Composite material damage imaging method based on temporal and spatial cross-correlation of ultrasonic guided wave fields

Abstract

The present invention relates to the field of nondestructive testing technology and provides a composite material damage imaging method based on the spatiotemporal cross-correlation of ultrasonic guided wave wavefields. The method comprises the following steps: S1, acquiring a wavefield signal of the plate being tested; S2, performing narrowband signal extraction on the wavefield signal; S3, filtering the wavefield signal after narrowband signal extraction; S4, shifting the wavefield signal to obtain wavefield signal cross-correlation features; and S5, using a convex hull algorithm to obtain damage imaging results. This method innovatively extracts the difference coefficient between wavefield signals as a basis for damage location, providing a new approach to feature extraction algorithms based on the time-frequency domain. This method enables automated identification of composite material delamination damage and is sensitive to both surface and internal damage.

Description

Technical Field

[0001] The present invention relates to the technical field of nondestructive testing, and in particular to a composite material damage imaging method based on the temporal and spatial cross-correlation of ultrasonic guided wave fields. Background Art

[0002] Composite materials are a new type of material that is formed by fusing metals, polymers, inorganic non-metallic materials in various ways. They not only inherit the main functions of the original materials, but also possess excellent properties that cannot be matched by the original materials, such as light weight, corrosion resistance, and fatigue resistance. Advanced composite materials meet the aviation industry's demand for lightweight, high reliability, and long life, and are therefore widely used in the aviation field. Their usage has gradually become an important indicator of the advancement of aircraft. However, during the use and manufacturing process, composite materials may produce damage such as delamination, scratches, cracks, pitting, pits, and debonding. Such structural damage is highly hidden and difficult to detect. To ensure the safe use of equipment, quality inspection and structural safety monitoring of composite parts must be carried out through various means to promptly detect defects in equipment operation.

[0003] Nondestructive evaluation (NDE) technology aims to assess material quality by measuring specific performance parameters without destroying composite materials. Guided waves are a type of ultrasonic wave. They are waves generated by the boundaries of a medium and propagate along the waveguide structure. Guided waves can propagate in various waveguide structures, such as plates, rods, pipes, and multilayer structures. Guided waves propagating in plates are called Lamb waves. When guided waves propagate in a medium, damage and defects will change the propagation mechanism of the guided wave signal. Guided wave detection technology is based on this principle to achieve damage detection. In the past decade, with the development of scanning laser Doppler vibrometry (SLVD) technology, algorithms for measuring the full wavefield of elastic waves in thin-walled structures and localizing damage have begun to emerge, providing many new solutions for intuitively displaying damage morphology. Compared with other damage detection technologies such as ultrasonic testing and acoustic emission testing, the advantages of SLVD technology lie in its non-contact measurement and high measurement accuracy. It avoids problems such as friction, heat transfer or mechanical interference that may be caused by contact, thereby ensuring measurement accuracy and reliability. It can achieve extremely high displacement and velocity resolution, reaching an amplitude resolution of picometers, providing strong support for the precise analysis of guided wave full-wavefield signals.

[0004] The full-wavefield signal collected by the vibrometer contains very complex information. It is difficult to intuitively determine damage information solely from the evolution of the time-domain wavefield. Therefore, it is necessary to construct a damage identification algorithm to extract damage characteristics. In signal analysis, cross-correlation measures the similarity between two sets of signals. During data acquisition, if there is no structural change between two adjacent scan points, the signals at these two points will have a high degree of similarity due to the small scan step size. Conversely, the signal characteristics of two scan points at the damage boundary will differ significantly due to structural changes. Based on this characteristic, damage location and damage boundary representation can be achieved. Summary of the Invention

[0005] The purpose of the present invention is to provide a detection method that can realize the rapid assessment of composite material delamination damage in a complex field working environment, which can more accurately obtain the quality characteristics of the composite material structure to be tested, while greatly reducing the complexity of the test process, and can conveniently and quickly realize the defect detection of the specimen.

