A single crystal superalloy heterocrystal automatic detection system based on transmission X-ray diffraction

Abstract

This invention discloses an automated detection system for impurities in single-crystal high-temperature alloys based on transmission X-ray diffraction, comprising: an X-ray source, an aperture, a sample stage, an area array detector, a beam deflector, and a control and data processing unit. This invention employs laboratory transmission Laue diffraction technology, which has strong penetration capabilities and is suitable for non-destructive testing of thick samples and complex single-crystal high-temperature alloy components. By defining the illumination area with an aperture and combining it with automatic sample translation scanning, automated and systematic acquisition of diffraction data is achieved, significantly improving detection efficiency and repeatability. Based on an unsupervised classification method, the diffraction images are analyzed, enabling objective identification of impurity regions without prior manual annotation, reducing the influence of subjective human factors. This invention is applicable to the quality inspection of single-crystal high-temperature alloy samples and engineering components, and has good engineering application prospects and promotional value.

Description

Technical Field

[0001] This invention relates to the field of material structure characterization and non-destructive testing technology, specifically to an automated detection and identification system for impurities in single-crystal high-temperature alloys based on laboratory X-ray transmission Laue diffraction images. Background Technology

[0002] Single-crystal superalloys are widely used in aerospace turbine blades, guide vanes, and hot-end components of gas turbines due to their excellent high-temperature mechanical properties and creep resistance. Their design and manufacturing typically require crystals to maintain a single orientation on a macroscopic scale to prevent grain boundaries from becoming failure sources for crack initiation and propagation under high temperatures and complex stresses. However, during actual fabrication or service, fluctuations in solidification processes, uneven thermal gradients, or compositional segregation can easily lead to the formation of heterogeneous or heterooriented grains in the matrix, thereby reducing the service performance and reliability of the components. Furthermore, in regions with complex geometries, localized plastic strain tends to accumulate during solid-state cooling and contraction. Due to the constraint of the ceramic core or cast shell on the alloy matrix, stress concentrations occur in different areas. During subsequent solution heat treatment, these localized stresses can induce the nucleation and growth of strain-free new grains, forming recrystallized grain defects and further compromising the integrity of the single-crystal structure.

[0003] Existing methods for detecting heterocrystalline defects in single-crystal superalloys mainly include macroscopic corrosion, electron backscatter diffraction (EBSD), and X-ray diffraction. Macroscopic corrosion relies on surface chemical reactions, is destructive, and makes it difficult to obtain quantitative information such as crystal orientation. EBSD requires fine polishing, limiting its detection depth and area, and is unsuitable for thick-walled or complex structural components. In the detection of internal defects in blades, these methods often require cutting the casting to expose the internal structure. For multi-channel or multi-layer configurations, multiple castings need to be repeatedly cut, resulting in high costs, long development cycles, and limiting the research and application of new single-crystal components.

[0004] While synchrotron X-ray technology can achieve non-destructive testing, it relies on large-scale research facilities, resulting in high costs and limited field applications. Although laboratory X-ray Laue diffraction methods have been studied, they are mostly reflection-based or localized measurements, typically requiring manual selection of the irradiation area and diffraction spot index. This makes it difficult to automate and systematically test large samples or complex components, and also fails to obtain comprehensive spatial distribution information of impurities. Furthermore, current Laue diffraction image analysis largely relies on manual interpretation or experience-based matching, lacking an automatic analysis and region reconstruction mechanism that incorporates sample spatial location, making it difficult to obtain impurity crystallographic parameters while ensuring detection efficiency.

[0005] Therefore, there is an urgent need for a detection method based on laboratory X-ray transmission Laue diffraction, which can efficiently acquire diffraction data of the entire region of single-crystal superalloys through beam confinement and automatic translation scanning, and combine automatic analysis and physical model calibration technology to quickly locate and analyze the heterocrystalline region, thereby meeting the needs of rapid and non-destructive testing of single-crystal superalloy samples and engineering components. Summary of the Invention

[0006] The technical objective of this invention is to provide an automatic detection system for impurities in single-crystal high-temperature alloys based on X-ray diffraction. By integrating an aperture and a beamstop into a laboratory X-ray imaging system and using an area array detector to acquire two-dimensional spatial distribution images of transmission Laue diffraction spots, combined with automatic sample translation scanning, automatic diffraction image acquisition, unsupervised classification, and forward simulation index analysis, the system can automatically identify and locate impurity regions in single-crystal high-temperature alloys, thereby overcoming the problems of high manual intervention, low detection efficiency, and poor applicability in existing technologies.

