Complex structure blade defect detection capability quantitative characterization method

Abstract

The embodiment of the invention provides a quantitative characterization method for defect detection capability of a blade with a complex structure, and relates to the technical field of turbine blade detection characterization. The objective of the invention is to improve the problem that the capability of detecting defects at different positions in a blade with a complex structure cannot be effectively represented. Comprising the steps of obtaining a three-dimensional grayscale image of a to-be-measured blade Then calculating gray levels, standard deviations and noises of different cross sections on the three-dimensional gray level image; then, carrying out intrinsic contrast calculation on the first selected region position with the minimum noise, and calculating to obtain a modulation transfer function (MTF); performing grid division of different sizes on the three-dimensional grayscale image, and calculating noise of a second selected area corresponding to different grid sizes; and finally, the minimum recognizable defect size of different second selected areas of each section is calculated through a defect recognition criterion, quantitative characterization of the defect detection capability of the complex structure blade is finally realized, and the technical bottleneck that the traditional defect detection capability of the complex structure blade lacks an effective evaluation method is broken through.

Description

Technical Field

[0001] This invention relates to the field of turbine blade inspection and characterization technology, and more specifically, to a quantitative characterization method for the defect detection capability of complex structure blades. Background Technology

[0002] Complex blade structures, such as turbine blades, are critical components in aero-engines and gas turbines. During manufacturing and service, they require non-destructive testing to ensure they are free of defects and guarantee operational safety. For internal defect detection in complex blade structures, industrially used methods primarily include radiographic testing and CT (Computed Tomography) testing. Radiographic testing, however, is often affected by the blade profile and complex internal structure, resulting in blind spots that significantly impact the reliability of defect detection. CT testing, on the other hand, offers high density contrast, is sensitive to density changes and internal defects, and provides intuitive results, making it increasingly used for internal defect detection in aero-engine and gas turbine blades.

[0003] Due to the high density and complex profile of turbine blades, which require significant penetration thickness, the penetration thickness varies along different paths during CT inspection. Furthermore, the complex internal structure easily generates X-ray scattering, leading to beam hardening and scattering artifacts, reducing image quality and affecting defect detection capabilities. Because of differences in penetration ability and the location and degree of scattering influence, the defect recognition capability varies across different locations in turbine blade CT images. Therefore, quantitative characterization of defect detection capabilities in different regions of the blade is necessary. Characterization of CT detection capabilities typically involves measuring the density and spatial resolution of standard parts under specific inspection conditions, or designing defect comparison samples and assessing the defect detection capability by detecting defects on these samples. However, standard parts are usually simple in structure and cannot represent the actual inspection conditions of real turbine blades; designing defect comparison samples is difficult, making it hard to prefabricate real defects that meet inspection requirements, thus failing to accurately evaluate the CT defect detection capability of real turbine blades. Therefore, there is currently no effective method to characterize the defect detection capability at different locations within complex turbine blades. Summary of the Invention

[0004] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form to prepare for the more detailed descriptions that follow.

[0005] The present invention aims to provide, for example, a method for quantitatively characterizing the defect detection capability of blades with complex structures, which can improve the problem of the inability to effectively characterize the defect detection capability at different locations inside blades with complex structures.

[0006] The embodiments of the present invention can be implemented as follows:

[0007] This invention provides a method for quantitatively characterizing the defect detection capability of complex structure blades, comprising the following steps: Step a, acquiring a three-dimensional grayscale image of the blade to be tested, and cropping multiple cross-sections at intervals from the three-dimensional grayscale image; Step b, calculating the noise of multiple different first selected regions within each cross-section, using a single pixel as the unit, according to the noise calculation formula; selecting the position of the first selected region corresponding to the minimum noise among all calculated noises for intrinsic contrast calculation, and performing function calculation on the contour boundary of the cross-section where the first selected region corresponding to the minimum noise is located to obtain the modulation transfer function (MTF); Step c, calculating the second selected region within the cross-section, using a single square as the unit, according to the noise calculation formula. The noise level of the region and the average gray value of the second selected region are combined with the average gray value of the first selected region corresponding to the minimum noise and the intrinsic contrast to calculate the lower limit of the defect modulation index. Then, the pixel size corresponding to the modulation transfer function (MTF) is read, and the identifiable defect size of the second selected region is calculated. This is compared with the preset defect size to obtain the minimum identifiable defect size within the second selected region. The area of ​​the square is equal to K pixels × K pixels, where K is a positive integer obtained based on the preset defect size. In step d, different second selected regions and different cross sections are selected, and step c is repeated to obtain the minimum identifiable defect size of the second selected region on all cross sections of the blade under test.