[0006] The present invention proposes a composite material damage imaging method based on the spatiotemporal cross-correlation of ultrasonic guided wave fields, which comprises the following steps:

[0007] Step 1: Obtain the wave field signal W(x, y, n) of the composite plate under test;

[0008] The guided wave propagates in the composite plate under test, and the scanning laser Doppler vibrometer obtains the wave field signal W(x, y, n) of the composite plate under test, where x = 0, 1, 2, ..., M-1, y = 0, 1, 2, ..., N-1, respectively representing the coordinates of two directions in the two-dimensional coordinate system, M and N are the image pixel values ​​of the row and column respectively, and n is the signal length, n = 1, 2, ..., T, and M, N, and T are all positive integers;

[0009] Step 2: Extract narrowband signals from wavefield signals;

[0010] Firstly, the time domain expression of wave field signal, the time domain expression of broadband excitation signal and the time domain expression of narrowband excitation signal are transformed into the frequency domain expression of wave field signal, the frequency domain expression of broadband excitation signal and the frequency domain expression of narrowband excitation signal respectively through Fourier transformation;

[0011] Then, use formula (1) to obtain the frequency domain expression of the wavefield signal after narrowband signal extraction. The specific formula is as follows:

[0012]

[0013] Where ω is the frequency, R b (ω) is the frequency domain expression of the wavefield signal after narrowband signal extraction, R c (ω) is the frequency expression of the wave field signal, S c(ω) is the frequency domain expression of the broadband excitation signal, S b (ω) is the frequency domain expression of the narrowband excitation signal;

[0014] Finally, the frequency domain expression of the wave field signal after narrowband signal extraction is R b (ω) The time domain expression W of the wavefield signal after narrowband signal extraction is obtained through inverse Fourier transform b (x,y,n).

[0015] Step 3: Filter the wavefield signal after narrowband signal extraction;

[0016] First, the time domain expression of the wave field signal after narrowband signal extraction is expressed in W b The slice on the (x, y, n) time axis is divided into T parts, then W b (x,y,n) is converted to two-dimensional f n (x,y), n=1,2,…,T;

[0017] Then, the sliced ​​wave field signal f is transformed into n (x,y) is transformed from the spatial domain to the wave number domain F n (u,v), by F n f represented by (u,v) n The two-dimensional discrete Fourier transform of (x,y) is given by:

[0018]

[0019] Among them, u and v are two variables in the frequency domain, corresponding to the frequency components of the two spatial dimensions x and y of the signal respectively;

[0020] Secondly, the frequency-wavenumber domain filtering algorithm is used to selectively suppress the wavenumber component of the reflected wave in the frequency-wavenumber domain;

[0021] Finally, the filtered spectrum is inverse Fourier transformed to convert the signal back to the time domain. The inverse discrete Fourier transform is in the form of:

[0022]

[0023] Finally, the wave field signal f of each time slice after filtering is obtained by n The (x, y) data is accumulated to obtain the filtered wave field signal W c (x,y,n);

[0024] Step 4: Move the wavefield signal to obtain the cross-correlation characteristics of the wavefield signal;

[0025] Select a wavefield signal W from multiple filtered wavefield signals c(x, y, n), moves the wave field signal in different directions to solve the cross-correlation characteristics of the signal between adjacent scanning points in different directions;

[0026] When the x,y values ​​are determined, according to the filtered wave field signal W c (x, y, n) to obtain a single scanning point signal f(t), and the cross-correlation feature solution formula of adjacent scanning point signals f(t) and g(t) is as follows:

[0027]

[0028] Where f(t) and g(t) are both single scan point signals based on time t, τ is the integral variable; f(t) and g(t) are adjacent scan point signals;

[0029] Step 5: Use the convex hull algorithm to obtain the damage imaging results;

[0030] The results of cross-correlation solutions at different center frequencies and in different directions are superimposed, and the convex envelope of discrete edge points is calculated using the convex hull algorithm to obtain the damage range of the tested composite plate.

