A phased array ultrasonic adaptive detection method for curved surface components

Abstract

The present application belongs to the field of high-end equipment manufacturing detection, and discloses a phased array ultrasonic adaptive detection method for curved surface components. The method comprises the following steps: building a phased array ultrasonic detection system for curved surface components; reconstructing point clouds based on single plane wave transmission of the curved surface components by using a linear array ultrasonic probe; perceiving the point clouds by an adaptive gradient weighted Hough circle detection algorithm, and automatically identifying the geometric parameters of the curved surface components; adjusting the pose of the linear array ultrasonic probe according to the geometric parameters of the curved surface components, and calculating a delay rule to make the incident sound wave adapt to the curved surface; performing adaptive two-dimensional interpolation and wavelet denoising collaborative processing on the collected high-fidelity array signals, reconstructing the high-fidelity array signals, and generating high-resolution and high signal-to-noise ratio ultrasonic images. The method organically integrates geometric perception, pose adaptation, data acquisition and signal processing, improves the signal-to-noise ratio and resolution of ultrasonic imaging of curved surface components, and provides support for high-quality detection of curved surface components.

Description

Technical Field

[0001] This invention belongs to the field of high-end equipment manufacturing testing and relates to a phased array ultrasonic adaptive testing method for curved components. Background Technology

[0002] With the pursuit of high-performance structural design in modern industry, components with complex geometric features (such as curved surfaces) are increasingly widely used in aerospace and high-end manufacturing. During manufacturing and use, these curved surface structures are prone to intrinsic defects such as delamination due to stress concentration and complex forming processes. To ensure manufacturing quality and service safety, it is essential to develop reliable non-destructive testing technologies for curved surface components. Currently, phased array ultrasonic technology shows great potential in the inspection of complex curved surface components, but it suffers from poor shape adaptation and severe sound energy attenuation when facing surfaces with unknown geometric information. Surface Adaptive Ultrasound (SAUL) utilizes an iterative calculation of the time delay law to allow the sound beam to adapt to the curved surface of the component and be incident perpendicularly. However, this method requires 4-5 iterations to complete the inspection, resulting in low efficiency. Furthermore, limited by the physical element spacing, its lateral resolution in ultrasonic imaging is insufficient, leading to significant quantitative errors in defects. Therefore, overcoming the bottleneck of data acquisition efficiency for curved surface components and achieving high signal-to-noise ratio and high-resolution imaging is of great significance for high-quality inspection of curved surface components. Summary of the Invention

[0003] This invention proposes a phased array ultrasonic adaptive detection method for curved surface components. This method can rapidly sense the geometric parameters of the curved surface, improve the data acquisition frame rate, generate high-fidelity array signals, and improve imaging resolution, thereby supporting high-quality detection of curved surface components.

