DVC calculation method and device for carbon fiber composite material
By processing the X-ray tomographic image set of carbon fiber composite materials, a reference coordinate system and mesh model were established, enabling precise monitoring of dynamic deformation and damage evolution during material loading. This solved the problems of inability to quantify dynamic deformation and poor mesh adaptability in existing technologies, and provided high-precision three-dimensional displacement and strain field data.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SICHUAN UNIV
- Filing Date
- 2026-01-23
- Publication Date
- 2026-05-12
AI Technical Summary
Existing technologies cannot accurately quantify the dynamic deformation characteristics and internal damage evolution of carbon fiber composites during loading, and the DVC method suffers from poor rigid body displacement and mesh adaptability.
By acquiring an X-ray tomographic image set of carbon fiber composite materials, noise reduction and brightness adjustment are performed, a reference coordinate system is established, and three-dimensional coordinate calibration and reconstruction are carried out. Threshold segmentation and mesh model generation are adopted, combined with DVC calculation, to achieve accurate monitoring of three-dimensional displacement and strain fields.
It enables precise, continuous, and full-field monitoring of the internal deformation of carbon fiber composite materials throughout the loading process, breaking through the limitations of traditional methods, providing reliable three-dimensional displacement and strain field data support, deeply revealing the material damage mechanism and optimizing material design.
Smart Images
Figure CN122016458A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of mechanical property testing and non-destructive testing technology for composite materials, and particularly to a method and apparatus for calculating the DVC of carbon fiber composite materials. Background Technology
[0002] Fiber-reinforced composites are widely used in aerospace, automotive, and new energy fields due to their high specific strength, high specific stiffness, and designability. Their mechanical properties are highly dependent on their microstructure characteristics. To gain a deeper understanding of the damage mechanism of carbon fiber reinforced polymer matrix composites and optimize material design, it is necessary to accurately quantify their internal three-dimensional displacement and strain fields, especially their dynamic response under complex loads. Traditional research methods often employ strain gauges, DIC, or SEM techniques, but these only provide load-displacement curves during the failure process, sample strength, surface strain, and the macro- and micro-morphological features of the fracture surface after failure. However, traditional methods cannot capture the dynamic evolution of defects during tensile testing, thus failing to reveal the material's damage evolution mechanism.
[0003] Digital Volume Correlation (DVC) is an image-based non-destructive deformation measurement technique. Its advantage lies in its ability to characterize the internal deformation of materials non-contactly, across the entire field, and in three dimensions, making it particularly suitable for multiphase, non-homogeneous materials. Despite its great potential, DVC is prone to uncontrolled rigid body displacement during experiments, causing global or local translations and rotations in the images before and after deformation, leading to insufficient measurement accuracy or unreliable results. Furthermore, if the mesh is not generated according to the actual size of the sample, the edges of the cube may not accurately match the fiber surfaces or interfaces, resulting in "step-like" artifacts in the interface regions.
[0004] Existing invention patent CN114419284B discloses a method for three-dimensional reconstruction modeling of fiber-reinforced composite materials based on CT slice images, including the following steps: acquiring microscopic slice images of fiber-reinforced composite materials using CT technology and converting them into grayscale images; binarizing the grayscale images of the composite material microscopic CT slices; identifying the fiber bundle contours in the binarized grayscale images; smoothing the fiber bundle contours in the grayscale images; and three-dimensional reconstruction of the microscopic structure of the fiber-reinforced composite material. This invention patent provides a method for three-dimensional reconstruction modeling at the microscopic scale, which greatly reduces the cost and error of manual three-dimensional reconstruction modeling. It can accurately and efficiently quantify and model changes in the morphology and size of the material's microscopic structure, and has good engineering applicability. However, this invention can only reconstruct the microscopic fiber bundle structure morphology of the composite material, and cannot reflect the dynamic deformation characteristics of the material during loading, nor can it establish a quantitative relationship with the macroscopic mechanical properties of the material, nor can it achieve quantitative analysis of the internal deformation behavior of the material at different loading stages.
[0005] Existing invention patent CN118602969B discloses a method and system for calculating three-dimensional strain of deformable objects. The method includes the following steps: For a set of three-dimensional images of an input object after strain deformation, a zero-normalized cross-correlation function is used to calculate the pixel displacement and deformation gradient of all pixels in the three-dimensional images after deformation; the alternating minimization algorithm is used to decompose the object into two sub-problems, and an augmented Lagrangian function is defined; the first sub-problem is solved, and its solution is obtained; the second sub-problem is solved, and its solution is obtained; the dual variable in the Lagrangian function is updated based on the solutions to the two sub-problems; the Lagrangian function after variable updates is calculated using alternating minimization iteration, and iteration stops when a preset iteration stopping condition is met, obtaining the three-dimensional strain random field of the object's strain deformation as the global displacement field of the last iteration. This invention provides a highly efficient three-dimensional strain calculation method for deformable objects, improving computational efficiency, enhancing noise resistance, and offering the advantage of fast convergence. However, this invention cannot obtain the true microstructural information of the material, cannot monitor the deformation evolution of the material during loading, and lacks the ability to build a mesh model based on image entities. Summary of the Invention
[0006] The purpose of this invention is to overcome the core shortcomings of existing technologies, such as only being able to obtain material surface and static information, lacking dynamic deformation and mechanical correlation in modeling, lacking real structural support and multi-stage evolution monitoring in existing DVC applications, and poor rigid body displacement and mesh adaptability in existing DVC applications, and to provide a DVC calculation method and device for carbon fiber composite materials.
