A method, system, electronic device and medium for generating random distribution of fibers

Abstract

The present invention relates to a method, system, electronic device and medium for generating a random distribution of fibers, and particularly relates to the technical field of composite materials. The method includes taking the first fiber as the center, and repeatedly using the artificial fish swarm algorithm to generate an initial fiber cluster within the fiber generation window; moving a search box within the fiber generation window where the initial fiber cluster is generated, and when the fiber volume fraction within the search box after each movement is less than a preset volume fraction, then based on the hardcore model, scatter points within the search box after each movement to obtain new fibers until the search box finishes moving within the fiber generation window, and deleting the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers. The present invention can effectively generate a random distribution structure of fibers and can solve the cluster problem.

Description

Technical Field

[0001] The present invention relates to the technical field of composite materials, and in particular to a method, a system, an electronic device and a medium for generating a random distribution of fibers. Background Art

[0002] Fiber-reinforced composite materials are increasingly widely used in the aerospace and automotive fields due to their high specific strength, specific stiffness, light weight, excellent fatigue resistance and corrosion resistance. Predicting the elastic properties of fiber-reinforced composite materials by establishing a representative volume element, and predicting the mechanical behaviors such as the fracture of composite materials by considering factors such as the constitutive relations of the matrix, fibers, fiber-matrix interfaces, pore defects, and thermal residual stresses have become important directions for the current analysis of the mechanical properties and structural design of composite materials.

[0003] The characteristics of the component materials, the geometric shapes and spatial distributions of the inclusions strongly affect the mechanical actions inside the composite materials, and thus affect the mechanical properties of the macroscopic structure. Therefore, the representative volume element (RVE) considering the random distribution of fibers has been favored by scholars. In recent years, researchers have proposed various algorithms to generate RVEs with random fiber distributions. At the beginning of the development of the algorithms, considering that fibers cannot collide, a hard-core model was proposed. However, the fiber content generated by this model cannot exceed 50%, so it is not suitable for the research and application of current composite materials with high fiber volume fractions. To increase the content, new RVE generation algorithms have been proposed, such as the method based on image processing. By obtaining the fiber distribution image of the composite material and using image technology to truly reproduce the fiber distribution, this method requires powerful software and hardware support and is difficult to be popularized and applied. The method based on fiber perturbation moves the fibers and continues to sprinkle points when a blank area is left. This method is very novel, but complex fiber movement conditions need to be considered. Liu Zhao et al. proposed the Random Distribution Particle Swarm Optimization (RDPSO) algorithm and put forward an important idea of fiber jumping to change the clustering phenomenon of fiber generation, which has achieved very good results, but there are also certain deficiencies: when the fiber content of the obtained RVE is small, if the fiber spacing is set unreasonably, a multi-clustering phenomenon is likely to occur; when the fiber content of the obtained RVE is high, due to the setting of fiber jumping, this jumping is likely to cause waste of space. For example, when the fiber spacing between the jumping fiber and a certain fiber is slightly less than at this time, no next fiber can be generated between these two fibers, ultimately resulting in the failure of RVE generation. Summary of the Invention

[0004] The objective of the present invention is to provide a method, system, electronic device and medium for generating a random distribution of fibers, which can effectively generate a random distribution structure of fibers and solve the cluster problem.

[0005] To achieve the above objective, the present invention provides the following solutions:

[0006] A method for generating a random distribution of fibers, comprising:

[0007] Randomly generate the first fiber within the fiber generation window;

[0008] Taking the first fiber as the center, based on the periodic boundary condition, use the artificial fish swarm algorithm multiple times to generate an initial fiber cluster within the fiber generation window;

[0009] Move the search box within the fiber generation window where the initial fiber cluster is generated. When the fiber volume fraction within the search box after each movement is less than the preset volume fraction, then based on the hardcore model, scatter points within the search box after each movement to obtain new fibers until the search box finishes moving within the fiber generation window, and delete the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers.

[0010] Optionally, the step of taking the first fiber as the center and using the artificial fish swarm algorithm multiple times based on the periodic boundary condition to generate an initial fiber cluster within the fiber generation window specifically includes:

[0011] Based on the periodic boundary condition, taking the first fiber as the center, use the artificial fish swarm algorithm to generate a fiber set at the first iteration number within the fiber generation window;

[0012] Judge whether the fiber volume fraction within the fiber generation window is greater than or equal to the preset volume fraction to obtain a first judgment result;

[0013] If the first judgment result is negative, then taking each fiber within the fiber set at the current iteration number as the center, based on the periodic boundary condition, use the artificial fish swarm algorithm to generate a fiber set at the next iteration number within the fiber generation window, update the iteration number and return to the step of "judging whether the fiber volume fraction within the fiber generation window is greater than or equal to the preset volume fraction to obtain a first judgment result";

[0014] If the first judgment result is positive, then stop the iteration.