[0004] The technical solution of this invention: An adaptive phased array ultrasonic testing method for curved surface components comprises the following steps: (1) Construct a phased array ultrasonic testing system for curved components; A phased array ultrasonic testing system for curved components was constructed: The phased array ultrasonic host, control computer, three-axis stepper platform, and water tank were arranged on the testing platform. The control computer was communicatively connected to both the phased array ultrasonic host and the three-axis stepper platform to control the ultrasonic transceiver and the movement of the linear array ultrasonic probe. The linear array ultrasonic probe was mounted on a probe holder and adjusted to be horizontal, and electrically connected to the phased array ultrasonic host. The curved component was fixed in the testing position using auxiliary fixtures inside the water tank. Water was injected into the water tank until the curved component was submerged. (2) Draw the point cloud of the curved surface component; Based on the phased array ultrasonic testing system for curved components constructed in step (1), in a water-immersed environment, all elements of the linear array ultrasonic probe are simultaneously excited to emit ultrasonic waves toward the curved component. Each element receives the A-scan signal in parallel, and a delay summation algorithm is used to quickly reconstruct the surface contour of the curved component to generate a point cloud. I The calculation formula is: (1) in, t ij For sound waves from the imaging point j To the i The propagation time of each array element S i For the first i A-scan signal acquired by each array element n Indicates the number of array elements; (3) Surface perception and pose adjustment; The adaptive gradient-weighted Hough circle detection algorithm is used to perform perception recognition on the point cloud generated in step (2) and extract the geometric parameters of the curved surface components, including the center position and radius of curvature: (2) in, For point Adaptive voting weights; For point The local gradient vector at that point; For point Pointing to the center of the candidate circle The radial vector, where the candidate circle center is the assumed circle center position in the current traversal; R Represents the radius of the candidate circle, which is formed by... and R A jointly determined virtual circle to be verified; Gradient vector With radial vector The angle between them; For point The amplitude at that point, The gradient consistency sharpening index; This is the allowable distance error; This is an indicator function that takes the value 1 if the condition is true, and 0 otherwise. For the point cloud subsets that participated in the voting; The output contains the coordinates of the center of the circle and the estimated radius of curvature. Based on the geometric parameters of the curved surface component, adjust the pose of the linear array ultrasonic probe so that its central axis passes through the center of the circle in the vertical direction. ; Calculate the delay rule for all array elements to make the emitted ultrasonic waves adapt to the curved surface components, and collect the reflected echoes from the curved surface components in parallel by all array elements to obtain the array signal. The longitudinal wave velocity of curved components was calculated using the ultrasonic pulse reflection method. (3) in, The thickness of the curved surface component. When the surface of a curved component reflects sound waves, When the bottom surface of a curved component reflects sound waves, The longitudinal wave velocity of the curved component; (4) Array signal feature extraction; The array signals acquired in step (3) are decoupled, and the spatiotemporal gradient, amplitude gradient and phase information are extracted to form a multidimensional physical feature set describing the physical characteristics of the sound field. The spatiotemporal gradient is as follows: (4) in, This indicates the decoupling form of the received signal; t This is a time variable, representing the time it takes for the signal to propagate. x , is a spatial variable representing the spatial position of the array elements in a linear array ultrasound probe; The magnitude gradient is as follows: (5) in, Indicates the amplitude of the received signal; The phase information is as follows: (6) in, This indicates the phase information of the received signal; (5) High-fidelity array signal reconstruction; Based on the multidimensional physical features extracted in step (4), the array signal is subjected to adaptive two-dimensional interpolation and wavelet denoising combined processing to obtain a high-fidelity array signal; the adaptive two-dimensional interpolation method is as follows: (7) in, For Akima processing based on magnitude gradient, For Akima functions, Indicates the signal amplitude. These are the magnitude gradient weights; Akima processing based on phase information, Represents phase information, Phase weighting factor; The specific steps for wavelet denoising are as follows: (8) Where Re represents the real-valued signal transformation function, and WPCCD represents the wavelet denoising function. These are wavelet packet coefficients; (6) Curved phased array ultrasound imaging; Using the array signal obtained in step (5) as input, combined with the geometric parameters of the component surface and the longitudinal wave velocity extracted in step (3), phased array ultrasonic imaging is performed on the curved component to improve the imaging quality and detection capability of defects in the curved component.

[0005] The beneficial effects of this invention are as follows: Compared with existing methods, this invention organically integrates geometric perception, pose adaptation, data acquisition, and signal processing, thereby improving the signal-to-noise ratio and resolution of phased array ultrasonic imaging of curved components. This method overcomes the problems of low efficiency, low imaging resolution, and low signal-to-noise ratio in ultrasonic surface adaptation methods, eliminates the dependence on prior geometry of the curved surface, improves the defect detection capability of curved components, and has great potential for engineering applications. Attached Figure Description

[0006] Figure 1 This is a schematic diagram of a phased array ultrasonic testing system for curved components.

[0007] Figure 2 It is a point cloud diagram of a single plane wave propagation surface.

[0008] Figure 3 It is a surface perception map obtained by the adaptive gradient-weighted Hough circle detection algorithm.

[0009] Figure 4 This is a schematic diagram of ultrasonic testing using a curved adaptive phased array.

[0010] Figure 5 This is an example of an A-scan signal before signal processing.

[0011] Figure 6 This is an example of an A-scan signal after signal processing.

[0012] Figure 7 This is the SAUL imaging result of a 3 mm circumferential layered defect on a CFRP surface.

[0013] Figure 8 This is the imaging result of a 3 mm circumferential layered defect on a CFRP surface using this method.

[0014] In the diagram: 1 Phased array ultrasonic main unit, 2 Control computer, 3 Triaxial stepping platform, 4 Water tank, 5 Probe bracket, 6 Linear array ultrasonic probe, 7 Auxiliary tooling, 8 Curved surface component. Detailed Implementation

[0015] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings and technical solutions.