[0007] In a first aspect, the present invention provides a method for calculating the DVC of carbon fiber composite materials, the method comprising: S1. Obtain a set of X-ray tomographic images of carbon fiber composite material samples; the set of X-ray tomographic images includes X-ray tomographic images of the in-situ experiment before loading, during loading, and after loading. S2. Preprocess the above X-ray tomographic image set to obtain a two-dimensional grayscale image set of the carbon fiber composite material sample; Based on the spatial parameters scanned by the above-mentioned X-ray tomography technique, the above-mentioned two-dimensional grayscale image set is stacked in an orderly manner and three-dimensional coordinates are assigned to generate the three-dimensional spatial data corresponding to the above-mentioned two-dimensional grayscale image set. Establish a reference coordinate system, and use the reference coordinate system to calibrate and reconstruct the coordinates of the three-dimensional spatial data to obtain three-dimensional spatial data with unified coordinates. Threshold segmentation is applied to the above two-dimensional grayscale image set to obtain image entities; a mesh model is then established based on these image entities. S3. Perform DVC calculations on the three-dimensional spatial data with the unified coordinates and the mesh model to obtain the three-dimensional displacement field distribution data and three-dimensional strain field distribution data inside the carbon fiber composite sample at each stage of the in-situ experiment.
[0008] Preferably, in S2 above: The above preprocessing includes noise reduction and image brightness adjustment; The above noise reduction is achieved through a bilateral filtering algorithm or through the Filter Sandbox component in the AVIZO software; The above image brightness adjustment is achieved through a combination of image background subtraction and grayscale normalization, or a method based on linear brightness adjustment. When performing noise reduction using the above bilateral filtering algorithm, the parameters are set as follows: KERNEL SIZE X=3, KERNEL SIZEY=3, SIMILARITY=20; When adjusting image brightness using a combination of background subtraction and grayscale normalization, the approximate background brightness of each X-ray tomographic image in the aforementioned X-ray tomographic image set is first calculated using the average grayscale value method or adaptive threshold method in the image edge region. The pixel grayscale value of each X-ray tomographic image is then subtracted from the corresponding approximate background brightness value to eliminate the interference of uneven background brightness. Next, grayscale normalization processing is performed on the X-ray tomographic image sequence of the same stage of the in-situ experiment, mapping the grayscale values of the aforementioned two-dimensional grayscale image sequence to the range of 0-255.
[0009] The core formula for adjusting image brightness using the above linear brightness adjustment method is:
[0010] in, The two-dimensional coordinates of pixels in the image after brightness adjustment are: The grayscale value at that time; This is the gain coefficient. Increase image contrast. Reduce image contrast; The two-dimensional coordinates of the pixels in the image before brightness adjustment are: The grayscale value at that time; This is the bias parameter. The image brightness increases. The image brightness has decreased.
[0011] Bilateral filtering, through dual constraints of spatial domain and gray-level similarity weights, efficiently removes CT scan noise while accurately preserving fine features such as carbon fiber and resin interfaces and pores, avoiding the loss of structural information. Algorithmic noise reduction using AVIZO's FilterSandbox component balances noise reduction efficiency with feature preservation, accurately distinguishing noise from the real structure and reducing blurring of carbon fiber and resin component boundaries. A method combining image background subtraction and gray-level normalization first eliminates uneven background brightness through edge gray-level averaging or adaptive thresholding, then maps the gray levels of images at the same stage to 0-255 to ensure gray-level consistency and avoid affecting subsequent thresholding for sample entity extraction. Linear brightness adjustment optimizes image contrast and brightness through gain and bias parameters, highlighting the gray-level differences of carbon fiber, resin, and pores, supporting subsequent analysis.
[0012] Preferably, the three-dimensional coordinate assignment in S2 above adopts the FDK reconstruction algorithm (Feldkamp-Davis-Kress Reconstruction Algorithm), as follows: The formula for FDK refactoring is:
[0013] in, For the reconstructed spatial location The grayscale value of the image at that location; The distance from the X-ray source to the center of sample rotation; The vertical distance from the sample rotation center to the virtual detector plane; The rotation angle of the X-ray; The coordinate system in which the virtual detector is located; This represents the weighted and corrected projection data; Indicates the filtering function; This is the original projection data; The angular frequency of the signal; It is an imaginary number; This is the cutoff angular frequency.
[0014] The weighted projection 3D image function can then be obtained using the following formula:
[0015] in, To reconstruct the three-dimensional spatial position of the sample The grayscale value of the image at that location; This is the magnification factor; This is the corrected projection data.