[0008] In addition, the quantitative characterization method for detecting defects in complex blade structures provided in the embodiments of the present invention may also have the following additional technical features:

[0009] Optionally, the area S1 of the first selected region is equal to a pixels × b pixels.

[0010] Optionally, the area S2 of the second selected region is equal to n × the area of ​​the squares, where n is the number of squares included in the second selected region.

[0011] Optionally, the noise formula includes:

[0012]

[0013]

[0014] Where, N jThe noise level in the selected area is represented by σ, where σ is the standard deviation of the grayscale values ​​in the selected area. μ is the average gray value of the selected area. i This represents the grayscale value of a single cell within the selected area, where a×b is the number of single cells within the selected area.

[0015] Optionally, the formula for calculating the intrinsic contrast ratio includes:

[0016] Where, Δμ 本征 For intrinsic contrast, The average gray value of the first selected region corresponding to the minimum noise. The average gray value of the air region near the first selected region with the least noise.

[0017] Optionally, the step of performing function calculation on the contour boundary of the cross section where the first selected region with minimum noise is located to obtain the modulation transfer function (MTF) includes:

[0018] Take the grayscale value upwards from the contour boundary of the cross section where the first selected region corresponding to the minimum noise is located, and draw the edge response curve; differentiate the edge response curve to obtain the point spread function; perform a fast Fourier transform on the point spread function to obtain the modulation transfer function (MTF).

[0019] Optionally, the step of calculating the identifiable defect size of the second selected area and comparing it with a preset defect size to obtain the minimum identifiable defect size within the second selected area includes:

[0020] If the difference between the identifiable defect size L_lim in the second selected region and the preset defect size L is within half a pixel, then the current identifiable defect size L_lim in the second selected region is the smallest identifiable defect size in the second selected region.

[0021] If the difference between the identifiable defect size L_lim of the second selected area and the preset defect size exceeds half a pixel, then the preset defect size L is set to the current identifiable defect size L_lim of the second selected area. Step c is repeated until the difference between the identifiable defect size L_lim of the second selected area and the preset defect size L is within half a pixel.

[0022] Optionally, the step of selecting K as a positive integer based on the preset defect size includes:

[0023] Based on the preset defect size L and the single pixel size Pix, the number of single pixels D included in the preset defect size is obtained, where D = [L / Pix], K = D, and [] indicates rounding down to the nearest integer.

[0024] Optionally, the formula for calculating the lower limit of the defect modulation index includes:

[0025] Where MTF_D_lim is the lower limit of defect modulation, N ROI-DD The noise is calculated using the noise formula for the second selected region when K=D. The average gray value of the second selected region calculated according to the noise formula when K=D. Δμ is the average gray value corresponding to the first selected region with the least noise. 本征 This is the intrinsic contrast ratio.

[0026] Optionally, the calculation formula for reading the pixel size corresponding to the modulation transfer function (MTF) and calculating the identifiable defect size of the second selected region includes:

[0027] L_lim = 1 / D_lim / 2, where D_lim is the pixel size corresponding to the lower limit of the defect modulation on the modulation transfer function (MTF), and L_lim is the defect size that can be identified in the second selected region.

