A CFRP surface and subsurface damage detection system and method

Abstract

The application discloses a CFRP surface and subsurface damage detection system and a detection method, and relates to the technical field of engineering nondestructive testing. The system comprises a probe assembly, a displacement device, a signal generator and a data acquisition card. The probe assembly comprises a fiber probe, a probe skeleton, an excitation coil and two receiving coils. The axis of the fiber probe is located at the midpoint of the line connecting the center of the two receiving coils and the midpoint of the line connecting the center of the excitation coil. The displacement device is used for driving the probe assembly to move in the direction perpendicular to the surface of the CFRP workpiece to be detected. The signal generator is electrically connected with the excitation coil and is used for providing an excitation signal. The data acquisition card is electrically connected with the fiber probe and the two receiving coils and is used for synchronously collecting the fiber signal output by the fiber probe and the induced voltage signals output by the two receiving coils when the probe assembly moves. The application effectively solves the problem of low detection sensitivity of single eddy current method under the weak conductive characteristics of CFRP, and provides high-precision and synchronous registration original data for subsequent image processing.

Description

Technical Field

[0001] This invention relates to the field of engineering nondestructive testing technology, and in particular to a damage detection system and method for CFRP surfaces and subsurfaces. Background Technology

[0002] Carbon fiber reinforced polymer (CFRP) has become a key material for high-performance structures and lightweight designs in modern industry due to its high strength-to-stiffness ratio, excellent corrosion resistance, fatigue performance, and designability. However, due to the instability of process parameters during CFRP manufacturing and the effects of mechanical loads, corrosion aging, and other physicochemical factors during service, CFRP inevitably suffers from surface and subsurface defects such as pits, delamination, cracks, and interface debonding. Among these, surface morphology defects and crack damage are particularly common and pose risks to the structural safety and service life of the material.

[0003] Researchers both domestically and internationally have developed instruments employing various non-destructive testing methods, such as ultrasound, X-ray, eddy current, and infrared imaging, to detect surface morphology defects and cracks. Compared to other methods, eddy current testing offers advantages such as non-contact, speed, and high sensitivity, making it particularly adept at detecting surface and near-surface defects in conductive materials or those containing conductive layers / interfaces. However, in CFRP, the carbon fiber reinforcement is conductive while the resin matrix is ​​non-conductive, resulting in weaker conductivity compared to ordinary metals, making damage detection more difficult and less sensitive. When facing the challenge of high-precision synchronous and in-situ characterization and decoupling of surface and subsurface damage in CFRP, simply using the eddy current method's phase lag principle can lead to a situation where the phase lag values ​​caused by a small surface damage and a large internal damage are identical. Therefore, the detection results can easily confuse deeper surface damage with shallower subsurface damage, making rapid and accurate decoupling impossible. Summary of the Invention

[0004] Based on the shortcomings of the existing technology, the present invention provides a damage detection system and method for CFRP surface and subsurface, which solves the problems of low sensitivity of existing eddy current nondestructive testing technology for CFRP damage detection and inability to quickly and synchronously characterize and decouple surface / subsurface damage.

[0005] The present invention adopts the following technical solution: In a first aspect, the present invention provides a damage detection system for CFRP surfaces and subsurfaces, comprising: The probe assembly, positioned opposite the CFRP surface to be tested, includes an optical fiber probe, a probe frame, an excitation coil, and two receiving coils. The axis of the optical fiber probe is located at the midpoint of the straight line formed by the receiving midpoint and the center of the excitation coil, wherein the receiving midpoint is the midpoint of the line connecting the centers of the two receiving coils. The detection end faces of the optical fiber probe, the excitation coil, and the two receiving coils are located on the same horizontal plane. A displacement device is used to drive the probe assembly to move in a direction perpendicular to the surface of the CFRP workpiece to be measured. A signal generator, electrically connected to the excitation coil, is used to provide an excitation signal; The data acquisition card is electrically connected to the fiber optic probe and the two receiving coils, and is used to synchronously acquire the fiber optic signal output by the fiber optic probe and the induced voltage signal output by the two receiving coils when the probe assembly is in motion.

