Metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis

Abstract

The invention discloses a metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis, and aims to position defect positions and invert defect shapes by using metal plate surface wave echoes. The position of detection light of a laser vibration meter is determined on the surface of a fixed metal plate, and multiple groups of ultrasonic data are obtained by performing field scanning on the surface of the same side of the metal plate through a field scanning system. Performing wavelet transform on the obtained surface wave data to remove a low-frequency component in the signal, and performing Fourier transform on a time component to obtain a frequency spectrum; and then extracting signal data of a specific frequency and carrying out two-dimensional Fourier transform to obtain a wave number spectrum. And finally, filtering and inverse Fourier transform are carried out on the wave number spectrum to obtain a defect image, and defect positioning and defect shape imaging are realized.

Description

Technical Field

[0001] The invention relates to laser ultrasonic nondestructive testing, and in particular to a metal plate defect imaging method based on surface wave number filtering analysis. Background Art

[0002] With the further development of the world's industry, the service environment faced by metal structural materials has become more severe and the processing requirements have become more stringent. Surface and near-surface defects can easily cause permanent damage to structural parts under cyclic loads. Accumulated damage can cause defects to further expand until they break, causing fatigue damage. Therefore, timely and effective detection of surface and near-surface defects plays an important role in structural safety assessment.

[0003] The most significant advantage of non-destructive testing technology is that it can monitor the quality of materials at all stages of the material service cycle and even during operation, prevent the material from further deteriorating during service and correct it when possible. In surface defect detection, conventional non-destructive testing methods are magnetic particle testing, radiographic testing and ultrasonic testing. However, the detection object of magnetic particle testing can only be ferromagnetic materials, which is not suitable for detecting complex-shaped surfaces, and the detection depth generally does not exceed 2mm, the detection sensitivity is not high, and it may be adversely affected by the surface coating, making it difficult to distinguish the type and shape of defects; the rays used in radiographic testing have radiation, which is harmful to the human body, and the equipment is expensive, the detection cycle is long, and the cost is high in actual application; traditional ultrasonic testing technology generally uses ultrasonic probes to excite and receive ultrasonic signals, which requires coupling agents, may contaminate the surface of the material, and due to the limitations of the probe surface, it is not conducive to detecting complex-shaped surfaces. The excitation and reception of laser ultrasound does not require water immersion and coupling agents, is not limited by the surface shape, and is suitable for non-destructive testing in harsh environments such as high temperature, high pressure, and toxicity, which can expand the detection technology to occasions where ultrasonic technology cannot be used. By controlling the pulse time of the laser, extremely short ultrasonic pulses can be generated on the surface of the test piece, increasing the ultrasonic frequency and greatly improving the ability to detect tiny defects. Therefore, it is suitable for high-precision non-destructive testing.

[0004] Laser surface acoustic waves propagate along the surface contour of the material, can only penetrate a wavelength deep, and the intensity decays exponentially with depth. It is very sensitive to small defects and has unique advantages in detecting surface defects and sub-surface defects. Defect characterization methods can be divided into two types according to different imaging methods, namely static imaging and dynamic imaging. Static imaging has traditional B-scan and C-scan imaging methods, which have good positioning and imaging effects, but this method requires the laser excitation spot and the laser receiving spot to maintain a set distance and move synchronously for scanning. However, this method has a slow scanning efficiency, and the ultrasonic signal strength depends on the return quality of the receiving spot on the surface of the material. The movement of the receiving spot will cause the ultrasonic signal to be unstable. In dynamic imaging, one of the excitation spot and the receiving spot is fixed, and the other performs raster scanning. For the case where the receiving spot is fixed and the excitation spot scans the field, based on the principle of acoustic reciprocity, spatial interchange can be performed to achieve dynamic reconstruction of the wave field. However, the defect image obtained by this method relies on the dynamic propagation image of the wave field. The wave field image changes with time, and the defect edge area is not clear. It is impossible to perform quantitative analysis of the defects, and it only has a positioning effect. Other imaging methods such as the time-domain synthetic aperture method are easily interfered by a series of clutter such as boundary echoes, resulting in poor imaging effect. Summary of the invention

[0005] Purpose of the invention: In order to solve the shortcomings of the prior art, the present invention proposes a metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis. While ensuring efficiency, the position of the surface defect of the plate is detected and located and the defect shape is imaged.