[0028] The beneficial effects of the quantitative characterization method for detecting defects in complex blade structures according to embodiments of the present invention include, for example:

[0029] A quantitative characterization method for the defect detection capability of complex structure blades includes the following steps: obtaining a three-dimensional grayscale image of the blade to be tested; then calculating the grayscale, standard deviation, and noise of different sections on the three-dimensional grayscale image; then calculating the intrinsic contrast of the first selected region with the least noise and obtaining the modulation transfer function (MTF); dividing the three-dimensional grayscale image into grids of different sizes and calculating the noise of the second selected region corresponding to different grid sizes; finally, calculating the minimum identifiable defect size of different second selected regions on each section using defect identification criteria, thereby achieving a quantitative characterization of the defect detection capability of complex structure blades. This method overcomes the technical bottleneck of the lack of effective evaluation methods for the defect detection capability of traditional complex structure blades; it enables the quantitative characterization of the minimum identifiable defects at different locations on the blade; and it characterizes the blade detection capability without the need for defect comparison samples or defective parts. Attached Figure Description

[0030] The above-described features and advantages of the present invention will be better understood after reading the following detailed description of embodiments of the present disclosure in conjunction with the accompanying drawings. In the drawings, components are not necessarily drawn to scale, and components having similar related characteristics or features may have the same or similar reference numerals.

[0031] Figure 1The CT scan image of the leaf under test provided in an embodiment of the present invention is shown;

[0032] Figure 2 A cross-sectional view of a three-dimensional grayscale image of a blade under test provided in an embodiment of the present invention is shown;

[0033] Figure 3 A schematic diagram of a local region of a three-dimensional grayscale image cross-section provided in an embodiment of the present invention is shown;

[0034] Figure 4 The local contour of the cross section of the first selected region corresponding to the minimum noise in the three-dimensional grayscale image provided in the embodiment of the present invention is shown;

[0035] Figure 5 A schematic diagram of the edge response curve provided in an embodiment of the present invention is shown;

[0036] Figure 6 A schematic diagram of the point extension function provided in an embodiment of the present invention is shown;

[0037] Figure 7 A schematic diagram of the modulation transfer function (MTF) curve provided in an embodiment of the present invention is shown;

[0038] Figure 8 This diagram illustrates the process of dividing a three-dimensional grayscale image into networks of different sizes, as provided in an embodiment of the present invention.

[0039] Figure 9 This diagram illustrates the selection location of the second selected region (ROI) of a cross section at a certain height Hi according to an embodiment of the present invention.

[0040] Figure 10 The following is a flowchart illustrating the steps of the method for quantitatively characterizing the defect detection capability of complex structure blades provided in an embodiment of the present invention;

[0041] Figure 11 The flowchart of the calculation process for the quantitative characterization method of defect detection capability of complex structure blades provided in the embodiments of the present invention is shown. Detailed Implementation

[0042] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should be noted that the aspects described below with reference to the accompanying drawings and specific embodiments are merely exemplary and should not be construed as limiting the scope of protection of the present invention in any way.

[0043] In the description of this invention, it should be noted that if terms such as "upper," "lower," "inner," "outer," or "vertical" appear, the orientation or positional relationship indicated is based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the product of this invention is usually placed when in use, and does not indicate or imply that the device or element referred to must have a specific orientation or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0044] At the same time, it should be noted that the terms "first" and "second" are used only for distinguishing descriptions and should not be interpreted as indicating or implying relative importance.

[0045] In the description of this invention, it should also be noted that, unless otherwise explicitly specified or limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, an integral connection, or a detachable connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium, or a connection within two components, etc. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0046] The following is combined Figures 1 to 11 The quantitative characterization method for detecting defects in complex blade structures provided in this embodiment is described in detail.

[0047] Please refer to Figures 1 to 11 This embodiment provides a method for quantitatively characterizing the defect detection capability of blades with complex structures. The method includes the following steps:

[0048] Reference Figure 1 , Figure 10 and Figure 11 Step a: Obtain a three-dimensional grayscale image of the blade to be tested, and extract multiple cross sections with interval distribution from the three-dimensional grayscale image.

[0049] Specifically, the blade under test is scanned and reconstructed using CT to obtain a three-dimensional grayscale image of the blade. Here, "CT" stands for CT (x-ray computed tomography), a computer-aided imaging method. The "blade under test" is a crucial component of the turbine section in an engine or gas turbine. High-speed airflow drives the compressor or turbine blades to perform work, providing power to the engine or gas turbine. "Multiple spaced cross-sections" refers to multiple cross-sections spaced apart with a step size of Δ. The value of Δ is determined based on the number of cross-sections to be captured within the defect detection range of the blade under test. In this embodiment, specifically, multiple spaced cross-sections are spaced apart along the height direction. The "height direction" refers to the direction of the turbine blade from the tenon to the blade tip. In other embodiments, multiple spaced cross-sections can also be captured along other directions, as long as they cover and characterize the entire blade.

