Nondestructive testing device for cracks at root of wind turbine blade

Abstract

The application discloses a wind turbine blade root crack nondestructive testing device, relates to the technical field of nondestructive testing, and mainly aims to eliminate a detection blind area and avoid crack missed detection, so that the detection precision of the detection device is improved. The main technical scheme of the application is as follows: the detection device comprises a self-adaptive curved surface coupling assembly, which comprises a flexible air bag array and a plurality of ultrasonic probes, the flexible air bag array comprises a plurality of flexible air bags arranged in a ring array, and adjacent two flexible air bags are communicated through a micro flow channel; each ultrasonic probe is connected with each flexible air bag in a corresponding mode; an edge computing analysis unit is connected with the self-adaptive curved surface coupling assembly, a deep learning algorithm is built in the edge computing analysis unit; a mobile platform is connected with the self-adaptive curved surface coupling assembly and is used for driving the self-adaptive curved surface coupling assembly to move along a preset path; and a central processing unit is connected with the edge computing analysis unit and the mobile platform.

Description

Technical Field

[0001] This invention relates to the field of non-destructive testing technology, and more specifically, to a non-destructive testing device for root cracks in wind turbine blades. Background Technology

[0002] As the core component of wind power generation systems, wind turbine blades are subjected to complex alternating loads (such as aerodynamic loads, gravity loads, centrifugal forces, etc.) and harsh environments (such as wind and sand erosion, ultraviolet radiation, salt spray corrosion, etc.) for a long time, which makes the root area a high-incidence site for crack initiation and propagation. Therefore, it is necessary to detect and maintain the root cracks of the blades in a timely manner.

[0003] Then, because the curved surface at the root of the blade is usually quite complex, conventional testing equipment is difficult to fit closely to the root of the blade, which can easily create blind spots and lead to missed cracks. Summary of the Invention

[0004] In view of this, the present invention provides a non-destructive testing device for root cracks of wind turbine blades, the main purpose of which is to eliminate blind spots in the detection and avoid missed cracks, thereby improving the detection accuracy of the device.

[0005] To achieve the above objectives, the present invention mainly provides the following technical solutions: This invention provides a non-destructive testing device for root cracks in wind turbine blades, comprising: An adaptive curved surface coupling component includes a flexible airbag array and multiple ultrasonic probes. The flexible airbag array comprises multiple flexible airbags arranged in a ring array, with adjacent flexible airbags connected by microchannels. The flexible airbag array is used to fit over the root of the blade. Each ultrasonic probe is correspondingly connected to each flexible airbag. The flexible airbag array is used to inflate and expand to make each ultrasonic probe conform to the curved surface of the blade root. An edge computing analysis unit is connected to the adaptive curved surface coupling component, and the edge computing analysis unit has a built-in deep learning algorithm. A mobile platform, connected to the adaptive surface coupling component, is used to drive the adaptive surface coupling component to move along a preset path; The central processing unit is connected to the edge computing analysis unit and the mobile platform.

[0006] Optionally, the surface of the flexible airbag is coated with a conductive elastic material.

[0007] Optionally, the surface roughness of the conductive elastic material is less than or equal to 0.8 micrometers.

[0008] Optionally, the non-destructive testing device for root cracks in wind turbine blades further includes: A phased array ultrasound module is connected to the edge computing analysis unit, and the phased array ultrasound module has a built-in 64-element probe array; The edge computing analysis unit includes a dynamic focusing algorithm module; the phased array ultrasonic module is used to perform layered imaging in the crack depth direction using the dynamic focusing algorithm.

[0009] Optionally, the non-destructive testing device for root cracks in wind turbine blades further includes: A multimodal sensing module is connected to the edge computing analysis unit and the mobile platform. The multimodal sensing module includes an infrared thermal imaging probe, an acoustic emission sensor, and an eddy current detection unit.

[0010] Optionally, the mobile platform includes a three-axis robotic arm and a magnetic adsorption track. The three-axis robotic arm and the magnetic adsorption track are respectively connected to the central processing unit. The magnetic adsorption track is disposed at the root of the blade and extends circumferentially along the root of the blade. The three-axis robotic arm is movably disposed on the magnetic adsorption track. The adaptive curved surface coupling component, the phased array ultrasonic module, and the multimodal sensing module are mounted on the three-axis robotic arm.

[0011] Optionally, an encoder is installed inside the magnetic adsorption track, which is used to provide real-time feedback on the position information of the three-axis robotic arm.

[0012] Optionally, the surface of the magnetically attached track is provided with a self-cleaning coating.