[0015] Optionally, move the search box within the fiber generation window for generating the initial fiber cluster. When the fiber volume fraction within the search box after each movement is less than the preset volume fraction, new fibers are scattered within the search box after each movement based on the hardcore model until the search box finishes moving within the fiber generation window. Then, the fibers within the fiber generation window are pruned according to the preset volume fraction to obtain the finally generated fibers. Specifically, it includes:

[0016] Set the length, width, and movement step size of the search box;

[0017] Move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the search box after the movement;

[0018] Judge whether the fiber volume fraction within the search box after the movement is less than the preset volume fraction to obtain a second judgment result;

[0019] If the second judgment result is yes, scatter points within the search box based on the hardcore model and periodic boundary conditions, and return to "Move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the search box after the movement" until the search box finishes moving within the fiber generation window;

[0020] If the second judgment result is no, return to "Move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the search box after the movement" until the search box finishes moving within the fiber generation window;

[0021] Based on the periodic boundary conditions, prune the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers.

[0022] A fiber random distribution generation system, including:

[0023] The first fiber generation module is used to randomly generate the first fiber within the fiber generation window;

[0024] The initial fiber cluster generation module is used to generate an initial fiber cluster within the fiber generation window based on the periodic boundary conditions by using the artificial fish swarm algorithm multiple times with the first fiber as the center;

[0025] A final fiber generation module is used to move a search box within a fiber generation window for generating the initial fiber cluster. When the fiber volume fraction within the search box after each movement is less than the preset volume fraction, new fibers are scattered within the search box after each movement based on a hardcore model until the search box finishes moving within the fiber generation window. Then, the fibers within the fiber generation window are trimmed according to the preset volume fraction to obtain the finally generated fibers.

[0026] Optionally, the initial fiber cluster generation module specifically includes:

[0027] A fiber set generation unit for the first iteration number is used to generate a fiber set for the first iteration number within the fiber generation window with the first fiber as the center based on periodic boundary conditions using an artificial fish swarm algorithm;

[0028] A first judgment unit is used to judge whether the fiber volume fraction within the fiber generation window is greater than or equal to the preset volume fraction to obtain a first judgment result;

[0029] A fiber set generation unit is used to, if the first judgment result is no, generate a fiber set for the next iteration number within the fiber generation window with each fiber within the fiber set for the current iteration number as the center based on periodic boundary conditions using an artificial fish swarm algorithm, update the iteration number, and return "judge whether the fiber volume fraction within the fiber generation window is greater than or equal to the preset volume fraction to obtain a first judgment result";

[0030] A stop iteration unit is used to, if the first judgment result is yes, stop the iteration.

[0031] Optionally, the final fiber generation module specifically includes:

[0032] A setting unit is used to set the length, width, and moving step size of the search box;

[0033] A fiber volume fraction calculation unit is used to move the search box within the fiber generation window for generating the initial fiber cluster and calculate the fiber volume fraction within the search box after the movement;

[0034] A second judgment unit is used to judge whether the fiber volume fraction within the search box after the movement is less than the preset volume fraction to obtain a second judgment result;

[0035] A scattering unit, which is used to, if the second judgment result is yes, scatter points within the search box based on the hardcore model and periodic boundary conditions, and return "move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the moved search box" until the search box finishes moving within the fiber generation window;

[0036] A repeating unit, which is used to, if the second judgment result is no, return "move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the moved search box" until the search box finishes moving within the fiber generation window;

[0037] A deletion unit, which is used to, based on the periodic boundary conditions, delete the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers.

[0038] An electronic device, comprising:

[0039] A memory and a processor, where the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the fiber random distribution generation method according to the above.

[0040] A computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, it implements the fiber random distribution generation method as described above.

[0041] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: The present invention takes the first fiber as the center, and repeatedly uses the artificial fish swarm algorithm to generate an initial fiber cluster within the fiber generation window; move the search box within the fiber generation window for generating the initial fiber cluster, and when the fiber volume fraction within the moved search box is less than the preset volume fraction each time, scatter points within the moved search box each time based on the hardcore model to obtain new fibers until the search box finishes moving within the fiber generation window, and delete the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers, which can effectively generate a random distribution structure of fibers and can solve the cluster problem. Description of the Drawings

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can be obtained according to these drawings without creative efforts.