[0007] To achieve the above objectives, the technical solution adopted by this invention is an automatic detection system for impurities in single-crystal high-temperature alloys based on X-ray diffraction, comprising:

[0008] An X-ray source used to emit continuous spectrum X-rays;

[0009] An aperture is disposed in the output light path of the X-ray source to define the irradiation area of ​​the X-rays and form local illumination on the single-crystal high-temperature alloy sample.

[0010] The sample stage is used to fix single-crystal superalloy samples and enables the single-crystal superalloy samples to perform automatic translational movements in at least one plane.

[0011] A planar array detector is used to acquire Laue diffraction images generated after the light passes through a single-crystal superalloy sample.

[0012] A beam deflector is positioned between the single-crystal high-temperature alloy sample and the area array detector to block direct X-rays.

[0013] The control and data processing unit is configured to perform the following operations:

[0014] Step a: Control the sample stage to perform automatic array translation scanning according to the preset scanning step size and scanning range;

[0015] Step b: At each translation position, the area array detector is synchronously controlled to acquire the corresponding transmission Laue diffraction image, and the spatial position coordinates of the single crystal high-temperature alloy sample corresponding to the image are recorded.

[0016] Step c: Preprocess and extract features from the acquired Laue diffraction images to construct diffraction feature vectors;

[0017] Step d: Based on the diffraction feature vector, an unsupervised clustering algorithm is used to automatically classify the diffraction images corresponding to different translation positions;

[0018] Step e: Based on the classification results and the spatial coordinates of the sample, reconstruct the spatial distribution map of different crystal orientation regions in the single crystal high-temperature alloy sample, and identify the impurity crystal region.

[0019] When performing step c, the control and data processing unit is specifically configured to: perform denoising, background subtraction, and speckle enhancement processing on the Laue diffraction image, and extract the spatial distribution, quantity, and / or intensity features of the diffraction specks to construct the diffraction feature vector.

[0020] The unsupervised clustering algorithms include K-means clustering, hierarchical clustering, or DBSCAN density clustering.

[0021] An automated method for detecting and calibrating impurities in single-crystal superalloys based on transmission Laue diffraction includes the following steps:

[0022] Step S1: Set up a transmission Laue diffraction imaging system so that continuous spectrum X-rays pass sequentially through the aperture, the single crystal high-temperature alloy sample, and the beam deflector, and are finally received by the area array detector.

[0023] Step S2: Configure the geometric parameters and X-ray source parameters for Laue diffraction;

[0024] Step S3: Control the single crystal high-temperature alloy sample to perform automatic array translation scanning relative to the aperture, and simultaneously acquire transmission Laue diffraction images at each scanning position, while recording the corresponding spatial position coordinates of the single crystal high-temperature alloy sample.

[0025] Step S4: Preprocess and extract features from the acquired Laue diffraction images to construct diffraction feature vectors;

[0026] Step S5: Based on the diffraction feature vector, an unsupervised clustering algorithm is used to automatically classify the diffraction images corresponding to different positions, and the spatial distribution map of the impurity region in the single crystal high-temperature alloy sample is reconstructed according to the classification results and spatial position coordinates.

[0027] Step S6: Output the detection, identification, and spatial calibration results of impurity crystals.

[0028] Further, in step S2, the geometric parameters include source distance Lss and geophysical distance Lsd, and the X-ray source parameters include tube voltage and tube power.

[0029] Furthermore, in step S3, the automatic array translation scanning is performed point by point or line by line, and the illumination area defined by the aperture covers the entire area to be detected of the single crystal high-temperature alloy sample during the scanning process.

[0030] Furthermore, the preprocessing in step S4 includes image denoising, background subtraction, and speckle enhancement; the feature extraction includes extracting at least one of the spatial distribution features, quantity features, and intensity features of the diffraction specks.

[0031] Furthermore, the unsupervised clustering algorithm in step S5 is used to group regions with the same or similar crystal orientation characteristics into the same category, and to identify categories that are significantly different from the orientation of the single crystal substrate as heterocrystalline.

[0032] A computer-readable storage medium having a computer program stored thereon, characterized in that the program, when executed by a processor, implements steps S3 to S6 of the method.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] 1. This invention uses laboratory transmission Laue diffraction technology, which has strong penetration ability and is suitable for non-destructive testing of thick samples and complex single-crystal high-temperature alloy parts.

