbow section of the pipe at predetermined intervals. Note that Figure 10 shows an example in which multiple sensors 20-1 to 20-5 are arranged in a one-dimensional array along the longitudinal direction of the elbow section of the pipe, but the number of sensors 20 can be three or more. In Figure 10, multiple sensors 20-1 to 20-5 are located on the same plane.

[0074] Next, it is necessary to excite elastic waves in the pipe. By using the elastic waves as guide waves that propagate through the pipe, they can be propagated over long distances with reduced attenuation, which is suitable for this inspection. In this case, elastic waves (AE waves) generated when the object under inspection M performs its normal function may be used. In particular, minute vibrations generated when droplets D flowing through the pipe collide with the pipe may be used. Thus, in the damage detection system 100a, unlike the first embodiment, droplets D can be used as a means of exciting elastic waves.

[0075] Multiple sensors 20-1 to 20-5 detect minute vibrations that occur when a droplet D collides with the pipe. Multiple sensors 20-1 to 20-5 convert the detected minute vibrations into analog signals and output them to the amplifier 25. The subsequent processing is the same as in the first embodiment. Here, as shown in Figure 10, even when multiple sensors 20 are arranged at the elbow of the pipe, the spatiotemporal distribution can be generated in the same way as in the first embodiment by representing the horizontal axis (position) in the spatiotemporal distribution with a one-dimensional coordinate in the longitudinal direction of the pipe.

[0076] Then, the interpolation unit 513 interpolates the spatiotemporal distribution. The damage detection unit 514 uses the interpolated spatiotemporal distribution to detect damage to the object M being inspected. For example, the damage detection unit 514 detects damage to the object M being inspected when there is a disturbance in the signal intensity shown in the spatiotemporal distribution. Furthermore, the damage detection unit 514 detects the location where the disturbance in signal intensity occurs as the location where thinning has occurred.

[0077] In this configuration, there is a possibility that elastic waves propagating from the top of the piping and elastic waves generated by droplets colliding with the inner wall may coexist. In this case, by observing the arrival time of the elastic waves at sensors 20-1 to 20-5 and focusing only on elastic waves propagating from a predetermined direction, for example, the direction of sensor 20-1, it is possible to obtain clear inspection results with less wave mixing.

[0078] With the damage detection system 100a configured as described above, the same effects as in the first embodiment can be obtained even when applied to a cylindrical member instead of a plate-shaped member.

[0079] (Third embodiment) The first and second embodiments describe a configuration in which multiple sensors are arranged on an object to be inspected as a one-dimensional sensor array. In contrast, multiple sensors can also be arranged on a two-dimensional plane at predetermined intervals and used as a two-dimensional sensor array. Therefore, the third embodiment describes a configuration in which multiple sensors are arranged on an object to be inspected as a two-dimensional sensor array.

[0080] The system configuration and the configuration of each device in the third embodiment are the same as in the first embodiment. In the damage detection system 100 of the third embodiment, the arrangement of the sensors 20 and the processing of the functional parts of the damage detection device 50 differ from those of the first embodiment. The differences from the first embodiment will be explained below.

[0081] Figure 11 shows an example of the arrangement of multiple sensors 20 in the third embodiment. As shown in Figure 12, the multiple sensors 20 are arranged on a two-dimensional plane at predetermined intervals. For example, if the first axis direction of the object to be inspected M is x, the multiple sensors 20 are arranged at predetermined intervals along the x-axis such that the interval d between adjacent sensors 20 is less than half the wavelength λ (λ / 2), and the sum of the intervals in the x-axis direction (Σd) is greater than the wavelength λ. Furthermore, if the second axis direction of the object to be inspected M is y, the multiple sensors 20 are arranged at predetermined intervals along the y-axis such that the interval d between adjacent sensors 20 is less than half the wavelength λ (λ / 2), and the sum of the intervals in the y-axis direction (Σd) is greater than the wavelength λ. In this way, the multiple sensors 20 are arranged such that the interval d between adjacent sensors 20 is less than half the wavelength λ (λ / 2), and the sum of the intervals in any axis direction (Σd) is greater than the wavelength λ. The predetermined intervals in the x-axis and the predetermined intervals in the y-axis may be the same or different. In this way, the multiple sensors 20 in the third embodiment are installed as a two-dimensional sensor array with respect to the object M to be inspected.

