Method for measuring in-situ corrosion fatigue crack growth rate based on synchrotron radiation light source

Abstract

The invention discloses an in-situ corrosion fatigue crack growth rate measurement method based on a synchrotron radiation light source. The method comprises the following steps: carrying out an in-situ corrosion fatigue test based on three-dimensional imaging of the synchrotron radiation light source; after a preset number of cycle stress loading is carried out on the sample, a synchrotron radiation light source is adopted to collect projection images of coaxial phase contrast CT of different faults of a sensitive area of the sample; judging whether the sample is fractured and invalid, if so, entering the next step, otherwise, returning to the previous step; pre-processing the projection images, and extracting cracks with the maximum length in all the pre-processed projection images corresponding to the same cycle as fatigue cracks; and calculating the average crack growth rate during the two adjacent projection image acquisition periods according to the corresponding fatigue crack length and the interval cycle number during the two adjacent projection image acquisition periods.

Description

Technical Field

[0001] The invention relates to a fatigue crack testing technology, and in particular to an in-situ corrosion fatigue crack growth rate measuring method based on a synchrotron radiation light source. Background Art

[0002] When metal components serve in important fields such as marine machinery, aerospace, oil and gas extraction, and rail transportation, they usually experience the combined effects of various forms of loads and corrosive environments. These loads include static and dynamic stresses, and the service environment covers a variety of situations such as atmospheric corrosion, seawater corrosion, and salt spray. Under such complex service conditions, metal components must not only withstand mechanical stress, but also resist damage caused by environmental media. Fracture mechanics parameters, such as fatigue crack growth rate and crack growth threshold, are key indicators for the engineering application of metal materials. Among them, the corrosion fatigue crack growth rate is one of the important parameters for evaluating the corrosion fatigue fracture performance of metal materials.

[0003] In salt spray or salt solution environments, fatigue crack problems of metal materials become extremely complicated. This is mainly because these corrosive environments react with the crack tips and crack surfaces of metal materials to produce a large number of corrosion products and hydrogen bubbles. The crack closure effect caused by corrosion products may cause the crack growth rate to decrease, while the anodic dissolution will cause the crack growth rate to accelerate. The extension of hydrogen atoms to the crack tip will reduce the plasticity of the crack tip and thus increase the corrosion fatigue crack growth rate. In addition, due to the existence of secondary cracks (secondary cracks), the crack growth model becomes more complicated. The appearance of secondary cracks will make the crack growth path complicated and even form a crack network, which increases the difficulty of measuring the crack growth length. At the same time, secondary cracks will cause the internal stress field of the material to be redistributed, thereby changing the stress concentration state near the main crack and affecting the growth rate and direction of the main crack. The superposition of these factors affects the accuracy of the corrosion fatigue crack growth rate measurement.

[0004] At present, in laboratory and engineering applications, methods such as potential difference method, replication method (or replica method) and infrared thermal imaging technology are often used to measure fatigue crack growth rate. The potential difference method applies a constant current to the sample and uses the principle that the change in the cross-sectional area of ​​the conductor causes the change in the current field to monitor the crack length. This method requires extremely high insulation measures, otherwise the leakage of the fixture will cause parallel resistance between the test piece and the specimen, which will seriously affect the accuracy of the test results. In addition, the potential difference method can only provide data at discrete points and cannot obtain information on the entire crack growth area, so it is difficult to fully describe the occurrence and development mechanism of the crack. Some researchers also measured the fatigue crack growth rate under salt spray conditions by using the DC potential difference method, but this method deposits salt on the surface of the sample by salt printing and salt pipetting. However, both methods are difficult to ensure the uniformity and accuracy of salt deposition, which directly affects the consistency of the experimental results. The replication method can provide higher measurement accuracy, but it requires multiple interruptions to the experiment at different stages of the experiment to measure the crack length, which not only prolongs the total duration of the experiment, but also affects the corrosion rate and the accuracy of the corrosion fatigue crack growth rate measurement. As a non-contact detection method, infrared thermal imaging technology can estimate the position of the crack tip by analyzing the temperature distribution image on the sample surface and combining it with a two-dimensional heat conduction model, and then calculate the crack growth. However, this method requires a series of complex pretreatments of the sample, and the temperature fluctuations of the external environment may significantly affect the test results. In particular, when the temperature difference on the sample surface is not large and the contrast of the thermal image is low, the measurement accuracy will be greatly reduced, which is obviously not suitable for the measurement of fatigue crack growth rate in a corrosive environment.