[0050] Reference Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 and Figure 7 Step b: According to the noise calculation formula, calculate the noise of multiple different first selected regions in each cross section, taking a single pixel as the unit; select the position of the first selected region corresponding to the minimum noise among all the calculated noises for intrinsic contrast calculation, and perform function calculation on the contour boundary of the cross section where the first selected region corresponding to the minimum noise is located to obtain the modulation transfer function (MTF).

[0051] First, different first selected regions represent different locations. A noise value is calculated within each first selected region. Multiple first selected regions on a cross section yield noise values ​​at multiple different locations. The location with the minimum noise on one cross section is selected. Then, the above calculation is repeated for each cross section to obtain the minimum noise location for each cross section. The location corresponding to the minimum noise among all minimum noise values ​​is then used to calculate the intrinsic contrast. This location is also the minimum noise location for the entire blade under test. The intrinsic contrast calculation is used in subsequent steps. Furthermore, for the cross section where the minimum noise location of the entire blade under test is located, a function calculation is performed on the cross section contour boundary to obtain the modulation transfer function (MTF), which is also used in subsequent steps. The noise in the first selected region is calculated according to the noise calculation formula. When calculating the noise within each first selected region, a single pixel is used as the smallest unit for calculation.

[0052] Reference Figure 8 and Figure 9Step c: According to the noise calculation formula, using a single square as a unit, calculate the noise of the second selected region and the average gray value of the second selected region within the cross section. Combine the average gray value of the first selected region corresponding to the minimum noise and the intrinsic contrast to calculate the lower limit of the defect modulation index. Then, read the pixel size corresponding to the modulation transfer function (MTF) to calculate the identifiable defect size of the second selected region. Compare it with the preset defect size to obtain the minimum identifiable defect size within the second selected region. Wherein, the area of ​​the square is equal to K pixels × K pixels, and K is selected as a positive integer obtained based on the preset defect size.

[0053] The "noise in the second selected area" is calculated using the same formula as the noise in the first selected area, except that the "noise and average gray value in the second selected area" are calculated using a single square as the smallest unit, and each square contains multiple pixels. Based on the noise in the second selected area, the average gray value in the second selected area, the average gray value in the first selected area corresponding to the minimum noise, and the intrinsic contrast, the lower limit of the defect modulation index can be calculated. On the modulation transfer function (MTF) curve, the pixel size corresponding to the lower limit of the defect modulation index is selected, and the identifiable defect size in the second selected area is calculated based on this pixel size. This identifiable defect size is then compared with a preset defect size. If the comparison meets the condition, the current identifiable defect size in the second selected area is the minimum identifiable defect size within the second selected area. If the comparison does not meet the condition, the preset defect size is set to a different value, and the calculation is repeated until the comparison between the identifiable defect size in the second selected area and the preset defect size meets the condition, thus yielding the minimum identifiable defect size within the second selected area.

[0054] Step d: Select different second selected regions and different cross sections, and repeat step c to obtain the smallest identifiable defect size of the different second selected regions on all cross sections of the blade under test.

[0055] "Selecting different second selected regions and different cross sections" includes changing the two variables of the second selected region and / or cross section. This includes repeatedly executing step c on different second selected regions on the same cross section to obtain the minimum identifiable defect size at different locations on the same cross section; and repeatedly executing step c on different cross sections, plus repeatedly executing step c on different second selected regions on the same cross section, to finally obtain the minimum identifiable defect size of different second selected regions on different cross sections, and finally obtain the minimum identifiable defect size of each part of the entire blade.

[0056] In this embodiment, the area S1 of the first selected region is equal to a pixels × b pixels. The noise and grayscale value of the first selected region are calculated using a single pixel as the smallest unit. The values ​​of a and b can be arbitrarily chosen.

