Controllable parameterized model generation method of three-dimensional braided composite material

Abstract

The invention relates to a method for generating a controllable parameterized model of a three-dimensional braided composite material, which comprises the following steps of: generating a three-dimensional braided preform track model by utilizing kinematics analysis of a four-step three-dimensional braiding process, and implementing geometric reconstruction of a yarn section on the basis of track parameterized modeling; secondly, generating multi-scale pore elements through ellipsoid geometric parameter randomization, and randomly putting after Euler angle random rotation transformation to enhance spatial heterogeneity characteristics of pore morphology; then, physical accurate simulation of densification behaviors of the yarns and the pores in the compaction process is achieved through a model dynamic compaction algorithm; and finally, utilizing a three-dimensional crack random generation calculation method to accurately simulate the crack characteristics in the yarns of the three-dimensional braided composite material in the compaction process so as to obtain a three-dimensional braided parameterized random model. According to the method, the topological structure, the initial pore distribution and the initial damage characteristics of the yarns in the three-dimensional woven composite material sample are accurately reproduced.

Description

Technical Field

[0001] The invention relates to a model generation method, in particular to a controllable parameterized model generation method for a three-dimensional braided composite material, and belongs to the technical field of image processing. Background Art

[0002] Three-dimensional braided composite materials are advanced structural materials with excellent properties such as high strength, high modulus, and light weight. They are widely used in aerospace, automobiles, ships, sports equipment, and other fields. Accurate characterization of three-dimensional braided composite materials is the key to understanding their performance, optimizing their design, improving their reliability, and promoting their application. Through multi-scale experimental analysis and modeling, we can fully understand the relationship between the structure and performance of materials, providing important support for materials science and engineering applications.

[0003] In current research, the construction of parametric random geometric models of three-dimensional woven composites still faces many challenges in terms of accurate description of real structures, multi-scale modeling, uncertainty quantification, and multi-physics field coupling. In particular, complex structures such as yarn fluctuations, defects, and cracks in real materials are difficult to simulate accurately. At the same time, how to efficiently achieve multi-scale modeling from micro to macro and quantify the impact of uncertainty in geometric and material parameters on mechanical properties is also a difficulty in current research. In addition, the establishment of a multi-physics field coupling model to fully reflect the behavior of materials under actual working conditions still needs further exploration.

[0004] Therefore, in order to solve the above problems, it is necessary to provide an innovative controllable parameterized model generation method for three-dimensional woven composite materials to overcome the above defects in the prior art. Summary of the invention

[0005] The purpose of the present invention is to provide a controllable parameterized model generation method for a three-dimensional braided composite material, which can accurately reproduce the internal yarn topology, initial pore distribution and initial damage characteristics in the three-dimensional braided composite material.

[0006] To achieve the above object, the technical solution adopted by the present invention is: a method for generating a controllable parameterized model of a three-dimensional braided composite material, which comprises the following steps: 1) Based on the kinematic analysis of the four-step three-dimensional weaving process, a three-dimensional weaving preform geometric model is constructed, and on the basis of trajectory parameterized modeling, the yarn cross-section geometry reconstruction is implemented, and the optimized cross-section is spatially mapped along the trajectory path to ensure that the cross-section center strictly coincides with the trajectory node and there is no geometric interference between the yarns, and finally a loose yarn random model is obtained; 2) Based on the loose yarn random model, multi-scale pore primitives are generated by randomizing the ellipsoid geometric parameters, and then the Euler angle random rotation transformation and random placement are used to enhance the spatial heterogeneity of the pore morphology. The optimized pore model is embedded in the topological structure of the loose yarn random model to obtain a loose yarn random model with pores; 3) A physical engine based on the Taichi framework is used to construct a yarn dynamic compaction algorithm. After the random model of loose yarn with pores is input into the algorithm, the physical and precise simulation of the densification behavior of yarn and pores during the compaction process is achieved, and a random model of compact yarn with pores is obtained; 4) Using the three-dimensional crack random generation algorithm, three-dimensional random cracks are generated inside the yarn topology structure of the random model of the porous compact yarn, simulating the crack characteristics inside the yarn of the three-dimensional woven composite material during the compaction process, and finally obtaining a three-dimensional woven parameterized random model.