[0016] Employing the FDK reconstruction algorithm, it can correlate X-ray CT scan parameters, ensuring that the 3D coordinates accurately correspond to the actual physical location of the sample, thus avoiding structural distortion. Simultaneously, it can completely restore the fine structures such as carbon fibers, resin, and pores without losing minute features, providing high-precision 3D data for subsequent coordinate calibration. Furthermore, it can efficiently process large numbers of 2D slices, generating 3D data without missing layers or edge distortion, ensuring full coverage of the effective sample area and providing a reliable data foundation for subsequent threshold segmentation, mesh modeling, and DVC calculation.
[0017] Preferably, in S2 above: Establishing a reference coordinate system specifically includes: selecting an experimental reference stage; loading the 3D spatial data of the reference stage into the AVIZO software; calling the VOLUME RENDERING component in the AVIZO software; setting the grayscale mapping mode and transparency of 0.7-0.9 based on the voxel grayscale values of the 3D spatial data; using a volume rendering algorithm to analyze and render each voxel to generate a volume rendering model; using the FIT TO POINT command in the SLICE component to click on three non-collinear feature points on the volume rendering model; confirming the XY plane and XYZ axis directions based on the coordinates of these feature points to define the reference coordinate system; then activating the RESAMPLE TRANSFORMED IMAGE component in the AVIZO software, setting the MODE to CROPPED, resampling the 3D spatial data of the reference stage, reconstructing the standard 3D spatial data under the reference coordinate system, and outputting the first 3D visualization model; using the first 3D visualization model to confirm the XY plane position and XYZ axis directions of the reference coordinate system. The specific process for calibrating the coordinates of the aforementioned three-dimensional spatial data includes: loading the preprocessed three-dimensional spatial data from all stages of the in-situ experiment into the AVIZO software; calculating the affine transformation matrix using the ITERATIVE OPTIMIZATION ALGORITHM algorithm; and aligning the aforementioned three-dimensional spatial data from the remaining stages of the in-situ experiment with the standard three-dimensional spatial data in the aforementioned reference coordinate system.
[0018] The affine transformation is a linear transformation model, which can be represented by a transformation matrix M:
[0019] in, These are the aligned 3D coordinates after the affine transformation. ~ These are the parameters for linear transformation; ~ These are translation parameters; Let be the three-dimensional coordinates before alignment after the affine transformation.
[0020] The main iterative process is as follows: (1) Transformation: Perform affine transformation and interpolation operations on all floating images to obtain the transformed floating images; (2) Measurement: Calculate the similarity measurement value between the transformed floating images and the reference images; (3) Solve for a set of affine transformation parameter increments based on the similarity measurement value; (4) Update: Update the current affine transformation parameters using the parameter increments. Then, resample all model data except for the first stage, and reconstruct new data with the spatial coordinates of the first stage as a reference; The specific process of reconstructing the coordinates of the aforementioned three-dimensional spatial data includes: using the standard three-dimensional spatial data in the aforementioned reference coordinate system as a reference, resampling the aforementioned three-dimensional spatial data from the remaining in-situ experimental stages after coordinate alignment.
[0021] The design utilizes a reference coordinate system established by combining volume rendering visualization assistance with feature point localization. AVIZO software is selected to load 3D spatial data, with an opacity of 0.7-0.9 to clearly present the internal structure. The FIT TOPOINT command of the SLICE component is used to precisely define the XY plane and XYZ axis directions through three non-collinear feature points, solving the problems of traditional coordinate definitions relying on experience and being prone to misalignment with the loading direction of in-situ experiments. The processing sequence of calibration and alignment followed by resampling and reconstruction is optimized. First, the REGISTER IMAGE component uses the pre-loading reference data as a reference, and the ITERATIVE OPTIMIZATIONAL GORITHM algorithm is used to calculate the affine transformation matrix, achieving rigid body displacement correction for multi-stage data. Then, resampling ensures that the reconstructed data retains only the effective sample area, avoiding redundant background interference. The entire process, through complementary component functions, progressive processing sequence, and targeted parameter settings, achieves coordinate unification for multi-stage data, providing a standard-scale deformation comparison basis for DVC calculation and significantly improving the accuracy of displacement and strain field measurements.
[0022] Preferably, the above-mentioned experimental baseline stage is the stage before the above-mentioned experiment is loaded.
[0023] Before loading, the carbon fiber composite sample is in an initial undeformed state with a complete microstructure and no damage accumulation. Using this as a benchmark can minimize the interference of initial defects on deformation measurement. The three-dimensional spatial data at this stage can completely preserve the original microstructure of the sample, providing an undeformed state reference for deformation comparison during and after loading. This facilitates accurate quantification of the entire process of displacement and strain evolution from the initial state to stress deformation and then to damage failure, and clearly reveals the dynamic laws of defect initiation and development.
