Defect reconstruction method applied to phased array ultrasonic detection of spinning shell

Abstract

The invention belongs to the technical field of defect detection, and particularly relates to a defect reconstruction method applied to phased array ultrasonic detection of a spinning shell. Aiming at the problems of difficulty in feature extraction and multi-direction detection and reconstruction of spinning textures in phased array ultrasonic detection of spinning shell defects, noise interference caused by the spinning textures is effectively inhibited by adopting a multi-channel blind separation processing method provided by the invention, and the feature echo extraction effect is improved; and in combination with data in four directions, a position reconstruction algorithm based on a geometrical relationship is established, and reconstruction of a workpiece structure and defects is realized.

Description

Technical Field

[0001] This invention belongs to the field of defect detection technology, specifically relating to a defect reconstruction method applied to phased array ultrasonic testing of spun shells. Background Technology

[0002] Spin-formed cylindrical shell workpieces are widely used in aerospace and other fields. During the production process, defects such as cracks and holes can form. These defects are affected by various factors such as temperature, vibration, aging, internal gas pressure, and chemical migration, and can continue to expand, directly affecting product performance. Therefore, it is necessary to inspect the products before production and use. Currently, engineering applications mainly rely on manual scanning using shear wave probes to determine the presence of defects based on wave height. This method suffers from problems such as being affected by human factors, not being able to visually display the defect situation, not being able to store the detection status, and having a low level of automation, failing to meet the requirements of intelligent and information-based systems. Phased array ultrasonic defect detection for cylindrical workpieces can meet the needs of large workpieces, small defects, and high-efficiency detection. Summary of the Invention

[0003] To address the challenges of feature extraction and multi-directional detection and reconstruction of spinning textures in phased array ultrasonic testing of spun shell defects, this invention proposes a defect reconstruction method for phased array ultrasonic testing of spun shells. The proposed multi-channel blind separation processing method effectively suppresses noise interference caused by spinning textures and improves the feature echo extraction effect. By combining data from four directions, a position reconstruction algorithm based on geometric relationships is established, realizing the reconstruction of workpiece structure and defects.

[0004] To achieve the above objectives, the present invention employs the following technical solution:

[0005] This invention provides a defect reconstruction method for phased array ultrasonic testing of spun shells, comprising the following steps:

[0006] Step 1: Obtain the standard defect echo signal sd of the calibration plate using a phased array ultrasonic testing system for spun shells;

[0007] Step 2: Use a phased array ultrasonic testing system for spun shells to acquire effective data arrays of ultrasonic waves in the four directions (X+, X-, Y+, Y-). , , , Where X+ indicates that the ultrasonic wave propagates counterclockwise along the circumference of the workpiece, X- indicates that the ultrasonic wave propagates clockwise along the circumference of the workpiece, Y+ indicates that the ultrasonic wave propagates upward along the workpiece, and Y- indicates that the ultrasonic wave propagates downward along the workpiece.

[0008] Step 3, targeting , , , Using SD as the reference signal, a multi-channel blind separation processing method is employed to extract the defect echo, and the defect echo time information is calculated. , , , ;

[0009] Step 4, create A two-dimensional matrix Q of size is reconstructed using a position reconstruction algorithm. , , , The information is mapped to Q to form a two-dimensional defect detection planar map, and then a three-dimensional point cloud defect map is reconstructed using coordinate transformation.

[0010] Furthermore, by rotating the workpiece, the phased array ultrasonic testing system for the spun shell acquires ultrasonic full-scan data in four directions (X+, X-, Y+, Y-) through one or multiple tests, and obtains an effective data array. , , , ;

[0011] Phased array probes in the X+ or X- directions Ultrasonic signals were acquired at several array locations, with a total of K sets of signals acquired at each location. Each signal has a length of [length missing]. , , These represent the number of points of the phased array probe in the X direction along the axial and circumferential directions, respectively, and K is the number of array elements of the phased array probe.

[0012] Phased array probes in the Y+ or Y- directions Ultrasonic signals were acquired at several array locations, with a total of K sets of signals acquired at each location. Each signal has a length of [length missing]. , , These represent the number of phased array probes in the Y direction, both circumferentially and axially.

[0013] , , , All are three-dimensional matrices, with dimensions of respectively , , , .