[0035] 2. By defining the illumination area with an aperture and combining it with automatic sample translation scanning, the acquisition of diffraction data is automated and systematic, significantly improving detection efficiency and repeatability;

[0036] 3. The analysis of diffraction images based on unsupervised classification can achieve objective identification of impurity regions without the need for manual prior annotation, thus reducing the influence of subjective human factors;

[0037] 4. The method of this invention is applicable to the quality inspection of single-crystal high-temperature alloy samples and engineering components, and has good engineering application prospects and promotional value. (See attached figures for details.)

[0038] Figure 1 This is a flowchart for the automatic detection and identification of impurities in high-temperature alloys.

[0039] Figure 2 This is a schematic diagram of a laboratory transmission white light Laue diffraction system.

[0040] In the diagram: 1. X-ray source; 2. Conical X-ray beam; 3. X-ray beam passing through the aperture; 4. Aperture; 5. Aperture aperture; 6. Sample to be tested; 7. Illuminated area of ​​the X-ray beam on the sample; 8. Beam damper; 9. Transmitted X-ray; 10. Diffracted X-ray; 11. Diffracted spot; 12. Detector; 13. First direction of sample movement (horizontal direction); 14. Second direction of sample movement (vertical direction).

[0041] Figure 3 Flowchart for configuring Laue diffraction geometry parameters.

[0042] Figure 4 The sample pairs are "matrix-heteromorphic crystals" predicted by unsupervised learning.

[0043] Figure 5 The prediction results for the heterogeneous crystal distribution on a 14×8 array are shown, with a single-map prediction accuracy of 97.50%. Detailed Implementation

[0044] The present invention will be further described below with reference to specific embodiments, but the scope of protection of the present invention is not limited thereto.

[0045] The technical solution adopted in this invention is an automatic detection system and calibration method for impurities in single-crystal high-temperature alloys based on X-ray diffraction, the method comprising the following steps:

[0046] (1) Construction of the transmission Laue diffraction imaging system. The system uses a continuous spectrum (i.e., "white light") X-ray source as the illumination source. An aperture and a beam blocker are arranged sequentially between the laboratory X-ray source and the area array detector. The aperture is used to define the irradiation area of ​​X-rays and form local illumination on the sample. The beam blocker is used to block direct X-rays to avoid detector saturation. The single crystal high-temperature alloy sample is placed between the aperture and the detector, and the X-rays pass through the sample in transmission mode.

[0047] (2) Laue diffraction geometry configuration. The spatial distribution and signal intensity of diffraction spots in a Laue diffraction image are mainly controlled by two factors: experimental geometry configuration and X-ray parameters. The main parameters include the source-object distance Lss and the object-detector distance Lsd, the size of the aperture and beam deflector, and X-ray source parameters such as tube voltage and tube power. Before the detection begins, the above parameters need to be calculated to formulate a reasonable experimental plan.

[0048] (3) Automatic array scanning and diffraction data acquisition of samples. By preset scanning step size and scanning range, the sample is controlled to translate point by point or line by line relative to the aperture, so that the white light illuminated area defined by the aperture covers the area to be detected of the sample during the translation scanning process. At each translation position, the corresponding transmission Laue diffraction image is acquired synchronously, and the spatial position coordinates of the sample corresponding to the image are recorded.

[0049] (4) Diffraction image preprocessing and feature extraction. The acquired Laue diffraction images are denoised, background subtracted and speckle enhanced. The spatial distribution features, quantity features and intensity features of the diffraction specks are extracted to construct diffraction feature vectors for subsequent analysis.

[0050] (5) Unsupervised classification and identification of impurity crystal regions. Based on the diffraction feature vectors, an unsupervised clustering algorithm is used to automatically classify the diffraction images corresponding to different translation positions, and regions with the same or similar crystal orientation features are grouped into the same category. According to the classification results and the spatial coordinates of the sample, the spatial distribution map of different crystal orientation regions in the sample is reconstructed, thereby identifying the impurity crystal distribution regions that are significantly different from the orientation of the single crystal substrate.

[0051] (6) Output of high-temperature alloy impurity detection results. Based on the classification results and spatial distribution information, the distribution of impurities in the parts is determined, realizing the automated detection, identification and calibration of impurities in single-crystal high-temperature alloys.

