Phased array ultrasonic shaft component defect detection method based on grid adaptive optimization

Abstract

The present invention belongs to the field of ultrasonic nondestructive testing technology, and specifically relates to a phased array ultrasonic shaft component defect detection method based on grid adaptive optimization, which includes: S1, establishing a polar coordinate system mapping model for sampling points of shaft components; S2, iteratively obtaining a mapping statistical matrix of the number of sampling points of shaft components, and adaptively optimizing the mapping matrix grid size; S3, using phased array ultrasound to detect shaft components, and obtaining ultrasonic imaging data based on the matrix grid size optimization result; S4, using Hilbert transform and Gaussian filtering to process the ultrasonic imaging data, identify and extract defect areas, and obtain a submatrix to quantify the size of the defects. The present invention optimizes the mapping grid size based on the sampling point distribution and the polar coordinate system, reduces the phase information superposition error caused by the sampling point distribution in the ultrasonic detection of curved surface components, thereby interfering with the defect size quantification, improves the accuracy of the results, and improves the service life of curved surface components in the industrial field.

Description

Technical Field

[0001] The present invention belongs to the technical field of ultrasonic nondestructive testing, and in particular relates to a phased array ultrasonic shaft component defect detection method based on grid adaptive optimization. Background Art

[0002] Ultrasonic nondestructive evaluation (UDE) has become increasingly popular as a standardized tool for defect detection in high-reliability components. Currently, UDE for large components with complex surface geometries still primarily relies on water immersion testing. However, with the growing demand for improved component lifespan and reliability in many industries, such as power generation and aviation, damage tolerance design and assessment are becoming increasingly important. Consequently, detection of smaller defects, as well as improved spatial resolution and quantification accuracy, are required. The emergence of new manufacturing methods, such as additive manufacturing, can produce materials with microstructures that differ from traditional materials to meet the evolving demands of the manufacturing industry. Due to manufacturing process limitations, components may contain microstructures such as non-uniform large grains, pores, and microcracks that negatively impact component service life. The presence of these microstructures results in a lower signal-to-noise ratio for ultrasonic signals compared to traditional materials. Furthermore, due to the complex surface geometries of components, existing methods have limitations in detecting and quantifying submillimeter defects. Optimizing data processing and analysis methods is necessary to improve the signal-to-noise ratio and detect smaller defects despite the noise scattering caused by the material's microstructural characteristics. Furthermore, improved spatial resolution is required to accurately identify closely spaced defects, enabling more precise design assessments.

[0003] There are two main approaches to improving spatial resolution and signal-to-noise ratio. The first involves acquiring inspection data using a large probe and incorporating the full focusing method for data post-processing. This method combines the transmit and receive modes of the array elements to generate a complete time-domain signal dataset. The process involves initially transmitting from one element while all elements receive simultaneously. Then, through electronic movement, ultrasound waves are transmitted from each element in the entire array until all elements have completed a transmission. The full focusing method is a data processing technique used to analyze data acquired from a large array probe. An ultrasound image is constructed by calculating the time it takes for ultrasound waves to reach each focal point within the inspection area from each transmitting element. The value at each focal point corresponds to the signal feedback at that location. In this method, the focusing range and signal-to-noise ratio are directly dependent on the probe size; therefore, a large probe with a large number of array elements is required to acquire data. For components with curved surfaces, a custom-designed probe is required to effectively utilize all array elements under contact inspection conditions. However, this method has limitations when testing components with varying curvatures, such as blades. The second method is based on synthetic aperture focusing technology (SAFT), using an ordinary small-sized phased array probe to obtain echo signals from different scanning angles and positions by moving along the detection path. The basic idea is to process the signals collected at different positions as equivalent to the signals collected by a large-aperture probe, thereby effectively improving the spatial resolution of ultrasonic imaging. By delaying and coherently superimposing the data collected at different positions, the effect of focusing the virtual large-aperture probe at various locations in the detection area is achieved. The signal at the defect position shows a high phase correlation, resulting in phase superposition. In contrast, there is a zero-mean random Gaussian noise component at other positions of the component, and the signal generated by this random scattering and noise can be averaged and reduced. This method can effectively improve the signal-to-noise ratio and the lateral resolution of imaging.