[0057] In this embodiment, the area S2 of the second selected region is equal to n × the area of ​​the squares, where n is the number of squares included in the second selected region, and S2 includes multiple K pixels × K pixels. K is a positive integer such as 1, 2, ...

[0058] Reference Figure 9 In this embodiment, when calculating the minimum identifiable defect size for each second selected region, K is taken as a positive integer based on the preset defect size. Prior to this, using a single pixel size (Pix) as the unit, 1×1, 2×2, 3×3, ..., k×k squares are used to re-grid and reorganize the grayscale image of the cross-sectional image IMG_Hi. The re-grid images are denoted as IMG_Hi_k×k. In the second selected region, the noise of the re-grid image is calculated and denoted as N. ROI-kk Record the average gray value of the second selected area. Both the K value and the second selected region are variables. Different values ​​of K result in different ranges of the second selected region, leading to different databases of noise and grayscale values. When the K value is a positive integer derived from the preset defect size, the average grayscale value already calculated in the database can be directly retrieved. and noise.

[0059] In this embodiment, the noise formula includes:

[0060]

[0061]

[0062] Where, N j The noise level in the selected area is represented by σ, where σ is the standard deviation of the grayscale values ​​in the selected area. μ is the average gray value of the selected area. i This represents the grayscale value of a single unit within the selected area, where a × b is the number of single units within the selected area. The "selected area" can be either a first or a second selected area; when calculating the first selected area, a single unit is a single pixel, and the first selected area includes a pixels × b pixels; when calculating the second selected area, a single unit is a single square, and the second selected area includes n single squares.

[0063] In this embodiment, the formula for calculating intrinsic contrast includes: Where, Δμ 本征 For intrinsic contrast, The average gray value of the first selected region corresponding to the minimum noise. This is the average gray value of the air region near the first selected region with the least noise. The intrinsic contrast ratio is calculated for the location of the first selected region with the least noise.

[0064] Reference Figure 4 , Figure 5 , Figure 6 and Figure 7 In this embodiment, the step of calculating the modulation transfer function (MTF) by performing function calculation on the contour boundary of the cross section where the first selected region with minimum noise is located includes: taking the grayscale value in the normal direction of the contour boundary of the cross section where the first selected region with minimum noise is located, and plotting the edge response curve; differentiating the edge response curve to obtain the point spread function; and performing a fast Fourier transform on the point spread function to obtain the MTF. The MTF reflects the density and resolution of the selected region.

[0065] In this embodiment, the step of selecting K as a positive integer obtained from the preset defect size includes: obtaining the number of individual pixels D included in the preset defect size based on the preset defect size L and the individual pixel size Pix, where D = [L / Pix], K = D, and [] indicates rounding down to the nearest integer.

[0066] The preset defect size is the estimated defect size. After calculating the D value, the K value is substituted to calculate the identifiable defect size L_lim in the second selected area. Then, it is compared with L. If the comparison result is not within half a pixel, the value L = L_lim is taken and the calculation and verification are continued until the comparison result falls within half a pixel.

[0067] In this embodiment, the formula for calculating the lower limit of the defect modulation index includes:

[0068] Where MTF_D_lim is the lower limit of defect modulation, N ROI-DD The noise level for the second selected region is calculated using the noise formula when K=D. The average gray value of the second selected region calculated according to the noise formula when K=D. Δμ is the average gray value corresponding to the first selected region with minimum noise. 本征 This is the intrinsic contrast. The lower limit of the defect modulation is calculated based on the condition that the second selected region includes D pixels × D pixels.

[0069] Reference Figure 7In this embodiment, the calculation formula for reading the pixel size corresponding to the modulation transfer function (MTF) and calculating the identifiable defect size of the second selected region includes: L_lim = 1 / D_lim / 2, where D_lim is the pixel size corresponding to the lower limit of the defect modulation on the MTF, and L_lim is the identifiable defect size of the second selected region.

