A defect implantation method for ceramic matrix composites based on grid space mapping

Abstract

The present invention discloses a defect implantation method for ceramic-based composite materials based on grid space mapping, which belongs to the field of simulation model defect implantation. The implementation method of the present invention is as follows: material defects are realistically mapped to a simulated finite element model through CT image data, thereby increasing the accuracy of the real model of the material, being able to accurately obtain the position information of the defects inside the material and locate them, thereby improving the prediction accuracy of ceramic-based composite materials containing defects. The constructed material model is meshed by finite element calculation software, and the node and unit information in the ideal model is obtained in turn. The grid node information of the composite material matrix model located between the image slices is statistically calculated, a node data set is created, and a binary defect image is used as the background. The bwboundaries function is combined to identify the defect boundary, and the inpolygon algorithm is used to determine the defect node information within the defect area, so as to accurately determine the location of the defect. By defining the k value judgment, the original data of the defect unit is set to blank, thereby improving the efficiency of defect implantation.

Description

Technical Field

[0001] The present invention relates to a defect implantation method for a ceramic matrix composite material finite element model, in particular to a real defect space mapping method based on a material CT image, and belongs to the field of simulation model defect implantation. Background Art

[0002] Ceramic matrix composites (CMCs) currently possess excellent material properties, such as high strength, high temperature resistance, oxidation resistance, light weight, and low density, and hold great promise for high-temperature applications. However, due to the material preparation and molding processes, woven / unidirectional composites often contain initial manufacturing defects, such as holes, microcracks, low-density areas, and fiber bundle breakage. These manufacturing defects can affect various properties of the composite material, affecting the transfer path of force and heat loads within the material, significantly altering the oxygen diffusion path during oxidation, and consequently changing the distribution of thermal stress within the material and the form of structural damage.

[0003] In some simulation studies, some researchers failed to consider the impact of initial manufacturing defects on the material, and instead proposed idealized finite element simulation models. Later research led to the use of random defect implantation for simulation, but this still resulted in uncertainty and unreality. Therefore, to ensure accurate calculations, a geometric model that reflects the material's true microstructure and fully captures internal defects is required to ensure the accuracy of subsequent simulations. Summary of the Invention

[0004] The main purpose of the present invention is to provide a defect implantation method for ceramic matrix composite materials based on grid space mapping, which can realize the real mapping of original material defects into the simulation finite element model through CT image data, thereby improving the prediction accuracy of defective ceramic matrix composite materials.

[0005] The above-mentioned purpose of the present invention is achieved through the following technical solutions:

[0006] The present invention discloses a method for defect implantation in a ceramic matrix composite material based on grid space mapping, comprising the following steps:

[0007] Step 1: Scan the actual material structure using an X-ray source CT scanner. By selecting the scanning resolution and scanning area, clear CT scan data of the sample's internal structure is obtained. The CT scan data is output in the form of image slices. The scanning resolution corresponds to the voxel size.

[0008] Step 2: Use the image threshold segmentation method to divide the grayscale of the image into several parts according to the grayscale histogram of the image. Use the optimal threshold to divide the grayscale value of the image into two parts so that the variance between the two parts is maximized, ensuring maximum separation, dividing the defect area to the greatest extent, and obtaining a binary image containing only defects. Then use the bwareaopen function to delete all connected pixel components with less than P pixels in the binary image and generate a new binary image to filter out defect areas with too small an area.

[0009] Step 3: Use finite element calculation software to mesh the ideal material model, obtain the node and unit information in the ideal model in turn, and count the grid node information of the composite material matrix model located between the image slices to create a node data set. Then, use the binary defect image as the background and combine the CT image resolution of iμm to project all matrix grid nodes between (n-1)×iμm and (n)×iμm onto the nth defect image.

[0010] The bwboundaries function is used to search for defect boundaries in the processed binary image, and the size of the connected area is determined by determining the pixel connectivity: the pixel p at the coordinate (x, y) has 4 horizontal and vertical adjacent pixels. This group of pixels is called the 4-neighborhood of p, denoted by N4(p). The 4 diagonal adjacent pixels of p are denoted by N D (p). Together they form the 8-neighborhood of p, denoted by N8(p):

[0011] N4(p)+N D (p) = N8(p) (1)

[0012] And define V as the set of adjacent grayscale values. In a binary image, pixels with a value of 1 are attributed to adjacent pixels, that is, V = {1}.