[0005] In summary, these existing technical means are still difficult to observe and record crack propagation behavior in real time, and at the same time, they are unable to accurately capture crack growth and measure its length. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the in-situ corrosion fatigue crack growth rate measurement method based on a synchrotron radiation light source provided by the present invention solves the problem that the prior art cannot accurately capture the growth of cracks and measure their length.

[0007] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0008] A method for measuring in-situ corrosion fatigue crack growth rate based on a synchrotron radiation light source is provided, which comprises the following steps:

[0009] S1. Install the in-situ corrosion fatigue testing machine on the three-axis displacement turntable of the synchrotron radiation source, and install the hourglass-shaped specimen on the in-situ corrosion fatigue testing machine so that it is coaxial with the three-axis displacement turntable;

[0010] S2. Set the environment of the sample to a corrosive medium environment, control the heating of the ceramic heating plate of the in-situ corrosion fatigue testing machine, and keep the temperature constant to the target temperature;

[0011] S3. After the sample is subjected to a preset number of cyclic stress loadings, the preset loading force is maintained, and while the sample is rotating, a synchrotron radiation light source is used to collect projection images of coaxial phase contrast CT of different faults in the sensitive area of ​​the sample;

[0012] S4, determine whether the sample is broken and failed, if so, proceed to step S5, otherwise return to step S3;

[0013] S5, preprocessing the projection images of the coaxial phase contrast CT, and extracting the maximum length crack in all preprocessed projection images corresponding to the same cycle as the fatigue crack;

[0014] S6. Calculate the average crack growth rate between two adjacent projection image acquisitions according to the lengths of the fatigue cracks corresponding to two adjacent projection image acquisitions and the number of cycles between the two intervals.

[0015] Furthermore, the method for extracting the maximum length crack corresponding to the same cycle number includes:

[0016] S51, for all pre-processed projection images corresponding to the same cycle, according to the order of projection image acquisition, the grayscale values ​​corresponding to the pixels of the projection images are sequentially stored in each row of a three-dimensional matrix;

[0017] S52, using an edge detection algorithm to obtain pixel points containing the sample contour in the three-dimensional matrix row by row to form multiple contour matrices;

[0018] S53, for each contour matrix, respectively extracting pixel points with the same abscissa and pixel points with the same ordinate, and respectively calculating the abscissa mean of the pixel points with the same abscissa and the ordinate mean of the pixel points with the same ordinate;

[0019] S54, selecting a pixel point whose horizontal and vertical coordinates of the pixel point in each contour matrix are equal to the maximum horizontal coordinate mean and the maximum vertical coordinate mean as the center of the circle;

[0020] S55, calculating the distance between each contour pixel point in the contour matrix and its corresponding circle center, and performing probability statistics, and using the distance with the highest probability of occurrence as the radius r of the circle;

[0021] S56, for each circle corresponding to the preprocessed projection image, assign a value of 0 to the pixel points whose distance from the center of the circle exceeds the radius, and assign a value of 1 to the remaining pixel points, thereby obtaining a matrix 1;

[0022] S57, performing threshold segmentation processing on the preprocessed projection image, assigning a value of 0 to the pixel points above the preset threshold in the same projection image, and assigning a value of 1 to the remaining pixel points, to obtain a matrix 2;

[0023] S58, performing an element-by-element multiplication operation on the matrix 1 and the matrix 2 corresponding to the same projection image to obtain a crack area, then performing edge detection on the crack area, retaining the grayscale value of the pixel points at the crack edge, and assigning the remaining pixel points to 0;

[0024] S59, calculating the distance l from the edge pixel point to the center of the circle in each projection image, the crack length a=rl, and selecting the maximum length crack in all projection images corresponding to the same cycle as the fatigue crack.