[0007] The controllable parameterized model generation method of the three-dimensional braided composite material of the present invention is further as follows: the step 1) is specifically: 1-1), based on the kinematic analysis of the four-step three-dimensional weaving process, a discrete space node coordinate model is established, and then the adjacent weaving step nodes are connected by a piecewise linear interpolation algorithm to generate the initial yarn space trajectory; further, a high-order polynomial interpolation algorithm is used to optimize the curvature continuity of the trajectory to obtain a smooth space path, and a three-dimensional weaving preform trajectory model with a complex spatial interlaced structure is constructed; 1-2), randomly generate an initial hexagonal contour based on geometric parameters; then, introduce a Perli noise field to apply random displacement perturbations to the hexagonal edge nodes; finally, use a piecewise Bezier curve to fit the perturbed contour to generate a high-fidelity yarn cross-sectional geometry; 1-3), the cross section performs spatial mapping along the trajectory path, and the geometric interference state is monitored in real time during this process; when interference is detected, the system automatically triggers the non-interference optimization algorithm, and achieves non-interference adjustment of the yarn spatial arrangement through iterative optimization.

[0008] The controllable parameterized model generation method of the three-dimensional braided composite material of the present invention is further as follows: in the steps 1-3), the non-interference optimization algorithm is specifically: 1-3-1), calculate the plane Euclidean distance according to the coordinates of the center points of the two yarns on the same cross section, and compare it with the sum of the radii of the circumscribed circles to determine whether the yarn cross sections interfere; 1-3-2), when interference occurs, simulated force calculation is performed to optimize the trajectory; 1-3-3), based on the results of geometric interference detection, the global curvature continuity optimization algorithm of the yarn trajectory is triggered, and the spatial path refitting is achieved through parameterized spline reconstruction, eventually forming a closed-loop feedback mechanism of "interference detection-trajectory optimization-refitting".

[0009] The controllable parameterized model generation method of the three-dimensional braided composite material of the present invention is further as follows: in the step 2), the three-dimensional reconstruction algorithm is specifically: 2-1), generate multi-scale pore primitives by randomizing the major semi-axis, middle semi-axis and minor semi-axis parameters of the ellipsoid; then, use the Euler angle random rotation transformation to realize the random directional distribution of these primitives in three-dimensional space, so as to construct a multi-scale pore structure model with anisotropic characteristics; 2-2), using the pore-yarn multi-level spatial interference detection algorithm to ensure that the pores can be accurately placed outside the yarn array and that the pores do not interfere with each other, and finally the pore structure model is placed in the loose yarn random model; 2-3), by iterating the above steps in a loop, the simulated pores are continuously placed in the loose yarn random model.

[0010] The controllable parameterized model generation method of the three-dimensional braided composite material of the present invention is further as follows: In the step 2-2), the pore-yarn multi-level spatial interference detection algorithm is specifically: 2-2-1), during the placement of the pore model, first detect whether the center point of the pore interferes with the yarn array or the placed pore space. If interference is detected, the center point position is randomly re-determined; if there is no interference, proceed to the next step of detection; 2-2-2), detect whether the six main axis endpoints of the pore after rotation interfere with the yarn array or the placed pore space. If any endpoint detects interference, the center point position is re-randomly determined; if all endpoints have no interference, proceed to the next step of detection; 2-2-3), detect whether the eight 45° endpoints of the rotated pore interfere with the yarn array or the placed pore space; if interference is detected at any endpoint, the center point position is re-randomly determined; if there is no interference at all endpoints, the pore space is placed in the loose yarn random model.

[0011] The controllable parameterized model generation method of the three-dimensional woven composite material of the present invention is further as follows: in the steps 2-3), after each successful placement, the pore volume is accumulated and added to the total pore volume; when the accumulated pore volume reaches a preset threshold, the pore placement process is terminated, and finally a random model of loose yarns containing pores is obtained; wherein the volume calculation formula of the ellipsoid and the volume porosity calculation formula are as follows:

[0012]

[0013] Among them, v i Pores represents the volume of pore i, a i , b i , c i Indicates the major axis, middle axis, and minor axis of the current pore; v model represents the volume of the entire random model, F 3D Pores Represents the volume porosity of the current random model.