[0043] Figure 1Schematic diagram of window areas divided for random deletion;

[0044] Figure 2 Principle diagram of generating fibers by the artificial fish swarm algorithm;

[0045] Figure 3 Schematic diagram of periodic boundary conditions;

[0046] Figure 4 Schematic diagram of the results of blank area retrieval and hardcore scattering;

[0047] Figure 5 Flowchart of a method for generating random fiber distribution provided by an embodiment of the present invention;

[0048] Figure 6 Diagram of the generation process of RVEs with three fiber volume fractions, where Figure 6 (a) Diagram of the generation process of an RVE with a 50% fiber volume fraction, Figure 6 (b) Diagram of the generation process of an RVE with a 60% fiber volume fraction, Figure 6 (c) Diagram of the generation process of an RVE with a 65% fiber volume fraction;

[0049] Figure 7 Abaqus three-dimensional solid model diagram of an RVE, where Figure 7 (a) Represents the three-dimensional solid model of the RVE, Figure 7 (b) Represents the three-dimensional model diagram of the resin matrix of the RVE, Figure 7 (c) Is the three-dimensional model diagram of the fibers of the RVE. Detailed implementation manners

[0050] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0051] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0052] The present invention refers to the fitness function 1643-1653 of the journal "Polymer composites" Volume 40, Issue 4, April 2019, and the improvement measures of the artificial fish school algorithm 1158-1163 of "Control Engineering" Volume 27, Issue 7, July 2020 by Liu Zhao et al., and proposes a fiber random distribution generation method, which is particularly suitable for unidirectional long fiber reinforced composite materials, and is often referred to as the generation method of RVE. The general steps are to first use the artificial fish school algorithm to generate initial fiber clusters, then establish a search box to search the blank area, carry out fiber filling, and finally perform a random fiber deletion method, and finally obtain the fiber random distribution RVE of the unidirectional long fiber composite material, which specifically includes:

[0053] Generates the first fiber randomly within the fiber generation window.

[0054] Taking the first fiber as the center, an artificial fish swarm algorithm is repeatedly used to generate initial fiber clusters in the fiber generation window based on periodic boundary conditions.

[0055] The search box is moved in the fiber generation window for generating the initial fiber cluster. When the fiber volume fraction in the search box after each movement is less than the preset volume fraction, new fibers are obtained by scattering points in the search box after each movement based on the hard-core model until the search box is completely moved in the fiber generation window. The fibers in the fiber generation window are deleted according to the preset volume fraction to obtain the finally generated fibers.

[0056] In practical applications, the first fiber is randomly generated within the fiber generation window, specifically including:

[0057] Enter the parameters of the fiber generation window area, such as the length a, width b of the RVE, and the radius r of the fiber f , preset volume fraction V f , the minimum and maximum spacing of fibers (l min ,l max ).

[0058] The first fiber is randomly generated in the given area of the window (x1,y1∈(a2-r,a2+r)).

[0059] In practical applications, the method of generating an initial fiber cluster in the fiber generation window by using an artificial fish swarm algorithm multiple times based on periodic boundary conditions with the first fiber as the center specifically includes:

[0060] Based on periodic boundary conditions, an artificial fish swarm algorithm is used to generate a fiber set with a first number of iterations within the fiber generation window with the first fiber as the center.

[0061] Determine whether the fiber volume fraction within the fiber generation window is greater than or equal to a preset volume fraction to obtain a first determination result.

[0062] If the first determination result is negative, then, with each fiber in the fiber set at the current iteration number as the center, and based on the periodic boundary condition, use the artificial fish swarm algorithm to generate a fiber set for the next iteration number within the fiber generation window, update the iteration number, and return "Determine whether the fiber volume fraction within the fiber generation window is greater than or equal to a preset volume fraction to obtain a first determination result".

[0063] If the first determination result is positive, then stop the iteration.

[0064] In practical applications, the fitness function in the artificial fish swarm algorithm is as follows:

[0065]

[0066] Among them, S0 is the cross-sectional area of a single fiber, r f is the radius of the fiber, l is the target distance of the fiber at the current iteration step, l = 2×r f +l0, l0 is a random number between (l min , l max ), l min and l max respectively represent the set minimum and maximum fiber spacings, l ij represents the center distance between fiber i and fiber j, and Y ij represents the fitness function.

[0067] In practical applications, move the search box within the fiber generation window for generating the initial fiber cluster. When the fiber volume fraction within the search box after each movement is less than the preset volume fraction, then, based on the hardcore model, scatter points within the search box after each movement to obtain new fibers until the search box has completed its movement within the fiber generation window, and delete the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers. Specifically, it includes:

[0068] Set the length, width, and movement step size of the search box.