[0006] Preferably, the displacement device includes: A flexible hinge, one end of which is connected to the probe assembly; An electrically controlled displacement stage is connected to a countersunk hole in a flexible hinge.

[0007] Preferably, the flexible hinge is implemented by a piezoelectric ceramic stack, which is electrically connected to a power amplifier, which drives the piezoelectric ceramic stack to extend and retract.

[0008] Preferably, a lock-in amplifier is provided between the two receiving coils and the data acquisition card. The lock-in amplifier is used to demodulate the induced voltage signal and output amplitude and phase signals.

[0009] Preferred options also include: A probe fixing device includes a probe fixing base and a probe fixing cover plate, used to fix the probe assembly; On the PC side, it is electrically connected to the data acquisition card and is used to process the fiber optic signal to generate the original grayscale image of the fiber optic cable. It is also used to process the induced voltage signals output by the two receiving coils to generate the corresponding amplitude grayscale image and phase grayscale image, respectively.

[0010] Preferably, the other end of the flexible hinge is provided with a hexagonal head bolt for providing preload to the piezoelectric ceramic stack.

[0011] Secondly, the present invention provides a method for detecting damage to CFRP surfaces and subsurfaces, comprising: The voltage-displacement calibration of the non-destructive and damaged regions of the CFRP under test is performed using a damage detection system to obtain non-destructive calibration data and damaged calibration data. Polynomial fitting is then performed on the non-destructive and damaged calibration data to obtain the calibration functions for the non-destructive and damaged regions. Set the initial lift-off distance of the fiber optic probe, perform a C-scan on the CFRP under test, and simultaneously acquire the original grayscale image, amplitude grayscale image, and phase grayscale image of the fiber optic probe. The mask image of the damaged region is obtained by comparing the voltage field and the segmentation threshold in the original grayscale image of the optical fiber; the calibration function is partitioned and inverted using the mask image to obtain the depth information of the undamaged region and the damaged region; the relative depth of the undamaged region and the damaged region is obtained based on the difference between the depth information of the undamaged region and the initial lift-off distance, and the depth inversion image of the CFRP under test is obtained. For each receiving coil, the orthogonal component of its amplitude is obtained based on its amplitude grayscale image and phase grayscale image, and the orthogonal component of the amplitude is gated; the orthogonal component of the amplitude after gated processing of the two receiving coils is differentially divided to generate the eddy current damage detection image of the CFRP under test. By comparing the depth inversion image of the CFRP under test with the eddy current damage detection image, surface damage and subsurface damage are decoupled.

[0012] Preferably, the voltage-displacement calibration of the undamaged and damaged areas of the CFRP under test using the damage detection system specifically includes the following steps: The position of the probe assembly is adjusted by an electrically controlled displacement stage so that the detection end face of the fiber optic probe contacts the surface of the CFRP to be tested. A power amplifier actuates a piezoelectric ceramic stack to drive a flexible hinge, performing multi-step lifting of the fiber optic probe. Each lifting step is recorded, and the corresponding fiber optic signal is acquired by a data acquisition card and uploaded to a PC, enabling voltage-displacement calibration in both lossless and lossy regions.

[0013] Preferably, the mask image of the damaged region is obtained based on the comparison result between the voltage field in the original grayscale image of the optical fiber and the segmentation threshold, specifically including the following steps: Initial candidate masks are obtained based on the comparison results between the voltage field and the segmentation threshold in the original grayscale image of the optical fiber. The initial candidate mask is converted into multiple connected regions and area filtering is performed to obtain the target connected region. Boundary regularization is applied to the target connected component to obtain the symbolic distance field; Gaussian smoothing and reconstruction of the symbolic distance field yields a mask image of the damaged region.

[0014] Preferably, the polynomial fitting uses a sixth-order polynomial.