[0025] Furthermore, the method for extracting the maximum length crack corresponding to the same cycle number includes:

[0026] A1. Import all pre-processed projection images corresponding to the same cycle into the 3D reconstruction software, and filter out points in the projection images with brightness greater than a preset value;

[0027] A2. By using the brightness inversion command, the matrix and the crack area are assigned the first grayscale value and the second grayscale value respectively, and the crack and pore parts are selected by the grayscale threshold, wherein the first grayscale value is less than the second grayscale value;

[0028] A3. Perform numerical analysis on the extracted cracks and pores to obtain the volume, surface area, length and width of the cracks and pores, retain the crack data with a size greater than the size threshold, and generate a crack solid model;

[0029] A4. Use the ruler tool in the 3D reconstruction software to obtain the length of the crack solid model, and select the crack with the maximum length in all projection images corresponding to the same cycle as the fatigue crack.

[0030] Furthermore, the method for preprocessing the projection image of the coaxial phase contrast CT includes:

[0031] Input the projection image of the coaxial phase contrast CT into the image algorithm software PITER or PITRE_BM, select FlatDark+ for interval background sampling, and adjust the parameters to achieve phase recovery of the projection image;

[0032] Select a projection image in the middle of the sensitive area of ​​the sample to generate a single sinusoidal image, and adjust the rotation axis of the sample;

[0033] The Ring Width value is set to correct the ring artifacts of all projection images, and then the 32-bit projection image is converted into 8-bit data to obtain the preprocessed projection image.

[0034] Furthermore, each cycle refers to the complete stress change process from minimum stress to maximum stress and then back to minimum stress during the stress loading process.

[0035] Furthermore, the calculation expression of the average crack growth rate is:

[0036]

[0037] in, is the average crack growth rate between the i-1th moment and the ith moment; the ith moment corresponds to the ith completion of the preset number of cycles of loading; a i and a i-1 are the fatigue crack lengths at the i-1th moment and the ith moment respectively; N i and N i-1 They are the cumulative number of cycles completed by the sample at the i-th moment and the i-1-th moment respectively.

[0038] Furthermore, the preset loading force is an average value of the sum of the maximum stress and the minimum stress of each cycle.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. This scheme uses the existing in-situ corrosion fatigue testing machine and synchrotron radiation light source to collect projection images of coaxial phase contrast CT in the sensitive area of ​​the sample. Compared with the commonly used potential difference method, replication method, and infrared thermal imaging measurement method, this method saves the construction of observation equipment during the measurement process. This method has a wide range of applicability and can be extended to the measurement of crack growth rate of metals and other materials in complex environments.

[0041] 2. This solution collects projection images under the rotating sample. The three-dimensional imaging data obtained in this way can calculate the crack growth rate of the sample in multiple directions;

[0042] In addition, this solution uses the distance from the edge pixel point to the center of the circle in each projection image to obtain the crack length. The entire process can be automatically processed and analyzed without manual annotation or software adjustment, which can greatly reduce manual intervention and improve accuracy.

[0043] 3. This scheme applies synchrotron radiation light source technology to corrosion fatigue crack growth tests, and uses X-ray phase contrast imaging to accurately obtain crack growth data. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 Schematic diagram of the structure of an hourglass-shaped sample.

[0045] Figure 2This is a flow chart of the in-situ corrosion fatigue crack growth rate measurement method based on synchrotron radiation light source.

[0046] Figure 3 This is the schematic diagram of in-situ corrosion fatigue forming.

[0047] Figure 4 A three-dimensional image formed by stacking all projection images corresponding to the same cycle.

[0048] Figure 5 This is the schematic diagram of the center and radius measurement corresponding to the contour matrix.