[0031] Preferably, in step 1, the waveguide is generated in the following manner:

[0032] The composite plate as the composite plate to be tested is fixed on the bracket. A reflective film is pasted on the surface of the composite plate to enhance the energy reflected to the scanning head. The PZT piezoelectric piece is pasted on the upper left corner of the reflective film. The signal generator generates a broadband frequency modulation signal, which is then amplified by the power amplifier to obtain a broadband excitation signal. The broadband excitation signal is transmitted to the PZT piezoelectric piece. The PZT piezoelectric piece converts the electrical signal into a mechanical signal to cause the composite plate to vibrate and generate guided waves.

[0033] Preferably, in step 2, the broadband excitation signal is a broadband chirp signal, and the frequency band of the broadband chirp signal is 5-500 kHz.

[0034] Preferably, in step 2, the narrowband excitation signal is a sinusoidal signal modulated by a Hanning window, and the time domain expression of the narrowband excitation signal is as follows:

[0035]

[0036] Where, f c is the center frequency of the Hanning window modulated sinusoidal signal.

[0037] Preferably, in step three, the frequency-wavenumber domain filtering algorithm preferably adopts a filtering method of setting the reflected wave to zero.

[0038] Preferably, in step 4, the wavefield signal is moved in different directions, preferably in four directions of 0°, 45°, 90°, and 135°.

[0039] Preferably, the convex hull algorithm is used in step 5 to calculate the convex hull of the discrete edge points as follows:

[0040] Let S={p1,p2,…,p s} is a set of points in a plane or space, S is a point set, p j is a point in S. The convex hull is defined as the set of convex combinations of all these points. The convex combination of the points in the point set S refers to the points that satisfy the following form:

[0041]

[0042] Among them, λ j is the coefficient, λ j ≥0, and satisfies x is the convex combination of the points in the point set S;

[0043] For s edge points, the convex hull Conv(S) is given by:

[0044]

[0045] The damage range of the composite plate is obtained according to formula (6).

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] 1. The present invention takes guided wave detection as its theoretical basis, and has the advantages of rich modal information, sensitivity to damage on the surface and inside of the structure, and flexible guided wave generation.

[0048] 2. The present invention avoids the limitation of manual comparison to find damage characteristics, gets rid of the dependence on engineering experience, and can realize the automatic identification of composite material delamination damage location.

[0049] 3. This invention innovatively extracts the difference coefficient between wave field signals as the basis for damage location, providing a new idea for feature extraction algorithms based on the time-frequency domain. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A test system for implementing a composite material damage imaging method based on spatiotemporal cross-correlation of ultrasonic guided wave fields according to the present invention;

[0051] Figure 2 Schematic diagram of the process of composite material damage imaging method based on spatiotemporal cross-correlation of ultrasonic guided wave field according to the present invention;

[0052] Figure 3AIt is an image of the one-dimensional wave field in the frequency-wavenumber domain of the present invention without removing the reflected wavefront;

[0053] Figure 3B The image is obtained by removing the reflected wave from the one-dimensional wave field in the frequency wavenumber domain of the present invention;

[0054] Figure 4A This is a graph showing the cross-correlation characteristics between the wavefield signal after filtering and the wavefield signal obtained after moving in the 0° direction;

[0055] Figure 4B This is a graph showing the cross-correlation characteristics between the wavefield signal after filtering and the wavefield signal obtained after moving in a 45° direction;

[0056] Figure 4C This is a graph showing the cross-correlation characteristics between the wavefield signal after filtering and the wavefield signal obtained after 90° shifting.

[0057] Figure 5 The damage imaging result diagram is obtained according to the convex hull algorithm of the present invention;

[0058] Figure 6 Schematic diagram of the generation of wave field signals of the present invention. DETAILED DESCRIPTION

[0059] To better understand the technical solutions of the present invention, the specific embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and embodiments. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0060] like Figure 2 As shown, a composite material damage imaging method based on the spatiotemporal cross-correlation of ultrasonic guided wave fields specifically includes the following steps:

[0061] Step 1: Obtain the wave field signal W(x,y,n) of the plate under test.