[0052] The aperture and beam deflector are sequentially placed between the X-ray source and the area array detector. The sample is a single-crystal high-temperature alloy specimen or a single-crystal high-temperature alloy component. The aperture defines the X-ray irradiation area, forming localized illumination on the sample to ensure spatial controllability of the diffraction signal region. The beam deflector blocks direct X-rays, preventing detector saturation and reducing background interference. The single-crystal high-temperature alloy sample is placed between the aperture and the detector, allowing X-rays to pass through the sample in transmission mode while ensuring the sample's position remains stable during the experiment; it can be fixed by the clamping device of the sample stage. To ensure X-rays can penetrate the high-temperature alloy sample and receive a sufficiently strong diffraction signal, the X-ray source energy should be higher than 100 keV and have high throughput. To ensure the beam deflector effectively blocks direct X-rays, it is made of a material with a high absorption attenuation coefficient and maintains sufficient thickness.

[0053] (2) The configuration of Laue diffraction geometric parameters includes a corresponding parameter calculation process. The calculated parameters include: source-object distance (Lss) and object-detector distance (Lsd), the size of the aperture and beam deflector, and X-ray source parameters such as tube voltage and tube power. The maximum Lsd, maximum tube voltage, and detector pixel and size parameters need to be input. The source-object distance (Lss) and object-detector distance (Lsd) together determine the position and scale of the diffraction spot on the detector. Their changes will directly change the magnification and geometric distortion of the diffraction pattern. The Lss only affects the size of the effective irradiation area and will not cause scaling of the diffraction image. As the Lsd increases, the geometric magnification increases accordingly, the contrast of the absorption image is magnified accordingly, the diffraction signal distribution becomes more discrete, the size of the diffraction spot increases, and the diffraction image is magnified. The size of the aperture and beam deflector will limit the divergence of the incident beam and the effective irradiation area of ​​the X-rays on the sample to be inspected, thereby affecting the sharpness of the diffraction spot and the background noise level. The larger the aperture size, the larger the effective illumination area; the larger the beam deflector size, the more diffraction signals are blocked at the center of the detector.

[0054] The selection of the object-detector distance (Lsd) needs to take into account the diffraction angle 2θ to ensure that the detector's field of view can cover a reasonable diffraction spot. The formula for calculating the diffraction angle is:

[0055]

[0056] Where: h is Planck's constant, h = 6.626 × 10 -34 J·s, where c represents the speed of light in a vacuum, c = 299,792,458 m / s. Let hkl be the interplanar spacing of a set of crystal planes in the crystal, and E be the X-ray energy. The maximum 2θ range that the detector can collect under different Lsd values ​​is calculated based on geometric relationships using the following formula:

[0057]

[0058] in: It is half the size of the short side of the detector. When selecting parameters, for diffraction spots with high or low energy, they will either be outside the detector range or be blocked by the beam deflector, and appropriate trade-offs should be made.

[0059] The aperture size affects the effective illumination area in a linear relationship. The aperture size directly influences the size of the diffraction spot on the single crystal; to maintain a sharp diffraction spot, a smaller aperture should be selected. The size of the beam deflector is related to both the aperture size and the Lsd size; its selection depends on the aperture selection, aiming to completely block the transmitted beam without obstructing excessive diffraction signals.

[0060] The tube voltage and tube power directly affect the energy and luminous flux of X-rays. As the tube voltage increases, more diffraction signals are excited, indicating that higher-energy X-rays can activate more crystal planes to participate in diffraction; however, these newly added diffraction spots are mostly concentrated near the central region and are more easily obscured by the beam blocker. As the tube power increases, the diffraction signal intensifies.

[0061] (3) Based on the geometric parameters and X-ray source parameters in step (2), diffraction data is acquired on the transmission Laue diffraction imaging system in step (1). The sample is mounted on a high-precision three-axis motion platform (ball screw + servo drive, positioning accuracy ±2 μm). By presetting the scanning step size and scanning range, the sample is controlled to translate point by point or line by line relative to the aperture, so that the white light illuminated area defined by the aperture covers the area to be detected of the sample during the translation scanning process. At each translation position, the corresponding transmission Laue diffraction image is acquired synchronously, and the spatial position coordinates of the sample corresponding to the image are recorded.

[0062] (4) Perform image preprocessing and feature extraction on the transmission Laue diffraction images acquired at different sample translation positions in step (3) to improve the identifiability of diffraction spots and provide quantitative feature information for subsequent automatic analysis.