[0013] Optionally, the non-destructive testing device for root cracks in wind turbine blades further includes: A laser marking device, connected to the central processing unit, is used to project a positioning mark onto the location of the crack when a crack is detected at the root of the blade.

[0014] Optionally, the edge computing analysis unit includes a transfer learning module, which supports rapid adaptation of crack features for different blade models by updating the crack feature database online.

[0015] By employing the above technical solution, the present invention has at least the following beneficial effects: In the wind turbine blade root crack non-destructive testing device provided in this embodiment of the invention, the flexible airbag array of the adaptive curved surface coupling component includes multiple flexible airbags arranged in a ring array, and adjacent airbags are connected through microchannels. Each ultrasonic probe is connected to each flexible airbag, so that the flexible airbag array forms a cavity design, with each independent air cavity corresponding to an ultrasonic probe, and the air cavities are connected through microchannels. Thus, by adjusting the air pressure in different zones, each flexible airbag can drive the corresponding ultrasonic probe to accurately fit the blade root to adapt to the curvature change region at the blade root, thereby completely eliminating the detection blind zone, avoiding the phenomenon of missed crack detection, and improving the detection accuracy of the testing device. Attached Figure Description

[0016] Figure 1 This is a structural block diagram of a non-destructive testing device for root cracks in wind turbine blades, provided in an embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the preferred embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Some embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0018] like Figure 1 As shown, this embodiment of the invention provides a non-destructive testing device for root cracks in wind turbine blades, including an adaptive curved surface coupling component 1. The adaptive curved surface coupling component 1 includes a flexible airbag array 11 and multiple ultrasonic probes 12. The flexible airbag array 11 includes multiple flexible airbags arranged in a ring array, with adjacent flexible airbags connected by microchannels. The flexible airbag array 11 is used to fit over the root of the blade. Each ultrasonic probe 12 is correspondingly connected to each flexible airbag. The flexible airbag array 11 is used to inflate and expand so that each ultrasonic probe 12 fits against the curved surface of the blade root. An edge computing analysis unit 2 is connected to the adaptive curved surface coupling component 1 and has a built-in deep learning algorithm. A moving platform 3 is connected to the adaptive curved surface coupling component 1 and is used to drive the adaptive curved surface coupling component 1 to move along a preset path. A central processing unit 4 is connected to the edge computing analysis unit 2.

[0019] The edge computing analysis unit 2 incorporates a deep learning-based crack feature extraction algorithm to analyze the data collected by the adaptive curved surface coupling component 1 and output crack features. The central processing unit 4 coordinates the movement of the mobile platform 3 and the inflation of the flexible airbag, ensuring that the mobile platform 3 drives the adaptive curved surface coupling component 1 to accurately scan and detect the blade root according to the preset detection path. Furthermore, the edge computing analysis unit 2 feeds back the analyzed results to the central processing unit 4, which then adjusts the control strategy accordingly.

[0020] The air pump inflates the flexible airbags. Multiple flexible airbags are connected in series through microchannels to form a unified pressure system. During inflation, the pressure of each airbag can be automatically adjusted, thereby driving each ultrasonic probe 12 to adaptively conform to the curved surface at the root of the blade.

[0021] In the wind turbine blade root crack non-destructive testing device provided in this embodiment of the invention, the flexible airbag array 11 of the adaptive curved surface coupling component 1 includes multiple flexible airbags arranged in a ring array, and adjacent airbags are connected through microchannels. Each ultrasonic probe 12 is connected to each flexible airbag, so that the flexible airbag array 11 forms a cavity design, with each independent air cavity corresponding to one ultrasonic probe 12, and the air cavities are connected through microchannels. Thus, by adjusting the air pressure in different zones, each flexible airbag can drive the corresponding ultrasonic probe 12 to accurately fit the blade root, so as to adapt to the curvature change area of ​​the blade root (such as the flange and bolt sleeve joint), thereby completely eliminating the detection blind zone, avoiding the phenomenon of missed crack detection, and improving the detection accuracy of the testing device.

[0022] In some embodiments, the surface of the flexible airbag may be covered with a conductive elastic material.

[0023] In some embodiments, the surface roughness of the conductive elastic material is less than or equal to 0.8 micrometers.

