A lamb wave defect characterization method based on weighted phase mismatch analysis

Abstract

The application discloses a Lamb wave defect characterization method based on weighted phase mismatch analysis, and according to the characteristics that Lamb wave instantaneous phase is relatively sensitive to defect geometric parameter change and exists certain correlation, weighted phase mismatch analysis is carried out on Lamb wave detection signals in the presence or absence of defects, and a weighted phase cumulative error characteristic index is proposed. The influence of Lamb wave dispersion characteristics and sensor diameter factors on detection signals is comprehensively considered to determine the time window for phase mismatch analysis. The instantaneous phase is weighted according to the wave packet amplitude weight and the wave packet width weight, the contribution of different wave packets to the weighted phase cumulative error is balanced, and the defect characterization accuracy is improved. The method carries out weighted phase mismatch analysis on the Lamb wave defect scattering signal through wave packet extraction and wave packet weighting, determines the corresponding relationship between the Lamb wave instantaneous phase characteristic parameter and the defect geometric parameter, and realizes the characterization of the defect depth and width.

Description

Technical Field

[0001] This invention relates to a Lamb wave defect characterization method based on weighted phase mismatch analysis. This method can characterize the geometric dimensions of small defects in metal plates and belongs to the field of nondestructive testing. Background Technology

[0002] Characterizing the geometry of real defects is one of the main challenges in ultrasonic nondestructive testing, especially for small-sized defects. Due to the complexity of the defect-scattered signal waveform, interpreting measurement data and extracting useful information about the defect is challenging. To address this problem, many researchers have attempted to consider using the Lamb wave transmission coefficient... [1] or reflection coefficient [2] Defects are characterized. However, for small or complex-shaped defects, characterizing them using Lamb wave transmission or reflection coefficients introduces certain errors. (Complete scattering coefficient matrix) [3,4] Containing rich defect information, it is widely used for characterizing small or complex defects. However, obtaining complete information about defects from the scattered sound field is challenging because a complete scattering matrix is ​​usually unavailable. To highlight the scattered sound field generated by defects, Li Li et al. [5] The paper proposes to use decorrelation coefficients to characterize defect scattering information, but the randomness of time window selection has a significant impact on the extraction and characterization of scattering signals.

[0003] With further research, the extraction of instantaneous phase from received signals and its application in signal analysis has received widespread attention. Ebru et al. [6] Extracting instantaneous phase and wave envelope from seismic waves and analyzing the entire signal can yield a significant amount of information from a single seismogram. However, the extracted signal contains a considerable amount of useless information, and the analysis of instantaneous phase does not distinguish between wave packets with high and low amplitudes, reducing the utilization rate of low-amplitude wave packets. Barbara [7] Some researchers have proposed weighting signal packets to balance the contributions of high-amplitude and low-amplitude packets, which can improve the utilization of low-amplitude packets. However, this method involves subjectivity in packet truncation and introduces certain errors in the truncation of low-amplitude packets. As mentioned above, existing literature has studied this by extracting the instantaneous phase from the signal, but the application of instantaneous phase analysis to defect characterization is still incomplete. For example, extracting the instantaneous phase of the received signal with and without defects... [8,9] It can determine whether there are defects. However, it does not propose specific instantaneous phase evaluation indicators, does not conduct a quantitative analysis of the difference between the presence and absence of defects, and cannot make a rough judgment on the size of defects.

[0004] To address the limitations of the aforementioned defect characterization methods, this invention proposes a Lamb wave defect characterization method based on weighted phase mismatch analysis. Summary of the Invention

[0005] The purpose of this invention is to provide a Lamb wave defect characterization method based on weighted phase mismatch analysis. This method leverages the sensitivity of Lamb wave phase to changes in defect geometric parameters and the existence of a certain correlation between phase and defect information. It performs weighted phase mismatch analysis on received Lamb wave signals with and without defect information, proposing a weighted phase cumulative error characteristic index. By comprehensively considering the influence of Lamb wave dispersion characteristics and sensor diameter on the detected signal, a time window for phase mismatch analysis is determined. Instantaneous phase is weighted according to wave packet amplitude weight and wave packet width weight, balancing the contribution of different wave packets to the weighted phase cumulative error and improving the accuracy of defect characterization.