[0006] Technical solution: The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis of the present invention comprises the following steps:

[0007] (1) Stably set up the metal plate, place and start a laser vibrometer on one side of the metal plate, and adjust the position of the laser vibrometer, the laser incident angle, and the energy output until a stable and high-amplitude detection light return signal is obtained;

[0008] (2) The pulsed laser emits a light beam into the galvanometer scanning system, which then reflects the incident laser onto the metal plate to be tested. The computer controls the galvanometer scanning system to scan the surface of the metal plate with excitation light to obtain spatial and temporal sampling matrix data;

[0009] (3) Perform wavelet transform on broadband surface wave signals to remove low-frequency components, then perform three-dimensional Fourier transform on the time component to convert the time domain signal into a frequency domain signal, select the central peak frequency as the characteristic frequency, and after determining the characteristic frequency, perform two-dimensional Fourier transform on the spatial component of the wave field signal to obtain the wave number image at the characteristic frequency;

[0010] (4) Calculate the theoretical wave number of the surface wave, remove the wave number outside the theoretical wave number curve, reduce the noise interference imaging effect, filter the wave number of the surface wave forward wave, retain the surface wave echo wave number, perform two-dimensional inverse Fourier transform on the remaining wave number, and obtain a single frequency wave number inversion defect image;

[0011] (5) The laser-excited ultrasonic wave is a broadband signal. Selecting a single frequency will cause information loss. Therefore, the wave field data of all the frequencies where the peaks are located are selected to repeat step (4). Finally, all the defect images obtained are normalized and added together to obtain the final defect image.

[0012] In the metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis, the expression of the three-dimensional Fourier transform in step (3) is as follows:

[0013]

[0014] Where f is the frequency, is the wave number, t is the time, For different directions, and is the signal amplitude;

[0015] In the metal plate defect imaging method based on laser ultrasound and surface wave wave number filtering analysis, the wave number calculation formula for calculating the theoretical wave number of the surface wave in step (4) is as follows:

[0016]

[0017]

[0018] Where f is the frequency; c is the ultrasonic wave velocity, which can be obtained by dividing the distance between the excitation point and the detection point by the propagation time; f 0 is the characteristic frequency selected in the spectrum; is the wave number; For different directions and positions; is the signal amplitude;

[0019] In the metal plate defect imaging method based on laser ultrasound and surface wave wave number filtering analysis, in the steps (1) and (2), the position of the laser vibrometer detection light is kept unchanged, and the galvanometer system is used to move the pulse laser excitation light position for scanning.

[0020] In the metal plate defect imaging method based on laser ultrasound and surface wave wavenumber filtering analysis, the step (4) only retains the surface wave echo wavenumber in the theoretical wavenumber curve and filters out other interfering wavenumbers. If there is no defect, there will be no echo wavenumber in the wavenumber spectrum, and no defect will appear after inversion.

[0021] In the metal plate defect imaging method based on laser ultrasound and surface wave wave number filtering analysis, the sampling matrix is ​​a sampling point matrix.

[0022] The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis comprises a device for implementing the method, which includes a nanosecond pulse laser, a galvanometer scanning system, a laser vibrometer, and a computer.

[0023] In the metal plate defect imaging method based on laser ultrasound and surface wave wave number filtering analysis, step (3) uses computer software Matlab to perform data processing and wave number imaging.

[0024] The metal plate defect imaging method based on laser ultrasound and surface wave wave number filtering analysis is characterized in that the characteristic frequency range of step (3) is 2MHz to 4MHz; the frequency of all wave peaks in step (5) is selected from 1MHz to 5MHz.

[0025] The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis is applied in detecting and / or locating surface defects of the plate.

[0026] Furthermore, a metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis, wherein the implementation device comprises a nanosecond pulse laser, a galvanometer scanning system, a laser vibrometer, and a computer, is characterized in that the method comprises the following steps:

[0027] (1) Stably set up the metal plate, place and start a laser vibrometer on one side of the metal plate, and adjust the position of the laser vibrometer, the laser incident angle, and the energy output until a stable and high-amplitude detection light return signal is obtained;

[0028] (2) The pulsed laser emits a light beam into the galvanometer scanning system, which then reflects the incident laser onto the metal plate to be tested. The galvanometer scanning system is controlled by a computer to scan the surface of the metal plate with an excitation light field to obtain matrix data of M×N rows and L columns, where M is the number of spatial sampling points in the x-direction, N is the number of spatial sampling points in the y-direction, and L is the number of temporal sampling points; the x-direction and y-direction are the horizontal and vertical coordinates of the plane of the metal plate.