[0014] Furthermore, the multi-channel blind separation processing method in step 3 is specifically as follows:

[0015] Step 3.1, Extract An ultrasonic signal ,in 0 ,right Three-level empirical mode decomposition is performed to obtain Among them, IMF1, IMF2, and IMF3 are the three components of the empirical mode decomposition, all of which are long-term components. Horizontal quantity;

[0016] Step 3.2: Generate a random matrix ,in , Calculations yielded Calculation obtained Given the autocovariance matrix C, find the eigenvectors V and eigenvalues ​​of C. ,make ,Pick ;

[0017] Step 3.3: Generate a uniformly distributed random matrix. ,make , , , and according to , , Normalization is performed, and S3 is used as the initial signal SU for blind signal iteration. , , Substitute these initial coefficients into the matrix and iterate in step 3.4.

[0018] Step 3.4: Iterate using the following three sets of formulas to obtain new matrix data. , , Update the W matrix to obtain ;

[0019] ;

[0020] ;

[0021] ;

[0022] Where i is the number of iterations, and T represents the matrix transpose;

[0023] Step 3.5, will Using SU as the initial signal for iteration, repeating step 3.4 yields the following result. This process is repeated iteratively until... Using SU as the initial signal for iteration, repeating step 3.4 yields the following result. Eight sets of cyclically separated signals can be obtained after eight iterations. ;

[0024] Step 3.6: Calculate the cross-correlation coefficient between each component of the 8 sets of cyclically separated signals and sd. The component with the largest correlation coefficient is the blind separation result s4.

[0025] Step 3.7: Determine the defect echo based on the amplitude and transit time of the ultrasonic test, and obtain the defect echo time information. If there is no defect echo, ;

[0026] Step 3.8, for All signals are processed through steps 3.1 to 3.7 to obtain... ;

[0027] Step 3.9, will Change to , , Perform steps 3.1 to 3.8 to obtain... , , .

[0028] Furthermore, the aforementioned , , , All are two-dimensional matrices, with dimensions as follows: , , , .

[0029] Furthermore, the location reconstruction algorithm in step 4 is specifically as follows:

[0030] Step 4.1, create four Two-dimensional zero matrix of size , , , ;

[0031] Step 4.2, for A non-zero element in The center position of its probe is the coordinate. According to the formula Defect location calculated Where d0 is the element spacing of the phased array probe, and d1 is the distance from the probe to the workpiece surface. The angle between the probe and the workpiece normal. The speed of ultrasound in water. The ultrasonic longitudinal wave velocity in the workpiece is... Set the element to 1;

[0032] Step 4.3, for all non-zero elements Perform the processing and calculation according to step 4.2 to obtain the two-dimensional zero matrix. ;

[0033] Step 4.4, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to:

[0034] ;

[0035] Step 4.5, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to:

[0036] ;

[0037] Step 4.6, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to:

[0038] ;

[0039] Step 4.7, for the two-dimensional zero matrix , , , OR operation yields a two-dimensional matrix. .

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] (1) A multi-channel blind separation processing method was proposed to effectively suppress noise interference caused by spinning texture and improve the effect of feature echo extraction;

[0042] (2) The geometric relationship of the four directions of the detection echo information was established, and a position reconstruction algorithm for the spinning shell structure was designed to realize the reconstruction of the workpiece structure and defects. Attached Figure Description

[0043] Figure 1 The flowchart is for the defect reconstruction method;

[0044] Figure 2 A flowchart for blind signal processing;

[0045] Figure 3 This is a diagram showing the propagation direction of ultrasound.

[0046] Figure 4 This is the result of blind signal processing;

[0047] Figure 5 This is a schematic diagram of the two-dimensional plane reconstruction result;

[0048] Figure 6 This is a schematic diagram of the 3D reconstruction results. Detailed Implementation

[0049] To further illustrate the technical solution of the present invention, the present invention will be further described below through embodiments.

[0050] like Figure 1 As shown in this embodiment, a defect reconstruction method applied to phased array ultrasonic testing of spun shells includes the following steps:

[0051] Step 1: Obtain the standard defect echo signal sd of the calibration plate using a phased array ultrasonic testing system for spun shells;

[0052] Step 2: Use a phased array ultrasonic testing system for spun shells to acquire effective data arrays of ultrasonic waves in the four directions (X+, X-, Y+, Y-). , , , Where X+ indicates that the ultrasonic wave propagates counterclockwise along the circumference of the workpiece, X- indicates that the ultrasonic wave propagates clockwise along the circumference of the workpiece, Y+ indicates that the ultrasonic wave propagates upward along the workpiece, and Y- indicates that the ultrasonic wave propagates downward along the workpiece.

[0053] By rotating the workpiece, the phased array ultrasonic testing system of the spun shell acquires full ultrasonic scanning data in four directions (X+, X-, Y+, Y-) through one or multiple tests (e.g., Figure 3 (as shown), and obtain the effective data array. , , , .