[0024] Preferably, in S2 above: threshold segmentation of the above two-dimensional grayscale image set specifically includes: loading the above three-dimensional spatial data with unified coordinates in AVIZO software, performing threshold segmentation on the above three-dimensional spatial data with unified coordinates, setting a threshold range based on the grayscale value differences of each component of the above carbon fiber composite material sample, segmenting the binary data of the above carbon fiber composite material sample based on the threshold range, and obtaining image entities.
[0025] By setting a threshold range based on the grayscale differences of the components in the composite material, precise segmentation can be achieved through interactive components. This effectively distinguishes target image entities from the background and pores, avoiding entity omissions or misjudgments of the background and pores caused by traditional threshold segmentation. The segmentation is performed using 3D spatial data with unified coordinates, ensuring consistency between the segmentation process and the data source for subsequent mesh modeling and DVC calculation, reducing errors caused by data conversion, and providing a reliable foundation for establishing a mesh model that fits the real structure.
[0026] Preferably, the mesh model in S2 above is a tetrahedral mesh model; establishing the mesh model based on the above image entities specifically includes: Based on the binarized data of the aforementioned image entities, the GENERALIZED MARCHINGCUBES ALGORITHM algorithm, which is based on non-binary classification, is used with CONSTRAINED SMOOTHING=2 to generate triangular approximations of the sample interface. Laplacian smoothing is then used for mesh smoothing. The advancing front method is employed to fill the area defined by the surface data with tetrahedrons, generating a tetrahedral mesh model for DVC calculation.
[0027] The specific process of using the GENERALIZED MARCHING CUBES ALGORITHM algorithm based on non-binary classification is as follows: (1) Construct an octree: Adaptively subdivide the three-dimensional space, use large voxels for areas with gentle data changes, and subdivide areas with drastic changes (such as near the interface).
[0028] (2) Calculate Hermite data: On the edge of each voxel, not only is the position of the isopleth point calculated, but also the normal (or gradient) of that point is calculated. This utilizes the gradient information of grayscale data, which is the key to processing "non-binary classification" data.
[0029] (3) Generating vertices within voxels: For each voxel, a quadratic error function is fitted using all Hermite data (position and normal) on its edges. The point that minimizes this error function is found and used as the vertex of the mesh to be generated for that voxel. This step ensures that the generated vertices are optimal.
[0030] (4) Connecting into a grid: Based on the adjacency relationship of the octree, connect these vertices into triangular patches to form the final grid.
[0031] Mesh smoothing aims to improve the shape of triangles and reduce the jagged appearance of surfaces ("staircase" effect). The most commonly used algorithm is Laplace smoothing, whose basic idea is to move each vertex toward the average position of its neighboring vertices.
[0032] The Laplace smoothing described above is specifically as follows:
[0033] in, These are the 3D coordinates of the vertices of the mesh after Laplacian smoothing; The original 3D coordinates of the vertices of the mesh before smoothing; As a relaxation factor, ; This represents the number of neighboring vertices of the current vertex. It is the sum of the three-dimensional coordinates of all neighboring vertices of the current vertex.
[0034] The advancing front method is used to find a node location that can form a high-quality tetrahedron. The relevant formula is:
[0035] in, The center point of the base plane; Let the coordinates of the three vertices of the face be given.
[0036]
[0037] in, For ideal nodes; The height of an ideal tetrahedron; is the unit normal vector of the base plane.
[0038] The tetrahedral mesh model can flexibly adapt to the complex microstructure of carbon fiber composites, solving the problems of traditional cubic meshes failing to fit curved surfaces and easily producing "step-like artifacts". The GENERALIZED MARCHING CUBESALGORITHM algorithm combined with the advancing front method is used to generate the mesh, and the interface approximation accuracy is optimized by combining Laplacian smoothing. This ensures that the mesh elements are accurately matched with the actual contour and interface of the sample, and that there are no empty areas, no out-of-bounds, and no distortion. It provides a high-quality discretization calculation carrier for DVC calculation, improving the calculation accuracy of displacement and strain fields.
[0039] Preferably, in S3 above, the DVC calculation of the above-mentioned coordinate-unified three-dimensional spatial data and the above-mentioned mesh model specifically includes: loading the above-mentioned coordinate-unified three-dimensional spatial data and the above-mentioned mesh model of each stage of the in-situ experiment into the AVIZO software, activating the INCREMENTAL GLOBAL DVC calculation function, setting relevant parameters such as window size and convergence threshold, and starting the calculation to obtain the three-dimensional displacement field distribution data and three-dimensional strain field distribution data inside the above-mentioned carbon fiber composite material sample at each stage of the in-situ experiment.
[0040] When using INCREMENTAL GLOBAL DVC for computation, the algorithm must first satisfy the following formula:
[0041] in, For time step Spatial location is The grayscale value at that location; For time steps Time and location are grayscale values, where It is the incremental displacement vector.
[0042] The total displacement field under large deformation is obtained by calculating only the incremental displacement vectors of adjacent states and then summing them. The strain field at each stage is calculated, and finally the internal three-dimensional displacement field and strain field of the carbon fiber composite sample at each stage are obtained.