[0062] Taking the actual test system as an example, the entire test system includes a composite plate 1, a signal generator 2, a power amplifier 3 and a scanning laser Doppler vibrometer 4. Figure 1As shown. Composite plate 1 is fixed on the bracket as the composite plate to be tested. In this embodiment, composite plate 1 has been prefabricated with delamination damage. A reflective film is attached to the surface of composite plate 1 to enhance the energy reflected to the scanning head. A PZT (lead zirconate titanate) piezoelectric piece is attached to the upper left corner of the reflective film. Signal generator 2 generates a broadband frequency modulation signal with a frequency range of 5-500KHz, which is then amplified by power amplifier 3 to obtain a broadband excitation signal. The broadband excitation signal is transmitted to the PZT piezoelectric piece, which converts the electrical signal into a mechanical signal to cause the plate to vibrate and generate guided waves. When the guided wave propagates to the scanning area, the scanning laser Doppler vibrometer 4 obtains the wave field signal W(x, y, n) of the composite plate 1 under test, where x = 0, 1, 2, ..., M-1, y = 0, 1, 2, ..., N-1, respectively representing the coordinates of the two directions in the two-dimensional coordinate system, M and N are the image pixel values ​​of the row and column respectively, n is the signal length, n = 1, 2, ..., T, and M, N, T are all positive integers. At this time, the wave field signal W(x, y, n) is the signal time domain expression, and the generation of the wave field signal is as follows: Figure 6 shown.

[0063] Step 2: Extract narrowband signals from wavefield signals.

[0064] The signal generator uses a broadband chirp signal as a broadband frequency modulation signal. The broadband chirp signal has a frequency band range of 5-500kHz and complex signal components, which results in a complex wavefield signal. Therefore, in order to improve the efficiency of damage feature identification, a narrowband excitation signal is required to extract the narrowband signal from the wavefield signal obtained in step 1.

[0065] First, the time domain expression of the wavefield signal, the time domain expression of the broadband excitation signal and the time domain expression of the narrowband excitation signal are respectively obtained by Fourier transform to obtain the frequency domain expression of the wavefield signal, the frequency domain expression of the broadband excitation signal and the frequency domain expression of the narrowband excitation signal.

[0066] Then, use formula (1) to obtain the frequency domain expression of the wavefield signal after narrowband signal extraction. The specific formula is as follows:

[0067]

[0068] Where ω is the frequency, R b (ω) is the frequency domain expression of the wavefield signal after narrowband signal extraction, R c (ω) is the frequency expression of the wave field signal, S c (ω) is the frequency domain expression of the broadband excitation signal, S b (ω) is the frequency domain expression of the narrowband excitation signal. The narrowband excitation signal is a sinusoidal signal modulated by the Hanning window. The time domain expression of the narrowband excitation signal is as follows:

[0069]

[0070] Where, f c is the center frequency of the Hanning window modulated sinusoidal signal.

[0071] Finally, the frequency domain expression of the wave field signal after narrowband signal extraction is R b (ω) The time domain expression W of the wavefield signal after narrowband signal extraction is obtained through inverse Fourier transform b (x,y,n).

[0072] Step 3: Filter the wavefield signal after narrowband signal extraction.

[0073] In order to remove the influence of boundary reflection on the similarity of wavefield signals and retain only the wavefield signal differences caused by structural changes, the wavefield signals need to be filtered. The filtering process is implemented in the frequency and wavenumber domain.

[0074] First, the time domain expression of the wave field signal after narrowband signal extraction is expressed in W b The slice on the (x, y, n) time axis is divided into T parts, then W b (x,y,n) is converted to two-dimensional f n (x,y)(n=1,2,…,T).