[0014] The controllable parameterized model generation method of the three-dimensional braided composite material of the present invention is further as follows: in the step 3), the yarn dynamic compaction algorithm is specifically: 3-1), based on the extraction of the geometric contour of the yarn cross section, its edge is discretized into a mass point-spring system; in this system, each mass point is given a solid collision attribute to accurately simulate the non-penetrating contact behavior between yarns; at the same time, adjacent mass points are connected to each other through linear springs to effectively characterize the elastic recovery characteristics of the yarn material; 3-2), when the discretized contours of adjacent yarns approach each other due to contraction motion, the system will trigger the particle-level collision detection and response mechanism; 3-3), the system presets a global shrinkage ratio threshold as the termination condition of the compaction process; when the minimum spacing between yarns reaches the threshold, the system automatically terminates the shrinkage iteration process; then, the compacted mass-spring system is remapped to a continuous geometric contour to generate a random model of compacted yarns with pores.

[0015] The controllable parameterized model generation method of the three-dimensional braided composite material of the present invention is also as follows: in the step 4), the three-dimensional crack random generation algorithm is specifically: 4-1), in the internal space of the yarn, firstly, the starting position of the microcrack is selected, and then the propagation range of the microcrack inside the yarn is randomly determined according to the preset maximum penetration length value of the microcrack; the microcrack starts from the starting point and expands along the random path until it reaches the end point; 4-2), based on the three-dimensional coordinate starting point and end point of the microcrack, combined with the three-dimensional coordinate of the yarn centerline at the same z height as the microcrack, calculate the deviation beam, and use the yarn centerline as a reference to generate a microcrack trajectory that is consistent with the curvature of the yarn centerline; 4-3), on the corrected microcrack trajectory, select the z plane where the midpoint is located, and the microcrack generation algorithm based on Perlin noise will take the center point as the starting point to generate the corresponding microcrack section on this z plane; 4-4), the generated microcrack section is copied to all z heights based on the microcrack path trajectory to form a microcrack surface; 4-5), introduce the erosion factor, the erosion factor is 100% at the center point of the microcrack path, and then based on the cosine function, map the distance from one end point to the midpoint or the other end point to the [-π / 2, π / 2] period, and combine with Perlin noise perturbation to generate irregular erosion factors, and finally construct irregular microcrack voxels in the random model.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. The controllable parameterized model generation method of the three-dimensional woven composite material of the present invention utilizes the pore-yarn multi-level spatial interference detection algorithm to achieve efficient positioning and interference-free embedding of the pore space position after rotation, ensuring the statistical consistency of the pore distribution characteristics and the actual process defects.

[0017] 2. The controllable parameterized model generation method of the three-dimensional braided composite material of the present invention utilizes the yarn dynamic compaction algorithm of the physical engine based on the Taichi framework to achieve physically accurate simulation of the yarn densification process.

[0018] 3. The controllable parameterized model generation method of the three-dimensional braided composite material of the present invention accurately simulates the three-dimensional morphological characteristics of micro-cracks generated after yarn compaction through a three-dimensional crack random generation algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of the controllable parameterized model generation method of the three-dimensional braided composite material of the present invention.

[0020] Figure 2 It is a step diagram of the four-step three-dimensional weaving process in step 1) of the present invention.

[0021] Figure 3 It is a flow chart of constructing the three-dimensional braided preform model in step 1) of the present invention.

[0022] Figure 4 This is a schematic diagram of the generation principle of the random model of loose yarns containing pores in step 2) of the present invention.

[0023] Figure 5 It is a schematic diagram of the yarn dynamic compaction algorithm in step 3) of the present invention.

[0024] Figure 6 It is a flow chart of the three-dimensional crack random generation algorithm in step 4) of the present invention.

[0025] Figure 7 It is a step diagram of the method for generating a controllable parameterized model of a three-dimensional woven composite material of the present invention. DETAILED DESCRIPTION

[0026] Please refer to the instruction manual Figure 1 As shown, the present invention is a method for generating a controllable parameterized model of a three-dimensional braided composite material, which comprises the following steps: 1) As shown in the instruction manual Figure 2 As shown, based on the kinematic analysis of the four-step three-dimensional weaving process, a three-dimensional weaving preform geometric model is constructed, and on the basis of trajectory parameterized modeling, the yarn cross-section geometry reconstruction is implemented, and the optimized cross-section is spatially mapped along the trajectory path to ensure that the cross-section center strictly coincides with the trajectory node and there is no geometric interference between the yarns, and finally a loose yarn random model is obtained.

