A baseline-free ultrasonic Lamb defect positioning method based on path energy matching

Abstract

The present application relates to a kind of based on path energy matching's no baseline ultrasonic Lamb defect positioning method, comprising the following steps: S1, on aluminum plate arrangement circular sensor array and construct sector area, simulate defect by punching on plate, gather the damage signal of sector scanning area;S2, pre-processing is carried out to damage signal, and all signals are normalized according to boundary echo signal;S3, according to sensor path length, all signals are grouped, and the first arrival wave packet energy ratio is calculated according to group;S4, the two paths of the lowest energy ratio in each sector scanning area are found out, and the intersection of path straight line is calculated, to further determine the defect position, draw defect positioning chart.Compared with prior art, the present application has the advantages of rapid and accurate detection and positioning of defect, clear imaging result and the like.

Description

Technical Field

[0001] This invention relates to the field of ultrasonic Lamb wave nondestructive testing of isotropic plate materials, and in particular to a baselineless ultrasonic Lamb defect localization method based on path energy matching. Background Technology

[0002] Ultrasonic guided waves offer advantages such as long propagation distance, slow attenuation, and high sensitivity to defects. Furthermore, the equipment required for their detection is simple and easy to operate, making them commonly used in non-destructive testing (NDT) and structural health monitoring. For NDT of plate structures, ultrasonic Lamb waves are frequently employed. Traditional ultrasonic Lamb wave defect detection and localization of isotropic plate structures relies on baseline signal data, requiring the acquisition of health signal data beforehand while the object under inspection is in a non-damaged state. The presence and location of defects are determined by comparing the detected signal with the baseline health signal. Examples of mature algorithms include elliptic imaging, probabilistic elliptic imaging, and total focusing imaging. However, in actual industrial production, baseline data and health signals of the object under inspection are often not collected or are difficult to obtain, thus limiting their application in practical industrial settings. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a baseline-free ultrasonic Lamb defect localization method based on path energy matching.

[0004] The objective of this invention can be achieved through the following technical solutions:

[0005] A baseline-free ultrasonic lamb defect localization method based on path energy matching includes the following steps:

[0006] S1. Arrange a circular sensor array on an aluminum plate and construct a fan-shaped area. Simulate defects by drilling holes in the aluminum plate and collect damage signals from the fan-shaped scanning area.

[0007] S2. Preprocess the damage signal and normalize all signals based on the boundary echo signal;

[0008] S3. Group all signals according to the sensor path length, and calculate the first-arrival packet energy ratio for each group;

[0009] S4. Find the two paths with the lowest energy ratio in each sector scanning area and calculate the intersection of the straight lines of the paths to further determine the defect location and draw the defect location map.

[0010] Furthermore, the method for obtaining the damage signal includes the following steps:

[0011] S101. Arrange a circular array of n sensors with an array radius of r on an aluminum plate, with each sensor equidistant from the others, and label each sensor, where n is an even number and n≥8;

[0012] S102. Drill holes at random locations on the aluminum plate to simulate defects in actual industry.

[0013] S103. Select an excitation signal with a set frequency and period, and excite the odd-numbered sensors in sequence. The m sensors facing each other receive the signals in sequence, and obtain the corresponding m columns of received signals, forming a fan-shaped scanning area. A total of (n / 2)×m columns of damage signals are obtained.

[0014] Furthermore, the preprocessing of the damage signal and the normalization of all signals based on the boundary echo signal specifically include the following steps:

[0015] S201. Read all the acquired damage signals into MATLAB for signal processing;

[0016] S202. Calculate the arrival time t of the boundary echo in each signal. B ;

[0017] S203, according to the boundary echo arrival time t B Extract the peak amplitude of the boundary echoes of all signals, and normalize all damaged signals proportionally according to the length of the boundary echo propagation distance.

[0018] Furthermore, the MATLAB signal processing specifically involves using the fir filter function to perform bandpass filtering on all signals, using the wavelet function to perform noise reduction on all signals, and using the Hilbert function transform to obtain the envelope curve of the damaged signal.

[0019] Furthermore, the arrival time t of the boundary echo in each of the aforementioned signals B :

[0020]

[0021] Among them, V S0 To obtain the group velocity of the S0 mode at a set frequency for the Lamb wave dispersion curve, d TBR This is the nearest boundary reflection distance between each pair of sensors.

[0022] Furthermore, the normalization of the damage signal, for example, the normalized data sequence X of the i-th column signal. i ′ for:

[0023]

[0024] Among them, X i Let A be the signal data sequence of the i-th column. Bi Let d be the peak amplitude of the boundary echo in the i-th column. TBRi Let A be the propagation distance of the boundary echo in the i-th column, and let A be the maximum peak amplitude of the boundary echo among all signals. Bmax And the distance of this path is d. TBRmax .