[0013] At the same time, the connectivity of the pixel area is determined by calculating the Euclidean distance of the pixels. Given any pixel point p(x,y) and q(u,v), its Euclidean distance is defined as:

[0014]

[0015] Then the four adjacent regions D4 are defined as:

[0016] D4(p,q)=|xu|+|yv| (3)

[0017] Then the eight-adjacent region D8 is defined as:

[0018] D8(p,q)=max(|xu|,|yv|) (4)

[0019] The definition shows that the distance D4 from the four-adjacent area to point p forms a diamond area centered on p(x,y), and the distance D8 from the eight-adjacent area to point p forms a square area centered on p(x,y). The image processing process in the bwboundaries function uses the pixel connectivity judgment of the eight-adjacent area. This judgment method can scan out all defect connected areas and create a defect contour point coordinate data set, which is marked with a red contour line on the defect image. That is, combined with the CT image resolution of iμm, all matrix grid nodes between (n-1)×iμm and (n)×iμm are projected onto the nth defect image.

[0020] Step 4: Use the inpolygon algorithm to determine the node position information within the defect area, improve the efficiency of determining the relationship between the points and the inside and outside of the curve, and divide and determine all node information within the defect area in the shortest time.

[0021] The inpolygon algorithm uses an improved incremental angle method to calculate the winding number of a closed curve on a plane about a certain point, and then determines the internal and external relationship between the point and the curve.

[0022] Define the curve equation in the form of a parametric equation:

[0023] x=x(t),y=y(t)(0≤t≤1) (1)

[0024] Or define the curve in polar coordinates:

[0025] r=r(t),θ=θ(t)(0≤t≤1) (2)

[0026] The winding number ω is determined by the difference between θ(0) and θ(1):

[0027]

[0028] Define R = (0,0), then but:

[0029]

[0030] By discretizing the curve, the curve is discretized into a closed polygon consisting of n points P0, P1, ..., Pn, and the number of windings is calculated as follows:

[0031]

[0032] Since the inverse trigonometric function arccos and the square root sqrt in the algorithm for calculating the winding number are slow to call, the qi counting is determined for the closed polygon vertex P according to the quadrant in which it is located:

[0033]

[0034] By defining the 1 / 4 increment angle δ i =q i+1 -q i To determine whether the edge of the closed polygon passes through the quadrant boundary counterclockwise (ccw) or clockwise (cw); and according to Table 1, the corrected 1 / 4 increment angle δ i To optimize the calculation:

[0035] Table 1 Correction of 1 / 4 increment angle δ i

[0036]

[0037] For δ i ∈{2,-2}, the clockwise and counterclockwise direction determination is defined according to the sign function, that is:

[0038]

[0039] For the previous winding number calculation formula (5), the winding number is determined by calculating its 1 / 4 angle increment:

[0040]

[0041] And further improve and optimize:

[0042]

[0043] And by calculating δ i , δ i The difference gives In order to achieve the purpose of optimizing calculation,

[0044]

[0045] By calculation To obtain the winding number of any point about the closed curve:

[0046]

[0047] According to the calculation sum The final winding number w is obtained to determine whether the point is within the closed curve area. If w = 0, it is outside the selected area, and w ≠ 0, the selected point is within the area. This algorithm avoids the tedious calculation of the inverse trigonometric function arccos, shortens the calculation and judgment time, and divides and judges all node information within the defect area in the shortest time.

[0048] Step 5. Define an array to save the node numbers of the defective area and the defective unit numbers corresponding to the defective nodes in sequence. Since a unit contains multiple nodes, a k value is defined to determine: when there are greater than or equal to k defective nodes, the unit is determined to be a defective unit, and a new empty array is created, and the original data of the defective unit is cleared to achieve the purpose of deleting the defective unit grid.