[0015] Compared with the prior art, the above-mentioned at least one technical solution adopted by the present invention can achieve the following beneficial effects: This invention integrates the fiber optic probe and eddy current coil into a single probe assembly, employing a coaxial layout where the fiber optic probe axis is located at the midpoint of the line connecting the centers of the two receiving coils and the midpoint of the line connecting the centers of the excitation coils. Simultaneously, it ensures that the detection end faces of the fiber optic probe and each coil are coplanar, achieving synchronous, co-located, non-contact measurement of fiber optic ranging and eddy current detection. This system, combined with a precision displacement device using a flexible hinge, enables high-resolution voltage-displacement calibration in both non-destructive and destructive regions. It effectively solves the problem of low detection sensitivity of the single eddy current method under the weak conductivity characteristics of CFRP, and provides high-precision, synchronously registered raw data for subsequent image processing. The system boasts high integration and ease of operation.

[0016] Furthermore, this invention performs voltage-displacement calibration on both the non-destructive and damaged regions and establishes a partitioned calibration function. Combined with the reconstructed mask, it accurately obtains a depth-inversion image of the CFRP surface morphology. Simultaneously, phase-gated orthogonal feature construction and dual-channel differential processing are employed for the eddy current signal, significantly suppressing edge noise and highlighting damage characteristics. Finally, by comparing and decoupling the fiber optic depth-inversion image with the eddy current differential damage image, this invention achieves, for the first time, rapid, synchronous, and in-situ differentiation of CFRP surface damage and subsurface damage. This overcomes the technical bottleneck of traditional eddy current methods, which rely solely on phase lag and cannot distinguish damage depth layers, significantly improving decoupling accuracy and detection reliability. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method for detecting damage to the surface and subsurface of CFRP according to the present invention; Figure 2 This is an exploded hardware view and a schematic diagram of the system connection for a CFRP surface and subsurface damage detection system according to the present invention. Figure 3 This is a schematic diagram of the end of the probe assembly of the present invention; Figure 4 The CFRP workpiece under test with surface depressions and internal crack damage disclosed in the embodiments of the present invention; Figure 5 The displacement-voltage calibration curve and polynomial fitting curve of the fiber optic probe at the undamaged surface of the workpiece to be tested are disclosed in the embodiments of the present invention. Figure 6The displacement-voltage calibration curve and polynomial fitting curve of the fiber optic probe at the surface damage site of the workpiece to be tested are disclosed in the embodiments of the present invention. Figure 7 This is the original grayscale image of the optical fiber disclosed in the embodiments of the present invention; Figure 8 The mask image obtained by processing the original grayscale image of the optical fiber disclosed in the embodiments of the present invention; Figure 9 This is a two-dimensional depth image of the surface morphology of the workpiece to be tested, as disclosed in an embodiment of the present invention. Figure 10 This is a three-dimensional curved surface image of the workpiece surface to be tested, as disclosed in an embodiment of the present invention; Figure 11 This is the original grayscale image of the eddy current C-scan amplitude signal of the two receiving coils disclosed in an embodiment of the present invention; in, Figure 11 (a): A receiving coil, Figure 11 (b): Another receiving coil; Figure 12 This is the original grayscale image of the eddy current C-scan phase signal of the two receiving coils after denoising, as disclosed in an embodiment of the present invention. in, Figure 12 (a): A receiving coil, Figure 12 (b): Another receiving coil; Figure 13 The image comparison of optical fiber and eddy current after image processing is disclosed in the embodiment of the present invention, wherein the white box represents the location and size of surface depression damage and in-plane crack damage; in, Figure 13 (a): Image of eddy current damage detection. Figure 13 (b): Depth inversion image.

[0019] In the diagram: 1-Hexagonal head bolt, 2-Flexible hinge, 3-Fiber optic probe, 4-Probe mounting cover, 5-Probe frame, 6-Receiver coil, 7-Excitation coil, 8-Probe mounting base, 9-Piezoelectric ceramic stack, 10-Signal generator, 11-Lock-in amplifier, 12-Data acquisition card, 13-Power amplifier, 14-PC terminal, 15-Motion controller. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] This invention discloses a damage detection system for CFRP surfaces and subsurfaces, specifically relating to a damage detection system for CFRP surfaces / subsurfaces based on an optical fiber-eddy current probe, such as... Figure 2 and Figure 3 As shown, the detection system includes a probe assembly, a probe fixing device, a flexible hinge 2, a piezoelectric ceramic stack 9, and a hexagonal head bolt 1.

