A system for laser ultrasonic detection of surface damage in carbon fiber composites

Abstract

The application discloses a system for laser ultrasonic detection of surface damage of carbon fiber composite materials, which comprises an excitation module, an interference detection module and a signal processing unit; the excitation module is used for vertically incident laser on the sample surface; the interference detection module is used for capturing vibration information in the sample and propagation information of ultrasonic waves; the signal processing unit carries out noise reduction processing on the received signal to obtain smooth and interference-free ultrasonic wave information; the application adopts a full non-contact laser galvanometer, a laser interference defect detection system, a nanosecond pulse laser as an excitation source and a laser Doppler vibrometer as a receiver; the distance between the excitation module, the detection module and the carbon fiber composite material sample is changed according to the actual working environment within the working distance; and the sample is rapidly scanned in a non-contact and non-damage manner by using the propagation of low-order mode Lamb waves in the sample.

Description

Technical Field

[0001] This invention belongs to the field of non-destructive testing technology of materials using laser ultrasound, and in particular relates to a system for detecting surface damage of carbon fiber composite materials using laser ultrasound. Background Technology

[0002] Composite materials are novel materials created by combining two or more materials, possessing the combined advantages of multiple materials. Carbon fiber composites, for example, are manufactured using resin as the matrix and carbon fiber as the reinforcement through special processes such as pressure application and impregnation. However, during the preparation and use of composite materials, due to their inherent anisotropy, production processes, service environments, and loads, various damages and defects inevitably occur, including porosity, fracture, debonding, buckling, and delamination. To ensure the healthy and undamaged structure of the material, it is necessary to detect defects without damaging the material. Existing technologies mainly employ the following methods:

[0003] Radiographic testing methods have a wide range of applications and can effectively detect volumetric defects such as gaps and inclusions in advanced composite materials. X-ray computed tomography (CT) can accurately characterize the porosity distribution and fiber orientation of damage, but the process poses radiation safety risks. Furthermore, this method has poor ability to identify delamination defects and is insensitive to cracks parallel to the material surface. It also has limitations regarding the size of the test specimen, is time-consuming, and relatively expensive. Eddy current testing has low efficiency and is only suitable for testing conductive specimens such as metallic materials, limiting its application. Penetrant testing is mainly used to detect surface defects, offering advantages such as ease of operation, low cost, and intuitive results, making it suitable for rapid on-site testing. However, it only reflects shallow surface defects and cannot be used for complex materials, making it difficult to accurately determine defect depth.

[0004] Laser ultrasonic nondestructive testing technology uses laser pulses to irradiate the material surface, generating thermal stress through the thermoelastic effect. This stress excites volume waves and guided waves, and the excited ultrasonic waves propagate along the material surface and interior to the receiving system, obtaining signals containing defect information. This allows for the detection of defects in the material structure without causing additional damage to the test piece. Compared to other testing technologies, ultrasonic waves have strong penetrating power, minimal energy loss during propagation within materials, high detection speed, and high sensitivity. They can accurately detect minute defects in workpieces, and the process is safer. Traditional ultrasonic testing technologies are similar in principle to laser ultrasonic testing, but most cannot be used in harsh environments such as high radiation or high pressure. Furthermore, by adjusting laser parameters, the wavelength and amplitude of the excited ultrasonic waves can be controlled to the micrometer level, overcoming the limitations of traditional methods in detecting minute defects and achieving higher precision testing.

[0005] The following methods are mainly used for ultrasonic signal receiving technology:

[0006] Using piezoelectric transducers for receiving signals is a stable and efficient contact coupling method. Piezoelectric ceramics are directly or indirectly bonded to the sample surface via a coupling agent. When ultrasound waves excited by a pulsed laser propagate within the material and reach the receiving point where the piezoelectric transducer is located, the transducer converts the sensed mechanical vibration into an electrical signal. After amplification by a preamplifier, the signal is acquired and stored by a high-precision oscilloscope or data acquisition card. However, this method sacrifices the advantage of being completely non-contact and requires a high degree of surface flatness of the sample.