[0049] Step 6: Create a new material model based on the new node unit information. The new material model contains real material preparation defects. Submit the material model to the finite element calculation software to carry out finite element simulation prediction of composite materials, thereby improving the prediction accuracy and efficiency of defective ceramic matrix composites.

[0050] The composite material finite element prediction includes braided composite material strength analysis, braided composite material strength analysis, braided composite material failure analysis, braided composite material fracture analysis, and braided composite material impact damage analysis.

[0051] The application fields of the composite materials include aerospace, such as aircraft wings and fuselages, satellite antennas and supporting structures, launch vehicle shells, engine shells, etc.; and automotive industry, such as automobile bodies, load-bearing components, transmission shafts, engine frames, and parts.

[0052] When composite materials are used in the aerospace field, the strength analysis of composite materials includes tensile strength, shear strength and compressive strength.

[0053] Preferably, the finite element calculation software is abaqus finite element calculation software.

[0054] Preferably, the image processing software is Avizo or imageJ.

[0055] Beneficial effects:

[0056] 1. The present invention discloses a defect implantation method for ceramic-based composite materials based on grid space mapping. The defects of the original material are realistically mapped to the simulated finite element model through CT image data. The defect position information inside the material can be accurately obtained and located, thereby improving the prediction accuracy of defective ceramic-based composite materials.

[0057] 2. The present invention discloses a defect implantation method for ceramic-based composite materials based on grid space mapping, which combines the bwboundaries function to identify defect boundaries and uses the inpolygon algorithm to determine the defect node information within the defect area. It divides and determines all node information within the defect area in the shortest time and accurately determines the location of the defect.

[0058] 3. The present invention discloses a defect implantation method for ceramic-based composite materials based on grid space mapping. By defining an array, the node numbers of the defective areas and the defective unit numbers corresponding to the defective nodes are saved in sequence. Since a unit contains multiple nodes, a k value is defined to determine: when there are greater than or equal to k defective nodes, the unit is determined to be a defective unit, and a new empty array is established, and the original data of the defective unit is cleared to achieve the purpose of deleting the defective unit grid, thereby further improving the efficiency of defect implantation.

[0059] 4. The present invention discloses a defect implantation method for ceramic-based composite materials based on grid space mapping, which introduces defects into the simulation model through CT scanning, improves the uncertainty of the traditional random defect implantation method, and increases the accuracy of the real model of the material. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 is the CT scan image of the original material, where: Figure 1 (a) is a CT scan image slice of a unidirectional ceramic matrix composite material. Figure 1 (b) is the three-dimensional reconstruction of the CT scanning image of the unidirectional ceramic matrix composite material. Figure 1 (c) is a CT scan image slice of the woven ceramic matrix composite material. Figure 1 (d) Three-dimensional reconstruction of the CT scan image of the woven ceramic matrix composite material.

[0061] Figure 2 is the Avizo software defect threshold segmentation map, where: Figure 2 (a) is the 3D reconstruction of the scan slice of the unidirectional ceramic matrix composite material defect threshold segmentation. Figure 2 (b) Three-dimensional reconstruction of the scan slice of the woven ceramic matrix composite material defect threshold segmentation.

[0062] Figure 3 Slice map for defective mesh nodes.

[0063] Figure 4 Result plot of the finite element model for spatial mapping of the defective mesh.

[0064] Figure 5 This is a flow chart of a defect implantation method for a ceramic matrix composite material based on grid space mapping according to the present invention.

[0065] Figure 6 It is the inpolygon algorithm (a method for determining within a polygonal node area), where: Figure 6 (a) is the continuous angle of the curve and polygon discrete symbol angles Figure 6 (b) is the quadrant classification of vertex Pi, Figure 6 (c) are all edges from the bottom to the top of the ray (the arrows indicate the direction from P i to P i+1 direction) are counted as +1, Figure 6 (d) are all edges from the top to the bottom of the ray (the arrows indicate the direction from P i to P i+1 direction) are counted as -1 DETAILED DESCRIPTION

[0066] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0067] Example 1

[0068] Mini unidirectional SiC / SiC composite defect implantation

[0069] The initial manufacturing defects of the material itself are one of the factors affecting the material strength. Many models are idealized models. This invention takes its influence into consideration and adds defects to the finite element simulation model through image information space mapping. By observing the CT image information, obvious grayscale differences can be found, and the defect space mapping is performed on the defect image obtained through the first step of void defect threshold segmentation.