[0004] Existing methods process only raw data acquired when the surface of the component being inspected is planar, and then use a Cartesian matrix grid for imaging. The resolution of the matrix grid is determined based on inspection parameters such as the velocity of sound and the step size of the sampling frequency encoder to achieve the optimal signal-to-noise ratio. For curved surfaces, as the probe moves along the surface, its central axis continuously shifts, resulting in an uneven distribution of sampled data points. This uneven distribution of sampled data points leads to variations in the number of sampling points within each matrix grid cell, resulting in varying resolution in the reconstructed image. For example, during ultrasonic inspection of curved components, a phased array probe moves along the surface of a cylinder, acquiring data at a fixed step size. At each location, a dataset consisting of multi-angle A-scan data is acquired. Due to the incident angle and inspection path, the sampling points are unevenly distributed within the inspection cross-section. In this case, the actual distribution of sampling points does not match the Cartesian matrix grid, resulting in over- or under-sampling of certain matrix grid cells, ultimately affecting the ultrasonic imaging. These limitations can reduce resolution and signal-to-noise ratio when directly applying existing synthetic aperture methods for imaging. Because defect identification and quantification rely on ultrasonic imaging data, the limited applicability of existing methods can lead to deficiencies in the final assessment. To address this issue, a new ultrasonic phased array imaging algorithm for curved components is needed. This algorithm can be used for ultrasonic inspection of parts with varying curvatures in industrial manufacturing. The goal is to improve the accuracy of defect quantitative detection and the spatial resolution of imaging.

[0005] This invention provides a phased array ultrasonic synthetic aperture imaging method for quantifying defects within shaft components, relating to the field of nondestructive testing. The method includes: constructing a mapping statistical matrix, adaptively optimizing the matrix grid size, employing a synthetic aperture algorithm, and quantifying defect sizes. The mapping statistical matrix is ​​constructed by calculating the number of sampling points in each matrix grid cell using a mapping algorithm. The mapping algorithm calculates the position of each sampling point and places it into the corresponding matrix grid cell. A matrix grid cell represents a geometric cross-sectional model of the object being inspected. The matrix grid size is adaptively optimized using the mapping statistical matrix based on the minimum coefficient of variation. The phase information of the sampling points is coherently superimposed using a synthetic aperture algorithm to obtain ultrasonic imaging data. Defect size is quantified using the distance-size-gain (DGS) method. Based on ultrasonic phased array synthetic aperture focusing technology and a matrix grid size optimization method, the method effectively identifies and quantifies defects. Optimizing the matrix grid size using a mapping statistical matrix based on the number of sampling points improves defect quantification accuracy, providing a more accurate and reliable basis for defect size assessment in shaft components. Summary of the Invention

[0006] In response to the shortcomings of the existing technology, the present invention provides a phased array ultrasonic shaft component defect detection method based on grid adaptive optimization. The grid size is optimized based on the sampling point distribution and the polar coordinate system, thereby reducing the phase information superposition error caused by the non-uniform distribution of sampling points in the ultrasonic detection of curved surface components, thereby interfering with the defect size quantification, improving the accuracy of defect quantification, reducing uncertainty, and improving the service life and reliability assessment of curved surface components in the industrial field.

[0007] To achieve the above objectives, the present invention discloses a phased array ultrasonic shaft component defect detection method based on grid adaptive optimization, which includes:

[0008] S1: Establish a polar coordinate system mapping model for sampling points of shaft components;

[0009] S2: Iteratively select the mapping statistical matrix of the number of sampling points of the shaft components and adaptively optimize the grid size; calculate the sampling point position coordinates according to the polar coordinate system mapping model of the shaft components in step S1 and the corresponding matrix grid index , and obtain the initial mapping statistics matrix;

[0010] S21: Use the matrix grid size coefficient to iterate the matrix grid size of the sampling points of the shaft components, specifically:

[0011] ;

[0012] in, is the radial size of the matrix grid; is the circumferential size of the matrix grid; is the first matrix grid size coefficient; is the second matrix grid size coefficient; is the radial size of the initial matrix grid; is the circumferential size of the initial matrix grid; To detect the sound velocity of the material; is the device sampling frequency; To detect the radius of shaft components; Move the encoder step size;