[0063] Denoising is performed on the Laue diffraction image to suppress the effects of detector background noise, dark current noise, and random scattering noise, thereby improving the signal-to-noise ratio between the diffraction spots and the background. This denoising can employ spatial or frequency domain smoothing techniques to reduce image noise levels while ensuring no significant distortion of the diffraction spot geometry. Background subtraction is then performed on the denoised diffraction image to eliminate background intensity distributions caused by non-diffraction scattering, uneven detector response, or residual signals from direct radiation. Background subtraction makes the diffraction spots more prominent in the image, facilitating subsequent spot identification and quantitative analysis. After denoising and background subtraction, spot enhancement processing is applied to the diffraction image. By enhancing local intensity contrast, the intensity difference between the diffraction spots and the surrounding background is further increased, thereby improving the discernibility of the spot edges and center positions.

[0064] Based on the enhanced Laue diffraction image, features are extracted from the diffraction spots. These features include at least the spatial distribution characteristics of the diffraction spots on the detector plane, the number of spots, and the intensity of the spots. The spatial distribution features characterize the spot positions and their relative distribution relationships; the number of spots reflects the number of crystal planes satisfying the Laue diffraction conditions; and the intensity features characterize the differences in diffraction response under different crystal orientations and diffraction conditions. These features are combined to construct diffraction feature vectors corresponding to the translational positions of each sample. These vectors characterize the crystal orientation characteristics and structural differences in local regions of the sample and serve as input data for subsequent unsupervised classification and heterogeneous region identification.

[0065] (5) The diffraction feature vectors constructed in step (4) are automatically analyzed, and an unsupervised clustering algorithm is used to classify the diffraction feature vectors corresponding to different sample translation positions. By comparing the similarity between each diffraction feature vector, sample regions with the same or similar diffraction features, and thus reflecting the same or similar crystal orientation features, are automatically classified into the same category.

[0066] By combining the sample's spatial coordinates corresponding to each diffraction feature vector, the classification results are mapped back to the sample's actual spatial location, reconstructing a spatial distribution map of different crystal orientation regions within the sample. Analysis of this spatial distribution map identifies regions with orientations significantly different from those of the single-crystal substrate and classifies them as impurity crystal distribution regions, thus achieving automatic identification and location of impurity crystal regions.

[0067] (6) Based on the analysis results of step (5), the classification results of diffraction spots at each translation scanning position and the spatial position information of the sample are combined to uniformly summarize and calibrate the different crystal orientation regions within the detection area. The positional distribution of the impurity crystal region in the sample or component is obtained.

[0068] Through the above processing, the spatial distribution results of the impurity region within the entire sample are output, thereby realizing the automated detection and identification of impurity defects in single-crystal high-temperature alloys.

[0069] Example

[0070] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings. This embodiment uses a single-crystal high-temperature alloy sample as the detection object, and specifically illustrates the implementation process and technical parameters of the impurity crystal detection method.

[0071] like Figure 1 As shown, this invention provides a method for the automatic detection of impurities in single-crystal high-temperature alloys based on a laboratory X-ray imaging system and transmission Laue diffraction images. The Laue diffraction analysis described in this invention is based on two-dimensional diffraction images acquired by an area array detector, rather than on one-dimensional diffraction spectrum data where diffraction intensity varies with energy or angle.

[0072] Step (1): As Figure 2 As shown, a transmission Laue diffraction imaging system is constructed, with the aperture and beam deflector placed sequentially between the X-ray source and the area array detector. When imaging begins, X-rays are emitted from the X-ray source. After being blocked by the aperture, only a small portion of the rays pass through the aperture and irradiate the sample under test. The transmitted beam that passes through the sample is completely blocked by the beam deflector, and the diffracted beam forms diffraction spots on the detector.

[0073] Step (2): Many factors influence the diffraction pattern, including the source-object distance (Lss) and object-detector distance (Lsd), the size of the aperture and beam deflector, and X-ray source parameters such as tube voltage and tube power. Before imaging, the geometric parameters of the diffraction imaging system need to be calculated and configured to facilitate the application of different imaging systems. Initially, the maximum Lsd, maximum tube voltage, and detector pixel and size parameters are input. Figure 3 The flowchart shown will yield a suitable geometric parameter configuration scheme.

[0074] For example, for imaging system A, the known parameters are tube voltage of 225 kV, maximum distance of detector from the detector of 1200 mm (i.e., Lss+Lsd=1200 mm), detector pixel size of 290×230 mm, and pixel size of 49.5μm. The calculated detector pixel value is 5800×4600.