[0006] A Lamb wave defect characterization method based on weighted phase mismatch analysis is implemented through the following steps:

[0007] 1) Acquire the time-domain waveform data of the test piece, extract the instantaneous phase of the detection signal according to the Hilbert transform, and perform weighted phase mismatch analysis on the Lamb wave detection signal with and without defects;

[0008] 2) Taking into account the influence of Lamb wave dispersion characteristics and sensor diameter on the detection signal, calculate the time window used for phase mismatch analysis;

[0009] 3) Extract the instantaneous phase of different wave packets in the detected signal, and calculate the weights applied to different wave packets based on the wave packet width and wave packet amplitude;

[0010] 4) Propose a weighted phase cumulative error index, and change the width or depth of the defect to obtain the trend of the weighted phase cumulative error curve;

[0011] 5) Calculate the distinguishability of different defect depth or width changes based on the weighted phase cumulative error increment when the defect size changes.

[0012] When performing phase mismatch analysis on defect detection signals, the time window for phase mismatch analysis is determined by comprehensively considering the Lamb wave dispersion characteristics and sensor diameter, and the wave packets of the detection signal are extracted.

[0013] The weights of different wave packets are calculated based on their width and amplitude, thereby balancing the contribution of different wave packets to the weighted phase accumulation error. This allows for a more reasonable differentiation of defects of different sizes and improves the accuracy of defect characterization.

[0014] The present invention proposes a Lamb wave defect characterization method based on weighted phase mismatch analysis, the basic principle of which is as follows:

[0015] The study analyzes the degree of instantaneous phase mismatch in the received signal with and without defects. Different wave packets in the received signal are extracted using a time window, and the instantaneous phase is weighted according to the wave packet amplitude and width to balance the contributions of wave packets with higher and lower amplitudes. The weighted phase cumulative error index is... :

[0016] (1)

[0017] In the formula —The sequence number of the wave packet in the received signal;

[0018] —When there is a defect in the received signal The instantaneous phase of the wave packet;

[0019] —Received signal when there are no defects The instantaneous phase of the wave packet;

[0020] ——No. Instantaneous phase weight of wave packet;

[0021] —The width or depth of the defect;

[0022] —Number of received signal packets;

[0023] ——No. Wave packet time window.

[0024] (2)

[0025] In the formula —Received signal without defects Wave packet;

[0026] —Received signal when there are no defects Hilbert transform of wave packets.

[0027] For determining the instantaneous phase weight, this method considers the wave packet amplitude weight. Sum packet width weights Two factors.

[0028] (3)

[0029] (4)

[0030] In the formula ——No. Maximum amplitude of the wave packet.

[0031] (5)

[0032] In the formula ——No. Wave packet data points.

[0033] Regarding the determination of the time window, due to the wide diameter of the sensor, the sound wave propagation path differs when it is tilted, thus the Lamb wave will broaden within a certain time range. This method considers the dispersion characteristics of the Lamb wave and the broadening of the received signal packet caused by the large sensor diameter.

[0034] Given the center frequency of the excitation signal and aluminum plate thickness In this case, the propagation speed of different Lamb wave modes can be determined from the dispersion curve, thereby determining the sound wave propagation time. The received signal in this method contains both A0 and S0 mode Lamb waves, and the lower cutoff frequency is determined using a -3dB bandwidth. The upper cutoff frequency is Assume the distance from the center of the excitation sensor to the aluminum plate is... The distance from the center of the receiving sensor to the aluminum plate is The shortest propagation distance of the A0 mode Lamb wave in the aluminum plate is The longest distance that the S0 mode Lamb wave propagates in the aluminum plate is The diameter of the sensor is The speed of sound waves in air is The excitation signal width is .

[0035] For the A0 mode Lamb wave, its farthest propagation distance from the sensor to the aluminum plate is:

[0036] (6)

[0037] In the formula —The farthest distance that the excitation A0 mode Lamb wave can travel through the air;

[0038] —The farthest distance that an A0 mode Lamb wave can travel through the air;

[0039] —Incident angle of the Lamb wave in mode A0;

[0040] —Incident angle of the leaked Lamb wave in mode A0.

[0041] For the A0 mode Lamb wave, the shortest propagation distance from the sensor to the aluminum plate is:

[0042] (7)

[0043] In the formula —The closest distance from the excitation sensor to the aluminum plate;

[0044] —The closest distance from the receiving sensor to the aluminum plate.

[0045] For the S0 mode Lamb wave, its farthest propagation distance from the sensor to the aluminum plate is:

[0046] (8)

[0047] In the formula —The farthest distance that the excitation S0 mode Lamb wave can travel through the air;

[0048] —The farthest distance that the S0 mode Lamb wave can travel in the air;

[0049] —Incident angle of the S0 mode Lamb wave;

[0050] —Incident angle of the leaked Lamb wave in mode S0.