[0013] S22: Evaluate the mapping statistical matrix of the sampling points of the shaft components through the minimum coefficient of variation to obtain the first matrix grid size coefficient in the mapping statistical matrix , optimize the matrix grid size of the sampling points; according to the sampling point mapping statistical matrix of each layer of the shaft components, perform iterative calculation to select the adaptive optimization circumferential matrix grid size, adjust the matrix grid coefficient of the jth layer of the sampling point mapping statistical matrix of the shaft components, and obtain the sampling point mapping statistical matrix of the shaft components for:

[0014] ;

[0015] in, The statistical matrix of sampling point mapping for the jth layer of shaft components; The number of sampling points in the mapping statistics matrix grid; is the grid size coefficient of the second matrix in the j-th layer of the mapping statistics matrix; is the circumferential matrix grid index; is the radial matrix grid index;

[0016] S23: Mapping the statistical matrix based on the sampling points of the j-th layer of the shaft components in step S22 , calculate the coefficient of variation, and optimize the grid size coefficient of the second matrix of the jth layer in the mapping statistics matrix ; Substitute into step S21 to obtain the circumferential size of the j-th layer matrix grid in the mapping statistical matrix and the matrix grid radial size Perform matrix mesh reconstruction;

[0017] S3: Optimize the matrix grid and use phased array ultrasonic testing to detect shaft components and obtain ultrasonic imaging data.

[0018] S4: Use Hilbert transform and Gaussian filtering to process the ultrasonic imaging data obtained in step S3, identify and extract the defect area, establish a numerical characterization model for shaft component defects, and complete the detection of shaft component defects.

[0019] Preferably, in step S1, a polar coordinate system mapping model of sampling points of shaft components is established, specifically:

[0020] ;

[0021] in, is the horizontal coordinate of the phased array ultrasonic probe; is the ordinate of the phased array ultrasonic probe; is the distance from the probe location to the sampling point; The sampling point index for detecting depth direction; For and is the angle between the indexed phased array ultrasonic beam and the Y axis; Move the position index for the encoder;

[0022] The angle between the phased array ultrasonic beam and the Y axis is:

[0023] ;

[0024] in, is the angle between the ultrasound probe position and the starting line; For Indexed ultrasound beam scanning angle;

[0025] The distance from the probe position to the sampling point is:

[0026] ;

[0027] in, The initial sampling interval in the depth direction of the device;

[0028] The angle between the ultrasonic probe position and the starting line is:

[0029] ;

[0030] in, The initial movement interval of the encoder.

[0031] Preferably, the sampling point position coordinates in step S2 are for:

[0032] ;

[0033] in, The horizontal coordinate of the sampling point for internal detection of shaft components; The vertical coordinate of the sampling point for internal detection of shaft components.

[0034] Preferably, the matrix grid index in step S2 is for:

[0035] ;

[0036] in, is the circumferential matrix grid index; is the radial matrix grid index;

[0037] The scanning section of the shaft parts is divided according to the basic size and Perform matrix grid cell indexing; the internal value of each matrix grid cell represents whether there is a defect inside the cross-sectional position of the actual shaft component.

[0038] Preferably, in step S22, the statistical moment of the mapping of sampling points of each layer of the shaft components is:

[0039] ;

[0040] in, for The indexed sampling point maps the value in the statistical matrix, specifically the number of sampling points; For The number of sampling points in the matrix grid under the index; The fan scanning angle index of the phased array probe; The total number of phased array probe fan scan angle indexes.

[0041] Preferably, in step S3, matrix grid optimization is performed, and phased array ultrasonic detection of shaft components is used to obtain ultrasonic imaging data, specifically:

[0042] S31: Obtain the matrix grid size optimization result through the adaptive optimization process in step S2, perform matrix grid size optimization and overall matrix grid construction; the overall matrix grid has a fixed size in the radial direction, and is adaptively adjusted in the circumferential direction according to the size result of each layer, and the sampling point information is remapped;

[0043] S32: Using phased array ultrasound to detect shaft components and process ultrasound imaging data; obtaining the matrix grid after adaptive reconstruction in step S2, extracting ultrasound imaging data based on the shaft component sampling point coordinates and matrix grid index in step S2, and performing sampling point mapping;

[0044] S33: Using a synthetic aperture focusing algorithm, the ultrasonic testing data of the shaft components are coherently superimposed and calculated to obtain ultrasonic imaging data, specifically:

[0045] ;

[0046] in, After coherent superposition, is the value within the indexed grid cell; For is the echo amplitude of the indexed sampling point; To map to the matrix grid cells A collection of sampling points.