[0016] (1) Construct a phased array ultrasonic testing system for curved components; Construct a phased array ultrasonic testing system for curved surface components, such as... Figure 1 As shown, a phased array ultrasonic transceiver, a control computer, a three-axis stepper platform, and a water tank are arranged on the detection platform. The control computer is communicatively connected to both the phased array ultrasonic transceiver and the three-axis stepper platform to control the ultrasonic transceiver and the movement of the linear array ultrasonic probe. The linear array ultrasonic probe is mounted on a probe holder and adjusted to be horizontal, and is electrically connected to the phased array ultrasonic transceiver. The curved component is fixed to the detection position by auxiliary fixtures in the water tank. Water is injected into the water tank until the curved component is submerged. The phased array ultrasonic probe is an Olympus 2.25L32-A11 with 32 array elements, a center-to-center distance of 0.6 mm, a width of 0.52 mm per element, and an excitation center frequency of 2.25 MHz. The curved component is made of carbon fiber reinforced plastic (CFRP) composite material with a density of 1.59 g / cm³. 3 The total thickness is 11.84 mm, the total number of layers is 64, the radius of curvature is 8 mm, and the center angle is 90°; the delamination defect is located at the geometric center of the curved surface, with a depth of 5.92 mm and a circumferential length of 3 mm; the coupling medium is water. (2) Draw the point cloud of the curved surface component; Based on the phased array ultrasonic testing system for curved components in step (1), in a water-immersed environment, all elements of the linear array ultrasonic probe are simultaneously excited to emit plane waves to the curved component. Each element receives the A-scan signal in parallel. According to formula (1), the surface contour of the curved component is rapidly reconstructed using a delay summation algorithm to generate a high-precision point cloud, such as... Figure 2 As shown.

[0017] (3) Surface perception and pose adjustment; After obtaining the point cloud, the adaptive gradient-weighted Hough circle detection algorithm is used to perceive and identify the generated point cloud. The center coordinates of the curved surface component are extracted as (0 mm, 4.3 mm) and the radius of curvature is 8 mm using formula (2). Figure 3 As shown; based on the extracted center coordinates and radius of curvature, the pose of the phased array ultrasonic probe is adjusted so that its central axis passes through the center of the circle in the vertical direction. At the same time, the vertical distance between the linear array ultrasonic probe and the center of the circle is set to 9.3 mm (half of the array aperture) to ensure that the emitted sound waves can cover the contour of the curved surface component; then the delay rule of all array elements is calculated (where the maximum delay time is 2.6 μs) to make the emitted ultrasonic waves adapt to the geometry of the curved surface. Figure 4All array elements collect reflected echoes from the curved surface components in parallel to obtain array signals; the longitudinal wave velocity of the curved surface components is calculated to be 3033 m / s and the sound velocity of water is 1485 m / s using the ultrasonic pulse reflection method (formula (3)).

[0018] (4) Array signal feature extraction; The array signals acquired in step (3) are decoupled, and the spatiotemporal gradient, amplitude gradient, and phase information of the decoupled signals are extracted according to formulas (4), (5), and (6) respectively, forming a multidimensional feature set describing the physical characteristics of the sound field. This provides guidance for subsequent high-fidelity array signal reconstruction; (5) High-fidelity array signal reconstruction; Based on the multidimensional physical features extracted in step (4), the array signal is subjected to adaptive two-dimensional interpolation (Formula (7)) and wavelet denoising (Formula (8)) in a coordinated process to obtain a high-fidelity array signal; compared with Figure 5 The A-scan signal after noise reduction ( Figure 6 Noise is significantly suppressed, and defect echoes are clearly identifiable.

[0019] (6) Curved phased array ultrasound imaging; Using the high-fidelity array signal obtained in step (5) as input, and combining the geometric parameters of the curved component and the longitudinal wave velocity extracted in step (3), phased array ultrasonic imaging is performed on the curved component. Figure 7 and Figure 8 A comparison of defect imaging results between the SAUL method and the method of this invention is presented: SAUL imaging shows blurred layered defects and significant noise interference; the images obtained by the method of this invention show clear and intuitive layered defects, with a signal-to-noise ratio improvement of 20.3 dB compared to the SAUL method, and a quantitative error reduction of 3.3% to 0.1 mm. Furthermore, the data acquisition frame rate of the method of this invention is twice that of the SAUL method. In summary, this invention achieves a simultaneous improvement in imaging quality, detection efficiency, and detection capability in the detection of curved components.