[0043] To address the need for multi-stage dynamic deformation monitoring, the INCREMENTAL GLOBAL DVC component in AVIZO software was selected. This component supports continuous tracking of deformation at each stage, avoiding the cumulative errors caused by segmented calculations. By first loading three-dimensional spatial data with unified coordinates and a high-quality mesh model, and then flexibly setting relevant parameters such as window size and convergence threshold according to the sample's microscale and deformation characteristics, the calculation parameters are ensured to be accurately matched with experimental requirements. Through the above design, efficient and stable full-field displacement and strain field calculations can be achieved, and the reliability of the calculation results is improved by adapting the parameters to the experimental scenario.
[0044] Preferably, the above method further includes: The above methods also include: The Convert Scalar DVCOutput to Volume tool was used to associate the three-dimensional displacement field distribution data and the three-dimensional strain field distribution data with the three-dimensional coordinates of the three-dimensional spatial data with unified coordinates in S2 above. The grid model was used as a data mapping carrier by calling the grid view component of AVIZO software. It was confirmed that the grid model and the outline of the carbon fiber composite sample were not suspended or recessed, and that the grid model had no empty areas, no boundary crossings, and no element distortion. Then, adjust the color field tool to set the range to the correct interval of the above three-dimensional displacement field distribution data and three-dimensional strain field distribution data. Adjust the color map to convert the above three-dimensional strain field distribution data into a color gradient map and overlay it onto the volume rendering model. Output the second three-dimensional visualization model. Confirm the following using the above second three-dimensional visualization model: the displacement direction of the above carbon fiber composite sample is consistent with the loading direction of the in-situ experiment; the strain region variation law of the above carbon fiber composite sample corresponds to the defects of the above carbon fiber composite sample; and the strain distribution inside the above carbon fiber composite sample conforms to the experimental law.
[0045] In a second aspect, the present invention provides a DVC calculation device for carbon fiber composite materials, the device comprising at least one processor and a memory communicatively connected to the at least one processor; the memory storing instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the aforementioned DVC calculation method for carbon fiber composite materials.
[0046] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention provides a method for calculating the digital visual capacity (DVC) of carbon fiber composite materials. By performing noise reduction and brightness adjustment on X-ray tomographic images of carbon fiber composite samples before, during, and after experimental loading, image quality is optimized, eliminating scanning noise and brightness inconsistencies, and accurately preserving microstructural features. Through ordered stacking of CT scan spatial parameters and assignment of three-dimensional coordinates, two-dimensional images are transformed into three-dimensional spatial data that conforms to the actual structure of the sample, completing the three-dimensional digital characterization of the structure. By establishing a reference coordinate system and performing coordinate calibration and reconstruction on the three-dimensional spatial data, coordinate unification of data from each in-situ experimental stage is achieved, effectively mitigating the impact of noise and brightness inconsistencies. This method eliminates measurement interference caused by uncontrolled rigid body displacement; it acquires image entities through threshold segmentation and builds a mesh model based on these entities, enabling accurate extraction of the effective structure of the sample and construction of a discretized computational carrier adapted to the real structure; it performs DVC calculations on coordinate-unified three-dimensional spatial data and mesh models, achieving quantitative characterization of three-dimensional displacement and strain at any position inside the sample, breaking through the limitations of traditional methods that can only acquire surface or static information; thus, it enables accurate, continuous, and full-field monitoring of the internal deformation evolution of carbon fiber composite materials throughout the loading process, providing reliable three-dimensional displacement and strain field data support for in-depth revelation of material damage mechanisms and optimization of material design. Attached Figure Description
[0047] Figure 1 This is a schematic diagram of the DVC calculation method for a carbon fiber composite material in Example 1.
[0048] Figure 2 This is a schematic diagram of the X-ray tomographic image set of the carbon fiber composite material in Example 1.
[0049] Figure 3 This is a schematic diagram of the three-dimensional reconstruction of the carbon fiber composite material sample in Example 1.
[0050] Figure 4 This is a schematic diagram of image reconstruction operations at different experimental stages of the carbon fiber composite material sample in Example 1.
[0051] Figure 5 This is a schematic diagram of the tetrahedral mesh model of the carbon fiber composite material sample in Example 1.
[0052] Figure 6 This is a schematic diagram of the strain distribution along the Z-axis of the carbon fiber composite sample in Example 1. Detailed Implementation
[0053] The present invention will now be described in further detail with reference to specific embodiments. However, this should not be construed as limiting the scope of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.
[0054] Unless otherwise specified, the terms "upper," "lower," "left," "right," "center," "inner," and "outer," etc., used in the description of specific embodiments of the present invention to indicate orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings, or the orientation or positional relationship in which the product / equipment / device is usually placed during use. These terms are merely for the purpose of facilitating the description of the present invention or simplifying the description in specific embodiments, and for enabling those skilled in the art to quickly understand the solution, and do not indicate or imply that a particular device / component / element must have a specific orientation, or be constructed and operated in a specific positional relationship. Therefore, they should not be construed as limitations on the present invention.