[0047] Preferably, in step S4, the ultrasonic imaging data is processed using Hilbert transform, specifically:

[0048] ;

[0049] in, is the complex-valued signal after Hilbert transform; is the Hilbert transform process; is the complex-valued signal before Hilbert transform; Signals during Hilbert transform; is the time node; is the value range; is the circumference constant of pi.

[0050] Preferably, in step S4, the ultrasound imaging data is processed by Gaussian filtering, specifically as follows:

[0051] The values ​​in the imaging matrix are redistributed through the Gaussian convolution kernel to obtain a smoothed image. The method for obtaining the Gaussian filter convolution kernel is:

[0052] ;

[0053] in, is the Gaussian filter convolution kernel value; is the size of the first Gaussian convolution kernel, ; is the size of the second Gaussian convolution kernel, ; is the standard deviation;

[0054] The values ​​in the ultrasound imaging matrix are redistributed according to the values ​​in the convolution kernel to achieve a smoothing effect.

[0055] Preferably, the shaft component defect characterization model in step S4 is specifically:

[0056] After processing the ultrasonic detection data, the ultrasonic imaging matrix is ​​obtained, and the defect area is identified by combining the chromatogram, and the extreme value of the defect is determined by box selection. ; Known defect size The value below , then the actual size of the unknown defect for:

[0057] ;

[0058] in, is the actual size of the unknown defect; is the actual size of the known defect; is the extreme value at the defect; is the value when the defect size is known;

[0059] The size of the unknown defect is obtained by the numerical value corresponding to the known defect size, and the defect size is quantified.

[0060] Compared with the prior art, the present invention has the following beneficial effects:

[0061] (1) The present invention calculates the number of sampling points in each grid cell through a mapping algorithm to construct a mapping statistical matrix. The mapping algorithm calculates the position of each sampling point and places it in the corresponding grid cell. Based on the mapping statistical matrix, the matrix grid size of shaft components can be adaptively optimized, the spatial resolution can be improved, and the sampling point data can be effectively utilized. The optimized grid and synthetic aperture method are used to obtain ultrasonic imaging data, and the defects of the ultrasonic imaging data after signal processing are quantified, thereby improving the quantitative accuracy of cylindrical component defect detection.

[0062] (2) The present invention is based on ultrasonic phased array synthetic aperture focusing technology and grid size optimization method to effectively identify and quantify defects. The grid size is optimized by mapping the statistical matrix of the number of sampling points to improve the accuracy of defect quantification. The distance-size-gain (DGS) method is used to quantify the defect size, providing a more accurate and reliable basis for the defect size assessment of cylindrical components.

[0063] (3) The present invention can accurately quantify the internal defects of shaft components, improve the accuracy of defect quantification of shaft components and reduce uncertainty, thereby improving the service life and reliability assessment of cylindrical shaft components in the industrial field. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is a flow chart of the phased array ultrasonic shaft component defect detection method based on grid adaptive optimization of the present invention;

[0065] Figure 2 Schematic diagram of the polar coordinate mapping algorithm parameters for sampling points of shaft components of the present invention;

[0066] Figure 3 This is a diagram of the matrix grid adaptive optimization algorithm for shaft components of the present invention;

[0067] Figure 4 This is a schematic diagram of the ultrasonic testing process for shaft components of the present invention;

[0068] Figure 5 This is an imaging diagram of the ultrasonic testing area of ​​a shaft component of the present invention;

[0069] Figure 6 This is an imaging diagram of a defect sub-region of a shaft component ultrasonically detected according to the present invention;

[0070] Figure 7 This is a diagram showing the quantitative results of ultrasonic testing defects on shaft components of the present invention.

[0071] Main reference numerals: 1. detection component; 2. detection section; 3. ultrasonic wave; 4. probe and 5. detection path. DETAILED DESCRIPTION

[0072] The exemplary embodiments, features, and aspects of the present invention will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise indicated.