[0022] The probe assembly includes an optical fiber probe 3, a probe frame 5, an excitation coil 7, and two receiving coils 6. The optical fiber probe 3, the excitation coil 7, and the two receiving coils 6 are installed in the probe frame 5. The axis of the optical fiber probe 3 is located on the axis of symmetry of the longitudinal section of the probe frame 5. The axis of the optical fiber probe 3 is located at the midpoint of the straight line formed by the receiving midpoint and the center of the excitation coil 7. The receiving midpoint is the midpoint of the line connecting the centers of the two receiving coils 6, forming a coaxial structure. The end of the optical fiber probe 3 and the end of the eddy current coil (detection end face) are located on the same horizontal plane.

[0023] The probe fixing device includes a probe fixing base 8 and a probe fixing cover plate 4. During installation, the fiber optic probe 3 is first placed into the arc-shaped groove in the probe fixing base 8, and the probe fixing cover plate 4 is then placed on top. Subsequently, the probe frame 5 is installed on the probe fixing base 8 and the probe fixing cover plate 4 with screws. Finally, the position of the fiber optic probe 3 is adjusted so that its end and the end of the eddy current coil are on the same horizontal plane, and the screws on the probe fixing cover plate 4 are tightened to complete the fixed installation of the fiber optic probe 3.

[0024] The excitation signal of the excitation coil 7 is given by the signal generator 10. The induced voltage signals received by the two receiving coils 6 are processed by the lock-in amplifier 11 and then input to the data acquisition card 12, and finally input to the PC terminal 14.

[0025] The piezoelectric ceramic stack 9 is given a signal by the power amplifier 13 to drive the micro-motion of the flexible hinge 2.

[0026] Fiber optic probe 3 is used to generate fiber optic signals, which are acquired by the data acquisition card and finally input to the PC.

[0027] PC terminal 14 sends commands to motion controller 15 (which controls the electronically controlled displacement stage and piezoelectric ceramic stack 9) for overall motion control.

[0028] Reference Figure 1 As shown, this invention discloses a method for detecting damage to the surface and subsurface of CFRP, the method comprising the following steps: S1: Place the CFRP laminate workpiece to be tested flat on a marble countertop, with the two vertical edges of the workpiece aligned as follows: x shaft and y The axis, perpendicular to the workpiece surface, is zShaft. The probe assembly and flexible hinge 2 are fixed to the electrically controlled displacement stage through the countersunk hole on the flexible hinge. The probe assembly is fixedly installed at the front end of the flexible hinge 2, and the piezoelectric ceramic stack 9 is preloaded by screwing hexagonal head bolts 1 into the rear end of the flexible hinge.

[0029] On undamaged areas of the workpiece surface, an electrically controlled displacement stage is used for... z The position of the directional probe is adjusted so that the front end of the probe assembly just contacts the workpiece surface (lifting distance is 0). Then, the power amplifier 13 is used to actuate the piezoelectric ceramic stack 9 to drive the flexible hinge 2 to perform micro-movement, thereby performing multi-step micro-lifting of the fiber optic probe 3. Each lift-up record and the corresponding fiber optic signal are collected by the data acquisition card 12 and uploaded to the PC terminal 14. After the displacement reaches full scale, the electronically controlled displacement stage is adjusted to perform voltage-displacement calibration of the fiber optic probe 3 at the non-destructive position.

[0030] In this embodiment, the full scale of the piezoelectric ceramic stack is 50 micrometers, which is insufficient for the calibrated displacement range (the effective range of the fiber optic sensor). Therefore, after reaching full scale, the control console performs a large-scale (50 micrometers) displacement, and then the piezoelectric ceramic stack drives the flexible hinge micro-motion to calibrate the next displacement range.