[0049] Figure 6 This is the crack length principle diagram.

[0050] Figure 7 This is the evolution curve of corrosion fatigue crack length and growth rate with cycle number.

[0051] Figure 8 This is a graph showing the variation of corrosion fatigue crack growth rate with stress intensity factor.

[0052] Among them, 1. sample; 11. sensitive area; 2. synchrotron radiation source; 21. monochromator; 22. detector; 3. in-situ corrosion fatigue testing machine; 4. corrosive medium. DETAILED DESCRIPTION

[0053] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0054] refer to Figure 2 , Figure 2 A flow chart of an in-situ corrosion fatigue crack growth rate measurement method based on a synchrotron radiation light source 2 is shown; Figure 1 As shown, the method S includes steps S1 to S6.

[0055] In step S1, the in-situ corrosion fatigue tester 3 is installed on the three-axis displacement turntable of the synchrotron radiation source 2, and the hourglass-shaped sample 1 is installed on the in-situ corrosion fatigue tester 3 so that it is coaxial with the three-axis displacement turntable. The specific structure of the in-situ corrosion fatigue tester 3 of this scheme can refer to patent application CN202310083984.3. When installing the sample 1, the automatic centering function of the in-situ corrosion fatigue tester 3 is used to apply a preset load to ensure that the sample 1 is coaxial with the rotating shaft.

[0056] like Figure 1As shown, the hourglass-shaped specimen 1 of this scheme is designed with dimensions and surface roughness according to GB / T 6398-2017 national standard "Metallic material fatigue test fatigue crack growth method". The hourglass-shaped specimen 1 used in the present invention has a length of F, a clamping end diameter of G, a straight end diameter of C, and a length of E. Figure 1 The areas with relatively large diameters at the two ends are the clamping sections of the specimen 1, and the area with relatively small dimensions in the middle is the sensitive area 11, that is, the area where cracks are generated and eventually break during the test.

[0057] like Figure 3 As shown, the sensitive area 11 of the sample 1 is located on the same straight line as the monochromator 21 and the detector 22 of the synchrotron radiation light source 2 to ensure that the X-ray beam is focused on the sensitive area 11 of the sample 1 so as to accurately collect image information.

[0058] Before the test begins, the test parameters and imaging parameters need to be determined. The test parameters include the diameter d of the sample 1, the stress amplitude σ a , stress ratio R and fatigue frequency f, through the formula S = π (d / 2) 2 , T=1 / f,σ max =2σ m / (1-R),σ min =Rσ max Calculate cross-sectional area S, cycle period T, and peak stress σ max and stress valley value σ min ; The imaging parameters involve the settings of light source exposure time and sampling frequency.

[0059] In step S2, the environment of the sample 1 is set to the corrosive medium 4 environment (specifically, salt spray solution / salt spray), the ceramic heating plate of the in-situ corrosion fatigue testing machine 3 is controlled to heat (specifically, the corrosive medium 4 in which the sample 1 is located is heated), and the temperature is maintained constant to the target temperature.

[0060] In step S3, after the sample 1 is subjected to a preset number of cyclic stress loadings, the preset loading force is maintained, and the synchrotron radiation light source 2 is used to collect the projection images of the coaxial phase contrast CT of different faults of the sensitive area 11 of the sample 1 while the sample 1 is rotating. Each time the image is collected, the height of the sample 1 is adjusted by the three-axis displacement turntable to collect the projection image of each fault layer by layer from the upper end of the sensitive area 11 downwards. After the entire sensitive area 11 is collected, the next step is entered.

[0061] During implementation, this scheme preferably refers to a complete stress change process from minimum stress to maximum stress and then back to minimum stress during stress loading in each cycle; the preset loading force is the average value of the sum of the maximum stress and the minimum stress of each cycle.

