A method for reconstructing surface profile of bifurcated crack of metal part

Abstract

The application provides a metal piece bifurcated crack surface profile reconstruction method, comprising the following steps: collecting magnetic field distortion signals of a detection range; changing the position of a probe to be orthogonal to the position of the probe in the previous step, and collecting magnetic field distortion signals of the detection range; obtaining two gradient value matrices in two directions according to the direction of the excitation current of the probe based on the value matrix composed of the magnetic field Bz perpendicular to the metal surface direction collected in the two steps; adding the corresponding positions of the two gradient value matrices to obtain a fusion gradient value matrix of the magnetic field distortion signal Bz; obtaining the absolute value of each element in the obtained fusion gradient value matrix to form a positive value matrix; extracting a key value matrix area of the metal piece bifurcated crack surface profile reconstruction; obtaining a key value matrix point position of the metal piece bifurcated crack surface profile reconstruction; and reconstructing the metal piece bifurcated crack surface profile information.

Description

Technical Field

[0001] This invention belongs to the field of nondestructive testing technology, and specifically provides a method for reconstructing the surface contour of bifurcated cracks in metal parts. Background Technology

[0002] Alternating current electromagnetic testing (ACEM) is an emerging non-destructive testing (NDT) technology used to detect cracks and defects in metals. It employs electromagnetic induction to create a uniform current field on the workpiece surface, and by observing the magnetic field distortion caused by defects, it achieves non-contact detection of surface defects in metal structures. It uses low-frequency excitation, eliminating the need for saturation magnetization; induced eddy currents generate significant disturbances in small cracks and discontinuous areas of shallow pits; it can effectively assess defect size and offers fast detection speeds; and it eliminates the need to remove surface contaminants from the metal being tested, greatly reducing time and labor costs.

[0003] In the development of AC electromagnetic detection technology, continuous innovation and changes have occurred in probe design, signal processing, data acquisition, and defect contour reconstruction. Among these, defect contour reconstruction, as a crucial step in the post-processing, has a significant impact on defect detection due to the design of its reconstruction algorithm. Therefore, improving the accuracy of defect contour reconstruction has become a research and technological development direction for AC electromagnetic detection technology.

[0004] Traditional quantitative analysis of crack size in metallic materials typically utilizes magnetic field signals Bx parallel to the metal surface and Bz perpendicular to the metal surface to infer the depth and length of defects. Simultaneously, Bx and Bz signals can be used as the horizontal and vertical axes of a butterfly diagram, and the presence of defects can be determined by observing the butterfly diagram curve. This method is highly sensitive only when the excitation current direction is perpendicular to the crack direction. As the crack deflection angle increases, accurate quantitative analysis of crack size becomes impossible. Furthermore, for bifurcated cracks, the magnetic fields generated by induced eddy currents interfere with each other, making it impossible to accurately reconstruct the defect surface contour. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a more accurate algorithm for detecting and reconstructing the surface contour of bifurcation cracks in metal parts, based on improved detection methods. To achieve the above-mentioned objective, the technical solution adopted by this invention is as follows:

[0006] A method for reconstructing the surface contour of a bifurcated crack in a metal part, using an AC electromagnetic detection probe, comprising an excitation coil and a detection coil for collecting magnetic field distortion signals, characterized by the following steps:

[0007] (1) Determine the detection range and initial detection position that can cover the bifurcated crack, and collect the magnetic field distortion signal of the detection range point by point and line by line using C-scan.

[0008] (2) Change the probe position to be orthogonal to the probe placement position during step (1) detection, and collect the magnetic field distortion signal of the detection range point by point and line by line using C-scan method;

[0009] (3) The numerical matrix composed of the magnetic field Bz perpendicular to the metal surface collected in steps (1) and (2) is used to obtain the gradient numerical matrix in two directions according to the direction of the probe excitation current. The signal of the induced current gathering around the crack is obtained, and the gradient direction of the gradient numerical matrix is ​​orthogonal to the direction of the probe excitation current during scanning.