[0029] (3) Perform wavelet transform on the broadband surface wave signal to remove the low-frequency component, then perform Fourier transform on the time component to convert the time domain signal into the frequency domain signal, select the central peak frequency as the characteristic frequency, and after determining the characteristic frequency, perform two-dimensional Fourier transform on the spatial component of the wave field signal to obtain the wave number image under the characteristic frequency. The expression of three-dimensional Fourier transform is as follows:

[0030]

[0031] Where f is the frequency, is the wave number, t is the time, For different directions, and is the signal amplitude;

[0032] (4) Calculate the theoretical wave number of the surface wave, remove the wave number outside the theoretical wave number curve, reduce the noise interference imaging effect, filter the wave number of the surface wave forward wave, retain the surface wave echo wave number, perform two-dimensional inverse Fourier transform on the remaining wave number, and obtain a single frequency wave number inversion defect image. The wave number calculation formula and two-dimensional inverse Fourier transform formula are as follows:

[0033]

[0034] Where c is the ultrasonic wave velocity, which can be obtained by dividing the distance between the excitation point and the detection point by the propagation time, and f 0 is a selected frequency in the spectrum;

[0035] (5) The laser-excited ultrasonic wave is a broadband signal. Selecting a single frequency will cause information loss. Therefore, the wave field data of all the frequencies where the peaks are located are selected to repeat step (4). Finally, all the defect images obtained are normalized and added together to obtain the final defect image.

[0036] In the steps (1) and (2), the position of the detection light of the laser vibrometer is kept unchanged, and the position of the excitation light of the pulsed laser is moved by the galvanometer system to scan the field.

[0037] The step (4) only leaves the surface wave echo wavenumber in the theoretical wavenumber curve and filters out other interfering wavenumbers. If there is no defect, there will be no echo wavenumber in the wavenumber spectrum, and no defect will appear after inversion.

[0038] The step (5) normalizes and adds the defect images obtained by inverting the wave field data of all the frequencies where the wave peaks are located to obtain the final defect image.

[0039] The algorithm principle of the present invention is as follows Figure 1 shown.

[0040] Beneficial effects: Compared with the prior art, the present invention has the following advantages: the present invention converts the ultrasonic signal from the time domain to the wave number domain through Fourier transform, filters all the sound waves except the reflected echo at the spatial level, and the broadband signal of laser ultrasound can provide frequencies with different sensitivities to defects for normalized synthesis of defect images, thereby realizing true non-destructive testing, giving full play to the advantages of broadband and high spatial resolution of laser ultrasound, and through the rapid scanning of the galvanometer system, reducing the average number of times in the signal acquisition process, improving the scanning efficiency, and combining with the Matlab program to achieve the purpose of efficient and high-quality imaging after scanning, it is expected to be used in practical industrial applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 Schematic diagram of the algorithm principle;

[0042] Figure 2 Schematic diagram of the defect location of the aluminum plate and the laser scanning path;

[0043] Figure 3 Time domain signal diagram after surface wavelet transform;

[0044] Figure 4 Theoretical wave number diagram of circular defect (a) and wave number diagram after filtering (b);

[0045] Figure 5 Circular defect image;

[0046] Figure 6 Theoretical wave number diagram of rectangular defect (a) and wave number diagram after filtering (b);

[0047] Figure 7 Rectangular defect image. DETAILED DESCRIPTION

[0048] The present invention will now be further described with reference to examples and drawings:

[0049] like Figure 2 As shown, a circular countersink with a diameter of 2 mm and a rectangular countersink with a length of 8 mm and a width of 2.5 mm are respectively processed on an aluminum plate with a size of 180 mm × 100 mm × 5 mm.

[0050] Example 1

[0051] (1) Fix the laser vibrometer on the translation table and connect it to the oscilloscope. Adjust the angle between the vibrometer and the aluminum plate to 90°. Turn on the laser vibrometer and continuously fine-tune the position of the vibrometer to make the signal amplitude received by the oscilloscope the highest and most stable, and ensure that the return light signal received by the vibrometer has the best quality. At this time, the detection spot is about 30 mm away from the defect and is fixed. After the vibrometer is pulled out of the oscilloscope, it is connected to the digital oscilloscope of the computer together with the synchronous trigger source of the pulse laser, with a sampling rate of 50 MHz and a sampling point number of 1000.