[0051] For the S0 mode Lamb wave, the shortest propagation distance from the sensor to the aluminum plate is:

[0052] (9)

[0053] In the formula —The shortest distance at which the excitation S0 mode Lamb wave propagates through the air;

[0054] —The shortest distance at which the S0 mode Lamb wave propagates through the air.

[0055] The time window for the A0 mode Lamb wave, considering dispersion characteristics and sensor diameter, is as follows:

[0056] (10)

[0057] The time window for the S0 mode Lamb wave, considering dispersion characteristics and sensor diameter, is as follows:

[0058] (11)

[0059] This method records the increment of the weighted phase cumulative error for each change in defect size as . The total increment is denoted as This allows for the differentiation of defect sizes. Recorded as

[0060] (12)

[0061] In the formula — The order in which the depth or width of the defect changes.

[0062] The present invention has the following advantages: (1) It comprehensively considers the Lamb wave dispersion characteristics and sensor diameter to determine the time window for phase mismatch analysis, thereby reducing the influence of useless information on the results; (2) It uses wave packet amplitude weight and wave packet width weight to weight the instantaneous phase, thereby balancing the contribution of different wave packets to the cumulative error of the weighted phase and improving the accuracy of defect characterization. Attached Figure Description

[0063] Figure 1 This is a flowchart of the method.

[0064] Figure 2 This is a schematic diagram of the system.

[0065] Figure 3 The influence of defect depth on the instantaneous phase of the received signal is shown.

[0066] Figure 4 This describes the influence of defect width on the instantaneous phase of the received signal.

[0067] Figure 5 The waveform and spectrum of the excitation signal.

[0068] Figure 6 This is a graph of the weighted phase cumulative error.

[0069] Figure 7 This is a comparison chart showing the difference in defect size. Detailed Implementation

[0070] The following is combined with Figures 1-6 Taking a typical aluminum plate as an example, this paper details the implementation process of the Lamb wave defect characterization method based on weighted phase mismatch analysis. The flowchart of this method is shown below. Figure 1 As shown.

[0071] The specific implementation steps are given below.

[0072] S1. This invention uses an air-coupled Lamb wave detection system to acquire defect scattering signals. See the schematic diagram of the system device. Figure 2The system mainly includes a JPR-600N ultra-high power ultrasonic excitation receiver, a point-focusing air-coupled sensor, a sensor mounting bracket, a specimen support, and a computer. The computer controls the JPR-600N ultra-high power ultrasonic excitation receiver, which excites ultrasonic signals via the air-coupled excitation sensor above the specimen. Using air as the propagation medium, the ultrasonic waves are incident on the specimen at a fixed angle, and the Lamb wave propagates through the aluminum plate and passes through the defect location. The air-coupled receiving sensor above the aluminum plate receives the signal carrying defect information, which is then amplified and transmitted back to the computer via the ultrasonic excitation receiver, where the waveform is saved.

[0073] S2. Two sets of defective specimens were fabricated. All specimens were aluminum plates, 400 mm long and 3 mm thick. The material parameters are shown in Table 1. The first set of specimens had a fixed defect width of 4 mm and a defect depth ranging from 0 to 1.6 mm with a step size of 0.2 mm, totaling 9 aluminum plates. The second set of specimens had a fixed defect depth of 1.2 mm and a defect width ranging from 3 mm to 12 mm with a step size of 1 mm, totaling 10 aluminum plates.

[0074] Table 1. Parameters of aluminum plate materials used in the experiment

[0075]

[0076] S3. The aluminum plate is inspected using a single excitation and reception method, with the sensor located on the same side of the plate. The excitation frequency is 400kHz, the sampling frequency is 50MHz, the sensor angle is 8°, and the normal distance from the sensor center to the aluminum plate is 15mm. Defect-free information signals are collected. and received signals carrying different defect information .

[0077] S4. Investigate the change in the instantaneous phase of the received signal as the defect depth changes. Obtain the instantaneous phase versus time curve from the detected signal using Hilbert transform, such as... Figure 3 As shown in a), the instantaneous phase of the received signal from defects at different depths was not analyzed. instantaneous phase with defect-free received signal The relationship between the difference and time, such as Figure 3 As shown in b).

[0078] S5. Similarly, the variation law of the instantaneous phase of the received signal when the defect width changes is analyzed. The instantaneous phase versus time curve is obtained from the received signal through Hilbert transform, as shown below. Figure 4 As shown in a), the instantaneous phase of the received signal from defects of different widths was analyzed. instantaneous phase with defect-free received signal The relationship between the difference and time, such as Figure 4 As shown in b).