[0062] In step S4, it is determined whether the sample 1 is broken and failed. If so, the process proceeds to step S5, otherwise, it returns to step S3;

[0063] In step S5, the projection images of the coaxial phase contrast CT are preprocessed, and the maximum length cracks in all the preprocessed projection images corresponding to the same cycle are extracted as fatigue cracks;

[0064] In one embodiment of the present invention, a method for preprocessing a projection image of in-axial phase contrast CT includes:

[0065] Input the projection image of the coaxial phase contrast CT into the image algorithm software PITER or PITRE_BM, select FlatDark+ for interval background sampling, and adjust the parameters (here they can be manually adjusted to the appropriate parameters) to achieve phase recovery of the projection image.

[0066] A projection image in the middle of the sensitive area 11 of sample 1 is selected to generate a single sinusoidal image, and the rotation axis of sample 1 is adjusted. This is mainly due to technical problems with the light source line station and the stability of the testing machine during the imaging process, which cause semicircular artifacts and annular artifacts to appear in the image, affecting the imaging effect. Setting appropriate rotation axis parameters can effectively eliminate semicircular artifacts and display the internal defect characteristics of sample 1.

[0067] The Ring Width value is set to correct the ring artifacts of all projection images, and then the 32-bit projection image is converted into 8-bit data to obtain the preprocessed projection image.

[0068] In step S6, the average crack growth rate between two adjacent projection image acquisitions is calculated according to the lengths of the fatigue cracks corresponding to the two adjacent projection image acquisitions and the number of cycles between them:

[0069]

[0070] in, is the average crack growth rate between the i-1th moment and the ith moment; the ith moment corresponds to the ith completion of the preset number of cycles of loading; a i and a i-1 are the fatigue crack lengths at the i-1th moment and the ith moment respectively; N i and N i-1 They are the cumulative number of cycles completed by sample 1 at the i-th moment and the i-1-th moment respectively.

[0071] In this scheme, two methods are provided to extract the maximum length crack, which are described as follows:

[0072] The first method to extract the maximum length crack corresponding to the same cycle number includes:

[0073] S51. For all preprocessed projection images corresponding to the same cycle, the grayscale values ​​corresponding to the pixels of the projection images are sequentially stored in each row of a three-dimensional matrix according to the order of projection image acquisition. The three-dimensional matrix is ​​equivalent to a three-dimensional graph formed by stacking. For details, please refer to Figure 4 .

[0074] S52, using an edge detection algorithm to obtain pixel points containing the contour of the sample 1 in the three-dimensional matrix row by row to form a plurality of contour matrices;

[0075] S53, for each contour matrix, respectively extracting pixel points with the same abscissa and pixel points with the same ordinate, and respectively calculating the abscissa mean of the pixel points with the same abscissa and the ordinate mean of the pixel points with the same ordinate;

[0076] S54, selecting a pixel point whose horizontal and vertical coordinates of the pixel point in each contour matrix are equal to the maximum horizontal coordinate mean and the maximum vertical coordinate mean as the center of the circle;

[0077] S55. Calculate the distance between each contour pixel in the contour matrix and its corresponding circle center, and perform probability statistics. Use the distance with the highest probability of occurrence as the radius r of the circle. The principle diagram of measuring the circle center and radius corresponding to the contour matrix can be referred to. Figure 5 .

[0078] S56, for each circle corresponding to the preprocessed projection image, assign a value of 0 to the pixel points whose distance from the center of the circle exceeds the radius, and assign a value of 1 to the remaining pixel points, thereby obtaining a matrix 1;

[0079] S57, performing threshold segmentation processing on the preprocessed projection image, assigning a value of 0 to the pixel points above the preset threshold in the same projection image, and assigning a value of 1 to the remaining pixel points, to obtain a matrix 2;

[0080] S58, performing an element-by-element multiplication operation on the matrix 1 and the matrix 2 corresponding to the same projection image to obtain a crack area, then performing edge detection on the crack area, retaining the grayscale value of the pixel points at the crack edge, and assigning the remaining pixel points to 0;

[0081] S59, calculate the distance l from the edge pixel point to the center of the circle in each projection image, the crack length a=rl, and the principle diagram of the crack length can be referred to Figure 6 ; Then the crack with the maximum length in all projection images corresponding to the same cycle is selected as the fatigue crack.