[0027] Specifically, the step 1) is: 1-1), as shown in the instruction manual Figure 3 As shown in (a) in the figure, based on the kinematic analysis of the four-step three-dimensional weaving process, a discrete spatial node coordinate model is established. Then, the adjacent weaving step nodes are connected by a piecewise linear interpolation algorithm to generate the initial yarn space trajectory. The high-order polynomial interpolation algorithm is further used to optimize the curvature continuity of the trajectory, obtain a smooth spatial path, and construct a three-dimensional weaving preform trajectory model with a complex spatial interlaced structure.

[0028] 1-2), as shown in the instruction manual Figure 3 As shown in (b) in the figure, the initial hexagonal contour is randomly generated based on the geometric parameters (including the major axis, minor axis, rotation angle, and center point offset, etc.). Subsequently, the random displacement perturbation is applied to the hexagonal edge nodes by introducing the Perli noise field, thereby effectively simulating the microscopic irregular characteristics of the real yarn edge. On this basis, the perturbed contour is smoothed by using the piecewise cubic Bezier curve fitting algorithm, and the curve continuity is ensured by optimizing the control point position, and finally a yarn section with high fidelity is generated.

[0029] 1-3), the cross section performs spatial mapping along the trajectory path, and the geometric interference state is monitored in real time during this process. When interference is detected, the system automatically triggers the non-interference optimization algorithm, and achieves non-interference adjustment of the yarn spatial arrangement through iterative optimization. Specifically: 1-3-1), calculate the plane Euclidean distance D based on the coordinates of the center points of the two yarns on the same cross section i,j , and compare it with the sum of the radii of the circumscribed circles to determine whether the yarn cross section is interfering. That is, when the center distance between the two cross sections is less than the sum of the radii of the circumscribed circles, it is determined to be an interference state. The specific calculation formula is as follows:

[0030]

[0031] Among them, (x i , y i ), (x j , y j ) is the cross-sectional center coordinate point p of yarn i and yarn j at the same height i and point p j , D i,j To calculate the plane Euclidean distance, the distance threshold Δ given in the formula is the sum of the distances from the cross-section centers of the two yarns i and j to their respective farthest vertex coordinates.

[0032] 1-3-2), when interference occurs, simulated force calculation is performed to optimize the yarn trajectory. When two yarn sections are judged to interfere with each other, simulated force F is calculated. i, j Calculate the total simulated repulsive force vector F for yarn i for the optimized trajectory i The calculation formula is as follows:

[0033]

[0034] Among them, u i, j From point p j Point to point p i The unit direction vector of m i, j is the magnitude of the simulated repulsive force, ξ is the repulsive force strength parameter; F i, j For point p i and point p j The repulsive force vector between them. F i is the sum of all repulsive force vectors on yarn i (when multiple yarns interfere with yarn i).

[0035]

[0036] Among them, p i 'Used to replace the yarn trajectory center p at the current height i， Update the trajectory coordinates of the yarn.

[0037] 1-3-3), based on the results of geometric interference detection, the global curvature continuity optimization algorithm of the yarn trajectory is triggered, and the spatial path refitting is achieved through parameterized spline reconstruction, eventually forming a closed-loop feedback mechanism of "interference detection-trajectory optimization-refitting".

[0038] 2) Please refer to the instruction manual. Figure 4As shown in (a), based on the loose yarn random model, multi-scale pore primitives are generated by randomizing the ellipsoid geometric parameters, and then the Euler angle random rotation transformation and random placement are used to enhance the spatial heterogeneity characteristics of the pore morphology. The optimized pore model is embedded in the topological structure of the loose yarn random model to obtain a loose yarn random model with pores.

[0039] Specifically, the step 2) is: 2-1), generate multi-scale pore primitives by randomizing the major semi-axis, middle semi-axis and minor semi-axis parameters of the ellipsoid, as shown in the attached Figure 4 As shown in (b) in the figure, the random rotation transformation of Euler angles is then used to achieve the random orientation distribution of these primitives in three-dimensional space, thereby constructing a multi-scale pore structure model with anisotropic characteristics.

[0040] Combined with Figure 4 In (c), the Euler angle rotation formula is as follows:

[0041]

[0042]

[0043]

[0044] Among them, θ(α), φ(β) and ψ(γ) represent the rotation matrices around the x, y, and z axes respectively; R is the sum of the rotation matrices.