[0054] Phased array probes in the X+ or X- directions Ultrasonic signals were acquired at several array locations, with a total of K sets of signals acquired at each location. Each signal has a length of [length missing]. , , These represent the number of points of the phased array probe in the X direction along the axial and circumferential directions, respectively, and K represents the number of array elements of the phased array probe.

[0055] Phased array probes in the Y+ or Y- directions Ultrasonic signals were acquired at several array locations, with a total of K sets of signals acquired at each location. Each signal has a length of [length missing]. , , These represent the number of phased array probes deployed in the Y direction along the circumference and axial direction, respectively.

[0056] , , , All are three-dimensional matrices, with dimensions of respectively , , , .

[0057] Step 3, targeting , , , Using SD as the reference signal, a multi-channel blind separation processing method is employed to extract the defect echo, and the defect echo time information is calculated. , , , ;

[0058] like Figure 2 As shown, the multi-channel blind separation processing method in this embodiment is as follows:

[0059] Step 3.1, Extract An ultrasonic signal ,in 0 ,right Three-level empirical mode decomposition is performed to obtain Among them, IMF1, IMF2, and IMF3 are the three components of the empirical mode decomposition, all of which are long-term components. Horizontal quantity;

[0060] Step 3.2: Generate a random matrix ,in , Calculations yielded Calculation obtained Given the autocovariance matrix C, find the eigenvectors V and eigenvalues ​​of C. ,make ,Pick ;

[0061] Step 3.3: Generate a uniformly distributed random matrix. ,make , , , and according to , , Normalization is performed, and S3 is used as the initial signal SU for blind signal iteration. , , Substitute these initial coefficients into the iteration step 3.4;

[0062] Step 3.4: Iterate using the following three sets of formulas to obtain new matrix data. , , Update the W matrix to obtain ;

[0063] ;

[0064] ;

[0065] ;

[0066] Where i is the number of iterations, and T represents the matrix transpose;

[0067] Step 3.5, will Using SU as the initial signal for iteration, repeating step 3.4 yields the following result. This process is repeated iteratively until... Using SU as the initial signal for iteration, repeating step 3.4 yields the following result. Eight sets of cyclically separated signals can be obtained after eight iterations. ;

[0068] Step 3.6: Calculate the cross-correlation coefficient between each component of the 8 sets of cyclically separated signals and sd. The component with the largest correlation coefficient is the blind separation result s4. Figure 4 );

[0069] Step 3.7: Determine the defect echo based on the amplitude and transit time of the ultrasonic test, and obtain the defect echo time information. If there is no defect echo, ;

[0070] Step 3.8, for All signals are processed through steps 3.1 to 3.7 to obtain... ;

[0071] Step 3.9, will Change to , , Perform steps 3.1 to 3.8 to obtain... , , , , , , All are two-dimensional matrices, with dimensions as follows: , , , .

[0072] Step 4, create A two-dimensional matrix Q of size is reconstructed using a position reconstruction algorithm. , , , The information is mapped onto Q to form a two-dimensional defect detection planar diagram. Figure 5 Then, using coordinate transformation, a three-dimensional point cloud defect map is reconstructed. Figure 6 );

[0073] The location reconstruction algorithm in this embodiment is as follows:

[0074] Step 4.1, create four Two-dimensional zero matrix of size , , , ;

[0075] Step 4.2, for A non-zero element in The center position of its probe is the coordinate. According to the formula Defect location calculated Where d0 is the element spacing of the phased array probe, and d1 is the distance from the probe to the workpiece surface. The angle between the probe and the workpiece normal. The speed of ultrasound in water. The ultrasonic longitudinal wave velocity in the workpiece is... Set the element to 1;

[0076] Step 4.3, for all non-zero elements Perform the processing and calculation according to step 4.2 to obtain the two-dimensional zero matrix. ;

[0077] Step 4.4, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to:

[0078] ;

[0079] Step 4.5, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to:

[0080] ;

[0081] Step 4.6, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to:

[0082] ;

[0083] Step 4.7, for the two-dimensional zero matrix , , , OR operation yields a two-dimensional matrix. .

[0084] The foregoing has shown and described the main features and advantages of the present invention. It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.