[0075] The selection of Lsd needs to consider the diffraction angle 2θ to ensure that the detector's field of view can cover a reasonable diffraction spot. The maximum 2θ range that the detector can collect under different Lsd values ​​can be calculated based on geometric relationships, as shown in Table 1. Lsd is set to Lsd = 100, 200, 400, 600, 900, 1200 mm.

[0076] surface The maximum 2θ (°) range that the detector can acquire under different Lsd values

[0077] 230 mm (short side) 49.00 29.90 16.04 10.85 7.28 5.47 Long side 290 mm 55.41 35.94 19.93 13.59 9.15 6.89 370 mm diagonal side 61.61 42.77 24.82 17.14 11.62 8.76

[0078] Calculate the 2θ values ​​for diffraction of Ni-based superalloys {111}, {200}, {220}, and {131} crystal planes under diffraction conditions ranging from 40 to 225 keV (40, 80, 100, 120, 150, 180, 200, 225 keV). As shown.

[0079] surface The 2θ values ​​of the characteristic crystal plane family that can diffract at different energies

[0080] <![CDATA[2θ 111 (°) ]]> 8.76 5.84 4.38 3.50 2.92 2.33 1.84 1.75 1.56 <![CDATA[2θ 200(°) ]]> 10.12 6.74 5.05 4.04 3.37 2.70 2.25 2.02 1.80 <![CDATA[2θ 220(°) ]]> 14.33 9.54 7.15 5.72 4.77 3.81 3.18 2.86 2.54 <![CDATA[2θ 311(°) ]]> 16.82 11.19 8.39 6.71 5.59 4.47 3.72 3.35 2.98

[0081] To ensure that diffraction energies above the characteristic line are collected, it is necessary to encompass the K-series characteristic radiation of the target at 58 keV (W). , The diffraction points above 58 keV can be calculated based on angle comparison. Lss should be around 600 mm. Considering the weaker diffraction signal at the detector edge, diffraction points above 58 keV should be distributed within 4 / 5 of the detector center, and the value of Lss should be as small as possible to be less than 600 mm.

[0082] For X-rays with energies of 0-225 keV, the high-flux portion is mainly concentrated in the 50-200 keV range. The calculated Lss range is 50-100 mm (as close to the X-ray source as possible without colliding with it); the Lsd range is 400-500 mm; the magnification of the sample absorption contrast image is 5-12 times, resulting in low contrast in the diffraction pattern, facilitating the identification and classification of diffraction spots. The aperture size is 350×350 μm. The choice of aperture size requires a balance between the size of the diffraction spots and the size of the illumination area. An excessively large aperture will cause overlapping diffraction spots, hindering subsequent identification and classification; an excessively small aperture will reduce the illumination area of ​​the X-rays on the sample. For the same sample area, a larger number of diffraction data points require a longer scan time. Therefore, an appropriately sized aperture is chosen. Under the condition of a 350 μm × 350 μm aperture, the empirical values ​​for the relationship between Lss selection and the illumination area are shown in Table 3. For impurity crystals with a size of 100-500 μm, an Lss parameter of 30-100 mm can be selected.

[0083] Table 3 Relationship between Lss parameter and lighting area

[0084] Lighting area (a*a mm) 0.8 1.2 2 4 8

[0085] The beam blocker measures 30×30mm (placed flush against the detector). To be on the safe side, the beam blocker size needs to be tested beforehand to ensure it is appropriate. Testing procedure: Set the exposure time to 1 second, take an image using the highest power at a low voltage (e.g., 80 kV), and observe the beam blocker and aperture outline. If it fails to completely block the aperture, shorten the distance between the beam blocker and the sample, and repeat the test until the beam blocker completely covers the aperture outline on the detector.

[0086] Step (3): Select a single-crystal high-temperature alloy sample as the test object, and mount it on a sample stage with two-dimensional translation function parallel to the detector plane, such as... Figure 2 As shown in the diagram. Imaging system B uses a laboratory white light X-ray source with an X-ray energy spectrum ranging from 20 to 120 keV. A square aperture is placed at the X-ray source exit to define the irradiation area diameter as 0.375 × 0.375 mm, and a beam deflector is placed in front of the detector to block direct light. Lss and Lsd are clearly set to 15 mm and 240 mm, respectively. By controlling the sample stage, the sample is scanned in the XY plane in 0.5 mm increments, acquiring one transmission Laue diffraction image at each scanning position.