[0075] Then, the wave field signal is transformed from the spatial domain to the wave number domain through the two-dimensional Fourier transform. n f represented by (u,v) n The two-dimensional discrete Fourier transform of (x,y) is given by:

[0076]

[0077] Among them, u and v are two variables in the frequency domain, corresponding to the frequency components of the two spatial dimensions x and y of the signal respectively.

[0078] Secondly, the wavenumber component of the reflected wave is selectively suppressed in the frequency wavenumber domain. When a signal is reflected in space, its wavenumber changes. In the frequency wavenumber domain, the wavenumber of the reflected wave is usually represented as a negative wavenumber, so its position corresponds to quadrant 2 or 4 in the wavenumber domain. In this embodiment, the reflected wave is filtered by setting it to zero. Finally, the filtered spectrum is subjected to an inverse Fourier transform to convert the signal back to the time domain. The form of the inverse discrete Fourier transform is:

[0079]

[0080] Finally, by filtering each time slice f n The data of (x, y) (n=1, 2, …, T) are accumulated to obtain the filtered wave field signal Wc (x,y,n).

[0081] Because the narrowband excitation signals with different center frequencies are used to extract the narrowband signal from the wavefield signal in step 2, this step can obtain multiple filtered wavefield signals with different center frequencies. When the center frequency of the narrowband excitation signal is f c When , we get the filtered wave field signal W c (x,y,n).

[0082] For a more intuitive explanation, Figure 3A is the image of the one-dimensional wave field in the frequency-wavenumber domain without removing the reflected wavefront, Figure 3B This is the image after removing the reflected wave from the one-dimensional wave field in the frequency wavenumber domain. By comparing, it can be intuitively found that the imaging quality is improved.

[0083] Step 4: Move the wavefield signal to obtain the cross-correlation characteristics of the wavefield signal.

[0084] Select a wavefield signal W from multiple filtered wavefield signals c (x, y, n), moves the wavefield signal in multiple directions, including 0° (right), 45° (upper right), 90° (upper), and 135° (upper left), so as to solve the cross-correlation characteristics of the signals between adjacent scanning points in different directions.

[0085] When the x,y values ​​are determined, according to the filtered wave field signal W c (x, y, n) to obtain a single scanning point signal f(t), and the cross-correlation feature solution formula of adjacent scanning point signals f(t) and g(t) is as follows:

[0086]

[0087] Wherein, f(t) and g(t) are both single scan point signals based on time t, τ is the independent variable; f(t) and g(t) are adjacent scan point signals.

[0088] Adjacent scanning point signals refer to the scanning point signals of the current wavefield signal and the scanning point signals of the wavefield signal after the shift. For example, the current wavefield signal is the filtered wavefield signal W obtained in step 3. c (x, y, n), when moving towards 0°, the wave field signal W is obtained c 0 (x,y,n), the filtered wavefield signal W c (x, y, n) and the wave field signal W after movement c 0 (x, y, n) are adjacent wave field signals, and their corresponding scanning point signals are adjacent scanning point signals; then the wave field signal W cAfter (x, y, n) moves in the 90° direction, the wave field signal W c (x,y,n) and wave field signal W c 90 (x, y, n) is called the adjacent wave field signal. If it is the wave field signal W c 0 (x,y,n) moves in the 90° direction, then the wave field signal W c 0 (x,y,n) and wave field signal W c 90 (x, y, n) are called adjacent wave field signals, and their corresponding scanning point signals are adjacent scanning point signals.

[0089] In this embodiment, the filtered wave field signal W c The results of solving the wavefield movement cross-correlation characteristics between adjacent scanning points in three different directions (x, y, n) of 0° (right shift), 45° (upper right), and 90° (upper) are as follows Figures 4A-4C As shown, Figure 4A is the filtered wavefield signal W c Cross-correlation characteristic result diagram of the wave field signal obtained after (x, y, n) moves in the 0° direction; Figure 4B is the filtered wavefield signal W c Cross-correlation characteristic result diagram of wave field signal after (x, y, n) moves in the 45° direction; Figure 4C is the filtered wavefield signal W c The cross-correlation characteristic results of the wavefield signal obtained after (x, y, n) moves in the 90° direction.