Claims

1. A phased array ultrasonic adaptive detection method for curved surface components, characterized in that, The steps are as follows: (1) Construct a phased array ultrasonic testing system for curved components; A phased array ultrasonic testing system for curved components was constructed: The phased array ultrasonic host, control computer, three-axis stepper platform, and water tank were arranged on the testing platform. The control computer was communicatively connected to both the phased array ultrasonic host and the three-axis stepper platform to control the ultrasonic transceiver and the movement of the linear array ultrasonic probe. The linear array ultrasonic probe was mounted on a probe holder and adjusted to be horizontal, and electrically connected to the phased array ultrasonic host. The curved component was fixed in the testing position using auxiliary fixtures inside the water tank. Water was injected into the water tank until the curved component was submerged. (2) Draw the point cloud of the curved surface component; Based on the phased array ultrasonic testing system for curved components constructed in step (1), in a water-immersed environment, all elements of the linear array ultrasonic probe are simultaneously excited to emit ultrasonic waves toward the curved component. Each element receives the A-scan signal in parallel, and a delay summation algorithm is used to quickly reconstruct the surface contour of the curved component to generate a point cloud. I The calculation formula is: in, t ij For sound waves from the imaging point j To the i The propagation time of each array element S i For the first i A-scan signal acquired by each array element n Indicates the number of array elements; (3) Surface perception and pose adjustment; (4) Array signal feature extraction; (5) High-fidelity array signal reconstruction; (6) Curved phased array ultrasound imaging; Using the high-fidelity array signal obtained in step (5) as input, and combining the geometric parameters of the component surface and the longitudinal wave velocity extracted in step (3), phased array ultrasonic imaging is performed on the curved component to improve the imaging quality and detection capability of defects in the curved component.

2. The phased array ultrasonic adaptive detection method for curved surface components according to claim 1, characterized in that, The specific implementation process of step (3) is as follows: The adaptive gradient-weighted Hough circle detection algorithm is used to perform perception recognition on the point cloud generated in step (2) and extract the geometric parameters of the curved surface components, including the center position and radius of curvature: in, For point Adaptive voting weights; For point The local gradient vector at that point; For point Pointing to the center of the candidate circle The radial vector, where the candidate circle center is the assumed circle center position in the current traversal; R Represents the radius of the candidate circle, which is formed by... and R A jointly determined virtual circle to be verified; Gradient vector With radial vector The angle between them; For point The amplitude at that point, The gradient consistency sharpening index; This is the allowable distance error; This is an indicator function that takes the value 1 if the condition is true, and 0 otherwise. For the point cloud subsets that participated in the voting; The output contains the coordinates of the center of the circle and the estimated radius of curvature. Based on the geometric parameters of the curved surface component, adjust the pose of the linear array ultrasonic probe so that its central axis passes through the center of the circle in the vertical direction. ; Calculate the delay rule for all array elements to make the emitted ultrasonic waves adapt to the curved surface components, and collect the reflected echoes from the curved surface components in parallel by all array elements to obtain the array signal. The longitudinal wave velocity of curved components was calculated using the ultrasonic pulse reflection method. in, The thickness of the curved surface component. When the surface of a curved component reflects sound waves, When the bottom surface of a curved component reflects sound waves, Let be the longitudinal wave velocity of the curved component.

3. The phased array ultrasonic adaptive detection method for curved surface components according to claim 2, characterized in that, The specific implementation process of step (4) is as follows: The array signals acquired in step (3) are decoupled, and the spatiotemporal gradient, amplitude gradient and phase information are extracted to form a multidimensional physical feature set describing the physical characteristics of the sound field. The spatiotemporal gradient is as follows: in, This indicates the decoupling form of the received signal; t This is a time variable, representing the time it takes for the signal to propagate. x , is a spatial variable representing the spatial position of the array elements in a linear array ultrasound probe; The magnitude gradient is as follows: in, Indicates the amplitude of the received signal; The phase information is as follows: in, This indicates the phase information of the received signal.

4. The phased array ultrasonic adaptive detection method for curved surface components according to claim 1, characterized in that, The specific implementation process of step (5) is as follows: Based on the multidimensional physical features extracted in step (4), the array signal is subjected to adaptive two-dimensional interpolation and wavelet denoising combined processing to obtain a high-fidelity array signal; the adaptive two-dimensional interpolation method is as follows: in, For Akima processing based on magnitude gradient, For Akima functions, Indicates the signal amplitude. These are the magnitude gradient weights; Akima processing based on phase information, Represents phase information, Phase weighting factor; The specific steps for wavelet denoising are as follows: Where Re represents the real-valued signal transformation function, and WPCCD represents the wavelet denoising function. These are the wavelet packet coefficients.

Images