[0073] The present invention provides a phased array ultrasonic shaft component defect detection method based on grid adaptive optimization, such as Figure 1As shown, a polar coordinate system mapping model of sampling points of shaft components is established; a mapping statistical matrix of the number of sampling points of shaft components is iteratively obtained, and the grid size of the mapping matrix is ​​adaptively optimized; phased array ultrasonic testing of shaft components is performed, and ultrasonic imaging data is obtained based on the matrix grid size optimization result; Hilbert transform and Gaussian filtering are used to process the ultrasonic imaging data, identify and extract defect areas, and obtain a sub-matrix to quantify the size of the defects; which includes:

[0074] Step S1: Establish a polar coordinate system mapping model for sampling points of shaft components, such as Figure 2 The figure shows the parameters of the polar coordinate system mapping algorithm for sampling points of shaft components of the present invention. Figure 2 Where S is the probe position and W is the detection surface; specifically:

[0075] ;

[0076] in, is the horizontal coordinate of the phased array ultrasonic probe; is the ordinate of the phased array ultrasonic probe; is the distance from the probe location to the sampling point; The sampling point index for detecting depth direction; For and is the angle between the indexed phased array ultrasonic beam and the Y axis; Move the position index for the encoder.

[0077] The angle between the phased array ultrasonic beam and the Y axis is:

[0078] ;

[0079] in, is the angle between the ultrasound probe position and the starting line; For Indexed ultrasound beam scanning angle.

[0080] The distance from the probe position to the sampling point is:

[0081] ;

[0082] in, The initial sampling interval in the depth direction of the device.

[0083] The angle between the ultrasound probe position and the starting line is:

[0084] ;

[0085] in, The initial movement interval of the encoder.

[0086] Step S2: Iteratively select the mapping statistical matrix of the number of sampling points of the shaft components and adaptively optimize the grid size; calculate the sampling point position coordinates according to the polar coordinate system mapping model of the shaft components in step S1 and the corresponding matrix grid index , and obtain the initial mapping statistics matrix, the number is ,by index.

[0087] Sampling point location coordinates for:

[0088] ;

[0089] in, The horizontal coordinate of the sampling point for internal detection of shaft components; The vertical coordinate of the sampling point for internal detection of shaft components.

[0090] Matrix Grid Index for:

[0091] ;

[0092] in, is the circumferential matrix grid index; is the radial matrix grid index.

[0093] The scanning section of the shaft parts is divided according to the basic size and Perform matrix grid cell indexing; the internal value of each matrix grid cell represents whether there is a defect inside the cross-sectional position of the actual shaft component.

[0094] Step S21: Use the matrix grid size coefficient to iterate the matrix grid size of the sampling points of the shaft components, specifically:

[0095] ;

[0096] in, is the radial size of the matrix grid; is the circumferential size of the matrix grid; is the first matrix grid size coefficient; is the second matrix grid size coefficient; is the radial size of the initial matrix grid; is the circumferential size of the initial matrix grid; To detect the sound velocity of the material; is the device sampling frequency; To detect the radius of shaft components; Move the encoder step size.

[0097] Step S22: Construct the mapping statistics matrix of sampling points of each layer of shaft components as follows:

[0098]

[0099] in, for The indexed sampling point maps the value in the statistical matrix, representing the number of sampling points; For The number of sampling points in the matrix grid under the index; The fan scanning angle index of the phased array probe; is the total number of phased array probe fan scan angle indexes; is the circumferential matrix grid index; is the radial matrix grid index.

[0100] The minimum variation coefficient is used to evaluate the mapping statistical matrix of the sampling points of the shaft components to obtain the first matrix grid size coefficient in the mapping statistical matrix. , optimize the matrix grid size of the sampling points; such as Figure 3 The figure shows the adaptive optimization algorithm of the matrix grid of the shaft components of the present invention; according to the sampling point mapping statistical matrix of each layer of the shaft components, iterative calculation is performed to select the adaptive optimization circumferential matrix grid size, and the matrix grid coefficient of the jth layer of the sampling point mapping statistical matrix of the shaft components is adjusted to obtain the sampling point mapping statistical matrix of the shaft components. for:

[0101] ;

[0102] in, The statistical matrix of sampling point mapping for the jth layer of shaft components; The number of sampling points in the mapping statistics matrix grid; is the grid size coefficient of the second matrix in the j-th layer of the mapping statistics matrix.