[0031] Since the surface is rougher after processing (i.e., at the damaged area), its displacement-voltage curve is different from that of the undamaged surface. Therefore, this process is repeated at the damaged area to perform damaged location calibration.

[0032] S2: The calibration data obtained from S1 is used to adjust the probe lift-off according to the voltage signal (set according to experience and within the effective range of the fiber optic sensor). Then, the PC terminal 14 sends a command to the motion controller 15 to perform C-scan on the workpiece. The signal generator 10 provides an excitation signal to the eddy current excitation coil 7. The induced voltage signals of the two receiving coils 6 are demodulated by the lock-in amplifier 11 and synchronously acquired by the data acquisition card along with the fiber optic signal, and then uploaded to the PC terminal.

[0033] The voltage data collected by the C-scan is in an Excel spreadsheet. The original data is then presented as a grayscale image using MATLAB programming.

[0034] S3: Import the voltage data into MATLAB for programming, and obtain the original grayscale image of the optical fiber. To obtain the surface morphology information of the specimen more intuitively, signal processing based on partition calibration for surface morphology reconstruction was performed. Let the two-dimensional voltage field obtained by optical fiber ranging scanning be... ,but: (1); in, Indicates the distance between the probe and the surface. Indicates the state of local reflection. This represents fluctuations caused by noise and micro-texture. and These represent the undamaged and damaged areas, respectively. The surface reflects light at the point of no damage. The voltage-displacement calibration function for fiber optic cables at a non-destructive surface. Two-dimensional coordinates of the scanned area, The reflection state at the damaged surface This is the fiber voltage-displacement calibration function at the damaged surface.

[0035] Given that surface damage exhibits a higher voltage response, a quantile threshold is used to achieve robust segmentation: (2); in, The fiber voltage signal within the robustly segmented range. Voltage diagram for segmentation after slight denoising. For the first Quantile operator in some embodiments. The value ranges from 0.85 to 0.94.

[0036] When the voltage in the damaged area is higher, the initial candidate mask is defined as: (3); in, As the initial candidate mask, This is an indicator function.

[0037] Subsequently, relying solely on thresholding often results in numerous fragmented connected components. To ensure that the segmentation results conform to the physical prior that "the defect is a finite number of entity targets," the initial candidate mask is subjected to closing operations, hole filling, opening operations, and area filtering, transforming the initial candidate mask into a set of connected components. And sort by area, retaining the K largest regions: (4); in, For the target connected component, Let K be the initial set of connected components, and K represent the number of surface damages. In some embodiments, K =1-6.

[0038] To mitigate the jagged edges and local convexity of the binary mask boundaries, while avoiding the erosion of small defects by large-scale morphological operations, a signed distance field is used for boundary regularization in the continuous domain, thus connecting the target domain. Mapped to the symbolic distance field (Internal is positive, external is negative): (5); in, This is a Euclidean distance transformation. for The logical NOT of this is the background region. Then, for... Perform Gaussian smoothing and reconstruct the mask using the zero level set: (6); (7); in, This is the Gaussian smoothed sign distance field. This represents the Gaussian smoothing operator. Control the boundary smoothing intensity, This represents the smoothed mask, which means the new boundary is contour lines, p Let be any point in the mask plane.

[0039] S4: Based on the displacement-voltage data obtained in S1 at both the undamaged and damaged sites, generate the lift-off displacement-voltage signal curve and its polynomial fitting in MATLAB, and establish the voltage-displacement calibration function. Then, perform partitioned inversion under mask constraints. (8); in, For the depth information of the inversion, To calibrate the inverse function of fiber voltage-displacement at a lossless surface, The inverse function for calibrating the fiber voltage-displacement at the damaged surface is given.

[0040] relative depth Defined as: (9); in, c Indicates the initial lift. Z The unit is mm. D Units are m.

[0041] Finally, to improve the continuity of the two-dimensional / three-dimensional surface topography images, edge-preserving smoothing and resampling processing are performed on the relative depth to obtain two-dimensional depth inversion images and three-dimensional surface images.