[0045] Next, when the three-dimensional rotation matrix is ​​constructed, the endpoint coordinates of the initial ellipsoid are multiplied by R. In addition to the center point, there are 14 endpoints in total. The coordinate calculation formula of the endpoints after rotation is as follows:

[0046]

[0047] Among them, p represents the coordinate value of each endpoint before rotation; x0, y0, z0 represent the three-axis coordinates of the pore origin, x, y, z represent the offset values ​​of the 14 endpoints compared to the origin; p' represents the new coordinate value of each endpoint after three-dimensional rotation.

[0048] 2-2), please refer to the attached Figure 4 As shown in (d), a pore-yarn multi-level spatial interference detection algorithm is used to ensure that the pores can be accurately placed outside the yarn array and that the pores do not interfere with each other. Finally, the pore structure model is placed in a loose yarn random model.

[0049] The pore-yarn multi-level spatial interference detection algorithm is specifically as follows: 2-2-1), during the placement of the pore model, first check whether the center point of the pore interferes with the yarn array or the placed pore space. If interference is detected, the center point position is re-randomly determined; if there is no interference, proceed to the next step of detection.

[0050] 2-2-2), detect whether the six main axis endpoints of the rotated pores interfere with the yarn array or the placed pore space. If interference is detected at any endpoint, the center point position is randomly re-determined; if there is no interference at all endpoints, proceed to the next step of detection.

[0051] 2-2-3), detect whether the eight 45° endpoints of the rotated pore interfere with the yarn array or the placed pore space; if interference is detected at any endpoint, the center point position is re-randomly determined; if there is no interference at all endpoints, the pore space is placed in the loose yarn random model.

[0052] 2-3), by iterating the above steps, continue to place simulated pores in the loose yarn random model. After each successful placement, the pore volume is accumulated and added to the total pore volume. When the cumulative pore volume reaches the preset threshold, the pore placement process is terminated, and finally a loose yarn random model with pores is obtained. Among them, the volume calculation formula of the ellipsoid and the volume porosity calculation formula are as follows:

[0053]

[0054] Among them, v i Pores represents the volume of pore i (accumulated by the number of spatial voxels), a i , b i , c i Indicates the major axis, middle axis, and minor axis of the current pore; v model represents the volume of the entire random model (including yarn voxels, blank voxels between yarns, and pore voxels), F 3D Pores Indicates the volume porosity (%) of the current random model.

[0055] 3) A physical engine based on the Taichi framework is used to construct a yarn dynamic compaction algorithm. After the random model of loose yarn with pores is input into the algorithm, the physical and precise simulation of the densification behavior of yarn and pores during the compaction process is achieved, and the random model of compact yarn with pores is obtained.

[0056] Among them, the yarn dynamic compaction algorithm is specifically as follows: 3-1), as attached Figure 5As shown in (a), based on the extraction of the geometric profile of the yarn cross section, its edge is discretized into a mass-spring profile. In this system, each mass point is given a solid collision attribute to accurately simulate the non-penetrating contact behavior between yarns; at the same time, adjacent mass points are connected to each other through linear springs to effectively characterize the elastic recovery characteristics of the yarn material.

[0057] 3-2), when the discretized contours of adjacent yarns approach each other due to contraction movement, the system will trigger the particle-level collision detection and response mechanism; this mechanism is based on the law of conservation of momentum and the spring potential energy function, dynamically calculating the contact force and deformation displacement between the contours, thereby realizing the physical drive simulation of the extrusion deformation behavior between the yarns. Figure 5 As shown in (b) of the figure, comparing the cross-sectional changes of the random model of loose yarns with pores and the random model of compact yarns with pores, yarns 1', 2', 3' undergo obvious deformation compared with yarns 1, 2, 3, and pore 1' also undergoes obvious deformation compared with pore 1 due to the extrusion of the adjacent yarns.

[0058] 3-3), the system presets a global shrinkage ratio threshold as the termination condition of the compaction process; when the minimum spacing between yarns reaches the threshold, the system automatically terminates the shrinkage iteration process; then, the compacted mass-spring system is remapped to a continuous geometric contour to generate a random model of compacted yarns with pores.

[0059] 4) Using the three-dimensional crack random generation algorithm, three-dimensional random cracks are generated inside the yarn topology structure of the random model of the porous compact yarn, simulating the crack characteristics inside the yarn of the three-dimensional woven composite material during the compaction process, and finally obtaining a three-dimensional woven parameterized random model.