[0085] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A defect reconstruction method applied to phased array ultrasonic testing of spun shells, characterized in that, Includes the following steps: Step 1: Obtain the standard defect echo signal sd of the calibration plate using a phased array ultrasonic testing system for spun shells; Step 2: Use a phased array ultrasonic testing system for spun shells to acquire effective data arrays of ultrasonic waves in the four directions (X+, X-, Y+, Y-). , , , Where X+ indicates that the ultrasonic wave propagates counterclockwise along the circumference of the workpiece, X- indicates that the ultrasonic wave propagates clockwise along the circumference of the workpiece, Y+ indicates that the ultrasonic wave propagates upward along the workpiece, and Y- indicates that the ultrasonic wave propagates downward along the workpiece. Step 3, targeting , , , Using SD as the reference signal, a multi-channel blind separation processing method is employed to extract the defect echo, and the defect echo time information is calculated. , , , ; Step 4, create A two-dimensional matrix Q of size is reconstructed using a position reconstruction algorithm. , , , The information is mapped to Q to form a two-dimensional defect detection planar map, and then a three-dimensional point cloud defect map is reconstructed using coordinate transformation.

2. The defect reconstruction method for phased array ultrasonic testing of spun shells according to claim 1, characterized in that, The phased array ultrasonic testing system for the spun shell of the rotating workpiece acquires ultrasonic full-scan data in four directions (X+, X-, Y+, Y-) through one or multiple tests, and obtains an effective data array. , , , ; Phased array probes in the X+ or X- directions Ultrasonic signals were acquired at several array locations, with a total of K sets of signals acquired at each location. Each signal has a length of [length missing]. , , These represent the number of points of the phased array probe in the X direction along the axial and circumferential directions, respectively, and K is the number of array elements of the phased array probe. Phased array probes in the Y+ or Y- directions Ultrasonic signals were acquired at several array locations, with a total of K sets of signals acquired at each location. Each signal has a length of [length missing]. , , These represent the number of phased array probes in the Y direction, both circumferentially and axially. , , , All are three-dimensional matrices, with dimensions of respectively , , , .

3. The defect reconstruction method for phased array ultrasonic testing of spun shells according to claim 1, characterized in that, The multi-channel blind separation processing method in step 3 is as follows: Step 3.1, Extract An ultrasonic signal ,in 0 ,right Three-level empirical mode decomposition is performed to obtain Among them, IMF1, IMF2, and IMF3 are the three components of the empirical mode decomposition, all of which are long-term components. Horizontal quantity; Step 3.2: Generate a random matrix ,in , Calculations yielded Calculation obtained Given the autocovariance matrix C, find the eigenvectors V and eigenvalues ​​of C. ,make ,Pick ; Step 3.3: Generate a uniformly distributed random matrix. ,make , , , and according to , , Normalization is performed, and S3 is used as the initial signal SU for blind signal iteration. , , Substitute these initial coefficients into the iteration step 3.4; Step 3.4: Iterate using the following three sets of formulas to obtain new matrix data. , , Update the W matrix to obtain ; ； ； ； Where i is the iteration number and T represents the matrix transpose; Step 3.5, will Using SU as the initial signal for iteration, repeating step 3.4 yields the following result. This process is repeated iteratively until... Using SU as the initial signal for iteration, repeating step 3.4 yields the following result. Eight sets of cyclically separated signals can be obtained after eight iterations. ; Step 3.6: Calculate the cross-correlation coefficient between each component of the 8 sets of cyclically separated signals and sd. The component with the largest correlation coefficient is the blind separation result s4. Step 3.7: Determine the defect echo based on the amplitude and transit time of the ultrasonic test, and obtain the defect echo time information. If there is no defect echo, ; Step 3.8, for All signals are processed through steps 3.1 to 3.7 to obtain... ; Step 3.9, will Change to , , Perform steps 3.1 to 3.8 to obtain... , , .

4. A defect reconstruction method for phased array ultrasonic testing of spun shells according to claim 1 or 3, characterized in that, The , , , All are two-dimensional matrices, with dimensions as follows: , , , .

5. The defect reconstruction method for phased array ultrasonic testing of spun shells according to claim 1, characterized in that, The location reconstruction algorithm in step 4 is as follows: Step 4.1, create four Two-dimensional zero matrix of size , , , ; Step 4.2, for A non-zero element in The center position of its probe is the coordinate. According to the formula Defect location calculated Where d0 is the element spacing of the phased array probe, and d1 is the distance from the probe to the workpiece surface. The angle between the probe and the workpiece normal. The speed of ultrasound in water. The ultrasonic longitudinal wave velocity in the workpiece is... Set the element to 1; Step 4.3, for all non-zero elements Perform the processing and calculation according to step 4.2 to obtain the two-dimensional zero matrix. ; Step 4.4, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to: ； Step 4.5, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to: ； Step 4.6, for Follow steps 4.1 and 4.2 to obtain a two-dimensional zero matrix. At this point, for non-zero elements The calculation formula has been changed to: ； Step 4.7, for the two-dimensional zero matrix , , , OR operation yields a two-dimensional matrix. .

Images