[0079] S6. Extract the received signal packet using a time window. Due to the wide diameter of the sensor, the sound wave propagation path differs when it is tilted, causing the Lamb wave to broaden within a certain time range. This method considers the dispersion characteristics of the Lamb wave and the broadening of the received signal packet caused by the large sensor diameter to determine the excitation time of the excitation signal. Upper cutoff frequency and lower cutoff frequency ,like Figure 5 As shown. The time windows for the Lamb wave of mode A0 are calculated according to the broadening formula. and S0 mode Lamb wave time window .

[0080] S7. Determination of Instantaneous Phase Weights. This method performs weighted processing on the received signal wave packets, calculating the amplitude weights of the wave packets based on their amplitude and width. Sum packet width weights The weight of each part of instantaneous phase information is calculated using a weighting formula. .

[0081] S8. The instantaneous phase of the received signal containing defect information obtained above. and the instantaneous phase of the defect-free signal received signal Mismatch analysis is performed. A time window is calculated based on factors such as excitation frequency, plate thickness, sensor distance, and sensor diameter, and wave packets are truncated. The truncated wave packets are then weighted according to their width and amplitude. The weighted phase cumulative error under different defect sizes is calculated. The variation law of weighted phase cumulative error with defect depth and width was obtained, such as... Figure 6 As shown.

[0082] S9. The method of the present invention, by comparing with the unweighted phase method, obtains a discrimination comparison map of defects with different depths and widths, such as... Figure 7 As shown, the unweighted phase method used for defect characterization suffers from insufficient discrimination for small defects and excessively high discrimination for large defects, leading to significant differences in the criteria used for defect characterization. The method of this invention can appropriately balance the discrimination of defects of different sizes, that is, appropriately reduce the discrimination of larger defects and increase the discrimination of smaller defects, thereby providing a more reasonable distinction for different defect sizes and improving the accuracy of defect characterization.

[0083] The above is a typical application of the present invention, but the applications of the present invention are not limited thereto.

[0084] References

[0085] [1]Lowe M, Cawley P, Kao J, et al. The low frequency reflection characteristics of the fundamental antisymmetric Lamb wave A0 from arectangular notch in a plate[J]. Journal of the Acoustical Society ofAmerica, 2002, 112(6):2612-2622.

[0086] [2]Cawley P. The effect of complex defect profiles on the reflection of the torsional mode in pipes[J]. NDT & E International, 2012, 46(1):41-47.

[0087] [3]Bai L, Velichko A, Clare A, et al. The effect of distortion modelson characterization of defects using ultrasonic arrays[J]. NDT & EInternational, 2020, 113:102263.

[0088] [4]Moreau L, Caleap M, Velichko A, et al. Scattering of guided waves by flat-bottomed cavities with irregular shapes[J]. Wave Motion, 2012, 49(2):375-387.

[0089] [5] Li Li, Jiao Jingpin, Wu Bin, et al. Decorrelation analysis and Bayesian inversion method of ultrasonic scattering signal for defect quantification [J]. Acta Acustica, 2021, 46(02):281-291.

[0090] [6]Ebru B, Jeannot T, Jeroen T. Misfit functions for full waveforminversion based on instantaneous phase and envelope measurements[J].Geophysical Journal International, 2011(2):845-870.

[0091] [7]Romanowicz B. Global mantle shear velocity model developed using nonlinear asymptotic coupling theory[J]. Journal of Geophysical ResearchSolid Earth, 1996, 101(22)245-272.

[0092] [8] Sun Qinlei, Yang Haitao. Research on electromagnetic guided wave signal processing based on instantaneous phase method [J]. Foreign Electronic Measurement Technology, 2011, 30(11):27-29.

[0093] [9] Chu Lina, Li Jianzeng, Zhou Hailin. Research on defect detection of glass fiber skin based on multi-scale instantaneous phase analysis [J]. Fiberglass / Composite Materials, 2012, 225(04):9-12.