[0042] S5: In MATLAB, the amplitude and phase grayscale images of receiving coil 6 are generated from the acquired data. These images contain background noise (random noise, texture stripes, edge effects, etc.), which significantly reduces the discernibility of damage. First, image denoising is performed: a non-damaging reference is selected... y The band, within which each x List the median: (10); in, The median of the undamaged reference band. For the non-damaging reference band y scope, The observed amplitude or phase, where statistical analysis is performed using the non-destructive region. To approximate background noise (and constant drift within the reference band), perform background denoising on the image: (11); in, The value is the amplitude or phase of the scanned region after denoising.

[0043] After initial denoising, the phase image was observed. The damaged area showed obvious phase lag, while the residual noise at the upper and lower edges showed phase lead. To further highlight the damage features and suppress edge noise, phase-gated orthogonal features were constructed. (12); (13); (14); in, For the first k The quadrature components of the amplitude of each receiving coil For amplitude, For phase, These are the quadrature amplitude components after phase-gated processing. The equivalent phase after phase gating.

[0044] when When >0, =0, the edge response is significantly suppressed; when When <0 (damage-induced phase hysteresis), As the hysteresis increases, the damaged area becomes more prominent. Finally, the orthogonal amplitude components of the two channels are differentially analyzed and normalized for further noise reduction, resulting in an eddy current damage detection image.

[0045] S6: By comparing the depth inversion image with the eddy current damage detection image, the surface / subsurface damage is decoupled, and the two are finally detected in a high-precision synchronous and in-situ manner.

[0046] Example In one specific embodiment, [0,90] is used. 4s CFRP laminate workpieces such as Figure 4As shown, the overall thickness is 2.4 mm, with 16 layers and a single layer thickness of 0.15 mm. Three surface depressions are pre-set on the right side of the upper surface: surface depression A1, depth 0.1 mm; surface depression A2, depth 0.2 mm; surface depression A3, depth 0.3 mm. Three in-plane cracks are pre-set on the left side of the lower surface: in-plane crack B1, depth 1.5 mm; in-plane crack B2, depth 1.2 mm; in-plane crack B3, depth 0.9 mm. A C-scan is performed on the upper surface using a probe. The specific testing and image processing steps are as follows: Step 1: Place the CFRP laminate workpiece to be tested flat on the marble countertop, such as... Figure 4 As shown. An electrically controlled displacement stage is used on an undamaged area of ​​the workpiece surface. z The position of the directional probe is adjusted so that the front end of the probe assembly just contacts the workpiece surface (lifting distance is 0). Then, the power amplifier 13 actuates the piezoelectric ceramic stack 9 to drive the flexible hinge 2 in micro-motion, with each micro-motion being 0.01mm, thus performing a small lift. Each lift step and the corresponding fiber optic signal are recorded by the data acquisition card 12 and uploaded to the PC 14. After the displacement reaches full scale, the electronically controlled displacement stage is adjusted to perform voltage-displacement calibration of the fiber optic probe 3 at the undamaged position. Because the surface is rougher after processing (i.e., at the damaged surface), its displacement-voltage curve differs from that at the undamaged surface. Therefore, this process is repeated at the damaged surface for damaged position calibration. The calibration curve is shown below. Figure 5 and Figure 6 The original measurement data curve is shown in the figure.

[0047] Step 2: Based on the calibration data obtained in Step 1, adjust the probe lift to 0.3mm according to the fiber optic voltage signal. Then, PC 14 sends a command to motion controller 15 to perform C-scan on the workpiece. The signal generator 10 provides an excitation signal of 5V to eddy current excitation coil 7 at a frequency of 2.15MHz. The induced voltage signals of the two receiving coils 6 are demodulated by lock-in amplifiers and synchronously acquired by the data acquisition card along with the fiber optic signal, and then uploaded to the PC.