[0087] Step (4): Perform image preprocessing and feature extraction on the transmission Laue diffraction images acquired at different sample translation positions in step (3). Specifically, perform denoising, background subtraction, and spot enhancement processing on the Laue diffraction images in sequence to improve the signal-to-noise ratio and recognizability between the diffraction spots and the background.

[0088] After image preprocessing, feature information of the diffraction spots is extracted from the enhanced Laue diffraction image. This feature information includes at least the spatial distribution characteristics of the diffraction spots on the detector plane, the number of spots, and the intensity of the spots. These feature information are combined to construct diffraction feature vectors corresponding to the translational positions of each sample. These vectors characterize the crystal orientation features of local regions of the sample and serve as input data for subsequent unsupervised classification and identification of impurity regions.

[0089] Step (5): The diffraction feature vectors constructed in step (4) are automatically analyzed. An unsupervised clustering algorithm is used to classify the diffraction feature vectors corresponding to different sample translation positions, and sample regions with the same or similar diffraction features are automatically grouped into the same category. For example Figure 4 As shown, typical matrix images are automatically selected in the frequency domain based on feature response intensity and consistency to establish a benchmark set. Then, by calculating the feature distance between the remaining images and the benchmark, and selecting candidate samples of impurities according to the "maximum difference" principle, a highly discriminative "matrix-impurity" sample pair is constructed unsupervised.

[0090] The classification results are correlated with the spatial coordinates of the sample corresponding to each diffraction feature vector to reconstruct the spatial distribution map of different crystal orientation regions in the sample. Based on the spatial distribution map, regions with orientations significantly different from those of the single-crystal substrate are identified and determined as impurity crystal distribution regions, thereby achieving automatic identification and positioning of impurity crystal regions. Figure 5 As shown, in the actual test, unsupervised classification was performed on 112 sets of data from the 14×8 acquisition array. The distribution of impurities (blue for the impurities and red for the matrix) indicates that the impurities are located at the sample boundaries, a typical surface recrystallization. For a single diffraction image, the model achieved an overall accuracy of 97.50% in the test, demonstrating its extremely high comprehensive discrimination reliability.

[0091] Step (6): Based on the analysis results of step (5), the diffraction feature classification results corresponding to the translational scanning positions of each sample and the spatial position information of the sample are combined to uniformly summarize and calibrate the different crystal orientation regions in the detection area, and obtain the spatial position distribution of the impurity crystal region in the sample or component.

Claims

1. A single crystal superalloy inhomogeneity automatic detection system based on transmission X-ray diffraction, characterized in that, include: An X-ray source used to emit continuous spectrum X-rays; An aperture is set in the output light path of the X-ray source to define the irradiation area of ​​the X-rays and form local illumination on the single-crystal high-temperature alloy sample. The sample stage is used to fix single-crystal superalloy samples and enables the single-crystal superalloy samples to perform automatic translational movements in at least one plane. A planar array detector is used to acquire Laue diffraction images generated after the light passes through a single-crystal superalloy sample. A beam deflector is positioned between the single-crystal high-temperature alloy sample and the area array detector to block direct X-rays. The control and data processing unit is configured to perform the following operations: Step a: Control the sample stage to perform automatic array translation scanning according to the preset scanning step size and scanning range; Step b: At each translation position, the area array detector is synchronously controlled to acquire the corresponding transmission Laue diffraction image, and the spatial position coordinates of the single crystal high-temperature alloy sample corresponding to the image are recorded. Step c: Preprocess and extract features from the acquired Laue diffraction images to construct diffraction feature vectors; Step d: Based on the diffraction feature vector, an unsupervised clustering algorithm is used to automatically classify the diffraction images corresponding to different translation positions; Step e: Based on the classification results and the spatial coordinates of the sample, reconstruct the spatial distribution map of different crystal orientation regions in the single crystal high-temperature alloy sample and identify the impurity crystal region; The spatial distribution morphology and signal intensity of the diffraction spots in the Laue diffraction image are synergistically regulated by two types of factors: experimental geometry and X-ray parameters. These parameters include the source-object distance Lss and the object-detector distance Lsd, the size of the aperture and beam block, as well as the tube voltage and tube power.

2. The single crystal superalloy inclusions automatic detection system based on transmission X-ray diffraction according to claim 1, characterized in that, When performing step c, the control and data processing unit is specifically configured to: perform denoising, background subtraction, and speckle enhancement processing on the Laue diffraction image, and extract the spatial distribution, quantity, and / or intensity features of the diffraction specks to construct the diffraction feature vector.