Claims

1. A Lamb wave defect characterization method based on weighted phase mismatch analysis, characterized in that: This is achieved through the following steps: 1) Acquire the time-domain waveform data of the test piece, extract the instantaneous phase of the detection signal according to the Hilbert transform, and perform weighted phase mismatch analysis on the Lamb wave detection signal with and without defects; 2) Taking into account the influence of Lamb wave dispersion characteristics and sensor diameter on the detection signal, calculate the time window used for phase mismatch analysis; 3) Extract the instantaneous phase of different wave packets in the detected signal, and calculate the weights applied to different wave packets based on the wave packet width and wave packet amplitude; 4) Propose a weighted phase cumulative error index, and change the width or depth of the defect to obtain the trend of the weighted phase cumulative error curve; 5) Calculate the distinguishability of different defect depth or width changes based on the weighted phase cumulative error increment when the defect size changes; The study analyzes the degree of instantaneous phase mismatch in the received signal with and without defects; it extracts different wave packets from the received signal through a time window, and weights the instantaneous phase according to the wave packet amplitude and width to balance the contributions of wave packets with higher and lower amplitudes; the weighted phase cumulative error index is as follows. : (1)； In the formula —The sequence number of the wave packet in the received signal; —When there is a defect in the received signal The instantaneous phase of the wave packet; —Received signal when there are no defects The instantaneous phase of the wave packet; ——No. Instantaneous phase weight of wave packet; —The width or depth of the defect; —Number of received signal packets; ——No. Wave packet time window; (2)； In the formula —Received signal without defects Wave packet; —Received signal when there are no defects Hilbert transform of wave packets; For determining the instantaneous phase weight, wave packet amplitude weight is considered. Sum packet width weights Two factors; (3)； (4)； In the formula ——No. Maximum amplitude of wave packet; (5)； In the formula ——No. Wave packet data points; Given the center frequency of the excitation signal and aluminum plate thickness The propagation speed of different Lamb modes can be determined from the dispersion curve, thus determining the sound wave propagation time. The received signal contains two types of Lamb waves, A0 and S0 modes, and the lower cutoff frequency is determined with a bandwidth of -3dB. The upper cutoff frequency is Assume the distance from the center of the excitation sensor to the aluminum plate is... The distance from the center of the receiving sensor to the aluminum plate is The shortest propagation distance of the A0 mode Lamb wave in the aluminum plate is The longest distance that the S0 mode Lamb wave propagates in the aluminum plate is The diameter of the sensor is The speed of sound waves in air is The excitation signal width is ; The time window for the A0 mode Lamb wave, considering dispersion characteristics and sensor diameter, is as follows: (10)； The time window for the S0 mode Lamb wave, considering dispersion characteristics and sensor diameter, is as follows: (11)； In the formula, To excite the A0 mode Lamb wave to travel the farthest distance in the air; To receive the farthest distance that the A0 mode Lamb wave can travel in the air; To ensure the closest possible distance between the sensor and the aluminum plate; To receive the closest distance from the sensor to the aluminum plate; To excite the S0 mode Lamb wave to travel the farthest distance in the air; To receive the farthest distance that the S0 mode Lamb wave can travel in the air; The shortest distance at which the S0 mode Lamb wave propagates in the air; The shortest distance at which the S0 mode Lamb wave propagates in the air.

2. The Lamb wave defect characterization method based on weighted phase mismatch analysis according to claim 1, characterized in that: When performing phase mismatch analysis on defect detection signals, the time window for phase mismatch analysis is determined by comprehensively considering the Lamb wave dispersion characteristics and sensor diameter, and the wave packets of the detection signal are extracted.

3. The Lamb wave defect characterization method based on weighted phase mismatch analysis according to claim 1, characterized in that: The weights of different wave packets are calculated based on their width and amplitude, thereby balancing the contribution of different wave packets to the weighted phase accumulation error, so that defects of different sizes have a more reasonable degree of differentiation and the accuracy of defect characterization is improved.

4. The Lamb wave defect characterization method based on weighted phase mismatch analysis according to claim 1, characterized in that: For the A0 mode Lamb wave, its farthest propagation distance from the sensor to the aluminum plate is: (6)； In the formula, —Incident angle of the Lamb wave in mode A0; —Incident angle of the leaked Lamb wave in mode A0; For the A0 mode Lamb wave, the shortest propagation distance from the sensor to the aluminum plate is: (7)； In the formula, for the S0 mode Lamb wave, its farthest propagation distance from the sensor to the aluminum plate is: (8)； In the formula, —Incident angle of the S0 mode Lamb wave; —Incident angle of the leaked Lamb wave in mode S0; For the S0 mode Lamb wave, the shortest propagation distance from the sensor to the aluminum plate is: (9)； In the formula, the increment of the weighted phase cumulative error for each change in defect size is denoted as . The total increment is denoted as Then the distinguishability of defect size will be... Recorded as (12)； In the formula — The order in which the depth or width of the defect changes.

Images