[0024] Among them, the conductive elastic material can shield redundant and cluttered signals, reduce signal attenuation, and improve the ultrasonic echo signal-to-noise ratio, making it suitable for anisotropic detection of carbon fiber composite materials and further improving the detection accuracy of the ultrasonic probe 12. Specifically, the conductive elastic material can be a silver nanowire-silicone rubber composite conductive material, which has a low resistivity (≤10 Ω·cm). - ³Ω With high tensile deformation capacity (≥200%) and surface roughness Ra≤0.8μm, the ultrasonic probe 12 can achieve micron-level close contact with the curved surface at the root of the blade, thereby significantly improving the ultrasonic signal coupling efficiency to over 80%.

[0025] In some embodiments, see Figure 1The non-destructive testing device for root cracks in wind turbine blades may also include a phased array ultrasonic module 5, which is connected to the edge computing analysis unit 2. The phased array ultrasonic module 5 has a built-in 64-element probe array. The edge computing analysis unit 2 includes a dynamic focusing algorithm module. The phased array ultrasonic module 5 is used to perform layered imaging in the crack depth direction through the dynamic focusing algorithm.

[0026] The phased array ultrasonic module 5 incorporates a 64-element probe array with a center frequency of 5MHz and a resolution of 0.1mm, supporting total focusing (TFM) and plane wave imaging (PWI). The dynamic focusing algorithm employs adaptive beamforming technology. In phased array ultrasonic testing, the dynamic focusing algorithm uses adaptive beamforming technology to adjust the focusing parameters in real time according to the sound velocity differences in the blade material, thereby compensating for the acoustic distortion caused by the anisotropy of the composite material in the blade. This can improve the detection depth resolution to 0.1mm, significantly enhancing the detection accuracy of the device.

[0027] Understandably, the phased array ultrasonic module 5 incorporates a 64-element probe array. By controlling the time delay (phase) of the ultrasonic signal emitted by each element, the direction and focus of the sound beam can be synthesized and controlled. Adaptive beamforming technology dynamically calculates and adjusts these delay parameters based on the real-time received echo signals, ensuring that the sound beam is always optimally focused on the crack, thereby achieving higher detection resolution and signal-to-noise ratio.

[0028] Changes in sound speed can cause shifts in the propagation path and focal point of ultrasonic waves. Dynamic focusing algorithms analyze the received signals in real time, detect these differences in sound speed, and immediately recalculate the delay parameters required for focusing, ensuring that the sound beam can be accurately focused on the current detection area, unaffected by material differences.

[0029] Among them, the dynamic focusing algorithm: Adaptive beamforming technology adjusts delay parameters in real time to compensate for differences in sound velocity in composite materials (formula):

[0030] in, This refers to the time delay (the time difference in signal propagation). This refers to the propagation distance (the path length of a signal propagating in a medium). The incident angle (the angle between the direction of signal incidence and the direction of propagation); The propagation speed in the medium is the speed of sound in the blade material (carbon fiber: 3000 m / s, glass fiber: 2700 m / s).

[0031] In some embodiments, see Figure 1The non-destructive testing device for root cracks in wind turbine blades may also include a multimodal sensing module 6, which is connected to the edge computing analysis unit 2 and the mobile platform 3. The multimodal sensing module 6 may include an infrared thermal imaging probe 61, an acoustic emission sensor 62, and an eddy current detection unit 63.

[0032] The infrared thermal imaging probe 61 is used to capture thermal anomalies caused by cracks. The acoustic emission sensor 62 is used to monitor elastic stress wave signals generated by the propagation of cracks inside the blade. The eddy current detection unit 63 is used to detect changes in the conductivity of the blade surface. By integrating data from infrared thermal imaging, acoustic emission sensor 62, and eddy current detection, defects are identified from multiple dimensions, including thermal anomalies, stress wave signals, and changes in conductivity. Sensor data is transmitted to the central processing unit 4 in real time via a synchronous acquisition circuit. The synchronous acquisition circuit integrates a timestamp alignment module and adopts the IEEE 1588 precision clock protocol to ensure that the time synchronization error of the multimodal sensor data is ≤1μs, reducing the false alarm rate to 0.3%. IEEE 1588 PTPv2 protocol: the master clock uses an OCXO temperature-controlled crystal oscillator (±0.1ppb stability), and the slave clock uses a TCXO temperature-compensated crystal oscillator. Hardware timestamps (implemented by FPGA) ensure that the time synchronization error of the multimodal data is ≤1μs.

[0033] Edge computing analysis unit 2 incorporates a deep learning-based crack feature extraction algorithm to fuse and analyze the aforementioned multi-source sensor data, distinguishing between real cracks and noise interference, and outputting a three-dimensional model of crack length, depth, and propagation trend. Edge computing analysis unit 2, based on the ResNet-50 architecture, fuses multimodal data (infrared thermograms, acoustic emission spectra, eddy current impedance) to differentiate between cracks and noise (accuracy ≥ 95.2%).