[0069] Move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the search box after the movement.

[0070] Determine whether the fiber volume fraction within the search box after the movement is less than the preset volume fraction to obtain a second determination result.

[0071] If the second judgment result is yes, points are scattered in the search box based on the hard core model and periodic boundary conditions, and the method returns to "moving the search box in the fiber generation window for generating the initial fiber cluster, and calculating the fiber volume fraction in the search box after the move" until the search box has been moved within the fiber generation window.

[0072] If the second judgment result is no, the process returns to the step of "moving the search box within the fiber generation window for generating the initial fiber cluster, and calculating the fiber volume fraction within the search box after the move", until the search box is completely moved within the fiber generation window.

[0073] Based on the periodic boundary conditions, the fibers in the fiber generation window are deleted according to the preset volume fraction to obtain the finally generated fibers.

[0074] In practical applications, based on the periodic boundary conditions, the fibers in the fiber generation window are deleted according to the preset volume fraction to obtain the final generated fibers, specifically including:

[0075] Divide the window area into Figure 1 In the areas shown in the figure, the fibers in the entire area are randomly deleted. The deleted fibers should meet the periodic boundary conditions: if the deleted fibers are Figure 1 1 area, only the selected fiber will be deleted; if the deleted fiber is located in Figure 1 In addition to deleting the selected fiber, you also need to delete the fiber translated from region 4 (region 2). Figure 1 Regions 3 and 5 are similar to regions 2 and 4; if the deleted fibers are located at the four corners of the window ( Figure 1 6, 7, 8, 9), in addition to deleting the selected fiber, the fibers in the other three regions need to be deleted until the preset volume fraction is met to obtain the final generated fiber.

[0076] The present invention adopts a more specific embodiment to describe in detail the above-mentioned fiber random distribution generation method, using the artificial fish school algorithm to generate initial fiber clusters; then establish a blank area in the search box search window, if the fiber volume fraction in the search box is less than the preset volume fraction, then scatter points in the search box area based on the hard core model until the fiber fills the entire window; finally, randomly delete the fibers in the entire window area until the preset volume fraction is met, including the following steps:

[0077] Step (1): Input the length, width, fiber radius, preset volume fraction, maximum fiber spacing and minimum fiber spacing of the RVE, and select the fiber generation window in a given area (x1,y1∈(a / 2-r f ,a / 2+r f)) After randomly generating the first fiber, the subsequent fibers are generated according to the artificial fish swarm algorithm. The specific process of generating subsequent fibers according to the artificial fish swarm algorithm is as follows: According to the artificial fish swarm algorithm and referring to relevant literature, a fitness function is established. Starting from the first fiber as the center, the next fiber is generated around it in turn until no more redundant fibers can be generated around the first fiber. Then, taking the second fiber as the center, the next fiber is generated around it until the required preset volume fraction is met, obtaining the initial fiber cluster. The formation process is as Figure 2 shown, and the generated fibers satisfy the periodic boundary conditions: Figure 3 The following figure shows the schematic diagram of the periodic boundary conditions. If the generated fiber is in Figure 1 region 1 of Figure 3 i.e., fiber 9 of Figure 1 then no treatment is required; if the generated fiber is located in region 2 of Figure 3 i.e., fiber 6 of Figure 1 then a fiber 5 needs to be generated in region 4. Fiber 5 is obtained by translating the ordinate of fiber 6 by b (window width). Figure 1 Regions 3 and 5 of Figure 1 are similar to regions 2 and 4, except that the ordinate remains unchanged and the abscissa is translated by the window length a; if the generated fiber is located at the four corners of the window ( Figure 1 6, 7, 8, 9 of Figure 3 ), then translations need to be performed in the other three regions. Taking the generated fiber located in region 6 as an example, i.e., Figure 3 fiber 4 of

[0078] Table 1 Parameter settings of the artificial fish swarm algorithm

[0079]

[0080] There are four basic behaviors of artificial fish, namely foraging, aggregation, following, and random behavior. The movement of 15 fish follows these four behaviors. After each fish moves, it has a new coordinate (x, y) and a fitness value at this coordinate (the fitness value here is determined by the distance between (x, y) and the selected fiber center coordinate. For example, when generating new fibers around the first fiber, the fitness value is calculated with the first fiber as the center. When no new fibers can be generated around the first fiber, the fitness value is calculated with the second fiber as the center). When the first movement of the 15 fish ends, there will be 15 fitness values. The smallest fitness value is selected as the temporary global best fitness value for the time being, and the coordinate under this fitness is the global best coordinate. After experiencing the maximum number of iterations Gen_max, there will be a final global best fitness value and global best coordinate. This coordinate is the coordinate of the required fiber. It should be noted that after each Gen_max iterations, only one best fiber center coordinate is generated. If n fibers need to be generated, except for the first fiber which is randomly generated, n - 1 Gen_max iterations are required.