[0007] Air-coupled ultrasonic wave (ACWS) is an improvement on the piezoelectric method. This method completely eliminates the need for a coupling agent, utilizing air as the ultrasonic transmission medium. A highly sensitive air-coupled transducer receives acoustic signals leaking from the material surface into the air. A typical ACWS system consists of an air-coupled transducer, a data acquisition system, a power amplifier, a preamplifier, and mechanical components. Triggered by a synchronous signal source, the excitation probe of the air-coupled sensor acts on the sample surface, generating thermally excited ultrasonic waves. The receiving probe synchronously receives the ultrasonic waves propagating within the sample and then moves to the next excitation point on the material, repeating the scan. After all excitation points have been detected, the ultrasonic waves carrying sample information are transmitted to a noise amplifier and a low-pass filter. The resulting ultrasonic waveform is displayed on an oscilloscope, and image reconstruction processing is performed to obtain defect information about the sample material. However, due to the extremely low efficiency of solid vibration energy transmission into the air, ultrasonic waves undergo severe reflection and energy loss at the solid-gas interface, resulting in extremely weak signals at the receiving end, typically only one-thousandth or even less than those of the contact method. Furthermore, the micron-diameter focused spot of the air coupling method amplifies the influence of material anisotropy and structural complexity on the ultrasonic signal. The detection device and the surface of the sample being tested can only be a few hundred microns apart, making long-distance detection impossible.

[0008] Based on the analysis of existing technologies, the problems with current detection methods can be summarized as follows:

[0009] The detection direction and sample size are limited; the detection efficiency is low and the receiving distance is limited; the signal noise is high and makes quantitative analysis difficult. Summary of the Invention

[0010] The purpose of this invention is to solve the problems mentioned in the background art and to propose a system for laser ultrasonic detection of surface damage of carbon fiber composite materials.

[0011] To achieve the objective of this invention, this invention provides a system for laser ultrasonic detection of surface damage in carbon fiber composite materials. The system includes an excitation module, an interferometric detection module, and a signal processing unit.

[0012] The excitation module is used to vertically incident the laser onto the sample surface;

[0013] The interferometric detection module is used to capture vibration information in the sample, as well as the propagation information of ultrasonic waves;

[0014] The signal processing unit performs noise reduction processing on the received signal to obtain smooth and interference-free ultrasonic information;

[0015] The excitation module includes a nanosecond pulsed laser (1), optical glass (2), focusing lens (3) and galvanometer (4); the laser output from the pulsed laser (1) passes through the optical glass (2), and after the focusing lens (3) adjusts the spot size, it enters the galvanometer (4).

[0016] Compared with the prior art, the significant advancement of the present invention lies in the fact that the fully non-contact "laser galvanometer-laser interferometry" defect detection system uses a nanosecond pulse laser as the excitation source and a laser Doppler vibrometer as the receiver. It can change the distance between the excitation module, the detection module, and the carbon fiber composite sample within the working distance according to the actual working environment. It utilizes the propagation of low-order mode Lamb waves in the sample to achieve rapid scanning of the sample in a non-contact and non-destructive manner.

[0017] The optimal parameters in the wavelet thresholding denoising algorithm were determined. Based on the sym8 wavelet basis and the wden function, a heuristic threshold was adopted, using signal-to-noise ratio (SNR), root mean square error (RMSE), and correlation coefficient as evaluation indicators. After denoising, the SNR of the signal was improved to 19.04 dB, and the correlation coefficient reached 0.9930. While effectively preserving the characteristics of defect echoes, the influence of noise on the signal was reduced, which is beneficial to subsequent image processing.

[0018] Using the constructed experimental system, the B-scan image obtained after wavelet transform denoising clearly shows the influence of defects on the Lamb wave signal. Wavelet threshold denoising was applied to each extracted A-scan data point, and Hilbert transform was used to extract the signal envelope to generate a C-scan image. Comparison with the actual defect location showed good accuracy in locating the defect center and a 96% accuracy in locating the defect size. The image also exhibited a high degree of similarity to the prepared defect morphology. This demonstrates that the combined defect detection system based on wavelet transform, laser galvanometer excitation, and laser interferometry reception can effectively and rapidly scan samples, effectively locating defects and demonstrating superior accuracy and engineering applicability compared to traditional imaging methods.