[0070] In this embodiment, the step of calculating the identifiable defect size of the second selected area and comparing it with a preset defect size to obtain the minimum identifiable defect size within the second selected area includes:

[0071] If the difference between the identifiable defect size L_lim in the second selected area and the preset defect size L is within half a pixel, then the current identifiable defect size L_lim in the second selected area is the smallest identifiable defect size in the second selected area; if the difference between the identifiable defect size L_lim in the second selected area and the preset defect size exceeds half a pixel, then the preset defect size L is taken as the current identifiable defect size L_lim in the second selected area, and step c is repeated until the difference between the identifiable defect size L_lim in the second selected area and the preset defect size L is within half a pixel.

[0072] According to the quantitative characterization method for detecting defects in complex structure blades provided in this embodiment, the working principle of the quantitative characterization method for detecting defects in complex structure blades includes:

[0073] Reference Figure 10 and Figure 11 Defects are internal discontinuities that affect the integrity and service life of parts. Characterization is a method to quantitatively describe (represent) information such as the size of defects. In this embodiment, the blade under test is first scanned with a CT scanner to obtain a grayscale CT image of the blade. Then, the grayscale, standard deviation, and noise of different sections on the grayscale CT image are calculated. Next, the intrinsic contrast of the part with the least noise is calculated, and the modulation transfer function (MTF) is obtained. The CT image is divided into grids of different sizes, and the noise corresponding to different grid sizes is calculated. Finally, the minimum identifiable defect size of different regions of interest in each section is calculated using defect recognition criteria, thus achieving quantitative characterization of the CT defect detection capability of complex structure blades.

[0074] The specific process is as follows:

[0075] 1. Reference Figure 1 The blade under test was CT scanned and reconstructed to obtain a three-dimensional grayscale image of the blade under test.

[0076] 2. Reference Figure 2The cross-sectional slice image of the blade under test at height Hi of the three-dimensional grayscale image is denoted as IMG_Hi. Taking a single pixel as the smallest statistical unit, the average grayscale value of the blade under test within the first selected area S on this cross-section is calculated. Given the standard deviation σ, S = a pixels × b pixels, calculate the noise N at different positions j. j The calculation formulas are shown in equations (1)-(3), and the values ​​of a and b can be chosen arbitrarily;

[0077] in,

[0078] μ i This represents the grayscale value of a single pixel within the area S.

[0079] The calculation results of average gray value, standard deviation and noise in typical areas are shown in Table 1.

[0080] Table 1. Calculation results of grayscale, noise, and minimum identifiable defect for a certain cross section.

[0081]

[0082]

[0083] 3. Obtain the minimum noise N in the cross-sectional image of the blade height Hi. Hi-min For example, minimum noise N Hi-min

[0084] =0.85%;

[0085] 4. Using Δ as the step size, and taking different values ​​for the blade height Hi, repeat steps 2-3 to calculate the minimum noise of the grayscale image of the cross-section at all heights Hi of the blade under test. Here, the selected step size is Δ1mm. The minimum noise N on the cross-section of all blade heights Hi is then calculated. Hi-min The minimum value is denoted as N. MIN The location corresponding to the minimum noise is denoted as Area_MIN, which is also the location of a certain first selected area. The location corresponding to the minimum noise, Area_MIN, is S = a pixels × b pixels.

[0086] 5. Calculate the minimum noise N MIN The average gray value at the corresponding location Area_MIN is denoted as

[0087] In the air region near the leaf blade in this area, the average gray value of this air region is calculated and denoted as .

[0088] =1078; calculate the intrinsic contrast ratio Δμ using the following formula. 本征 ,

[0089]

[0090] 6. Reference Figure 4 and Figure 5 In the minimum noise N MIN On the wall thickness section at the corresponding location Area_MIN, take the grayscale value in the normal direction of the wall thickness contour boundary and draw the edge response curve;

[0091] 7. Reference Figure 6 The point spread function is obtained by differentiating the edge response curve.

[0092] 8. Reference Figure 7 Perform a Fast Fourier Transform on the point spread function to obtain the modulation transfer function (MTF); 9. Refer to Figure 9 Select the blade section height Hi;

[0093] 10. Reference Figure 9 The grayscale cells of the cross-sectional image IMG_Hi at the blade section height Hi are divided as follows:

[0094] 1) Using a single pixel size (Pix) as the unit, take 1×1, 2×2, 3×3, ..., k×k squares, and re-grid and reorganize the grayscale image of the cross-section image IMG_Hi. The images obtained after re-griding are denoted as IMG_Hi_k×k, where k is a positive integer such as 1, 2, ..., etc.