[0025] Furthermore, the determination of the energy amplitude ratio of the damage signal includes the following steps:

[0026] S301. Group all signals according to the distance of the sensor to the path;

[0027] S302. Calculate the arrival time t of the first wave packet in each signal.

[0028] S303. According to the arrival time t of the first arrival packet, find the first arrival packet of each signal and extract the peak amplitude of the first arrival packet. Compare it with the maximum energy amplitude in each group to obtain the energy ratio of each signal in each group.

[0029] Furthermore, the arrival time t of the first wave packet in each signal is:

[0030]

[0031] Among them, V S0 To obtain the group velocity of the S0 mode at a set frequency for the Lamb wave dispersion curve, d TR This represents the distance between each pair of sensors.

[0032] Furthermore, the energy ratio of each column of signals in each group is obtained. For example, for the i-th column of signals, the signals are grouped into the j-th group according to the path length of the sensor, and the energy ratio B of the i-th column of signals is calculated. i for:

[0033]

[0034] Among them, A i Let A be the peak amplitude of the first wave packet in the i-th column. jmax It represents the maximum energy amplitude in the group containing the i-th column.

[0035] Furthermore, determining the location of the defect includes the following steps:

[0036] S401. According to the sector scanning areas divided in step S1, find the two paths with the lowest energy ratio in each group of sector scanning areas, for a total of n paths.

[0037] S402. In MATLAB, plot the two straight lines with the lowest energy ratio in each sector region and calculate the intersection of all straight lines.

[0038] S403. A clustering algorithm is used to calculate the cluster center of all intersections of straight lines. The cluster center is the location of the defect, thus realizing the detection and localization of the defect.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. This invention simulates defects by drilling holes in an aluminum plate, collects damage signals, normalizes the boundary echo signals, and then locates the defects based on the data obtained by the sensors. It uses only damage signals throughout the process, enabling direct baseline-free detection and eliminating the dependence on baseline health data signals. Compared with existing technologies that require obtaining health signal data in advance when the object being detected is in a non-damaged state, this invention is more in line with the requirements of engineering practice.

[0041] 2. This invention can directly find the path of the sensor pair closest to the defect location by directly comparing the first arrival wave packet energy amplitude ratio of the normalized damage signal. This solves the problem of sensor excitation differences in actual experiments and engineering, making subsequent defect localization imaging more accurate. Compared with the existing technology, it is simpler and faster to calculate the defect elliptical trajectory.

[0042] 3. After obtaining the energy ratio of the first wave packet, this invention identifies the two paths with the lowest energy ratio in each sector scanning area, and uses a clustering algorithm to calculate the cluster center of all straight line intersections. This cluster center is the location of the defect, enabling the detection and localization of the defect. Compared with existing technologies, the imaging is clearer and has better practical engineering application value.

[0043] 4. This invention groups all signals according to the sensor path length and calculates the energy amplitude ratio of the first arrival wave packet in each group. This method is independent of the shape and size of the material and is applicable to the detection and location of defects in most isotropic materials. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the process of the present invention;

[0045] Figure 2 This is a schematic diagram showing the arrangement of the circular array on the aluminum plate and the location of defects in an embodiment of the present invention;

[0046] Figure 3 This is a schematic diagram of the sensor excitation receiving setup in an embodiment of the present invention;

[0047] Figure 4 These are the two time-domain plots of the signals before signal normalization in this embodiment of the invention;

[0048] Figure 5 This is a time-domain diagram of two signals after the signal is normalized using boundary echoes in an embodiment of the present invention;

[0049] Figure 6 This is a diagram showing the path connections with lower energy values ​​in each group of sector scanning areas in this embodiment of the invention;

[0050] Figure 7 This is a diagram showing the intersection points of paths with lower path energy values ​​in an embodiment of the present invention.

[0051] Figure 8 This is a probability imaging cloud map in an embodiment of the present invention;

[0052] Figure 9 This is a local cloud map of the probability imaging for defect localization in an embodiment of the present invention. Detailed Implementation

[0053] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0054] like Figure 1 The diagram shown illustrates the process of this invention, a baseline-free ultrasonic Lamb defect localization method based on path energy matching, comprising the following steps:

[0055] S1. Arrange a circular sensor array on an aluminum plate and construct a fan-shaped region. Simulate defects by drilling holes in the aluminum plate and collect damage signals from the fan-shaped scanning area. The specific steps are as follows:

[0056] S101. On a 1000mm×1000mm×3mm 6061 aluminum plate, arrange 32 piezoelectric crystal sensors (n=32) with a radius of 200mm (r=200mm), a diameter of 10mm, and a thickness of 1mm, at equal intervals to form a circular sensor array, and number the sensors (from number 1 to number 32). Figure 2 As shown;

[0057] S102. Drill holes at random positions on the aluminum plate and in the middle of the array to simulate defects. In this embodiment, a Cartesian coordinate system is established with the lower left corner of the aluminum plate as the origin. The defect is set at the coordinates (506, 550). Figure 2 As shown in the figure, Sensor is a sensor, and Defect is a defect;

[0058] S103. Select a sinusoidal pulse signal with a center frequency of 300kHz and a 5-cycle Hanning window modulation as the excitation signal. Excite the odd-numbered sensors sequentially, and the seven sensors facing each other receive the signals, resulting in seven columns of received signals, forming a sector-shaped scanning area. For example, when sensor 1 is excited, the seven sensors facing each other (sensors 14, 15, 16, 17, 18, 19, and 20) simultaneously receive seven columns of damage signals, forming the first sector-shaped scanning area; in the second sector-shaped scanning area, sensor 3 is excited, and sensors 16, 17, 18, 19, 20, 21, and 22 simultaneously receive the signals, and so on. Figure 3 As shown; a total of 112 damage signals were acquired;

[0059] S2. Preprocess the damage signal and normalize all signals based on the boundary echo signal;

[0060] S201. Read all 112 damaged signals into MATLAB. Use the fir filter function to perform bandpass filtering on all signals, with a high-pass cutoff frequency of 230kHz and a passband frequency of 270kHz, a low-pass cutoff frequency of 370kHz and a passband frequency of 330kHz. Use the dB40 wavelet function to perform noise reduction on all signals to remove unnecessary noise signals. Use the Hilbert function transform to obtain the envelope curves of all noise-reduced damaged signals.

[0061] S202. Obtain the group velocity V of the S0 mode at 300kHz based on the known Lamb wave dispersion curve. S0 =5180m / s, and the nearest boundary reflection distance d between each pair of signal excitation-receiving sensors. TBR The arrival time t of the boundary echo in each signal column is calculated. B for:

[0062]

[0063] S203, according to the boundary echo arrival time t B Extract the peak amplitude of the boundary echo of all signals, and normalize all damaged signals proportionally according to the length of the boundary echo propagation distance;

[0064] For example, the signal data sequence of column 77 is X 77 The peak amplitude of the boundary echo is A B77 The boundary echo propagation distance is d. TBR77 The maximum value of the peak amplitude of the boundary echo among all signals is A. B10 And the distance of this path is d. TBR106 Then the normalized data sequence X7 of the 77th column signal ′ 7 is:

[0065]

[0066] like Figure 4 and Figure 5 As shown.

[0067] S3. Group all signals according to the sensor path length, and calculate the first-arrival packet energy ratio for each group;

[0068] S301. Group all signals according to the distance of the path from the sensor. In this example, the path can be divided into 4 groups.

[0069] S302. Obtain the group velocity V of the S0 mode at this frequency based on the known Lamb wave dispersion curve. S0 =5180m / s, and the distance d between the excitation-receiver sensor pair. TR The arrival time t of the first wave packet in each signal is calculated as follows:

[0070]

[0071] S303. According to the arrival time t of the first arrival packet, find the first arrival packet of each signal and extract the peak amplitude of the first arrival packet. Compare it with the maximum energy amplitude in the group to obtain the energy ratio of each signal in each group.

[0072] For example, the signal in column 35, grouped into group 4 according to the path length of the sensor, has a peak amplitude of A for the first arriving wave packet. 35 =11.6mV, the maximum energy amplitude in this group is A 4max =A 56 =15.6mV, calculate the energy ratio B of the 35th column signal. 35 for:

[0073]

[0074] S4. Find the two paths with the lowest energy ratio in each sector scanning area and calculate the intersection of the straight lines of the paths to further determine the defect location and draw the defect location map.

[0075] A lower energy ratio indicates that, for the same sensor path distance, the energy damage on that path is greater and the path is closer to the defect location. Two paths with the lowest energy ratios are found in each sector. To better determine the defect location area, the probability of the defect falling in the middle area between the two paths is higher, making it easier to determine the defect location area. Two paths are found in each sector, for a total of 16 sectors, and a total of 32 paths are found.