[0082] The second method for extracting the maximum length crack corresponding to the same cycle number includes:

[0083] A1. Import all preprocessed projection images corresponding to the same cycle into the 3D reconstruction software (data processing is performed in Avizo), and filter the points in the projection image whose brightness is greater than the preset value to obtain a suitable grayscale range, that is, select the matrix part of sample 1 that does not contain defects.

[0084] A2. By using the brightness inversion command, the matrix and the crack area are assigned the first grayscale value and the second grayscale value respectively, and the crack and pore parts are selected by the grayscale threshold, wherein the first grayscale value is less than the second grayscale value;

[0085] A3. Perform numerical analysis on the extracted cracks and pores to obtain the volume, surface area, length and width of the cracks and pores, retain the crack data with a size greater than the size threshold, and generate a crack solid model;

[0086] A4. Use the ruler tool in the 3D reconstruction software to obtain the length of the crack solid model, and select the crack with the maximum length in all projection images corresponding to the same cycle as the fatigue crack.

[0087] Taking 7050-T7451 aluminum alloy as an example, the fatigue crack growth rate of pre-corroded 7050 aluminum alloy in 3.5wt.% NaCl solution was successfully measured using the corrosion fatigue crack growth rate method proposed in this scheme. The specific test conditions are as follows:

[0088] Based on the 2BL16U2 imaging beamline of the Shanghai Synchrotron Radiation Facility, the test was carried out using an in-situ corrosion fatigue testing machine 3. The diameter of the sample 1 was 1 mm, the corrosion environment was 3.5 wt.% NaCl solution, and the stress amplitude was 320 MPa.

[0089] In-situ imaging data of sample 1 is obtained by adopting steps S2 to S4, and then the data processing method of steps S51 to S59 is used to obtain the evolution of corrosion fatigue crack length and corrosion fatigue crack growth rate with the number of cycles. The results are as follows: Figure 7 and Figure 8 As shown; from these two figures, it can be seen that on the same specimen, this scheme simultaneously measures the crack growth rates of four corrosion fatigue cracks, and the lengths of these corrosion fatigue cracks are positively correlated with the number of cycles.

Claims

1. The in-situ corrosion fatigue crack growth rate measurement method based on synchrotron radiation source is characterized by: Includes steps: S1. Install the in-situ corrosion fatigue testing machine on the three-axis displacement turntable of the synchrotron radiation source, and install the hourglass-shaped specimen on the in-situ corrosion fatigue testing machine so that it is coaxial with the three-axis displacement turntable; S2. Set the environment of the sample to a corrosive medium environment, control the heating of the ceramic heating plate of the in-situ corrosion fatigue testing machine, and keep the temperature constant to the target temperature; S3. After the sample is subjected to a preset number of cyclic stress loadings, the preset loading force is maintained, and while the sample is rotating, a synchrotron radiation light source is used to collect projection images of coaxial phase contrast CT of different faults in the sensitive area of ​​the sample; S4, determine whether the sample is broken and failed, if so, proceed to step S5, otherwise return to step S3; S5, preprocessing the projection images of the coaxial phase contrast CT, and extracting the maximum length crack in all preprocessed projection images corresponding to the same cycle as the fatigue crack; S6. Calculate the average crack growth rate between two adjacent projection image acquisitions according to the lengths of the fatigue cracks corresponding to two adjacent projection image acquisitions and the number of cycles between the two intervals.