[0095] 2) In the second selected region ROI, repeat formulas (1)-(3). At this time, μ i This represents the grayscale value of a single k×k square within the second selected region of interest (ROI), where a×b is the number of squares included in the second selected region. The noise of the image after re-gridization is calculated and denoted as N. ROI-kk ;

[0096] 3) Record the average gray value of the second selected ROI.

[0097] 11. Preset a defect size, denoted as L = 0.3mm, and calculate the number of pixels per unit occupied by this length, as shown in the following formula:

[0098] D=[L / Pix],D=[L / Pix]=[0.3 / 0.026]=11 (4)

[0099] Where [] indicates rounding down to the nearest integer, and Pix is ​​the size of a single pixel;

[0100] 12. For the second selected region ROI, according to and And the noise level when the grid size in the second selected area is D, that is, when K=D, the lower limit of the identifiable defect modulation index MTF_D_lim is calculated according to formula (5):

[0101]

[0102] MTF_D_lim=3×1.06%×44793 / 35572 / 97.13%=0.04(5);

[0103] 13. Read the pixel size D_lim = 0.04 corresponding to the modulation index MTF_D_lim from the MTF curve;

[0104] Calculate the minimum identifiable defect size in this region using formula (6):

[0105] L_lim=1 / D_lim / 2=1 / 8.05 / 2=0.062mm(6);

[0106] 14. Determine if the difference between L_lim obtained in step 15 and L set in step 13 is within half a pixel. If the difference exceeds half a pixel, recalculate L = L_lim and repeat steps 13-16 until the difference between L_lim and L is within half a pixel. The smallest identifiable defect in this area is then L_lim obtained in the last calculation. Since 0.062mm and 0.3mm differ by more than half a pixel, select L = 0.062 and repeat steps 13-16. The final result is D_lim, which is close to L, at D_lim = 0.068mm. Therefore, the smallest identifiable defect in this area is L_lim = 0.068mm obtained in the last calculation.

[0107] 15. Repeat steps 10-14 to obtain the smallest identifiable defect size in different second selected regions on the cross section of blade height Hi, as shown in Table 1.

[0108] 16. Repeat steps 9-15 at different cross-sectional heights to obtain the smallest identifiable defect size in all parts of the entire blade, for example, 0.055 mm;

[0109] 17. Characterize the ability to detect CT defects in blades with complex structures.

[0110] The quantitative characterization method for detecting defects in complex blade structures provided in this embodiment has at least the following advantages:

[0111] A method for characterizing the CT defect detection capability of blades with complex structures is proposed, which can achieve quantitative characterization of the smallest identifiable defects at different locations on the blade; the CT detection capability of the blade can be characterized without the need for defect comparison samples or defective parts.

[0112] This method overcomes the bottleneck of traditional CT inspection methods, which cannot quantitatively and accurately characterize the CT inspection capability of leaves. It can quantitatively characterize the smallest identifiable defects at different locations on the leaf body. It can characterize the CT inspection capability of leaves without the need for defect comparison samples or defective parts. The defect identification accuracy can reach below 0.1mm. At the same time, it saves cycle costs and has broad application prospects and promotion value.

[0113] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for quantitatively characterizing the defect detection capability of blades with complex structures, characterized in that, Includes the following steps: Step a: Obtain a three-dimensional grayscale image of the blade to be tested, and crop multiple cross sections at intervals from the three-dimensional grayscale image; Step b: Based on the noise calculation formula, calculate the noise of multiple different first selected regions within each cross section, using a single pixel as the unit; The intrinsic contrast is calculated by selecting the position of the first selected region corresponding to the minimum noise among all the calculated noises, and the function is calculated on the contour boundary of the cross section where the first selected region corresponding to the minimum noise is located to obtain the modulation transfer function (MTF). Step c: Based on the noise calculation formula, using a single square as a unit, calculate the noise in the second selected region and the average gray value of the second selected region within the cross section. Combine the average gray value of the first selected region corresponding to the minimum noise and the intrinsic contrast to calculate the lower limit of the defect modulation index. Then, read the pixel size corresponding to the modulation transfer function (MTF) to calculate the identifiable defect size in the second selected region. Compare this with a preset defect size to obtain the minimum identifiable defect size in the second selected region. Wherein, the area of ​​the square is equal to K pixels × K pixels, and K is selected as a positive integer obtained based on the preset defect size. Step d: Select different second selected regions and different cross sections, and repeat step c to obtain the smallest identifiable defect size of the different second selected regions on all cross sections of the blade under test.