[0048] Step 3: Import the data into MATLAB to obtain the following results: Figure 7 After obtaining the original grayscale image of the optical fiber, signal processing based on partition calibration was performed to reconstruct the surface morphology of the specimen more intuitively. The final mask image is shown below. Figure 8 As shown.

[0049] Step 4: The obtained displacement-voltage data at both the undamaged and damaged sites are used to generate lift-off displacement-voltage signal curves, linear polynomial fitting, and sixth-order polynomial fitting in MATLAB, as shown below. Figure 5 and Figure 6As shown, a voltage-displacement calibration function for the fiber optic probe was established simultaneously. It can be seen that the linear fitting effect is acceptable for undamaged surfaces, but the fitting effect shows a significant deviation for damaged areas, while the sixth-order polynomial fitting works well for both cases. Subsequently, under mask constraints, sixth-order polynomial fitting was used for partitioned inversion.

[0050] Finally, to improve the continuity of the 2D / 3D surface topography images, edge-preserving smoothing and resampling were performed on the depth map, resulting in the following: Figure 9 The two-dimensional depth inversion image shown is similar to... Figure 10 The three-dimensional surface image shown clearly reveals the spatial location and depth information of the surface damage in 9-1 (A1), 9-2 (A2), and 9-3 (A3).

[0051] Step 5: In MATLAB, generate grayscale images of the amplitude and phase of the two receiving coils 6 from the acquired data. The original amplitude image is shown below. Figure 11 As shown, the image contains background noise (random noise, texture stripes, edge effects, etc.), which greatly reduces the discernibility of damage. First, denoising processing is performed on the image.

[0052] After the initial denoising is completed, observe the phase image as follows: Figure 12 As shown, the damaged area exhibits a significant phase lag, while the residual noise at the upper and lower edges shows a phase lead. To further highlight the damage characteristics and suppress edge noise, phase-gated orthogonal features are constructed.

[0053] Finally, the orthogonal amplitude components of the two channels are differentially analyzed and normalized for further noise reduction, resulting in the following: Figure 13 The eddy current damage detection image shown in (a) is shown.

[0054] Step Six: As Figure 13 As shown, the inverted image of fiber surface depth information is compared with the eddy current differential damage image to decouple surface / subsurface damage. The white dashed box in the figure represents the actual damage location. It can be seen that the eddy current probe responds to several damage locations, among which the control... Figure 13 (b) can be decoupled to show that the three cracks on the left are in-plane damage and the three on the right are surface morphology damage.

[0055] The embodiment successfully achieved decoupling of CFRP surface / subsurface damage information and simultaneous high-sensitivity imaging.

[0056] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0057] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for detecting damage to the surface and subsurface of CFRP, characterized in that, The damage detection system includes: The probe assembly, positioned opposite the CFRP surface to be tested, includes an optical fiber probe (3), a probe frame (5), an excitation coil (7), and two receiving coils (6); the axis of the optical fiber probe (3) is located at the midpoint of the straight line formed by the receiving midpoint and the center of the excitation coil (7), wherein the receiving midpoint is the midpoint of the line connecting the centers of the two receiving coils (6); the detection end faces of the optical fiber probe (3), the excitation coil (7), and the two receiving coils (6) are located on the same horizontal plane; A displacement device is used to drive the probe assembly to move in a direction perpendicular to the surface of the CFRP workpiece to be measured. A signal generator (10) is electrically connected to an excitation coil (7) to provide an excitation signal; The data acquisition card (12) is electrically connected to the fiber optic probe (3) and the two receiving coils (6) to synchronously acquire the fiber optic signal output by the fiber optic probe (3) and the induced voltage signal output by the two receiving coils (6) when the probe assembly is in motion. The detection method includes the following steps: The voltage-displacement calibration of the non-destructive and damaged regions of the CFRP under test is performed using a damage detection system to obtain non-destructive calibration data and damaged calibration data. Polynomial fitting is then performed on the non-destructive and damaged calibration data to obtain the calibration functions for the non-destructive and damaged regions. Set the initial lift-off distance of the fiber optic probe (3), perform a C-scan on the CFRP to be tested, and simultaneously acquire the original grayscale image, amplitude grayscale image and phase grayscale image of the fiber optic probe; The mask image of the damaged region is obtained by comparing the voltage field and the segmentation threshold in the original grayscale image of the optical fiber; the calibration function is partitioned and inverted using the mask image to obtain the depth information of the undamaged region and the damaged region; the relative depth of the undamaged region and the damaged region is obtained based on the difference between the depth information of the undamaged region and the initial lift-off distance, and the depth inversion image of the CFRP under test is obtained. For each receiving coil (6), the orthogonal component of its amplitude is obtained based on its amplitude grayscale image and phase grayscale image, and the orthogonal component of the amplitude is gated; the orthogonal component of the amplitude after gated processing of the two receiving coils (6) is differentially divided to generate the eddy current damage detection image of the CFRP under test. By comparing the depth inversion image of the CFRP under test with the eddy current damage detection image, surface damage and subsurface damage are decoupled.