[0055] Furthermore, the use of terms such as "horizontal," "vertical," "suspended," "parallel," and "coaxial" does not imply that the corresponding device / component / element must be absolutely horizontal, vertical, suspended, parallel, or coaxial. Slight tilt or deviation is permissible, as long as it does not affect the normal function of the relevant component. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," not that the structure must be perfectly horizontal; a slight tilt is acceptable. "Coaxial" means that two components are arranged as coaxially as possible, allowing them to move coaxially or approximately coaxially when their relative positions change. Alternatively, it can be simplified to mean that the corresponding device / component / element, when arranged in "horizontal," "vertical," "suspended," "parallel," or "coaxial" directions, can have an error / deviation of ±10% relative to the corresponding direction, more preferably within ±8%, more preferably within ±6%, more preferably within ±5%, and more preferably within ±4%. For example, the deviation in the "coaxial" direction is controlled within 0.2-1mm, preferably within 0.2-0.5mm. As long as the corresponding device / component / element is within the error / deviation range, it can still achieve its function in the solution of the present invention.
[0056] Furthermore, the use of terms such as "first," "second," and "third" in terminology is merely for distinguishing descriptions of identical or similar components and should not be interpreted as emphasizing or implying the relative importance of a particular component.
[0057] Furthermore, in the description of the embodiments of the present invention, "several", "more than", and "a number of" represent at least two. The number can be any number, such as two, three, four, five, six, seven, eight, or nine, and can even exceed nine.
[0058] Furthermore, in the description of the technical solution of this invention, unless otherwise explicitly specified / limited / restricted, the terms "set up," "install," "connect," "link," "provided with," "laid out," and "arranged" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to connection methods commonly used in the art, such as welding, riveting, bolting, and threaded connections. Such connections can be mechanical, electrical, or communication connections; they can be direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components.
[0059] Example 1 To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions provided by this invention will be further described in detail below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments.
[0060] This invention provides a method for calculating the DVC of carbon fiber composite materials, the flowchart of which is shown below. Figure 1 As shown, it includes the following steps: S1. Obtain a set of X-ray tomographic images of carbon fiber composite material samples; the set of X-ray tomographic images includes X-ray tomographic images of the in-situ experiment before loading, during loading, and after loading.
[0061] In this embodiment, a carbon fiber composite material sample with a length of approximately 40 mm, a width of approximately 10 mm, and a thickness of approximately 2.7 mm was selected, and an in-situ tensile CT scan experiment was conducted using X-ray computed tomography (CT) technology.
[0062] The obtained data area is approximately 3 mm long and 2.7 mm wide at the center of the carbon fiber composite sample. Each stage of data consists of 2500 grayscale images in 8-bit format, with grayscale values ranging from 0 to 255. Different grayscale values represent different microstructures (including pores, carbon fibers, and resin). Specific images are shown below. Figure 2 As shown.
[0063] The ability of X-rays from CT scanners to penetrate carbon fiber composite samples is inversely proportional to the density of the coal and rock mass sample itself. That is, the gray value is related to the density of the tissue components. This is manifested in the grayscale image as low-density areas appearing darker, such as pores or cracks, with gray values close to 0; and high-density areas appearing brighter, such as carbon fibers, with gray values close to 255. The higher the mineral density, the brighter the area.
[0064] S2. The above X-ray tomographic image set is denoised and the image brightness is adjusted to obtain a two-dimensional grayscale image set of the carbon fiber composite material sample. Based on the spatial parameters scanned by the above-mentioned X-ray computed tomography technique, the above-mentioned two-dimensional grayscale image set is stacked in an orderly manner, and the FDK reconstruction algorithm is used to assign three-dimensional coordinates to generate the three-dimensional spatial data corresponding to the above-mentioned two-dimensional grayscale image set. Establish a reference coordinate system, and use the reference coordinate system to calibrate and reconstruct the coordinates of the three-dimensional spatial data to obtain three-dimensional spatial data with unified coordinates. Threshold segmentation is applied to the above two-dimensional grayscale image set to obtain image entities; a mesh model is then established based on these image entities. A schematic diagram of 3D reconstruction using AVIZO software is shown below. Figure 3 As shown.
[0065] In the denoising process of the aforementioned two-dimensional grayscale image set, the BILATERAL filtering algorithm was employed to eliminate some of the noise generated during the shooting process. To eliminate the influence of uneven brightness, the volumetric grayscale image was normalized by calculating an approximate value of the background brightness and subtracting it from the image, thereby making the image brightness more uniform.