[0010] (4) The corresponding positions of the two obtained gradient numerical matrices are added together to obtain the fused gradient numerical matrix of the magnetic field distortion signal Bz;

[0011] (5) Take the absolute value of each element in the obtained fused gradient numerical matrix to form a positive value matrix;

[0012] (6) Set a threshold and binarize the smoothed positive matrix. If the numerical matrix element is greater than the threshold, set it to 1; otherwise, set it to 0. Extract the key numerical matrix region for reconstructing the surface contour of the bifurcated crack of the metal part.

[0013] (7) Find the centroid position of each independent region that is not connected within the key numerical matrix region to obtain the key numerical matrix point position for reconstructing the surface contour of the bifurcated crack of the metal part.

[0014] (8) Search and reconstruct the surface contour information of the bifurcated crack of the metal part based on the obtained key numerical matrix point positions.

[0015] Furthermore, the excitation coil is wound around the U-shaped magnetic yoke, and the detection coil is placed below the center of the U-shaped magnetic yoke.

[0016] Furthermore, the magnetic field information detection system for metal parts, in addition to the probe, also includes a signal generator, a power amplifier, a data acquisition card, and a computer. The signal generator is connected to the input of the power amplifier, the output of the power amplifier is connected to the excitation coil, the detection coil is connected to the input of the data acquisition card, and the output of the data acquisition card is connected to the computer.

[0017] Before step (6), the positive value matrix is ​​smoothed. Further, mean filtering is used to smooth the positive value matrix.

[0018] Furthermore, step (8) uses the Kruskal algorithm to search and reconstruct the surface contour information of the bifurcated crack of the metal part based on the obtained key numerical matrix point positions.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0020] Unlike traditional AC electromagnetic detection, which performs a single scan of the crack area, this method uses magnetic field signals Bx parallel to the metal surface and Bz perpendicular to the metal surface to infer the depth and length of the defect. However, it only has high sensitivity when the excitation current direction is perpendicular to the crack direction. As the crack deflection angle increases, it can no longer accurately quantitatively analyze the crack size. Furthermore, for bifurcated cracks, the magnetic fields generated by induced eddy currents interfere with each other, making it impossible to reconstruct the defect surface contour. This invention uses sensors placed orthogonally to perform two scans, which can cover bifurcated cracks. The rectangular detection range is collected point by point and row by row in the metal region using C-scan. The gradients of the magnetic field signal Bz numerical matrix perpendicular to the metal surface are calculated in two directions, the gradient matrices are fused, a positive numerical matrix is ​​generated, the positive numerical matrix is ​​filtered and smoothed, the key numerical matrix region for reconstructing the bifurcated crack surface contour is extracted using a threshold and binarized, the centroid coordinates of each key region of the numerical matrix are calculated, and the Kruskal algorithm is used to search for the reconstructed bifurcated crack surface contour of the metal part, which improves the accuracy of the reconstructed bifurcated crack surface contour of the metal part. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the metal part bifurcation crack surface contour reconstruction detection system of the present invention;

[0022] Figure 2 This is a top view and dimensional information of the bifurcated crack surface contour of the Y-shaped and T-shaped metal parts used in the experiments of this invention.

[0023] Figure 3 This is a schematic diagram of the scanning method used in the experiment of this invention;

[0024] Figure 4 The images show the numerical matrix of two initial Bz signals for the bifurcated crack of the Y-shaped metal part of the present invention (the left side is the initial Bz signal of the first scan, and the right side is the initial Bz signal of the second scan).

[0025] Figure 5 The images show the numerical matrix of two initial Bz signals for the bifurcated crack of the T-shaped metal part of the present invention (the left side is the initial Bz signal of the first scan, and the right side is the initial Bz signal of the second scan).