[0103] Step S23: Map the statistical matrix based on the sampling points of the j-th layer of the shaft components in step S22 , calculate the coefficient of variation, and optimize the grid size coefficient of the second matrix of the jth layer in the mapping statistics matrix ; Substitute into step S21 to obtain the circumferential size of the j-th layer matrix grid in the mapping statistical matrix and the matrix grid radial size Perform matrix mesh reconstruction.

[0104] Step S3: Optimize the matrix grid and use phased array ultrasound to detect shaft components to obtain ultrasound imaging data.

[0105] Step S31: Obtain the matrix grid size optimization result through the adaptive optimization process in step S2, perform matrix grid size optimization and overall matrix grid construction; the overall matrix grid has a fixed size in the radial direction, and is adaptively adjusted in the circumferential direction according to the size result of each layer, and the sampling point information is remapped.

[0106] Step S32: Use phased array ultrasound to detect shaft components and process ultrasound imaging data; Figure 4 The figure shows a schematic diagram of the ultrasonic testing process of shaft components of the present invention. Figure 4 Where 1 is the detection component, 2 is the detection section, 3 is the ultrasound, 4 is the probe, and 5 is the detection path; obtain the matrix grid after adaptive reconstruction in step S2, extract the ultrasonic imaging data according to the sampling point coordinates of the shaft component in step S2 and the matrix grid index, and perform sampling point mapping.

[0107] Step S33: Using a synthetic aperture focusing algorithm to perform coherent superposition calculation on the ultrasonic detection data of the shaft components to obtain ultrasonic imaging data; specifically:

[0108] ;

[0109] in, After coherent superposition, is the value within the indexed grid cell; For is the echo amplitude of the indexed sampling point; To map to the matrix grid cells A collection of sampling points.

[0110] Step S4: Based on the ultrasonic imaging data obtained in step S3, the ultrasonic imaging matrix is ​​processed using Hilbert transform and Gaussian filtering to obtain the ultrasonic imaging matrix, and the chromatogram is used for imaging to identify the defect area and locate it, extract the defect area, and extract the sub-matrix; Figure 5 Shown is an imaging diagram of the ultrasonic detection area of ​​the shaft component of the present invention.

[0111] Step S41: Processing the ultrasonic imaging using Hilbert transform, specifically:

[0112] ;

[0113] in, is the complex-valued signal after Hilbert transform; is the Hilbert transform process; is the complex-valued signal before Hilbert transform; Signals during Hilbert transform; is the time node; is the value range; is the circumference constant of pi.

[0114] Step S42: Redistribute the values ​​in the imaging matrix through the Gaussian convolution kernel to obtain a smoothed image; the method for obtaining the Gaussian filter convolution kernel is:

[0115] ;

[0116] in, is the Gaussian filter convolution kernel value; is the size of the first Gaussian convolution kernel, ; is the size of the second Gaussian convolution kernel, ; is the standard deviation.

[0117] The values ​​in the ultrasound imaging matrix are redistributed according to the values ​​in the convolution kernel to achieve a smoothing effect.

[0118] Step S43: quantify the defect size of the shaft component defect to obtain the quantified shaft component defect diameter, and establish a numerical characterization model for the shaft component defect, specifically:

[0119] like Figure 6 The figure shows the imaging diagram of the defect sub-region of the ultrasonic detection of shaft components of the present invention; after processing the ultrasonic detection data, the ultrasonic imaging matrix is ​​obtained, and the defect area is identified by combining the chromatogram, and the extreme value of the defect is determined by box selection. ; Known defect size The value below , then the actual size of the unknown defect for:

[0120] ;

[0121] in, is the actual size of the unknown defect; is the actual size of the known defect; is the extreme value at the defect; is the value when the defect size is known.

[0122] The defect size is determined based on the distance-size-gain (DGS) method with a known reflection size amplitude. Before conducting ultrasonic testing, the velocity and sensitivity calibration must be carefully performed. The material of the calibration sample must be the same as the test sample to obtain the same attenuation, acoustic impedance and sound wave propagation conditions. Sensitivity calibration is performed so that the light beam at each incident angle has the same signal response when detecting the same reflector. Through multi-point position movement detection, the actual sound velocity can be calculated based on the bottom surface echo time and the sample thickness. The defect quantification procedure is arranged as follows: locate and extract the defect area in the ultrasonic imaging results, and normalize the values ​​in the defect sub-area. Locate and extract the defect peak as the effective value. The effective local peak refers to the grid coordinate The maximum value of the defect indication area The defect amplitude in the calibration sub-area is related to the defect size.