[0066] Load the three-dimensional spatial data from the pre-load stage of the experiment into the AVIZO software. Using the VOLUME RENDERING component in AVIZO, set the grayscale mapping mode and transparency of 0.7-0.9 based on the voxel grayscale values of the three-dimensional spatial data. Employ a volume rendering algorithm to analyze and render each voxel individually, generating a volume rendering model. Use the FIT TO POINT command in the SLICE component to click on three non-collinear feature points on the volume rendering model. Based on the coordinates of these feature points, confirm the XY plane and XYZ axis directions to define the reference coordinate system. Then, activate the RESAMPLE TRANSFORMEDIMAGE component in AVIZO, set MODE to CROPPED, resample the three-dimensional spatial data from the reference stage, reconstruct the standard three-dimensional spatial data under the reference coordinate system, and output the first three-dimensional visualization model. Use the first three-dimensional visualization model to confirm the XY plane position and XYZ axis directions of the reference coordinate system. A detailed operation diagram is shown below. Figure 4 As shown; Load the preprocessed three-dimensional spatial data of all stages of the in-situ experiment into the AVIZO software, calculate the affine transformation matrix using the ITERATIVEOPTIMIZATION ALGORITHM algorithm, and align the above three-dimensional spatial data of the remaining in-situ experimental stages with the standard three-dimensional spatial data under the above reference coordinate system. Using the standard three-dimensional spatial data in the aforementioned reference coordinate system as a reference, the aforementioned three-dimensional spatial data of the remaining in-situ experimental stages after coordinate alignment are resampled.
[0067] The threshold segmentation of the above two-dimensional grayscale image set specifically includes: loading the above three-dimensional spatial data with unified coordinates in AVIZO software, performing threshold segmentation on the above three-dimensional spatial data with unified coordinates, setting a threshold range based on the grayscale value differences of each component of the above carbon fiber composite material sample, and segmenting the binary data of the above carbon fiber composite material sample based on the above threshold range to obtain image entities.
[0068] Based on the binarized data of the aforementioned image entities, the GENERALIZED MARCHINGCUBES ALGORITHM algorithm (based on non-binary classification) was used with CONSTRAINED SMOOTHING=2 to generate triangular approximations of the sample interface. Laplacian smoothing was then applied for mesh smoothing. The advancing front method was used to fill the area defined by the surface data with tetrahedrons, generating a tetrahedral mesh model for DVC calculation. A schematic diagram of this mesh is shown below. Figure 5 As shown.
[0069] S3. Perform DVC calculations on the three-dimensional spatial data with the unified coordinates and the mesh model to obtain the three-dimensional displacement field distribution data and three-dimensional strain field distribution data inside the carbon fiber composite sample at each stage of the in-situ experiment.
[0070] Specifically, the unified three-dimensional spatial data and mesh model of each stage of the in-situ experiment were loaded into the AVIZO software. The INCREMENTAL GLOBAL DVC calculation function was activated, and relevant parameters such as window size and convergence threshold were set. The calculation was then started to obtain the three-dimensional displacement field distribution data and three-dimensional strain field distribution data inside the carbon fiber composite sample at each stage of the in-situ experiment. A schematic diagram of the strain distribution along the Z-axis is shown below. Figure 6 As shown.
[0071] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for calculating the DVC of carbon fiber composite materials, characterized in that, The method includes the following steps: S1. Obtain a set of X-ray tomographic images of carbon fiber composite material samples; the set of X-ray tomographic images includes X-ray tomographic images of the in-situ experiment before loading, during loading, and after loading. S2. Preprocess the X-ray tomographic image set to obtain a two-dimensional grayscale image set of the carbon fiber composite material sample; Based on the spatial parameters scanned by the X-ray tomography technique, the two-dimensional grayscale image set is stacked in an orderly manner and three-dimensional coordinates are assigned to generate the three-dimensional spatial data corresponding to the two-dimensional grayscale image set. A reference coordinate system is established, and the coordinates of the three-dimensional spatial data are calibrated and reconstructed using the reference coordinate system to obtain three-dimensional spatial data with unified coordinates. Threshold segmentation is applied to the two-dimensional grayscale image set to obtain image entities; a mesh model is then established based on the image entities. S3. Perform DVC calculation on the three-dimensional spatial data with unified coordinates and the mesh model to obtain the three-dimensional displacement field distribution data and three-dimensional strain field distribution data inside the carbon fiber composite sample at each stage of the in-situ experiment.
2. The DVC calculation method for carbon fiber composite materials according to claim 1, characterized in that, In S2: The preprocessing includes noise reduction and image brightness adjustment; The noise reduction is achieved through a bilateral filtering algorithm or through the Filter Sandbox component in AVIZO software; The image brightness adjustment is achieved through a combination of image background subtraction and grayscale normalization, or a method based on linear brightness adjustment. When performing noise reduction using the bilateral filtering algorithm, the parameters are set as follows: KERNEL SIZE X=3, KERNEL SIZE Y=3, SIMILARITY=20; When adjusting image brightness using a combination of image background subtraction and grayscale normalization, the approximate background brightness of each X-ray tomographic image in the X-ray tomographic image set is first calculated using the average grayscale value method of the image edge region or the adaptive threshold method. Then, the pixel grayscale value of each X-ray tomographic image is subtracted from the corresponding approximate background brightness value to eliminate the interference of uneven background brightness. Then, the X-ray tomographic image sequences of the same stage of the in-situ experiment were subjected to grayscale normalization processing to map the grayscale values of the two-dimensional grayscale image sequences to the range of 0-255.