[0082] In this case, the spatiotemporal distribution generated by the generation unit 512 can be represented as a three-dimensional graph of xyt, where the two orthogonal axes on the installation plane of the sensor 20 are the x-axis and y-axis, and time is t. The interpolation unit 513 interpolates the three-dimensional spatiotemporal distribution generated by the generation unit 512. The interpolation method is the same as in the first embodiment. The damage detection unit 514 detects damage to the object to be inspected M based on the three-dimensional spatiotemporal distribution interpolated by the interpolation unit 513. The only difference is that the spatiotemporal distribution used by the damage detection unit 514 to detect damage is now three-dimensional; the specific method used by the damage detection unit 514 is the same as in the first embodiment.

[0083] With the damage detection system 100 in the third embodiment configured as described above, it is possible to obtain the same effects as in the first embodiment even when multiple sensors 20 are used as a two-dimensional sensor array.

[0084] (Modifications common to the first to third embodiments) In the embodiments described above, a configuration in which the BPF40 extracts a signal from one frequency band as a signal of a specific frequency band was explained as an example. The BPF40 may extract signals from two or more frequency bands as signals of a specific frequency band. Here, a configuration in which signals from two frequency bands are extracted will be described as an example. In this configuration, each BPF40 extracts a signal from a first frequency band (e.g., 10kHz) and a signal from a second frequency band (e.g., 5kHz) in the digital signal output from the ADC30. Here, the first frequency band is a high frequency band, and the second frequency band is a frequency band other than the high frequency band.

[0085] The generation unit 512 generates a first spatiotemporal distribution based on the signals of the first frequency band (e.g., 10 kHz) extracted by each BPF 40. Furthermore, the generation unit 512 generates a second spatiotemporal distribution based on the signals of the second frequency band (e.g., 5 kHz) extracted by each BPF 40.

[0086] The interpolation unit 513 uses the first spatiotemporal distribution to generate an interpolated first spatiotemporal distribution by interpolating the signal strengths between adjacent sensors 20 based on information indicating the installation position of each of the multiple sensors 20 and the signal strength of the first frequency band signals extracted by each BPF 40. Similarly, the interpolation unit 513 uses the second spatiotemporal distribution to generate an interpolated second spatiotemporal distribution by interpolating the signal strengths between adjacent sensors 20 based on information indicating the installation position of each of the multiple sensors 20 and the signal strength of the second frequency band signals extracted by each BPF 40.

[0087] The damage detection unit 514 detects damage to the object M being inspected based on the interpolated first spatiotemporal distribution and the interpolated second spatiotemporal distribution generated by the interpolation unit 513. For example, the damage detection unit 514 uses the interpolated first spatiotemporal distribution to identify the location where damage has occurred. Then, the damage detection unit 514 uses the interpolated second spatiotemporal distribution to identify the size of the damage at the identified location. In the high frequency band, the presence or absence of damage can be detected over a wide range. However, because the range is wide, even if the location of the damage can be identified, the size of the damage cannot be identified. Therefore, the damage detection unit 514 uses the interpolated second spatiotemporal distribution in a lower frequency band than the interpolated first spatiotemporal distribution to identify the size of the damage. This allows for more accurate damage detection.

[0088] According to at least one embodiment described above, damage can be detected in a short time by having a plurality of sensors 20 arranged at predetermined intervals to detect elastic waves generated in the object to be inspected, a generation unit 512 that generates a spatiotemporal distribution based on the elastic waves detected by each of the plurality of sensors 20, and a damage detection unit 514 that detects damage to the object to be inspected M based on the spatiotemporal distribution.

[0089] Some of the processing performed by the damage detection device 50 in the above-described embodiment may be implemented by a computer. In that case, a program for implementing this function may be recorded on a computer-readable recording medium, and the program recorded on this recording medium may be loaded into a computer system and executed. The term "computer system" here includes hardware such as an OS (Operating System) and peripheral devices. The term "computer-readable recording medium" refers to portable media such as flexible disks, magneto-optical disks, ROMs (Read Only Memory), CD-ROMs, and storage devices such as hard disks built into a computer system. Furthermore, the term "computer-readable recording medium" may also include those that dynamically hold programs for a short period of time, such as communication lines used when transmitting programs via networks such as the Internet or communication lines such as telephone lines, and those that hold programs for a certain period of time, such as volatile memory inside a computer system that acts as a server or client in such cases. The above-mentioned program may be for implementing some of the functions described above, or it may be a program that can implement the above-mentioned functions in combination with a program already recorded in the computer system, or it may be implemented using a programmable logic device such as an FPGA (Field Programmable Gate Array).