[0052] (2) Set the metal plate stably, adjust the distance between the pulse laser and the aluminum plate to 100 mm, and make the excitation light incident vertically at 90°. Set the pulse laser output energy to 4 mJ, adjust the galvanometer system so that the scanning area covers the surface defects of the aluminum plate, and the excitation spot focusing diameter is 2 mm. The galvanometer scanning area is as follows: Figure 2 As shown, the scanning area size is 18mm×12mm, the initial position of the excitation spot is at the upper left corner of the defect, and the surface wave signal is collected point by point with a step length of 0.3mm. In this example, a total of 2400 sets of data are taken to obtain a numerical matrix of 2400 rows and 3000 columns.

[0053] (3) Import the numerical matrix into Matlab and convert it into a 60×40×3000 three-dimensional matrix according to the zigzag scanning sequence. Fill 10 groups of zeros before and after the second dimension of the matrix to convert the numerical matrix into a 60×60×3000 numerical matrix. Use discrete wavelet transform to remove the low-frequency components in the signal and interpolate for smooth processing. Figure 3 After removing the low-frequency components, the third (time) dimension of the matrix is ​​Fourier transformed to convert the time domain signal into a frequency domain signal and determine the center frequency f of the signal. 0 The matrix is ​​then subjected to a two-dimensional (spatial) Fourier transform to obtain a wavenumber image at a frequency of 2 MHz.

[0054] (4) The theoretical wave number of the aluminum plate surface wave at this frequency is calculated, which is consistent with the actual measured wave number. Figure 4 As shown in (a), the wave numbers outside the theoretical wave numbers are removed to reduce noise interference. According to the law of symmetry of the amplitude in the wave number domain, the program can automatically identify the surface wave echo wave number and filter out the forward wave wave number. The image after filtering is as follows Figure 4 As shown in (b), the reflected wave number is subjected to a two-dimensional inverse Fourier transform to obtain a wave number inversion image in a single frequency domain.

[0055] (5) The signal excited by the laser is a broadband signal. In order to retain more information, the frequency signal corresponding to the peak with higher amplitude in step (3) is Fourier transformed, and step (4) is repeated. The wave number inversion images obtained at each single frequency are normalized and superimposed. The final circular defect image is as follows: Figure 5 shown.

[0056] Example 2

[0057] (1) Fix the laser vibrometer on the translation table and connect it to the oscilloscope. Adjust the angle between the vibrometer and the aluminum plate to 90°. Turn on the laser vibrometer and continuously fine-tune the position of the vibrometer to make the signal amplitude received by the oscilloscope the highest and most stable, and ensure that the return light signal received by the vibrometer has the best quality. At this time, the detection spot is about 20 mm away from the defect and is fixed. After the vibrometer is pulled out of the oscilloscope, it is connected to the digital oscilloscope of the computer together with the synchronous trigger source of the pulse laser, with a sampling rate of 50 MHz and a sampling point number of 1000.

[0058] (2) Set the metal plate stably, adjust the distance between the pulse laser and the aluminum plate to 100 mm, and make the excitation light incident vertically at 90°. Set the pulse laser output energy to 4 mJ, adjust the galvanometer system so that the scanning area covers the surface defects of the aluminum plate, and the excitation spot focusing diameter is 2 mm. The galvanometer scanning area is as follows: Figure 2As shown, the scanning area size is 24mm×12mm, the initial position of the excitation spot is at the upper left corner of the defect, and the surface wave signal is collected point by point with a step length of 0.3mm. In this example, a total of 3200 sets of data are taken to obtain a numerical matrix with 3200 rows and 3000 columns.

[0059] (3) Import the numerical matrix into Matlab and convert it into a 80×40×3000 three-dimensional matrix in a zigzag scanning sequence. Fill 20 groups of zeros before and after the second dimension of the matrix to convert the numerical matrix into a 80×80×3000 numerical matrix. Use discrete wavelet transform to remove the low-frequency components in the signal and interpolate for smooth processing. After removing the low-frequency components, perform Fourier transform on the third (time) dimension of the matrix to convert the time domain signal into the frequency domain signal and determine the center frequency f of the signal. 0 The matrix is ​​then subjected to a two-dimensional (spatial) Fourier transform to obtain a wavenumber image at a frequency of 2.7 MHz.