[0090] Step 5: Use the convex hull algorithm to obtain the damage imaging results.

[0091] The results of cross-correlation solutions at different center frequencies and directions are superimposed, and the convex hull algorithm is introduced to calculate the convex envelope of discrete edge points. Let S = {p1, p2, ..., p s} is a set of points in a plane or space, S is a point set, p j The convex hull can be defined as the set of convex combinations of all points in S. A convex combination of points in the point set S is a point that satisfies the following form:

[0092]

[0093] Among them, λ j is the coefficient, λ j ≥0, and satisfies x is the convex combination of the points in the point set S.

[0094] For s edge points, the convex hull Conv(S) can be given by the following formula:

[0095]

[0096] The damage range is obtained according to formula (6).

[0097] Figure 5 This is a damage imaging result diagram obtained according to the convex hull algorithm in this embodiment. It can be seen that the damage is marked by a red frame line.

[0098] This paper uses the full-wavefield signal collected by a scanning laser Doppler vibrometer as the analysis basis, explores the relationship between the guided wave signal and damage in composite materials, and introduces the characteristic changes of the guided wave signal at the damage boundary. The purpose is to extract the signal characteristics of the scanning point in the wavefield, explore the signal characteristic changes of adjacent points in the wavefield, and realize the damage location and damage boundary identification of pre-damaged composite materials.

[0099] Finally, it should be noted that the embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A composite material damage imaging method based on temporal and spatial cross-correlation of ultrasonic guided wave fields, characterized by: It includes the following steps: Step 1: Obtain the wave field signal W(x, y, n) of the composite plate under test; The guided wave propagates in the composite plate under test, and the scanning laser Doppler vibrometer obtains the wave field signal W(x, y, n) of the composite plate under test, where x = 0, 1, 2, ..., M-1, y = 0, 1, 2, ..., N-1, respectively representing the coordinates of two directions in the two-dimensional coordinate system, M and N are the image pixel values ​​of the row and column respectively, and n is the signal length, n = 1, 2, ..., T, and M, N, and T are all positive integers; Step 2: Extract narrowband signals from wavefield signals; Firstly, the time domain expression of wave field signal, the time domain expression of broadband excitation signal and the time domain expression of narrowband excitation signal are transformed into the frequency domain expression of wave field signal, the frequency domain expression of broadband excitation signal and the frequency domain expression of narrowband excitation signal respectively through Fourier transformation; Then, use formula (1) to obtain the frequency domain expression of the wavefield signal after narrowband signal extraction. The specific formula is as follows: Where ω is the frequency, R b (ω) is the frequency domain expression of the wavefield signal after narrowband signal extraction, R c (ω) is the frequency expression of the wave field signal, S c (ω) is the frequency domain expression of the broadband excitation signal, S b (ω) is the frequency domain expression of the narrowband excitation signal; Finally, the frequency domain expression of the wave field signal after narrowband signal extraction is R b (ω) The time domain expression W of the wavefield signal after narrowband signal extraction is obtained through inverse Fourier transform b (x,y,n); Step 3: Filter the wavefield signal after narrowband signal extraction; First, the time domain expression of the wave field signal after narrowband signal extraction is expressed in W b The slice on the (x, y, n) time axis is divided into T parts, then W b (x,y,n) is converted to two-dimensional f n (x,y), n=1,2,…,T; Then, the sliced ​​wave field signal f is transformed into n (x,y) is transformed from the spatial domain to the wave number domain F n (u,v), by F n f represented by (u,v) n The two-dimensional discrete Fourier transform of (x,y) is given by: Among them, u and v are two variables in the frequency domain, corresponding to the frequency components of the two spatial dimensions x and y of the signal respectively; Secondly, the frequency-wavenumber domain filtering algorithm is used to selectively suppress the wavenumber component of the reflected wave in the frequency-wavenumber domain; Finally, the filtered spectrum is inverse Fourier transformed to convert the signal back to the time domain. The inverse discrete Fourier transform is in the form of: Finally, the wave field signal f of each time slice after filtering is obtained by n The (x, y) data is accumulated to obtain the filtered wave field signal W c (x,y,n); Step 4: Move the wavefield signal to obtain the cross-correlation characteristics of the wavefield signal; Select a wavefield signal W from multiple filtered wavefield signals c (x, y, n), moves the wave field signal in different directions to solve the cross-correlation characteristics of the signal between adjacent scanning points in different directions; When the x,y values ​​are determined, according to the filtered wave field signal W c (x, y, n) to obtain a single scanning point signal f(t), and the cross-correlation feature solution formula of adjacent scanning point signals f(t) and g(t) is as follows: Where f(t) and g(t) are both single scan point signals based on time t, τ is the integral variable; f(t) and g(t) are adjacent scan point signals; Step 5: Use the convex hull algorithm to obtain the damage imaging results; The results of cross-correlation solutions at different center frequencies and in different directions are superimposed, and the convex envelope of discrete edge points is calculated using the convex hull algorithm to obtain the damage range of the tested composite plate.