3. The DVC calculation method for carbon fiber composite materials according to claim 1, characterized in that, The 3D coordinate assignment in S2 uses the FDK reconstruction algorithm.
4. The DVC calculation method for carbon fiber composite materials according to claim 1, characterized in that, In S2: Establishing a reference coordinate system specifically includes: selecting an experimental reference stage; loading the 3D spatial data of the reference stage into the AVIZO software; calling the VOLUME RENDERING component in the AVIZO software; setting the grayscale mapping mode and transparency of 0.7-0.9 based on the voxel grayscale values of the aforementioned 3D spatial data; using a volume rendering algorithm to analyze and render each voxel one by one to generate a volume rendering model; using the FIT TO POINT command in the SLICE component to click on three non-collinear feature points on the volume rendering model; confirming the XY plane and XYZ axis directions based on the coordinates of the feature points to define the reference coordinate system; then activating the RESAMPLE TRANSFORMED IMAGE component in the AVIZO software, setting MODE to CROPPED, resampling the 3D spatial data of the reference stage, reconstructing standard 3D spatial data under the reference coordinate system, and outputting a first 3D visualization model; using the first 3D visualization model to confirm the XY plane position and XYZ axis directions of the reference coordinate system. The calibration of the three-dimensional spatial data coordinates specifically includes: loading the preprocessed three-dimensional spatial data of all stages of the in-situ experiment into the AVIZO software, calculating the affine transformation matrix using the ITERATIVE OPTIMIZATION ALGORITHM algorithm, and aligning the three-dimensional spatial data of the remaining in-situ experimental stages with the standard three-dimensional spatial data under the reference coordinate system. The reconstruction of the three-dimensional spatial data coordinates specifically includes: using the standard three-dimensional spatial data in the reference coordinate system as a reference, resampling the remaining three-dimensional spatial data from the in-situ experimental stage after coordinate alignment.
5. The DVC calculation method for carbon fiber composite materials according to claim 4, characterized in that, The experimental baseline stage is the stage before the experimental load is applied.
6. The method for calculating the DVC of carbon fiber composite materials according to claim 1, characterized in that, In step S2: threshold segmentation of the two-dimensional grayscale image set specifically includes: loading the coordinate-unified three-dimensional spatial data in AVIZO software, performing threshold segmentation on the coordinate-unified three-dimensional spatial data, setting a threshold range based on the grayscale value differences of each component of the carbon fiber composite material sample, and segmenting the binary data of the carbon fiber composite material sample based on the threshold range to obtain image entities.
7. The DVC calculation method for carbon fiber composite materials according to claim 1, characterized in that, The mesh model in S2 is a tetrahedral mesh model; establishing the mesh model based on the image entity specifically includes: Based on the binarized data of the image entities, the GENERALIZED MARCHINGCUBES ALGORITHM algorithm based on non-binary classification is used, with CONSTRAINED SMOOTHING=2, to generate triangular approximations of the sample interface. Laplacian smoothing is then used for mesh smoothing. The advancing front method is employed to fill the area defined by the surface data with tetrahedrons, generating a tetrahedral mesh model for DVC calculation.
8. The DVC calculation method for carbon fiber composite materials according to claim 1, characterized in that, In step S3, the DVC calculation of the coordinate-unified three-dimensional spatial data and the mesh model specifically includes: loading the coordinate-unified three-dimensional spatial data and the mesh model of each stage of the in-situ experiment into the AVIZO software, activating the INCREMENTALGLOBAL DVC calculation function, setting relevant parameters such as window size and convergence threshold, and starting the calculation to obtain the three-dimensional displacement field distribution data and three-dimensional strain field distribution data inside the carbon fiber composite sample of each stage of the in-situ experiment.
9. The method for calculating the DVC of carbon fiber composite materials according to claim 1, characterized in that, The method further includes: The Convert Scalar DVC Output to Volume tool was used to associate the three-dimensional displacement field distribution data and the three-dimensional strain field distribution data with the three-dimensional coordinates of the coordinate-unified three-dimensional spatial data in S2; the grid model was used as a data mapping carrier by calling the grid view component of AVIZO software; it was confirmed that the grid model and the outline of the carbon fiber composite sample were not suspended or recessed, and that the grid model had no empty areas, no boundary crossings, and no element distortion; Then, adjust the color field tool to set the range to the correct interval of the three-dimensional displacement field distribution data and the three-dimensional strain field distribution data. Adjust the color map to convert the three-dimensional strain field distribution data into a color gradient map and overlay it onto the volume rendering model. Output the second three-dimensional visualization model. Use the second three-dimensional visualization model to confirm the following: the displacement direction of the carbon fiber composite sample is consistent with the loading direction of the in-situ experiment; the strain region variation law of the carbon fiber composite sample corresponds to the defect of the carbon fiber composite sample; and the strain distribution inside the carbon fiber composite sample conforms to the experimental law.
10. A DVC computing device based on carbon fiber composite material, characterized in that, It includes at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to perform a DVC calculation method for carbon fiber composite materials according to any one of claims 1 to 9.