2. The method for quantitatively characterizing the defect detection capability of complex structure blades according to claim 1, characterized in that, The area S1 of the first selected region is equal to a pixels × b pixels.

3. The method for quantitatively characterizing the defect detection capability of complex structure blades according to claim 1, characterized in that, The area S2 of the second selected region is equal to n × the area of ​​the squares, where n is the number of squares included in the second selected region.

4. The method for quantitatively characterizing the defect detection capability of complex structure blades according to claim 1, characterized in that, The noise formula includes: Where, N j The noise level in the selected area is represented by σ, where σ is the standard deviation of the grayscale values ​​in the selected area. μ is the average gray value of the selected area. i This represents the grayscale value of a single cell within the selected area, where a×b is the number of single cells within the selected area.

5. The method for quantitatively characterizing the defect detection capability of complex structure blades according to claim 1, characterized in that, The formula for calculating the intrinsic contrast ratio includes: Where, Δμ 本征 For intrinsic contrast, The average gray value of the first selected region corresponding to the minimum noise. The average gray value of the air region near the first selected region with the least noise.

6. The method for quantitatively characterizing the defect detection capability of complex structure blades according to claim 1, characterized in that, The step of performing function calculation on the contour boundary of the cross section where the first selected region with minimum noise is located to obtain the modulation transfer function (MTF) includes: Take the grayscale value upwards from the contour boundary of the cross section where the first selected region corresponding to the minimum noise is located, and plot the edge response curve; Differentiating the edge response curve yields the point spread function; The point spread function is subjected to a fast Fourier transform to obtain the modulation transfer function (MTF).

7. The method for quantitatively characterizing the defect detection capability of complex structure blades according to any one of claims 1-6, characterized in that, The step of calculating the identifiable defect size of the second selected area and comparing it with a preset defect size to obtain the minimum identifiable defect size within the second selected area includes: If the difference between the identifiable defect size L_lim in the second selected region and the preset defect size L is within half a pixel, then the current identifiable defect size L_lim in the second selected region is the smallest identifiable defect size in the second selected region. If the difference between the identifiable defect size L_lim of the second selected area and the preset defect size exceeds half a pixel, then the preset defect size L is set to the current identifiable defect size L_lim of the second selected area. Step c is repeated until the difference between the identifiable defect size L_lim of the second selected area and the preset defect size L is within half a pixel.

8. The method for quantitatively characterizing the defect detection capability of complex structure blades according to claim 7, characterized in that, The step of selecting K as a positive integer based on the preset defect size includes: Based on the preset defect size L and the single pixel size Pix, the number of single pixels D included in the preset defect size is obtained, where D = [L / Pix], K = D, and [] indicates rounding down to the nearest integer.

9. The method for quantitatively characterizing the defect detection capability of complex structure blades according to claim 7, characterized in that, The formula for calculating the lower limit of the defect modulation index includes: Where MTF_D_lim is the lower limit of defect modulation, N ROI-DD The noise is calculated using the noise formula for the second selected region when K=D. The average gray value of the second selected region calculated according to the noise formula when K=D. Δμ is the average gray value corresponding to the first selected region with the least noise. 本征 This is the intrinsic contrast ratio.

10. The method for quantitatively characterizing the defect detection capability of complex structure blades according to claim 7, characterized in that, The calculation formula for reading the pixel size corresponding to the modulation transfer function (MTF) and calculating the identifiable defect size of the second selected region includes: L_lim = 1 / D_lim / 2, where D_lim is the pixel size corresponding to the lower limit of the defect modulation on the modulation transfer function (MTF), and L_lim is the defect size that can be identified in the second selected region.

Images