2. The composite material damage imaging method based on temporal and spatial cross-correlation of ultrasonic guided wave fields according to claim 1, characterized in that: In step 1, the waveguide is generated in the following manner: The composite plate as the composite plate to be tested is fixed on the bracket. A reflective film is pasted on the surface of the composite plate to enhance the energy reflected to the scanning head. The PZT piezoelectric piece is pasted on the upper left corner of the reflective film. The signal generator generates a broadband frequency modulation signal, which is then amplified by the power amplifier to obtain a broadband excitation signal. The broadband excitation signal is transmitted to the PZT piezoelectric piece. The PZT piezoelectric piece converts the electrical signal into a mechanical signal to cause the composite plate to vibrate and generate guided waves.

3. The composite material damage imaging method based on temporal and spatial cross-correlation of ultrasonic guided wave fields according to claim 1, characterized in that: In the step 2, The broadband excitation signal is a broadband chirp signal, and the broadband chirp signal frequency band range is 5-500kHz.

4. The composite material damage imaging method based on temporal and spatial cross-correlation of ultrasonic guided wave fields according to claim 1, characterized in that: In the step 2, The narrowband excitation signal is a sinusoidal signal modulated by the Hanning window. The time domain expression of the narrowband excitation signal is as follows: Where, f c is the center frequency of the Hanning window modulated sinusoidal signal.

5. The composite material damage imaging method based on temporal and spatial cross-correlation of ultrasonic guided wave fields according to claim 1, characterized in that: In the step three, The frequency-wavenumber domain filtering algorithm preferably adopts a filtering method of setting the reflected wave to zero.

6. The composite material damage imaging method based on temporal and spatial cross-correlation of ultrasonic guided wave fields according to claim 1, characterized in that: In the step 4, The wavefield signal is moved in different directions, preferably in four directions of 0°, 45°, 90°, and 135°.

7. The composite material damage imaging method based on temporal and spatial cross-correlation of ultrasonic guided wave fields according to claim 1, characterized in that: In step 5, the convex hull algorithm is used to calculate the convex envelope of discrete edge points. Specifically, Let S={p1,p2,…,p s } is a set of points in a plane or space, S is a point set, p j is a point in S. The convex hull is defined as the set of convex combinations of all these points. The convex combination of the points in the point set S refers to the points that satisfy the following form: Among them, λ j is the coefficient, λ j ≥0, and satisfies x is the convex combination of the points in the point set S; For s edge points, the convex hull Conv(S) is given by: The damage range of the composite plate is obtained according to formula (6).

Images