[0019] To more clearly illustrate the functional characteristics and structural parameters of the present invention, further explanation is provided below in conjunction with the accompanying drawings and specific embodiments. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the non-contact "laser galvanometer-laser interferometry" defect detection system provided in this application;

[0021] Figure 2 This application provides a schematic diagram of the noise reduction effect and local magnification of the received ultrasonic signal;

[0022] Figure 3 This is a B-scan image of the crack defect provided in this application;

[0023] Figure 4 This is an image showing the imaging inversion results of borehole defects and crack defects provided in this application. Detailed Implementation

[0024] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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.

[0025] Optical methods utilize laser beams as "probes," enabling high-precision detection of targets several meters or even further away, significantly overcoming the spatial limitations of the former two methods. When using optical methods to detect ultrasonic signals, the construction and debugging of the optical path, as well as the selection of detection system components, are crucial. On the one hand, the detector's response bandwidth must completely cover the target's ultrasonic frequency, and the photosensitive surface must possess highly uniform photoelectric properties to ensure stable interference effects. On the other hand, the laser power must be appropriately selected to ensure sufficient light energy to overcome rough scattering from the material surface and obtain a high signal-to-noise ratio, while avoiding excessive power that could increase thermal noise or cause optical damage to the personnel and samples.

[0026] Traditional bulk wave and surface wave techniques have significant limitations for non-destructive testing of surface and near-surface defects in carbon fiber composites, while Lamb waves exhibit unique advantages. The main problems with bulk wave (longitudinal and transverse waves) testing are: the need for point-by-point grid scanning, resulting in low efficiency and difficulty in meeting the requirements for rapid large-area testing; and the weak resolution of minute defects within approximately one wavelength depth below the surface, creating a near-surface blind zone, precisely where impact microcracks and resin-rich layer pores in CFRP are located. Furthermore, the high attenuation and anisotropy of composite materials lead to complex bulk wave sound velocities and difficult signal interpretation. While surface waves are sensitive to surface-opening defects, their energy is concentrated within a single wavelength depth (approximately 0.5-1.5 wavelengths), offering little response to subsurface and internally buried defects, making them prone to missed detection. Simultaneously, CFRP surface roughness, curvature, and layup ends severely interfere with surface wave propagation, generating clutter and reducing the signal-to-noise ratio. Anisotropy also causes drastic changes in surface wave velocity with direction, requiring complex calibration.

[0027] like Figure 1As shown, this application includes an excitation module, an interferometric detection module, and a signal processing unit.

[0028] The excitation module is used to vertically incident the laser onto the sample surface;

[0029] The interferometric detection module is used to capture vibration information in the sample, as well as the propagation information of ultrasonic waves;

[0030] The signal processing unit performs noise reduction processing on the received signal to obtain smooth and interference-free ultrasonic information;

[0031] The excitation module includes a nanosecond pulsed laser (1), optical glass (2), focusing lens (3) and galvanometer (4); the laser output from the pulsed laser (1) passes through the optical glass (2), and after the focusing lens (3) adjusts the spot size, it enters the galvanometer (4);

[0032] Among them, the pulsed laser (1) is used to output a high-energy pulsed laser with a wavelength of 1064nm; the discharge current is set to 105A, the repetition frequency is 100Hz, the single pulse energy is 32.7mJ, and the pulse width is 9.7ns, to excite the low-order mode Lamb wave on the sample (8) and then detect the defect.

[0033] The optical glass (2) is located at the laser output end;

[0034] The focusing lens (3) is used to adjust the spot size to 5 mm, maximizing the energy while ensuring no ablation damage to the carbon fiber plate, and exciting a low-order Lamb wave with a high signal intensity. In order to detect the sample without damage, the laser needs to be lower than the ablation threshold of the sample. By changing the spot size irradiated on the sample surface while keeping the single pulse energy and pulse width constant, the laser power in contact with the sample (8) surface can be changed. Generally, the spot diameter on the sample surface is kept between the wavelength of the detection ultrasonic wave and half the wavelength. At the same time, the larger the spot, the less damage to the sample and the closer it is to non-destructive, but the energy will be lower. The smaller the spot, the higher the received signal intensity, which is more conducive to signal reception and subsequent defect inversion imaging. At the working distance of the example, the spot diameter is reduced to 5 mm without ablation damage to the sample, so that the energy at the excitation point on the sample is higher and the ultrasonic signal is more significant.