[0123] The size of the unknown defect is obtained by the numerical value corresponding to the known defect size, and the defect size is quantified to complete the defect detection of shaft components. Figure 7 The figure shows the quantitative results of ultrasonic detection of defects in shaft components of the present invention. The horizontal axis represents the actual size and the vertical axis represents the quantitative size. It can be seen from the figure that the calculation results of this method are accurate.

[0124] The beneficial effects of the present invention are: the present invention provides a phased array ultrasonic shaft component defect detection method based on grid adaptive optimization, which optimizes the mapping grid size based on the sampling point distribution and the polar coordinate system, and uses a mapping algorithm to calculate the position of each sampling point to achieve adaptive optimization of the matrix grid size of the shaft component, reduce the phase information superposition error caused by the uneven distribution of sampling points in the ultrasonic detection of curved components, thereby interfering with the defect size quantification, and use the optimized grid and synthetic aperture method to obtain ultrasonic imaging data, and quantify the defects of the ultrasonic imaging data after signal processing; the present invention can accurately quantify the internal defects of shaft components, improve the accuracy of the defect quantification of shaft components and reduce uncertainty, thereby improving the service life and reliability evaluation of cylindrical shaft components in the industrial field.

[0125] The embodiments described above are merely descriptions of preferred implementations of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should fall within the scope of protection determined by the claims of the present invention.

Claims

1. A phased array ultrasonic shaft component defect detection method based on grid adaptive optimization, characterized in that: It includes: S1: Establish a polar coordinate system mapping model for sampling points of shaft components; S2: Iteratively select the mapping statistical matrix of the number of sampling points of the shaft components and adaptively optimize the grid size; calculate the sampling point position coordinates according to the polar coordinate system mapping model of the shaft components in step S1 and the corresponding matrix grid index , and obtain the initial mapping statistics matrix; S21: Use the matrix grid size coefficient to iterate the matrix grid size of the sampling points of the shaft components, specifically: ； in, is the radial size of the matrix grid; is the circumferential size of the matrix grid; is the first matrix grid size coefficient; is the second matrix grid size coefficient; is the radial size of the initial matrix grid; is the circumferential size of the initial matrix grid; To detect the sound velocity of the material; is the device sampling frequency; To detect the radius of shaft components; Move the encoder step size; S22: Evaluate the mapping statistical matrix of the sampling points of the shaft components through the minimum coefficient of variation to obtain the first matrix grid size coefficient in the mapping statistical matrix , optimize the matrix grid size of the sampling points; according to the sampling point mapping statistical matrix of each layer of the shaft components, perform iterative calculation to select the adaptive optimization circumferential matrix grid size, adjust the matrix grid coefficient of the jth layer of the sampling point mapping statistical matrix of the shaft components, and obtain the sampling point mapping statistical matrix of the shaft components for: ； in, The statistical matrix of sampling point mapping for the jth layer of shaft components; The number of sampling points in the mapping statistics matrix grid; is the grid size coefficient of the second matrix in the j-th layer of the mapping statistics matrix; is the circumferential matrix grid index; is the radial matrix grid index; The statistical matrix of sampling point mapping of each layer of the shaft components in step S22 is: ； in, for The indexed sampling point maps the value in the statistical matrix, specifically the number of sampling points; For The number of sampling points in the matrix grid under the index; The fan scanning angle index of the phased array probe; is the total number of phased array probe fan scan angle indexes; S23: Mapping the statistical matrix based on the sampling points of the j-th layer of the shaft components in step S22 , calculate the coefficient of variation, and optimize the grid size coefficient of the second matrix of the jth layer in the mapping statistics matrix ; Substitute into step S21 to obtain the circumferential size of the j-th layer matrix grid in the mapping statistical matrix and the matrix grid radial size Perform matrix mesh reconstruction; S3: Perform matrix grid optimization and use phased array ultrasonic testing of shaft components to obtain ultrasonic imaging data. Specifically: S31: Obtain the matrix grid size optimization result through the adaptive optimization process in step S2, perform matrix grid size optimization and overall matrix grid construction; the overall matrix grid has a fixed size in the radial direction, and is adaptively adjusted in the circumferential direction according to the size result of each layer, and the sampling point information is remapped; S32: Using phased array ultrasound to detect shaft components and process ultrasound imaging data; obtaining the matrix grid after adaptive reconstruction in step S2, extracting ultrasound imaging data based on the shaft component sampling point coordinates and matrix grid index in step S2, and performing sampling point mapping; S33: Using a synthetic aperture focusing algorithm, the ultrasonic testing data of the shaft components are coherently superimposed and calculated to obtain ultrasonic imaging data, specifically: ； in, After coherent superposition, is the value within the indexed grid cell; For is the echo amplitude of the indexed sampling point; To map to the matrix grid cells The set of sampling points; S4: Use Hilbert transform and Gaussian filtering to process the ultrasonic imaging data obtained in step S3, identify and extract the defect area, establish a numerical characterization model for shaft component defects, and complete the detection of shaft component defects.