[0076] S402. In MATLAB, plot the two straight lines with the lowest energy ratios in each sector (a total of 32 lines), such as... Figure 6 As shown, calculate the intersection points of all the lines, as follows. Figure 7 As shown;

[0077] S403. Based on all the path intersections obtained above, a clustering algorithm is used to calculate the cluster centers of all path intersections. These cluster centers are the locations of the defects. A probabilistic imaging algorithm is then used to perform defect localization imaging, such as... Figure 8 and Figure 9 As shown, this enables the detection and location of defects.

[0078] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A baseline-free ultrasonic Lamb defect localization method based on path energy matching, characterized in that, Includes the following steps: S1. Arrange a circular sensor array on an aluminum plate and construct a fan-shaped region. Simulate defects by drilling holes in the aluminum plate and collect damage signals from the fan-shaped scanning region. The method for acquiring the damage signals includes the following steps: S101. Arrange a circular array of n sensors with an array radius of r on an aluminum plate, with each sensor equidistant from the others, and label each sensor, where n is an even number and n≥8; S102. Drill holes at random locations on the aluminum plate to simulate defects in actual industry. S103. Select an excitation signal with a set frequency and period, and excite the odd-numbered sensors sequentially. The m sensors facing each other receive the signals sequentially, obtaining the corresponding m columns of received signals, forming a fan-shaped scanning area, and acquiring a total of m signals. Series of damage signals; S2. Preprocess the damage signal and normalize all signals based on the boundary echo signal; S3. Group all signals according to the sensor path length, and calculate the first-arrival packet energy ratio for each group; S4. Find the two paths with the lowest energy ratio in each sector scanning area and calculate the intersection of the straight lines of the paths to further determine the defect location and draw the defect location map. Step S2 involves preprocessing the damage signal and normalizing all signals based on the boundary echo signal, specifically including the following steps: S201. Read all the acquired damage signals into MATLAB for signal processing; S202. Calculate the arrival time of the boundary echo in each signal column. ; S203, According to the arrival time of the boundary echo Extract the peak amplitude of the boundary echo of all signals, and normalize all damaged signals proportionally according to the length of the boundary echo propagation distance; Step S3, calculating the first wave packet energy ratio, includes the following steps: S301. Group all signals according to the distance of the sensor to the path; S302. Calculate the arrival time of the first wave packet in each signal. ; S303, according to the arrival time of the first wave packet Find the first arrival packet of each signal and extract the peak amplitude of the first arrival packet. Compare it with the maximum energy amplitude in each group to obtain the energy ratio of each signal in each group.

2. The baseline-free ultrasonic Lamb defect localization method based on path energy matching according to claim 1, characterized in that, The MATLAB signal processing specifically involves using the fir filter function to perform bandpass filtering on all signals, using wavelet functions to perform noise reduction on all signals, and using the Hilbert function transform to obtain the envelope curve of the damaged signal.

3. The baseline-free ultrasonic Lamb defect localization method based on path energy matching according to claim 1, characterized in that, The arrival time of the boundary echo in each of the aforementioned signals : ， in, To obtain the dispersion curve of the Lamb wave at a set frequency Modal group velocity, This is the nearest boundary reflection distance between each pair of sensors.

4. The baseline-free ultrasonic Lamb defect localization method based on path energy matching according to claim 1, characterized in that, The aforementioned normalization of the damage signal, for the first Normalized data sequence of the column signal for: ， in, For the first Column signal data sequence, For the first Peak amplitude of echo at column boundary For the first The propagation distance of the boundary echo is the maximum value of the peak amplitude of the boundary echo among all signals. And the distance of this path is .

5. The baseline-free ultrasonic Lamb defect localization method based on path energy matching according to claim 1, characterized in that, The arrival time of the first wave packet in each of the aforementioned signals for: ， in, To obtain the dispersion curve of the Lamb wave at a set frequency Modal group velocity, This represents the distance between each pair of sensors.

6. The baseline-free ultrasonic Lamb defect localization method based on path energy matching according to claim 1, characterized in that, The above describes obtaining the energy ratio of each column of signals in each group, for the first... The signals are grouped into the first group according to the path length of the sensor. Group, calculate the first Energy ratio of the signal for: ， in, For the first The peak amplitude of the first wave packet. For the first The maximum energy amplitude in the group to which the column belongs.

7. The baseline-free ultrasonic Lamb defect localization method based on path energy matching according to claim 1, characterized in that, Determining the defect location in step S4 includes the following steps: S401. According to the sector scanning areas divided in step S1, find the two paths with the lowest energy ratio in each group of sector scanning areas, for a total of n paths. S402. In MATLAB, plot the two straight lines with the lowest energy ratio in each sector region and calculate the intersection of all straight lines. S403. Using a clustering algorithm, the cluster center of all intersection points of straight lines is calculated. This cluster center is the location of the defect, thus enabling the detection and localization of the defect.

Images