[0026] Figure 6 The images show two sets of gradient numerical matrix images of the bifurcated crack of the Y-shaped metal part of the present invention (the left side is the gradient in the horizontal direction obtained from the initial Bz signal of the first scan, and the right side is the gradient in the vertical direction obtained from the initial Bz signal of the second scan).

[0027] Figure 7The images show two sets of gradient numerical matrix images of the bifurcated crack of the T-shaped metal part of the present invention (the left side is the gradient in the horizontal direction obtained from the initial Bz signal of the first scan, and the right side is the gradient in the vertical direction obtained from the initial Bz signal of the second scan).

[0028] Figure 8 The positive value matrix image after mean filtering and smoothing of the Y-type and T-type metal parts of this invention (left side originates from Y-type bifurcation crack, right side originates from T-type bifurcation crack);

[0029] Figure 9 This is an image of the key numerical matrix region of the bifurcation cracks in the Y-type and T-type metal parts of the present invention (the left side originates from the Y-type bifurcation crack, and the right side originates from the T-type bifurcation crack);

[0030] Figure 10 This is a schematic diagram showing the centroid positions of local areas of Y-type and T-type bifurcated cracks in this invention (the left side originates from the Y-type bifurcated crack, and the right side originates from the T-type bifurcated crack);

[0031] Figure 11 This is a schematic diagram of the surface contours of Y-type and T-type bifurcated cracks reconstructed by the algorithm of this invention (the left side originates from the Y-type bifurcated crack, and the right side originates from the T-type bifurcated crack);

[0032] In the diagram: 1-excitation coil; 2-U-shaped magnetic yoke; 3-detection coil; 4-aluminum metal part under test; 5-rectangular detection range of crack. Detailed Implementation

[0033] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The examples given herein are only for illustrating and explaining the present invention and are not intended to limit the present invention.

[0034] Example

[0035] The detection probe of this invention consists of a U-shaped magnetic yoke 2 made of manganese-zinc ferrite, an excitation coil 1 wound around the U-shaped magnetic yoke, and a detection coil 3 used to collect magnetic field distortion signals. The detection coil 3 is positioned directly below the center of the U-shaped magnetic yoke 2. A signal generator is connected to the input of a power amplifier, the output of the power amplifier is connected to the excitation coil 1, the detection coil 3 is connected to the input of a data acquisition card, and the output of the data acquisition card is connected to a computer. Furthermore, this system uses a sinusoidal signal as the excitation source signal, and its specific configuration is as follows: Figure 1 As shown.

[0036] In this implementation case, the tested metal part 4 is made of aluminum, and its surface has Y-type and T-type bifurcated cracks. The length of each individual branch crack that makes up the bifurcated crack is 30 mm, the width is 0.5 mm, and the depth is 4 mm. Detailed information on the crack dimensions is as follows: Figure 2As shown, a three-axis scanning platform was built to record the position of the probe during the scanning process. Using LabVIEW 2018, the probe's excitation acquisition was integrated with the motion function of the three-axis platform. The data can be displayed in real time on the computer, and all data from a single scan can be saved as a document.

[0037] In ACFM testing, high sensitivity is only achieved when the excitation current direction is perpendicular to the crack direction. As the crack deflection angle increases, accurate quantitative analysis of crack size becomes impossible. Furthermore, for bifurcated cracks, the magnetic fields generated by induced eddy currents interfere with each other, making accurate reconstruction of the defect surface contour impossible. This invention employs a double scan, differing in that the direction of the excitation current is changed, allowing for high-sensitivity magnetic field distortion characteristic signals to be obtained for cracks at any deflection angle. A rectangular detection range 5 covering the bifurcated crack is defined on the metal part 4 under test. The detection probe is placed at one corner of the rectangular detection range 5. In this embodiment, the detection probe starts from the lower left corner of the rectangular detection range 5 and scans it point-by-point and row-by-row using a C-scan method, collecting all magnetic field distortion signals during the scan. The probe placement is then changed to be orthogonal to the initial probe placement, and the same rectangular detection range 5 is scanned and collected again using the same detection method. The scanning method is as follows: Figure 3 As shown. In this embodiment, the probe scanning movement speed is 20mm / s, the excitation frequency is 10kHz, and the rectangular scanning range is 135mm×135mm.