2. The detection method of the CFRP surface and subsurface damage detection system as described in claim 1, characterized in that, The displacement device includes: A flexible hinge (2) is connected at one end to the probe assembly; The electrically controlled displacement stage is connected to the countersunk hole of the flexible hinge (2).

3. The detection method of the CFRP surface and subsurface damage detection system as described in claim 2, characterized in that, The flexible hinge (2) is realized by a piezoelectric ceramic stack (9), which is electrically connected to a power amplifier (13) for driving the piezoelectric ceramic stack to extend and retract.

4. The detection method of the CFRP surface and subsurface damage detection system as described in claim 1, characterized in that, A lock-in amplifier (11) is provided between the two receiving coils (6) and the data acquisition card (12). The lock-in amplifier (11) is used to demodulate the induced voltage signal and output the amplitude signal and the phase signal.

5. The detection method of the CFRP surface and subsurface damage detection system as described in claim 3, characterized in that, Also includes: The probe fixing device includes a probe fixing base (8) and a probe fixing cover plate (4) for fixing the probe assembly; The PC terminal (14) is electrically connected to the data acquisition card (12) and is used to process the fiber optic signal to generate the original grayscale image of the fiber optic signal. It is also used to process the induced voltage signal output by the two receiving coils (6) to generate the corresponding amplitude grayscale image and phase grayscale image respectively.

6. The detection method of the CFRP surface and subsurface damage detection system as described in claim 3, characterized in that, The other end of the flexible hinge (2) is provided with a hexagonal head bolt (1) for providing preload to the piezoelectric ceramic stack (9).

7. The detection method of the CFRP surface and subsurface damage detection system as described in claim 5, characterized in that, The voltage-displacement calibration of the non-damaged and damaged areas of the CFRP under test using a damage detection system includes the following steps: The position of the probe assembly is adjusted by an electrically controlled displacement stage so that the detection end face of the fiber optic probe (3) contacts the CFRP surface to be tested. The power amplifier (13) actuates the piezoelectric ceramic stack (9) to drive the flexible hinge (2) to perform multi-step lifting of the fiber optic probe (3). Each lifting step is recorded and the corresponding fiber optic signal is collected by the data acquisition card (12) and uploaded to the PC (14) to realize voltage-displacement calibration in the non-destructive and damaged areas.

8. The detection method of the CFRP surface and subsurface damage detection system as described in claim 1, characterized in that, The mask image of the damaged region is obtained based on the comparison between the voltage field in the original grayscale image of the optical fiber and the segmentation threshold. The specific steps include: Initial candidate masks are obtained based on the comparison results between the voltage field and the segmentation threshold in the original grayscale image of the optical fiber. The initial candidate mask is converted into multiple connected regions and area filtering is performed to obtain the target connected region. Boundary regularization is applied to the target connected component to obtain the symbolic distance field; Gaussian smoothing and reconstruction of the symbolic distance field yields a mask image of the damaged region.

9. The detection method of the CFRP surface and subsurface damage detection system as described in claim 1, characterized in that, The polynomial fitting uses a sixth-order polynomial.

Images