[0070] like Figure 5 As shown, the present embodiment discloses a method for defect implantation in a ceramic matrix composite material based on grid space mapping, and the specific implementation steps are as follows:

[0071] Step 1: Use the threshold segmentation method to obtain images containing only defects (a total of 480 images containing defects), and filter out defective areas with too small an area. The defective area information is attached. Figure 2 (a);

[0072] Step 2: Use finite element calculation software to mesh the ideal material model, obtain node and unit information in the ideal model in turn, and count the mesh node information of the composite material matrix model located between the image slices to create a node data set;

[0073] Step 3: Import the original image of the defect and the image after threshold segmentation, and determine the distance between slices according to the resolution of the CT image. Use the acquired defect image as the background and combine the CT image resolution of iμm to project all SiC matrix grid nodes between (n-1)×1μm and (n)×1μm onto the nth defect image. In this example, the CT image resolution is 1μm. The node projection slices are shown in the attached figure. Figure 3 ;

[0074] Step 4: For the 480 selected defect images, record the defect position of each image in turn, and sort the defect coordinates in counterclockwise order. The defect contour coordinates can be enclosed into a polygon, as shown by the red boundary contour line in the figure. The x and y coordinates of the contour are converted to the actual resolution and then judged with the node coordinates. If it is within the defect contour area, the judgment is recorded as 1. If it is not within the area, the judgment is recorded as 0, and the defect nodes judged as 1 are recorded and saved.

[0075] Step 5: By defining an array, save the node number of the defective area and the defective unit number corresponding to the defective node in sequence. Since a unit contains multiple nodes, it is necessary to define a k value to determine when there are greater than or equal to k defective nodes. The unit is considered to be a defective unit, and a new empty array is created to clear the original data of the defective unit to achieve the purpose of deleting the defective unit grid.

[0076] Step 6: Create a new material model based on the new node element information. The model contains real material preparation defects. The model can be submitted to the finite element calculation software for calculation. The real model after implanting defects is shown in the attached Figure 4 .

[0077] Attachment Figure 5 The 3D reconstruction of the material's real CT image, the original finite element model, and the finite element model with defects deleted are displayed in turn. It can be seen that the model truly reflects the original morphology of the material and can better restore the subsequent tensile strength simulation.

[0078] Example 2

[0079] Defect Implantation Mapping in C / SiC Braided Composites

[0080] The C / SiC composite material consists of three parts: fiber yarn, SiC matrix and matrix void defects. The first step in reconstructing the SiC matrix is ​​to identify the matrix void defects from the original μCT image. Due to the obvious grayscale difference between the defects and the other two material phases, we use the threshold segmentation method and some image processing operations to separate the matrix void defects from the original μCT image. In this segmentation, the threshold is set to 65. Thereafter, in order to maintain the processability of the reconstructed model, the extremely small gaps in the fiber yarn are filtered and removed, and only the defects in the SiC matrix are considered. Finally, the defects in the tiny SiC matrix (area less than 200 square pixels) are filtered to obtain the final defect image, and its 3D image is shown in the attached figure. Figure 2 (b) shown.

[0081] like Figure 5 As shown, the present embodiment discloses a method for defect implantation in a ceramic matrix composite material based on grid space mapping, and the specific implementation steps are as follows:

[0082] Step 1: Use threshold segmentation, Boolean operation, filtering and other methods to obtain images containing only defects (891 XY slices along the warp axis), as shown in the attached figure. Figure 3 The defect image is shown in .

[0083] Step 2: For the nth defect image, record the positions of all its defects and sort the defect coordinates in a counterclockwise order to form a polygon, as shown by the red boundary line in the fourth small picture. Then project all SiC matrix grid nodes between (n-1)×iμm and (n)×iμm onto the nth defect image. As shown in the figure, the orange and green in the figure are projected SiC matrix grid nodes, but the nodes inside the defect are marked in green, and the rest of the nodes are marked in yellow. In addition, in this example, the resolution of the μCT original image is 5.22μm, as shown in the attached figure. Figure 3 shown.