[0035] The galvanometer (4) is used to change the direction of the laser beam and reflect the beam that passes through the focusing lens (3) onto the surface of the sample (8). The range of the sample surface is scanned by means of the reflector inside the galvanometer (4).

[0036] The interferometric detection module includes a laser Doppler vibrometer (6), which outputs a probe light with a wavelength of 532nm that is incident on the sample surface. The ultrasonic signal propagating in the scanning plate carries the defect information triggered by the synchronization signal.

[0037] Because the inherent frequency of lasers is extremely high, it is technically difficult to directly measure the minute shift in laser frequency. Therefore, the interferometric detection module uses optical heterodyne interferometry to convert the high-frequency laser signal into a low-frequency beat frequency signal that can be accurately measured, thereby indirectly extracting Doppler frequency shift information.

[0038] The signal processing unit includes a photodetector (7) and a host computer (5); the photodetector (7) is connected to the oscilloscope in the host computer (5); the optical signal refracted by the optical glass (2) is connected to the oscilloscope in the host computer (5) through the photodetector (7) to trigger the synchronous detection of the ultrasonic signal;

[0039] The host computer (5) performs noise reduction processing on the received signal to obtain almost smooth and interference-free ultrasonic information. The host computer (5) selects a wavelet threshold denoising algorithm based on the sym8 wavelet basis and the wden function. With the help of noise reduction processing of the experimental signal, wavelet threshold denoising is performed by heuristic threshold principle through the improvement of the correlation coefficient and signal-to-noise ratio of the denoised waveform. The denoised signal is then separated into modes by continuous wavelet transform, which solves the aliasing problem in the time domain or frequency domain. The time-of-flight extraction and inversion of the full-field scanning points are performed using the denoised signal to generate a C-scan image and obtain defect information.

[0040] This application enables defect detection of carbon fiber sheet materials in a completely non-contact manner, and performs noise reduction processing on the received ultrasonic signals to achieve the identification and inversion imaging of surface defects of the sample.

[0041] The optical transmission path of this application is as follows:

[0042] The excitation module excites a laser beam that is incident perpendicularly onto the sample surface. When the laser energy density is below the material damage threshold, a thermoelastic effect occurs, causing the material surface to absorb light energy within nanoseconds. This instantaneous temperature rise generates localized thermal expansion, which, constrained by surrounding cooler particles, produces thermal stress, thus generating ultrasonic waves. Lamb waves propagate along the carbon fiber direction within the plate. At low frequencies, the A0 mode velocity is much lower than the S0 mode velocity. When the lamb wave encounters a defect in its propagation path, reflection, transmission, scattering, or mode conversion occurs, leading to signal amplitude attenuation and arrival time delay at the receiving point.

[0043] A laser Doppler vibration meter (6) is placed at a certain distance from the excitation point. The laser inside the vibration meter emits a continuous and stable laser beam, which is split into a probe beam and a reference beam by a beam splitter. The probe beam is focused on the test point on the material surface. When the Lamb wave reaches this point, it causes a small out-of-plane displacement of the plate surface, which is particularly significant for the A0 mode. According to the optical Doppler effect, the frequency of the probe beam reflected back from the moving surface will shift, and the shift is proportional to the surface vibration velocity. The vibration meter contains an interferometer, which causes interference between the received reflected light and the internal reference light. The intensity of the interference light changes periodically with the frequency difference between the two beams, forming a beat frequency signal. The internal demodulator converts this beat frequency signal into an analog voltage signal that is proportional to the vibration velocity in real time. After being processed by the built-in filter and amplifier, it is output to the oscilloscope in the host computer (5). The oscilloscope collects and records the acoustic wave signal for subsequent data processing and analysis. The received Lamb wave signal is denoised using wavelet transform, and the defect inversion imaging and precise positioning are obtained by calculating the flight time of the Lamb wave.