[0060] The three-dimensional crack random generation algorithm is specifically: 4-1), please refer to the attached Figure 6 As shown in the figure, in the internal space of the yarn topology structure of the random model of the compact yarn with pores, the starting position of the microcrack is first selected, and then the propagation range of the microcrack inside the yarn is randomly determined according to the preset maximum penetration length of the microcrack. The microcrack starts from the starting point and expands along the random path until it reaches the end point; 4-2), based on the three-dimensional coordinate starting point and end point of the microcrack, combined with the three-dimensional coordinate of the yarn centerline at the same z height as the microcrack, calculate the deviation beam, and use the yarn centerline as a reference to generate a microcrack trajectory that is consistent with the curvature of the yarn centerline; the specific implementation formula is:

[0061]

[0062]

[0063]

[0064]

[0065]

[0066] Among them, P-Start and P-End represent the starting and ending points of the three-dimensional coordinates of the microcracks randomly generated by the algorithm; P C -Start and P C -End represents the three-dimensional coordinates of the yarn centerline at the same z height; Δx, Δy represent the deviations between the starting and ending points of the microcrack and the starting and ending points of the yarn centerline on the x and y axes. Steps represents the total coordinate step length to be generated, which is determined by the z-axis height difference of the random starting and ending points; Step represents the current coordinate generation step, ranging from [0, Steps].

[0067] 4-3), on the corrected microcrack trajectory, select the z plane where the midpoint is located, and the microcrack generation algorithm based on Perlin noise will take the center point as the starting point to generate the corresponding microcrack section on this z plane. The specific implementation formula is as follows:

[0068]

[0069]

[0070] In the above formula, P-Mid is used to represent the three-dimensional coordinates of the midpoint of the microcrack trajectory; x and y represent the plane coordinates of the microcrack trajectory at the z-axis height, and they are also the starting point coordinates of the microcrack generation; x i and i Represent the x-axis and y-axis coordinates of the randomly generated end point; dx and dy represent the difference from the starting point to the end point in the x-axis and y-axis directions respectively; t is the normalized step length, and its value range is limited to the interval [0, 1]; D x (t) and D y (t) represents the value of the Perlin noise function in the x-axis and y-axis directions, respectively. Its main function is to control the degree of disturbance on the x-axis and y-axis. A represents the amplitude. This parameter can be appropriately adjusted according to actual needs. The larger its value, the more severe the fluctuation of the microcracks will be.

[0071] 4-4), the generated microcrack section is copied to all z heights based on the microcrack path trajectory to form a microcrack surface.

[0072] 4-5), introduce the erosion factor, the erosion factor is 100% at the center point of the microcrack path, and then based on the cosine function, map the distance from one end point to the midpoint or the other end point to the [-π / 2, π / 2] period, and combine with Perlin noise perturbation to generate irregular erosion factors, and finally construct irregular microcrack voxels in the random model. The specific implementation formula is as follows:

[0073]

[0074] Among them, L represents the distance from the starting point P-Mid to the end point P-Start or P-End; l represents the distance from the current point to the point P-Mid; D represents the Perlin noise function; seed represents the random seed, the purpose is to improve the randomness of the erosion factor and make the microcracks more diversified; A represents the amplitude, which regulates the fluctuation degree of the erosion factor; E represents the randomly generated corrosion factor curve, which will determine the corrosion degree of the crack section at different z-axis heights based on this parameter.

[0075] In summary, combined with the attached Figure 7 The visual parametric modeling process of the controllable parametric model generation method of the three-dimensional woven composite material of the present invention utilizes the kinematic analysis of the four-step three-dimensional weaving process to generate a three-dimensional woven preform trajectory model, and implements yarn cross-section geometry reconstruction based on the trajectory parametric modeling. Then, multi-scale pore primitives are generated by randomizing the geometric parameters of the ellipsoid, and the random placement after the Euler angle random rotation transformation is adopted to enhance the spatial heterogeneity characteristics of the pore morphology. Subsequently, the model dynamic compaction algorithm is used to realize the physical and accurate simulation of the densification behavior of the yarn and the pore during the compaction process; finally, the three-dimensional crack random generation algorithm is used to accurately simulate the crack characteristics inside the yarn of the three-dimensional woven composite material during the compaction process, and a three-dimensional woven parametric random model can be obtained, including four-phase characteristics such as yarn, matrix, initial bubble, and yarn microcrack.