[0084] Step 3: Count the number of green nodes in each SiC matrix mesh element; if there are more than two green nodes (three or four) in a cell, it will be deleted from the SiC matrix mesh element. The SiC matrix mesh with initial defects is shown in the figure.

[0085] The figure shows the results of removing the initial manufacturing defects at four different slicing positions. The first and last figures correspond to the starting and ending surfaces of the model. These two surfaces contain a large number of matrix mesh nodes, so they can capture the matrix mesh nodes within the defects very well. When generating the SiC matrix mesh in Abaqus, we set the mesh node seed size to 0.1mm, which is about 20 times the resolution of the CT image. Therefore, at other slicing positions, the number of matrix mesh nodes is small, and a part of the matrix mesh nodes within the defects can be captured. The smaller the mesh node seed size is set, the more mesh nodes within the defect can be captured, and the more the geometric morphology of the real defect can be reflected. However, too small a seed size will result in a huge number of SiC matrix meshes, making it difficult to carry out subsequent numerical simulations. After comprehensive consideration, the mesh node seed size is set to 0.1mm. Subsequent numerical calculations can be carried out while ensuring that the defects can be reflected. The final real model of the implanted defect is attached. Figure 4 .

[0086] The above specific description further illustrates the purpose, technical solutions and beneficial effects of the invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for defect implantation in ceramic matrix composite materials based on grid space mapping, characterized by: The following steps are included: Step 1: Scan the actual material structure itself using an X-ray source CT scanning device. By selecting the scanning resolution and scanning area, clear CT scanning data of the internal structure of the sample is obtained. The CT scanning data is output in the form of image slices; the scanning resolution corresponds to the voxel size; Step 2: Use the image threshold segmentation method to divide the grayscale of the image into several parts according to the grayscale histogram of the image. Use the optimal threshold to divide the grayscale value of the image into two parts so that the variance between the two parts is maximized, ensuring maximum separation, dividing the defect area to the greatest extent, and obtaining a binary image containing only defects. Then use the bwareaopen function to delete all connected pixel components with less than P pixels in the binary image and generate a new binary image to filter out defect areas with too small an area. Step 3: Use finite element calculation software to mesh the ideal material model, obtain node and unit information in the ideal model in turn, and count the mesh node information of the composite material matrix model located between the image slices to create a node data set. Then, using the binary defect image as the background and combining the CT image resolution of iμm, all matrix mesh nodes between (n-1)×iμm and (n)×iμm are projected onto the nth defect image; Step 4: Use the inpolygon algorithm to determine the node position information within the defect area, improve the efficiency of determining the relationship between the points and the inside and outside of the curve, and divide and determine all the node information in the defect area in the shortest time; Step 5: Define an array to sequentially save the node numbers of the defective area and the defective unit numbers corresponding to the defective nodes. Since a unit contains multiple nodes, a k value is defined to determine: when there are more than or equal to k defective nodes, the unit is determined to be a defective unit, and a new empty array is created, and the original data of the defective unit is cleared to achieve the purpose of deleting the defective unit grid; Step 6: Create a new material model based on the new node unit information. The new material model contains real material preparation defects. Submit the material model to the finite element calculation software to carry out finite element simulation prediction of composite materials, thereby improving the prediction accuracy and efficiency of defective ceramic matrix composites.