[0038] The basic principle of ACFM detection is to collect the magnetic field by collecting disturbances caused by induced eddy currents at the crack defect to evaluate the defect. The characteristic signal Bz is the spatial magnetic field distortion caused by the concentration of the electric field at the defect endpoint. When there is no crack, the amplitude of the characteristic signal Bz is 0. When a defect exists, it will show an extreme value at the defect endpoint, reflecting the surface morphology information of the defect. In addition, the magnetic field Bz perpendicular to the metal surface has the characteristic of being insensitive to lift-off and has strong anti-interference ability. Lift-off is the distance between the detection probe and the surface of the metal specimen being tested. In this invention, the Bz signal is used as the characteristic signal for the inversion of the surface profile of the bifurcation crack. After two scans, two sets of initial Bz signal numerical matrices can be obtained within the rectangular detection range. MATLAB has powerful matrix processing and plotting functions and has been widely used in various fields of scientific research and engineering technology. In this embodiment, MATLAB 2021 is used to process the detection signal. The images of the two sets of initial Bz signal numerical matrices for the Y-shaped bifurcation defect are shown below. Figure 4 As shown, the numerical matrix images of the two initial Bz signals for the T-shaped bifurcation defect are as follows: Figure 5 As shown.

[0039] The gradient field is a vector field obtained from the magnitude field, reflecting the direction and value of the greatest change in the magnitude field. Using the gradient field method for image processing on the characteristic signal Bz can obtain the location of maximum distortion in a certain direction of the image, effectively capturing the accumulation of induced current around the defect and reflecting the information of the crack defect surface endpoints and edges. The gradient numerical matrices in two directions are obtained by combining the numerical matrix of the magnetic field Bz collected from the first two scans perpendicular to the metal surface, according to the direction of the probe excitation current. This obtains the accumulation of induced current around the crack at any deflection angle. The gradient directions obtained from the numerical matrices are all orthogonal to the direction of the probe excitation current during scanning. The two sets of gradient numerical matrices obtained for the Y-shaped bifurcation defect are as follows: Figure 6 As shown, the two sets of gradient numerical matrices obtained for the T-shaped bifurcation defect are as follows: Figure 7 As shown.

[0040] After obtaining the gradient numerical matrices of the two directions of the bifurcation crack, the surface contour reconstruction information of the bifurcation crack in the metal part is fused, and the corresponding elements of the matrices are added to form a fused gradient numerical matrix. To facilitate unified operation of the surface contour reconstruction information of the bifurcation crack in the metal part, the fused gradient numerical matrix is ​​converted into a positive value matrix. The sign of each element in the fused gradient numerical matrix is ​​determined. If it is positive, it remains unchanged; if it is negative, it is replaced by ×(-1) to form a positive value matrix. The mean filtering method is used to smooth the positive value matrix. The smoothed positive value matrices of Y-type and T-type bifurcation cracks are as follows. Figure 8 As shown.

[0041] After obtaining the positive value matrix, threshold segmentation is used to focus on reconstructing the key regions of the bifurcation crack surface contour of the metal part. A threshold is set to binarize the positive value fusion gradient numerical matrix; if a numerical matrix element is greater than the threshold, it is set to 1; otherwise, it is set to 0. This extracts the key numerical matrix regions for reconstructing the bifurcation crack surface contour of the metal part. The key numerical matrix regions for Y-shaped and T-shaped bifurcation cracks are shown below. Figure 9 As shown. The centroid positions are determined for each independent, unconnected region within the key numerical matrix region. The centroid positions for the local regions of Y-type and T-type bifurcated cracks are shown in the figure. Figure 10 As shown, obtain the coordinates of the centroid in the entire matrix.