3. The single crystal superalloy inclusions automatic detection system based on transmission X-ray diffraction according to claim 1, characterized in that, The geometric parameter configuration for Laue diffraction includes a corresponding parameter calculation process. The calculated parameters include: source-object distance Lss and object-detector distance Lsd, the dimensions of the aperture and beam deflector, as well as tube voltage and tube power. The maximum Lsd, maximum tube voltage, and detector pixel and size parameters need to be input. The source-object distance Lss and object-detector distance Lsd together determine the position and scale of the diffraction spot on the detector; their changes directly alter the magnification and geometric distortion of the diffraction pattern. Lss only affects the size of the effective irradiation area; an increase in Lsd increases the geometric magnification, correspondingly amplifying the absorption image contrast, making the diffraction signal distribution more discrete, increasing the size of the diffraction spot, and amplifying the diffraction image. The dimensions of the aperture and beam deflector limit the divergence of the incident beam and the effective irradiation area of ​​the X-rays on the sample, thus affecting the sharpness of the diffraction spot and the background noise level. A larger aperture size results in a larger effective irradiation area; a larger beam deflector size results in more diffraction signals being blocked at the center of the detector.

4. An automatic detection and calibration method for impurities in single-crystal superalloys using the automatic detection system for impurities in single-crystal superalloys based on transmission X-ray diffraction as described in claim 1, characterized in that, Includes the following steps: Step S1: Set up a transmission Laue diffraction imaging system so that continuous spectrum X-rays pass sequentially through the aperture, the single crystal high-temperature alloy sample, and the beam deflector, and are finally received by the area array detector. Step S2: Configure the geometric parameters and X-ray source parameters for Laue diffraction; Step S3: Control the single crystal high-temperature alloy sample to perform automatic array translation scanning relative to the aperture, and simultaneously acquire transmission Laue diffraction images at each scanning position, while recording the corresponding spatial position coordinates of the single crystal high-temperature alloy sample. Step S4: Preprocess and extract features from the acquired Laue diffraction images to construct diffraction feature vectors; Step S5: Based on the diffraction feature vector, an unsupervised clustering algorithm is used to automatically classify the diffraction images corresponding to different positions, and the spatial distribution map of the impurity region in the single crystal high-temperature alloy sample is reconstructed according to the classification results and spatial position coordinates. Step S6: Output the detection, identification, and spatial calibration results of impurity crystals.

5. The automatic detection and calibration method for impurities in single-crystal high-temperature alloys according to claim 4, characterized in that, In step S2, the geometric parameters include source distance Lss and geophysical distance Lsd, and the X-ray source parameters include tube voltage and tube power.

6. The single crystal superalloy intercrystalline automatic detection and calibration method according to claim 4, characterized in that, In step S3, the automatic array translation scanning is performed point by point or line by line, and the illumination area defined by the aperture covers the entire area to be detected of the single crystal high-temperature alloy sample during the scanning process.

7. The single crystal superalloy intercrystalline automatic detection and calibration method according to claim 4, characterized in that, The preprocessing in step S4 includes image denoising, background subtraction, and speckle enhancement; the feature extraction includes extracting at least one of the spatial distribution features, quantity features, and intensity features of diffraction specks.

8. The single crystal superalloy intercrystalline automatic detection and calibration method according to claim 4, characterized in that, The unsupervised clustering algorithm in step S5 is used to group regions with the same or similar crystal orientation characteristics into the same category, and to identify categories that are significantly different from the orientation of the single crystal substrate as heterocrystalline.

9. The single crystal superalloy inclusions automatic detection system based on transmission X-ray diffraction according to claim 1, characterized in that, The selection of the object-detector distance Lsd needs to take into account the diffraction angle 2θ to ensure that the detector's field of view can cover the diffraction spot. The formula for calculating the diffraction angle is: ； Where h is Planck's constant and c represents the speed of light in a vacuum. Let Lsd be the interplanar spacing of a set of crystal planes, and E be the X-ray energy. The maximum 2θ range acquired by the detector under different Lsd values ​​is calculated based on geometric relationships using the following formula: ； wherein: is half the short dimension of the detector.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the program is executed by the processor, it implements steps S3 to S6 of the automatic detection and calibration method for heterocrystals in single-crystal high-temperature alloys as described in claim 4.

Images