[0076] The controllable parameterized model generation method of the three-dimensional woven composite material of the present invention utilizes the pore-yarn multi-level spatial interference detection algorithm to achieve efficient positioning and interference-free embedding of the pore space position after rotation; secondly, a high-performance physical engine based on the Taichi architecture is used to develop a yarn dynamic compaction algorithm, which can accurately simulate the physical behavior of the yarn densification process and pore morphological changes; finally, through the three-dimensional crack random generation algorithm, the accurate reconstruction of the three-dimensional morphological characteristics of the micro-cracks after compaction is achieved. This method effectively solves the technical problems of low geometric interference detection efficiency, overly ideal defect morphology, and insufficient yarn compactness in traditional modeling.

[0077] The above specific implementation methods are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for generating a controllable parameterized model of a three-dimensional braided composite material, characterized in that: The steps include: 1) Based on the kinematic analysis of the four-step three-dimensional weaving process, a three-dimensional weaving preform geometric model is constructed, and on the basis of trajectory parameterized modeling, the yarn cross-section geometry reconstruction is implemented, and the optimized cross-section is spatially mapped along the trajectory path to ensure that the cross-section center strictly coincides with the trajectory node and there is no geometric interference between the yarns, and finally a loose yarn random model is obtained; 2) Based on the loose yarn random model, multi-scale pore primitives are generated by randomizing the ellipsoid geometric parameters, and then the Euler angle random rotation transformation and random placement are used to enhance the spatial heterogeneity of the pore morphology. The optimized pore model is embedded in the topological structure of the loose yarn random model to obtain a loose yarn random model with pores; 3) A physical engine based on the Taichi framework is used to construct a yarn dynamic compaction algorithm. After the random model of loose yarn with pores is input into the algorithm, the physical and precise simulation of the densification behavior of yarn and pores during the compaction process is achieved, and a random model of compact yarn with pores is obtained; 4) Using the three-dimensional crack random generation algorithm, three-dimensional random cracks are generated inside the yarn topology structure of the random model of the porous compact yarn, simulating the crack characteristics inside the yarn of the three-dimensional woven composite material during the compaction process, and finally obtaining a three-dimensional woven parameterized random model.

2. The method for generating a controllable parameterized model of a three-dimensional braided composite material according to claim 1, characterized in that: The step 1) is specifically as follows: 1-1), based on the kinematic analysis of the four-step three-dimensional weaving process, a discrete space node coordinate model is established, and then the adjacent weaving step nodes are connected by a piecewise linear interpolation algorithm to generate the initial yarn space trajectory; further, a high-order polynomial interpolation algorithm is used to optimize the curvature continuity of the trajectory to obtain a smooth space path, and a three-dimensional weaving preform trajectory model with a complex spatial interlaced structure is constructed; 1-2), randomly generate the initial hexagonal outline according to the geometric parameters; Subsequently, the random displacement perturbation is applied to the hexagonal edge nodes by introducing the Perli noise field. Finally, the perturbed contour is fitted using a piecewise Bezier curve to generate high-fidelity yarn cross-sectional geometry. 1-3), the cross section performs spatial mapping along the trajectory path, and the geometric interference state is monitored in real time during this process; when interference is detected, the system automatically triggers the non-interference optimization algorithm, and achieves non-interference adjustment of the yarn spatial arrangement through iterative optimization.

3. The method for generating a controllable parameterized model of a three-dimensional braided composite material according to claim 2, characterized in that: In the steps 1-3), the non-interference optimization algorithm is specifically as follows: 1-3-1), calculate the plane Euclidean distance according to the coordinates of the center points of the two yarns on the same cross section, and compare it with the sum of the radii of the circumscribed circles to determine whether the yarn cross sections interfere; 1-3-2), when interference occurs, simulated force calculation is performed to optimize the trajectory; 1-3-3), based on the results of geometric interference detection, the global curvature continuity optimization algorithm of the yarn trajectory is triggered, and the spatial path refitting is achieved through parameterized spline reconstruction, eventually forming a closed-loop feedback mechanism of "interference detection-trajectory optimization-refitting".