[0044] In the experimental system built in this invention, the Nd:YAG pulsed laser emits a laser wavelength of 1064nm, and the laser interferometer in the detection part has a wavelength of 532nm. This combination is because the CFRP surface has a dark color and a high absorption rate of 1064nm infrared light, which can effectively excite the thermoelastic mechanism and generate ultrasound with a high signal-to-noise ratio. In the receiving system, under the same laser power, the single photon energy of 532nm visible light is higher, which can usually obtain a higher signal-to-noise ratio and detection sensitivity, which is beneficial for capturing weak ultrasonic signals caused by tiny defects.

[0045] Figure 2 This is a schematic diagram showing the noise reduction effect and local magnification of the received ultrasonic signal according to the present invention;

[0046] This invention uses signal-to-noise ratio, waveform similarity index, and root mean square error to determine a heuristic threshold selection principle. The wden function automatically selects and adapts the threshold based on the noise level in the signal, automatically calculates the threshold, and performs threshold processing on the noisy high-frequency wavelet coefficients. The wavelet basis sym8 has good symmetry, high reconstruction accuracy, and strong noise suppression capability. After determining the denoising principle and function, the original signal is analyzed and processed to obtain the denoised signal. By comparing the signals before and after denoising with local magnification, it can be seen that although threshold denoising loses a small amount of high-frequency information, it can effectively remove noise from the original signal. It can be applied to the denoising of low-order A0 mode Lamb waves of surface defects in carbon fiber composite materials, as well as the inversion imaging of defects after signal processing.

[0047] Figure 3 This is the B-scan image result of the crack defect according to the present invention;

[0048] To address crack defects in carbon fiber composite plates, the scanning direction is kept approximately parallel to the crack length direction, resulting in the following: Figure 3 The B-scan image shown reveals that acoustic waves scatter at the edge of the defect, while the acoustic waves at the center of the defect are completely reflected, with virtually no transmitted signal. Changing the scanning direction to be perpendicular to the crack length reveals that the transmitted signal is significantly weakened by the defect after the probe light passes through it. Scattered waves are present in the middle of the image, and the Lamb wave peak at the defect center disappears.

[0049] Figure 4 This is an image showing the imaging inversion results of borehole defects and crack defects according to the present invention;

[0050] Envelope extraction is performed on the A-scan signal at the scan point, and time-of-flight extraction is further performed using the signal envelope to obtain the following results: Figure 4 The Time-of-Flight (ToF) image of the borehole defect shows a bright red area at the defect location due to the delay effect of the defect on the acoustic signal. The circle in the image represents the outer contour of the defect in the actual sample, and the experimental system can effectively reflect the defect contour and basic size. The center coordinates of the borehole defect were obtained as (39.85, 40.12) mm through scanning imaging using the experimental system, indicating a relative error of approximately 4% in defect detection size and a center positioning error of 1.2%, demonstrating good center positioning accuracy. However, due to the irregularity of the borehole defect, the circular image edge exhibits a certain radial pattern. Compared to circular borehole defects, the contour imaging of rectangular crack defects deviates more from the actual defect boundary. When the probe passes perpendicularly through the crack, the acoustic wave passes through two closely spaced side edges of the crack, generating complex reflections and other signal interference, leading to severe signal fluctuations in the B-scan image. This results in a difference between the defect image presented in the C-scan and the actual prepared defect, although accurate positioning of the defect center can still be achieved.

[0051] This application has the following advantages:

[0052] (1) Solve the problem of limitations on detection direction and sample size.

[0053] This invention utilizes a laser galvanometer system as the excitation system for ultrasonic waves, achieving non-contact acoustic excitation. The galvanometer system can scan the excitation source at points, lines, and surfaces on the material surface. The core components of the galvanometer include a focusing lens, a servo motor, and a drive module, replacing the traditional bulky mechanical scanning platform. This allows the laser beam to rapidly scan materials at high speed and high precision. The motor driver operates the motor according to input control commands, and two sets of motors work together to adjust the angle of the focusing lens, enabling controllable displacement of the laser focus point in a two-dimensional plane. With the material fixed, the direction and position of the laser excitation point can be changed, effectively improving scanning efficiency compared to scanning methods that use stepper motors to move the material. Furthermore, it allows for applications in various scenarios, regardless of the distance or size of the sample being tested.

[0054] (2) Solve the problems of low detection efficiency and limited receiving distance.