[0034] In some embodiments, see Figure 1 The mobile platform 3 may include a three-axis robotic arm 31 and a magnetic adsorption track 32. The three-axis robotic arm 31 and the magnetic adsorption track 32 are respectively connected to the central processing unit 4. The magnetic adsorption track 32 is set at the root of the blade and extends circumferentially along the root of the blade. The three-axis robotic arm 31 is movably set on the magnetic adsorption track 32. The adaptive curved surface coupling component 1, the phased array ultrasonic module 5 and the multimodal sensing module 6 are set on the three-axis robotic arm 31.

[0035] In some embodiments, an encoder is provided inside the magnetic adsorption track 32, which is used to provide real-time feedback on the position information of the three-axis robotic arm 31.

[0036] The three-axis robotic arm 31 is driven by a servo motor, with a repeatability of ±0.05mm. It carries the detection components and moves circumferentially along the root of the blade via a magnetically attached track 32. The magnetically attached track 32 has a built-in absolute encoder (0.1mm resolution) that provides real-time feedback on the position of the three-axis robotic arm 31, ensuring that the detection path planning error is <0.5%. It supports automated grid scanning (overlap rate ≥10%, meeting the GB / T42592-2023 Class C inspection requirements).

[0037] The central processing unit 4 generates the detection path, and the three-axis robotic arm 31, equipped with the detection components, covers the root of the blade according to the generated detection path, achieving automated coverage of the detection path. The encoder corrects position deviations in real time to ensure precise focusing of the phased array probe array.

[0038] In some embodiments, the surface of the magnetic adsorption track 32 may be provided with a self-cleaning coating. The magnetic adsorption track 32 may be made of a high magnetic permeability alloy and have a self-cleaning coating on its surface, thereby ensuring that the magnetic adsorption track 32 can operate continuously and stably under harsh working conditions.

[0039] In some embodiments, see Figure 1 The non-destructive testing device for cracks at the root of the wind turbine blade may also include a laser marking device 7, which is connected to the central processing unit 4 and is used to project a positioning mark to the location of the crack when a crack is detected at the root of the blade.

[0040] The laser marking device 7 is linked with the central processing unit 4. When a crack is detected, a high-precision laser head (wavelength 532nm, power 50mW) is triggered to project a positioning mark at the defect location. The laser marking device 7 adopts high-precision laser marking technology (accuracy ±0.5mm), supports nighttime visual recognition, and includes the crack level and detection timestamp in the marking information, which is convenient for subsequent maintenance and tracking.

[0041] In some embodiments, see Figure 1 The edge computing analysis unit 2 may include a transfer learning module 21. The transfer learning module 21 supports online updates of the crack feature database (such as porosity, delamination, and crack propagation trend) to quickly adapt the crack features of different types of blades (such as carbon fiber and glass fiber composites). The model training cycle is shortened to less than 30 minutes. It can quickly adapt the features of different types of blades within 30 minutes and output a three-dimensional model in real time that includes crack length (accuracy ±0.2mm), depth (accuracy ±0.1mm), and propagation trend prediction, providing comprehensive and reliable data support for blade health assessment.

[0042] In this embodiment of the invention, a segmented flexible airbag array 11 is employed. Through a zoned air pressure regulation mechanism, it precisely conforms to the complex curved surface of the blade root, completely eliminating detection blind spots. The silver nanowire-silicone rubber composite conductive material possesses low resistivity (≤10). - ³Ω With high tensile deformation capacity (≥200%), the ultrasonic probe 12 achieves micron-level close contact with the curved surface, significantly improving coupling efficiency to over 80%. Simultaneously, the phased array ultrasonic module 5, equipped with a dynamic focusing algorithm and adaptive beamforming technology, effectively compensates for acoustic distortion caused by the anisotropy of composite materials, improving detection depth resolution to 0.1 mm and significantly enhancing detection accuracy.

[0043] In this embodiment of the invention, by integrating three modal data—infrared thermal imaging, acoustic emission sensor 62, and eddy current detection—defects are identified from multiple dimensions, including thermal anomalies, stress wave signals, and changes in conductivity characteristics. High-precision data association with a time synchronization error ≤1μs is achieved using the IEEE 1588 protocol, reducing the false alarm rate to 0.3%. The intelligent edge computing system incorporates a transfer learning module 21, which can quickly adapt to the characteristics of different blade models within 30 minutes, and output a three-dimensional model in real time containing crack length (accuracy ±0.2mm), depth (accuracy ±0.1mm), and propagation trend prediction, providing comprehensive and reliable data support for blade health assessment.