2. The in-situ corrosion fatigue crack growth rate measurement method according to claim 1, characterized in that: Methods for extracting the maximum length crack corresponding to the same cycle include: S51, for all pre-processed projection images corresponding to the same cycle, according to the order of projection image acquisition, the grayscale values ​​corresponding to the pixels of the projection images are sequentially stored in each row of a three-dimensional matrix; S52, using an edge detection algorithm to obtain pixel points containing the sample contour in the three-dimensional matrix row by row to form multiple contour matrices; S53, for each contour matrix, respectively extracting pixel points with the same abscissa and pixel points with the same ordinate, and respectively calculating the abscissa mean of the pixel points with the same abscissa and the ordinate mean of the pixel points with the same ordinate; S54, selecting a pixel point whose horizontal and vertical coordinates of the pixel point in each contour matrix are equal to the maximum horizontal coordinate mean and the maximum vertical coordinate mean as the center of the circle; S55, calculating the distance between each contour pixel point in the contour matrix and its corresponding circle center, and performing probability statistics, and using the distance with the highest probability of occurrence as the radius r of the circle; S56, for each circle corresponding to the preprocessed projection image, assign a value of 0 to the pixel points whose distance from the center of the circle exceeds the radius, and assign a value of 1 to the remaining pixel points, thereby obtaining a matrix 1; S57, performing threshold segmentation processing on the preprocessed projection image, assigning a value of 0 to the pixel points above the preset threshold in the same projection image, and assigning a value of 1 to the remaining pixel points, to obtain a matrix 2; S58, performing an element-by-element multiplication operation on the matrix 1 and the matrix 2 corresponding to the same projection image to obtain a crack area, then performing edge detection on the crack area, retaining the grayscale value of the pixel points at the crack edge, and assigning the remaining pixel points to 0; S59, calculating the distance l from the edge pixel point to the center of the circle in each projection image, the crack length a=rl, and selecting the maximum length crack in all projection images corresponding to the same cycle as the fatigue crack.

3. The in-situ corrosion fatigue crack growth rate measurement method according to claim 1, characterized in that: Methods for extracting the maximum length crack corresponding to the same cycle include: A1. Import all pre-processed projection images corresponding to the same cycle into the 3D reconstruction software, and filter out points in the projection images with brightness greater than a preset value; A2. By using the brightness inversion command, the matrix and the crack area are assigned the first grayscale value and the second grayscale value respectively, and the crack and pore parts are selected by the grayscale threshold, wherein the first grayscale value is less than the second grayscale value; A3. Perform numerical analysis on the extracted cracks and pores to obtain the volume, surface area, length and width of the cracks and pores, retain the crack data with a size greater than the size threshold, and generate a crack solid model; A4. Use the ruler tool in the 3D reconstruction software to obtain the length of the crack solid model, and select the crack with the maximum length in all projection images corresponding to the same cycle as the fatigue crack.

4. The in-situ corrosion fatigue crack growth rate measurement method according to any one of claims 1 to 3, characterized in that: The method for preprocessing the projection image of the coaxial phase contrast CT includes: Input the projection image of the coaxial phase contrast CT into the image algorithm software PITER or PITRE_BM, select FlatDark+ for interval background sampling, and adjust the parameters to achieve phase recovery of the projection image; Select a projection image in the middle of the sensitive area of ​​the sample to generate a single sinusoidal image, and adjust the rotation axis of the sample; The Ring Width value is set to correct the ring artifacts of all projection images, and then the 32-bit projection image is converted into 8-bit data to obtain the preprocessed projection image.

5. The in-situ corrosion fatigue crack growth rate measurement method according to claim 1, characterized in that: Each cycle refers to the complete stress change process from minimum stress to maximum stress and then back to minimum stress during stress loading.

6. The in-situ corrosion fatigue crack growth rate measurement method according to claim 1, characterized in that: The calculation expression of the average crack growth rate is: in, is the average crack growth rate between the i-1th moment and the ith moment; the ith moment corresponds to the ith completion of the preset number of cycles of loading; a i and a i-1 are the fatigue crack lengths at the i-1th moment and the ith moment respectively; N i and N i-1 They are the cumulative number of cycles completed by the sample at the i-th moment and the i-1-th moment respectively.

7. The in-situ corrosion fatigue crack growth rate measurement method according to claim 1, characterized in that: The preset loading force is the average value of the sum of the maximum stress and the minimum stress of each cycle.

Images