2. The method for defect implantation in ceramic matrix composite materials based on grid space mapping according to claim 1, characterized in that: In step three, The bwboundaries function is used to search for defect boundaries in the processed binary image, and the size of the connected area is determined by determining the pixel connectivity: the pixel p at coordinate (x, y) has four horizontal and vertical adjacent pixels. This group of pixels is called the 4-neighborhood of p, denoted by N4(p); the four diagonal adjacent pixels of p are denoted by N D (p); together they constitute the 8-neighborhood of p, denoted by N8(p): N4(p)+N D (p)=N8(p) (1) And define V as a set of adjacent grayscale values; in a binary image, pixels with a value of 1 are attributed to adjacent pixels, that is, V = {1}; At the same time, the connectivity of the pixel area is determined by calculating the Euclidean distance of the pixels. Given any pixel point p(x,y) and q(u,v), its Euclidean distance is defined as: Then the four adjacent regions D4 are defined as: D4(p,q)=|xu|+|yv| (3) Then the eight-adjacent region D8 is defined as: D8(p,q)=max(|xu|,|yv|) (4) The definition shows that the distance D4 from the four-adjacent area to point p forms a diamond area centered on p(x,y), and the distance D8 from the eight-adjacent area to point p forms a square area centered on p(x,y). The image processing process in the bwboundaries function uses the pixel connectivity judgment of the eight-adjacent area. This judgment method can scan out all defect connected areas and create a defect contour point coordinate data set, which is marked with a red contour line on the defect image. That is, combined with the CT image resolution of iμm, all matrix grid nodes between (n-1)×iμm and (n)×iμm are projected onto the nth defect image.

3. The method for defect implantation in a ceramic matrix composite material based on grid space mapping according to claim 2, characterized in that: In step four, The inpolygon algorithm uses an improved incremental angle method to calculate the winding number of a closed curve on a plane about a point, and then determines the internal and external relationship between the point and the curve. Define the curve equation in the form of a parametric equation: x=x(t),y=y(t)(0≤t≤1) (1) Or define the curve in polar coordinates: r=r(t),θ=θ(t)(0≤t≤1) (2) The winding number ω is determined by the difference between θ(0) and θ(1): Define R = (0,0), then but: By discretizing the curve, the curve is discretized into n points P0, P1, ..., P n The number of windings is calculated as follows: Since the inverse trigonometric function arccos and the square root sqrt in the algorithm for calculating the winding number are slow to call, the qi counting is determined for the closed polygon vertex P according to the quadrant in which it is located: By defining the 1 / 4 increment angle δ i =q i+1 -q i To determine whether the edge of the closed polygon passes through the quadrant boundary counterclockwise (ccw) or clockwise (cw); and according to Table 1, the corrected 1 / 4 increment angle δ i To optimize the calculation: Table 1 Correction of 1 / 4 increment angle δ i For δ i ∈{2,-2}, the clockwise and counterclockwise direction determination is defined according to the sign function, that is: For the previous winding number calculation formula (5), the winding number is determined by calculating its 1 / 4 angle increment: And further improve and optimize: And by calculating δ i , δ i The difference gives In order to achieve the purpose of optimizing calculation, By calculation To obtain the winding number of any point about the closed curve: According to the calculation sum To obtain the final winding number w, determine whether the point is within the closed curve area. If w = 0, it is outside the selected area, and if w ≠ 0, the selected point is within the area. This algorithm can avoid the tedious calculation of the inverse trigonometric function arccos, shorten the calculation and judgment time, and divide and judge all node information in the defect area in the shortest time.

4. The method for defect implantation in a ceramic matrix composite material based on grid space mapping according to claim 3, characterized in that: In step six, The composite material finite element prediction includes braided composite material strength analysis, braided composite material strength analysis, braided composite material failure analysis, braided composite material fracture analysis, and braided composite material impact damage analysis.

5. The method for defect implantation in ceramic matrix composite materials based on grid space mapping according to claim 4, characterized in that: The application fields of the composite materials include aerospace and automobile industries.

6. The method for defect implantation in a ceramic matrix composite material based on grid space mapping according to claim 5, characterized in that: When composite materials are used in the aerospace field, the strength analysis of composite materials includes tensile strength, shear strength and compressive strength.

7. The method for defect implantation in a ceramic matrix composite material based on grid space mapping according to claim 1, 2, 3, 4, 5 or 6, characterized in that: The finite element calculation software used is abaqus finite element calculation software.

8. The method for defect implantation in a ceramic matrix composite material based on grid space mapping according to claim 1, 2, 3, 4, 5 or 6, characterized in that: The image processing software is Avizo or imageJ.

Images