2. The phased array ultrasonic shaft component defect detection method based on grid adaptive optimization according to claim 1 is characterized in that: In step S1, a polar coordinate system mapping model of sampling points of shaft components is established, specifically: ； in, is the horizontal coordinate of the phased array ultrasonic probe; is the ordinate of the phased array ultrasonic probe; is the distance from the probe location to the sampling point; The sampling point index for detecting depth direction; For and is the angle between the indexed phased array ultrasonic beam and the Y axis; Move the position index for the encoder; The angle between the phased array ultrasonic beam and the Y axis is: ； in, is the angle between the ultrasound probe position and the starting line; For Indexed ultrasound beam scanning angle; The distance from the probe position to the sampling point is: ； in, The initial sampling interval in the depth direction of the device; The angle between the ultrasonic probe position and the starting line is: ； in, The initial movement interval of the encoder.

3. The phased array ultrasonic shaft component defect detection method based on grid adaptive optimization according to claim 1 is characterized in that: The sampling point position coordinates in step S2 for: ； in, The horizontal coordinate of the sampling point for internal detection of shaft components; The vertical coordinate of the sampling point for internal detection of shaft components.

4. The phased array ultrasonic shaft component defect detection method based on grid adaptive optimization according to claim 1 is characterized in that: Matrix grid index in step S2 for: ； in, is the circumferential matrix grid index; is the radial matrix grid index; The scanning section of the shaft parts is divided according to the basic size and Perform matrix grid cell indexing; the internal value of each matrix grid cell represents whether there is a defect inside the cross-sectional position of the actual shaft component.

5. The phased array ultrasonic shaft component defect detection method based on grid adaptive optimization according to claim 1 is characterized in that: In step S4, the ultrasonic imaging data is processed using Hilbert transform, specifically: ； in, is the complex-valued signal after Hilbert transform; is the Hilbert transform process; is the complex-valued signal before Hilbert transform; Signals during Hilbert transform; is the time node; is the value range; is the circumference constant of pi.

6. The phased array ultrasonic shaft component defect detection method based on grid adaptive optimization according to claim 1 is characterized in that: In step S4, the ultrasound imaging data is processed by Gaussian filtering, specifically: The values ​​in the imaging matrix are redistributed through the Gaussian convolution kernel to obtain a smoothed image. The method for obtaining the Gaussian filter convolution kernel is: ； in, is the Gaussian filter convolution kernel value; is the size of the first Gaussian convolution kernel, ; is the size of the second Gaussian convolution kernel, ; is the standard deviation; The values ​​in the ultrasound imaging matrix are redistributed according to the values ​​in the convolution kernel to achieve a smoothing effect.

7. The phased array ultrasonic shaft component defect detection method based on grid adaptive optimization according to claim 1 is characterized in that: The shaft component defect characterization model in step S4 is specifically: After processing the ultrasonic detection data, the ultrasonic imaging matrix is ​​obtained, and the defect area is identified by combining the chromatogram, and the extreme value of the defect is determined by box selection. ; Known defect size The value below , then the actual size of the unknown defect for: ； in, is the actual size of the unknown defect; is the actual size of the known defect; is the extreme value at the defect; is the value when the defect size is known; The size of the unknown defect is obtained by the numerical value corresponding to the known defect size, and the defect size is quantified.

Images