[0042] Kruskal's algorithm is a method for finding the minimum spanning tree of a connected network. Its basic idea is: Assume a connected network G = (V, E), and let the initial state of the minimum spanning tree be a disconnected graph T = (V, {}) with only n vertices and no edges, where each vertex forms a connected component. In E, select the edge with the minimum cost. If the vertices attached to this edge are in different connected components in T, then add this edge to T; otherwise, discard this edge and select the next edge with the minimum cost. Continue in this manner until all vertices in T form a connected component. In MATLAB 2021, the algorithm is implemented using the obtained centroid coordinates. Based on the edges, the edges are first arranged from smallest to largest, and then edges that do not form loops are added from smallest to largest. Finally, the surface profile of the bifurcated crack in the metal part is obtained. The Y-shaped and T-shaped bifurcated crack surface profiles reconstructed by the algorithm are shown below. Figure 11 As shown.

[0043] The above embodiments are merely a detailed description of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications, substitutions, improvements, etc., made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for reconstructing the surface contour of a bifurcated crack in a metal part, wherein the probe used is an AC electromagnetic detection probe, comprising an excitation coil and a detection coil for collecting magnetic field distortion signals, characterized in that, Includes the following steps: (1) Determine the detection range and initial detection position that can cover the bifurcated crack, and collect the magnetic field distortion signal of the detection range point by point and line by line using C-scan. (2) Change the probe position to be orthogonal to the probe placement position during step (1) detection, and collect the magnetic field distortion signal of the detection range point by point and line by line using C-scan method; (3) The numerical matrix composed of the magnetic field Bz perpendicular to the metal surface collected in steps (1) and (2) is used to obtain the gradient numerical matrix in two directions according to the direction of the probe excitation current. The signal of the induced current gathering around the crack is obtained, and the gradient direction of the gradient numerical matrix is ​​orthogonal to the direction of the probe excitation current during scanning. (4) Add the corresponding positions of the two obtained gradient numerical matrices to obtain the fused gradient numerical matrix of the magnetic field distortion signal; (5) Take the absolute value of each element in the obtained fused gradient numerical matrix to form a positive value matrix; (6) Set a threshold and binarize the smoothed positive matrix. If the numerical matrix element is greater than the threshold, set it to 1; otherwise, set it to 0. Extract the key numerical matrix region for reconstructing the surface contour of the bifurcated crack of the metal part. (7) Find the centroid position of each independent region that is not connected within the key numerical matrix region to obtain the key numerical matrix point position for reconstructing the surface contour of the bifurcated crack of the metal part. (8) Search and reconstruct the surface contour information of the bifurcated crack of the metal part based on the obtained key numerical matrix point positions.

2. The reconstruction method according to claim 1, characterized in that, The excitation coil is wound around the U-shaped magnetic yoke, and the detection coil is placed below the center of the U-shaped magnetic yoke.

3. The reconstruction method according to claim 1, characterized in that, The magnetic field information detection system for metal parts used includes, in addition to the probe, a signal generator, a power amplifier, a data acquisition card, and a computer. The signal generator is connected to the input of the power amplifier, the output of the power amplifier is connected to the excitation coil, the detection coil is connected to the input of the data acquisition card, and the output of the data acquisition card is connected to the computer.

4. The reconstruction method according to claim 1, characterized in that, Before step (6), the positive matrix is ​​smoothed.

5. The reconstruction method according to claim 4, characterized in that, The positive matrix is ​​smoothed using mean filtering.

6. The reconstruction method according to claim 1, characterized in that, Step (8) Use the Kruskal algorithm to search and reconstruct the surface contour information of the bifurcated crack of the metal part based on the obtained key numerical matrix point positions.

Images