[0090] While several embodiments of the present invention have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of Symbols]

[0091] 10…Impact application unit, 20, 20-1~20-8…Sensors, 25, 25-1~25-8…Amplifiers, 30…ADC, 40, 40-1~40-8…BPF, 50…Damage detection device, 51…Control unit, 52…Storage unit, 511…Acquisition unit, 512…Generation unit, 513…Interpolation unit, 514…Damage detection unit, 100, 100a…Damage detection system

Claims

1. Multiple sensors are arranged at predetermined intervals to detect elastic waves generated in the object being inspected, A generation unit generates a spatiotemporal distribution representing the signal intensity of the elastic wave at different times for each of the elastic waves detected by each of the plurality of sensors, based on the elastic waves detected by each of the plurality of sensors. A damage detection unit detects damage to the object to be inspected based on the spatiotemporal distribution, A damage detection system equipped with the following features.

2. The damage detection unit, Damage to the object being inspected is detected based on the change in signal intensity shown in the spatiotemporal distribution, the velocity estimate, or the information indicating reflection. The damage detection system according to claim 1.

3. The damage detection unit, Damage to the object being inspected is detected when disturbances occur in the signal intensity shown in the aforementioned spatiotemporal distribution. The damage detection system according to claim 2.

4. The damage detection unit, The propagation speed of elastic waves is estimated in the aforementioned spatiotemporal distribution, and if the estimated propagation speed of elastic waves is below a threshold, damage to the object being inspected is detected. The damage detection system according to claim 2.

5. The damage detection unit, If reflection characteristics are observed in the aforementioned spatiotemporal distribution, damage to the object being inspected is detected. The damage detection system according to claim 2.

6. The system further comprises an interpolation unit that generates the spatiotemporal distribution by interpolating the signal strengths between adjacent sensors based on information indicating the installation positions of the plurality of sensors and the signal strength of the elastic wave detected by each of the plurality of sensors, The damage detection unit detects damage to the object being inspected based on the interpolated spatiotemporal distribution. A damage detection system according to any one of claims 1 to 5.

7. The system further comprises an extraction unit that extracts signals in a specific frequency band from the elastic waves detected by each of the multiple sensors, The generating unit is The spatiotemporal distribution is generated based on the extracted signals in the specific frequency band. A damage detection system according to any one of claims 1 to 5.

8. The extraction unit is Extract signals from at least two frequency bands, The generating unit is The spatiotemporal distribution is generated for each of the two or more frequency band signals that have been extracted. The damage detection unit, Based on the spatiotemporal distribution generated in each of the two or more frequency bands, damage to the object being inspected is detected. The damage detection system according to claim 7.

9. The aforementioned multiple observations The following are installed at predetermined intervals based on wavelengths determined from the material and shape of the object to be inspected: A damage detection system according to any one of claims 1 to 5.

10. The aforementioned multiple observations The sensors are arranged along one axis at predetermined intervals such that the distance between adjacent sensors is less than half the wavelength, and the sum of the axial distances is greater than the wavelength. The damage detection system according to claim 9.

11. The aforementioned multiple observations The sensors are arranged on a two-dimensional plane at predetermined intervals such that the distance between adjacent sensors is less than half the wavelength, and the sum of the intervals in any axial direction is greater than the wavelength. The damage detection system according to claim 9.

12. The aforementioned multiple observations It is a sensor formed from a cantilever-type piezoelectric thin film. A damage detection system according to any one of claims 1 to 5.

13. The aforementioned multiple observations The aforementioned sensor is a non-contact sensor for the object being inspected. A damage detection system according to any one of claims 1 to 5.

14. A generation unit generates a spatiotemporal distribution representing the signal intensity of each elastic wave at different times, based on the elastic waves detected by each of a plurality of sensors arranged at predetermined intervals and detecting elastic waves generated in the object under inspection, A damage detection unit detects damage to the object to be inspected based on the spatiotemporal distribution, A damage detection device equipped with the following features.

15. Multiple sensors are arranged at predetermined intervals to detect elastic waves generated in the object under inspection. Based on the elastic waves detected by each of these sensors, a spatiotemporal distribution is generated representing the signal intensity of the elastic waves at different times for each elastic wave detected by each of the multiple sensors. Based on the aforementioned spatiotemporal distribution, damage to the object to be inspected is detected. Damage detection method.

Images