[0060] (4) The theoretical wave number of the aluminum plate surface wave at this frequency is calculated, which is consistent with the actual measured wave number. Figure 6 As shown in (a), the wave numbers outside the theoretical wave numbers are removed to reduce noise interference. According to the law of symmetry of the amplitude in the wave number domain, the program can automatically identify the surface wave echo wave number and filter out the forward wave wave number. The image after filtering is as follows Figure 6 As shown in (b), the reflected wave number is subjected to a two-dimensional inverse Fourier transform to obtain a wave number inversion image in a single frequency domain.

[0061] (5) The signal excited by the laser is a broadband signal. In order to retain more information, the frequency signal corresponding to the peak with higher amplitude in step (3) is Fourier transformed, and step (4) is repeated. The wave number inversion images obtained at each single frequency are normalized and superimposed. The final rectangular defect image is as follows: Figure 7 shown.

Claims

1. A metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis, characterized in that: The steps include: (1) Stably set up the metal plate, place and start a laser vibrometer on one side of the metal plate, and adjust the position of the laser vibrometer, the laser incident angle, and the energy output until a stable and high-amplitude detection light return signal is obtained; (2) The pulsed laser emits a light beam into the galvanometer scanning system, which then reflects the incident laser onto the metal plate to be tested. The computer controls the galvanometer scanning system to scan the surface of the metal plate with excitation light to obtain spatial and temporal sampling matrix data; (3) Perform wavelet transform on broadband surface wave signals to remove low-frequency components, then perform three-dimensional Fourier transform on the time component to convert the time domain signal into a frequency domain signal, select the central peak frequency as the characteristic frequency, and after determining the characteristic frequency, perform two-dimensional Fourier transform on the spatial component of the wave field signal to obtain the wave number image at the characteristic frequency; (4) Calculate the theoretical wave number of the surface wave, remove the wave number outside the theoretical wave number curve, reduce the noise interference imaging effect, filter the wave number of the surface wave forward wave, retain the surface wave echo wave number, perform two-dimensional inverse Fourier transform on the remaining wave number, and obtain a single frequency wave number inversion defect image; (5) The laser-excited ultrasonic wave is a broadband signal. Selecting a single frequency will cause information loss. Therefore, the wave field data of all the frequencies where the peaks are located are selected to repeat step (4). Finally, all the defect images obtained are normalized and added together to obtain the final defect image.

2. The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis according to claim 1 is characterized in that: The expression of the three-dimensional Fourier transform in step (3) is as follows: Where f is the frequency, is the wave number, t is the time, For different directions, and is the signal amplitude; 3. The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis according to claim 1 is characterized in that: The wave number calculation formula for calculating the theoretical wave number of the surface wave in step (4) is as follows: Where, f is the frequency; c is the ultrasonic wave velocity, which can be obtained by dividing the distance between the excitation point and the detection point by the propagation time; f0 is the characteristic frequency selected in the spectrum; is the wave number; For different directions and positions; is the signal amplitude; 4. The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis according to claim 1 is characterized in that: In the steps (1) and (2), the position of the detection light of the laser vibrometer is kept unchanged, and the position of the excitation light of the pulsed laser is moved by the galvanometer system to scan the field.

5. The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis according to claim 1 is characterized in that: The step (4) only leaves the surface wave echo wavenumber in the theoretical wavenumber curve and filters out other interfering wavenumbers. If there is no defect, there will be no echo wavenumber in the wavenumber spectrum, and no defect will appear after inversion.

6. The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis according to claim 1 is characterized in that: The sampling matrix is ​​a sampling point matrix.

7. The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis according to claim 1 is characterized in that: The implementation device includes a nanosecond pulse laser, a galvanometer scanning system, a laser vibrometer, and a computer.

8. The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis according to claim 1 is characterized in that: Step (3) Use computer software Matlab to perform data processing and wavenumber imaging.

9. The metal plate defect imaging method based on laser ultrasound and surface wave number filtering analysis according to claim 1 is characterized in that: The characteristic frequency in step (3) is in the range of 2 MHz to 4 MHz; the frequencies of all the peaks in step (5) are selected from 1 MHz to 5 MHz.

10. Application of the metal plate defect imaging method based on laser ultrasound and surface wave wave number filtering analysis as described in claim 1 in detecting and / or locating surface defects of plate materials.

Images