4. The method for generating a controllable parameterized model of a three-dimensional braided composite material according to claim 1, characterized in that: The step 2) is specifically as follows: 2-1), generate multi-scale pore primitives by randomizing the major semi-axis, middle semi-axis and minor semi-axis parameters of the ellipsoid; then, use the Euler angle random rotation transformation to realize the random directional distribution of these primitives in three-dimensional space, so as to construct a multi-scale pore structure model with anisotropic characteristics; 2-2), using the pore-yarn multi-level spatial interference detection algorithm to ensure that the pores can be accurately placed outside the yarn array and that the pores do not interfere with each other, and finally the pore structure model is placed in the loose yarn random model; 2-3), by iterating the above steps in a loop, the simulated pores are continuously placed in the loose yarn random model.

5. The method for generating a controllable parameterized model of a three-dimensional braided composite material according to claim 4, characterized in that: In step 2-2), the pore-yarn multi-level spatial interference detection algorithm is specifically: 2-2-1), during the delivery of the pore model, firstly detect whether the center point of the pore interferes with the yarn array or the delivered pore space, and if interference is detected, re-randomly determine the center point position; If there is no interference, proceed to the next step of detection; 2-2-2), detect whether the six main axis endpoints of the pore after rotation interfere with the yarn array or the placed pore space. If any endpoint detects interference, the center point position is re-randomly determined; if all endpoints have no interference, proceed to the next step of detection; 2-2-3), detecting whether the eight 45° end points of the rotated pores interfere with the yarn array or the placed pore space; If interference is detected at any endpoint, the center point position is randomly re-determined; If all endpoints have no interference, the void space is cast into a loose yarn random model.

6. The method for generating a controllable parameterized model of a three-dimensional braided composite material according to claim 4, characterized in that: In the step 2-3), after each successful placement, the pore volume is accumulated and added to the total pore volume; when the accumulated pore volume reaches a preset threshold, the pore placement process is terminated, and finally a random model of loose yarn containing pores is obtained; wherein, the volume calculation formula of the ellipsoid and the volume porosity calculation formula are as follows: Among them, v i Pores represents the volume of pore i, a i , b i , c i Indicates the major axis, middle axis, and minor axis of the current pore; v model represents the volume of the entire random model, F 3D Pores Represents the volume porosity of the current random model.

7. The method for generating a controllable parameterized model of a three-dimensional braided composite material according to claim 1, characterized in that: In step 3), the yarn dynamic compaction algorithm is specifically as follows: 3-1), based on the extraction of the geometric contour of the yarn cross section, its edge is discretized into a mass point-spring system; in this system, each mass point is given a solid collision attribute to accurately simulate the non-penetrating contact behavior between yarns; at the same time, adjacent mass points are connected to each other through linear springs to characterize the elastic recovery characteristics of the yarn material; 3-2), when the discretized contours of adjacent yarns approach each other due to contraction motion, the system will trigger the particle-level collision detection and response mechanism; 3-3), the system presets a global shrinkage ratio threshold as the termination condition of the compaction process; When the minimum spacing between yarns reaches this threshold, the system automatically terminates the shrinking iteration process; Subsequently, the compacted mass-spring system is remapped into a continuous geometric contour to generate a random model of a compact yarn with pores.

8. The method for generating a controllable parameterized model of a three-dimensional braided composite material according to claim 1, characterized in that: In step 4), the three-dimensional crack random generation algorithm is specifically: 4-1), in the internal space of the yarn, firstly, the starting position of the microcrack is selected, and then the propagation range of the microcrack inside the yarn is randomly determined according to the preset maximum penetration length value of the microcrack; the microcrack starts from the starting point and expands along the random path until it reaches the end point; 4-2), based on the three-dimensional coordinate starting point and end point of the microcrack, combined with the three-dimensional coordinate of the yarn centerline at the same z height as the microcrack, calculate the deviation beam, and use the yarn centerline as a reference to generate a microcrack trajectory that is consistent with the curvature of the yarn centerline; 4-3), on the corrected microcrack trajectory, select the z plane where the midpoint is located, and the microcrack generation algorithm based on Perlin noise will take the center point as the starting point to generate the corresponding microcrack section on this z plane; 4-4), the generated microcrack section is copied to all z heights based on the microcrack path trajectory to form a microcrack surface; 4-5), introduce the erosion factor, the erosion factor is 100% at the center point of the microcrack path, and then based on the cosine function, map the distance from one end point to the midpoint or the other end point to the [-π / 2, π / 2] period, and combine with Perlin noise perturbation to generate irregular erosion factors, and finally construct irregular microcrack voxels in the random model.

Images