[0055] Lamb waves are employed because they can propagate over long distances in thin plate structures. Their antisymmetric A0 mode is extremely sensitive to near-surface and internal micro-damage. Furthermore, when Lamb waves interact with defects such as delamination and cracks, they produce strong mode conversion, scattering, and energy trapping effects. The extracted characteristic signals are easily used to achieve quantitative imaging of defects. Interferometry is sensitive to Lamb wave propagation. By using a Doppler vibrometer based on the heterodyne interferometry principle to convert high-frequency laser signals into accurately measurable low-frequency beat signals, Doppler frequency shift information can be indirectly extracted, improving the accuracy of signal detection and solving the problem that the receiving device needs to be near the sample surface.

[0056] (3) Solve the problem of high signal noise making quantitative analysis difficult.

[0057] Existing technologies have high signal-to-noise ratios. This invention compares the noise reduction effects of different threshold selection principles and optimizes the parameters of the noise reduction scheme through quantitative evaluation indicators such as signal-to-noise ratio, root mean square error, and waveform similarity index. It suppresses background noise and accurately preserves the transient characteristics of defect echoes. The instantaneous energy features extracted by mode decomposition based on continuous wavelet transform significantly improve the identification accuracy of defect signals.

[0058] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A system for laser ultrasonic detection of surface damage in carbon fiber composite materials, characterized in that, The system includes an excitation module, an interferometric detection module, and a signal processing unit; The excitation module is used to vertically incident the laser onto the sample surface; The interferometric detection module is used to capture vibration information in the sample, as well as the propagation information of ultrasonic waves; The signal processing unit performs noise reduction processing on the received signal to obtain smooth and interference-free ultrasonic information; The excitation module includes a nanosecond pulsed laser (1), optical glass (2), focusing lens (3) and galvanometer (4); the laser output from the pulsed laser (1) passes through the optical glass (2), and after the focusing lens (3) adjusts the spot size, it enters the galvanometer (4).

2. The system according to claim 1, characterized in that, The pulsed laser (1) is used to output a high-energy pulsed laser with a wavelength of 1064nm; the single pulse energy is 32.7mJ and the pulse width is 9.7ns, which excites a low-order mode Lamb wave on the sample (8); The optical glass (2) is located at the laser output end; The focusing lens (3) is used to adjust the size of the light spot, so as to maximize the energy and excite a low-order Lamb wave with a high signal intensity while ensuring that the carbon fiber plate is not ablated and damaged. The galvanometer (4) is used to change the direction of the laser beam and reflect the beam that passes through the focusing lens (3) onto the surface of the sample (8). The range of the sample surface is scanned by means of the reflector inside the galvanometer (4).

3. The system according to claim 2, characterized in that, The focusing lens (3) is adjusted to a spot size of 5mm.

4. The system according to claim 1, characterized in that, The interferometric detection module includes a laser Doppler vibrometer (6); The interferometric detection module uses optical heterodyne interferometry to convert high-frequency laser signals into low-frequency beat signals, thereby indirectly extracting Doppler frequency shift information.

5. The system according to claim 4, characterized in that, The laser Doppler vibrometer (6) outputs a probe light with a wavelength of 532nm, which is incident on the sample surface. It is triggered by a synchronization signal, and the ultrasonic signal propagating in the scanning plate carries the defect information.

6. The system according to claim 4, characterized in that, The signal processing unit includes a photodetector (7) and a host computer (5); The photodetector (7) is connected to the oscilloscope in the host computer (5); The optical signal refracted by the optical glass (2) is connected to the oscilloscope in the host computer (5) through the photodetector (7) to trigger the synchronous detection of the ultrasonic signal; The host computer (5) performs noise reduction processing on the received signal to obtain interference-free ultrasonic information.

7. The system according to claim 6, characterized in that, The host computer (5) selects a wavelet threshold denoising algorithm based on the sym8 wavelet basis and the wden function. With the help of the denoising processing of the experimental signal, the wavelet threshold denoising is performed by the heuristic threshold principle through the improvement of the correlation coefficient and signal-to-noise ratio of the denoised waveform. The denoised signal is then separated into modes by continuous wavelet transform.

Images