[0044] In this embodiment of the invention, a high-precision automated inspection platform is constructed, with the magnetic adsorption track 32 achieving a positioning accuracy of 0.1mm. Combined with the three-axis robotic arm 31, it achieves full circumferential coverage inspection of the blade root. The encoder provides real-time feedback of position information, ensuring that the inspection path planning error is <0.5%. The intelligent marking system adopts high-precision laser marking technology (accuracy ±0.5mm), supports nighttime visual recognition, and includes crack level and inspection timestamp in the marking information, facilitating subsequent maintenance and tracking.

[0045] The non-destructive testing device for root cracks in wind turbine blades provided in this invention has excellent adaptability to extreme environments. The self-cleaning coating on the track ensures continuous and stable operation under harsh conditions. The surface roughness Ra of the conductive elastic material is ≤0.8μm, which effectively avoids signal attenuation. In terms of testing efficiency, compared with traditional manual testing, the single testing cycle is significantly shortened to within 20 minutes. Moreover, the edge computing system enables real-time on-site analysis, completely eliminating data transmission delays and significantly improving testing efficiency and work performance, providing a solid guarantee for the safe operation of wind turbine blades.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A non-destructive testing device for root cracks in wind turbine blades, characterized in that, include: An adaptive curved surface coupling component includes a flexible airbag array and multiple ultrasonic probes. The flexible airbag array comprises multiple flexible airbags arranged in a ring array, with adjacent flexible airbags connected by microchannels. The flexible airbag array is used to fit over the root of the blade. Each ultrasonic probe is correspondingly connected to each flexible airbag. The flexible airbag array is used to inflate and expand to make each ultrasonic probe conform to the curved surface of the blade root. An edge computing analysis unit is connected to the adaptive curved surface coupling component, and the edge computing analysis unit has a built-in deep learning algorithm. A mobile platform, connected to the adaptive surface coupling component, is used to drive the adaptive surface coupling component to move along a preset path; The central processing unit is connected to the edge computing analysis unit and the mobile platform.

2. The non-destructive testing device for root cracks in wind turbine blades according to claim 1, characterized in that, The surface of the flexible airbag is covered with a conductive elastic material.

3. The non-destructive testing device for root cracks in wind turbine blades according to claim 2, characterized in that, The surface roughness of the conductive elastic material is less than or equal to 0.8 micrometers.

4. The non-destructive testing device for root cracks in wind turbine blades according to claim 1, characterized in that, Also includes: A phased array ultrasound module is connected to the edge computing analysis unit, and the phased array ultrasound module has a built-in 64-element probe array; The edge computing analysis unit includes a dynamic focusing algorithm module; the phased array ultrasonic module is used to perform layered imaging in the crack depth direction using the dynamic focusing algorithm.

5. The non-destructive testing device for root cracks in wind turbine blades according to claim 4, characterized in that, Also includes: A multimodal sensing module is connected to the edge computing analysis unit and the mobile platform. The multimodal sensing module includes an infrared thermal imaging probe, an acoustic emission sensor, and an eddy current detection unit.

6. The non-destructive testing device for root cracks in wind turbine blades according to claim 5, characterized in that, The mobile platform includes a three-axis robotic arm and a magnetic adsorption track. The three-axis robotic arm and the magnetic adsorption track are respectively connected to the central processing unit. The magnetic adsorption track is located at the root of the blade and extends circumferentially along the root of the blade. The three-axis robotic arm is movably located on the magnetic adsorption track. The adaptive curved surface coupling component, the phased array ultrasonic module, and the multimodal sensing module are mounted on the three-axis robotic arm.

7. The non-destructive testing device for root cracks in wind turbine blades according to claim 6, characterized in that, An encoder is installed inside the magnetic adsorption track, which is used to provide real-time feedback on the position information of the three-axis robotic arm.

8. The non-destructive testing device for root cracks in wind turbine blades according to claim 6, characterized in that, The surface of the magnetically adsorbed track is coated with a self-cleaning coating.

9. The non-destructive testing device for root cracks in wind turbine blades according to claim 1, characterized in that, Also includes: A laser marking device, connected to the central processing unit, is used to project a positioning mark onto the location of the crack when a crack is detected at the root of the blade.

10. The non-destructive testing device for root cracks in wind turbine blades according to claim 1, characterized in that, The edge computing analysis unit includes a transfer learning module, which supports rapid adaptation of crack features for different blade models by updating the crack feature database online.

Images