[0081] Foraging behavior:

[0082] Set the current state of the artificial fish Randomly select another state X within its visual perception range j . If the fitness value of the other state is better than that of the current state, the artificial fish moves one step closer to the new state direction and generates a new state Otherwise, reselect a new state and continue to judge whether the fitness value is better than that of the current state. When the new state is reselected try_num times and the objective function condition is still not satisfied, the random behavior will be executed. The function description of the foraging behavior is as follows:

[0083] Artificial fish Randomly select a state X within the visual field j :

[0084]

[0085] Calculate respectively and X j 's fitness function values and make a comparison. If the fitness function value of X j is better than 's fitness function value, then move in the direction of X j to generate a new state

[0086]

[0087] Otherwise, Continue to select X within its visual perception range j , if it has tried try_num times and still does not meet the conditions, then perform random behavior.

[0088] Swarming behavior:

[0089] During the swarming process, they follow two rules: one is to approach neighboring fish, and the other is to avoid overcrowding. The artificial fish searches for the number of fish n within the current visual field f and calculates the center position X of the fish school c , calculates the fitness function value Y of the center position c , and compares it with the fitness function value Y of the current position i . If the fitness function value of the center position is better than that of the current position, and the non-crowding condition Y c / n f <δY i is satisfied, then move one step in the direction from the current position to the center position:

[0090]

[0091] If there are no fish in the current visual field, or the non-crowding condition is not met, then perform foraging behavior.

[0092] Following behavior:

[0093] When some fish in the fish school find food, nearby fish will all follow, and gradually, farther fish will also follow. The artificial fish counts the number of fish n within the visual field f and the best position of the fish. When the fitness function value Y j of the best position is better than the current fitness function value Y i , and there is no crowding phenomenon at the best position Y j / n f <δY i , move one step towards the best position:

[0094]

[0095] If there are no fish in the current visual field, or the non-crowding condition is not met, then perform foraging behavior.

[0096] Random behavior:

[0097] Move randomly within the visual field:

[0098]

[0099] Introduce the inertia factor ω of the particle swarm algorithm:

[0100] ω = ω max-(ω max -ω min )×α

[0101] ω max and ω min are taken as 0.9 and 0.4 respectively;

[0102] Step size:

[0103]

[0104] Field of view:

[0105]

[0106] Attenuation function:

[0107]

[0108] where and represent the next position and the current position of particle i respectively, Step and Visual represent the step size and the field of view length respectively, Step min and Visual min represent the minimum values of the step size and the field of view, Gen represents the current iteration number, Gen_max represents the maximum iteration number, and s = 3.

[0109] Based on step (1), the coordinate information of each fiber in the initial fiber cluster generated by the artificial fish swarm algorithm has been stored.

[0110] Step (2): Establish a search box of 15×15μm, and retrieve the blank area in the entire a×b fiber generation window area. The step size of the search box moving horizontally each time is 10, and the step size of the search box moving vertically is 15. If the fiber volume fraction in the area where the search box moves (the total fiber area in the search box divided by the area of the search box 15×15) is less than the preset volume fraction, 700 attempts of scatter points are made based on the hardcore model. The schematic diagram of the scatter point process is as Figure 4 shown. When scattering points, the fiber interference condition (not intersecting with other fibers) and the periodic boundary condition should be satisfied. Until the retrieval of the entire fiber generation window ends, there is no large - area blank area in the fiber generation window area, and the fiber fills the entire fiber generation window.

[0111] Step (3): Divide the entire window into regions, and randomly delete the fiber center points in the window following the periodic boundary condition until the preset volume fraction is satisfied.

[0112] Compared with the prior art, after generating the initial fiber clusters using the artificial fish swarm algorithm in the embodiments of the present invention, the blank area search, fiber filling, and random deletion methods are adopted to obtain the final RVE. (1) It avoids the clustering phenomenon that occurs when the fiber spacing is set unreasonably in methods such as RSE and RDPSO. Even if the set fiber spacing is not the most suitable, it can still generate a relatively ideal RVE; (2) When the fiber spacing is set to (0, 0.2), the highest content of RVE can reach 67%, which can meet the needs of most analyses. (3) The entire algorithm is implemented using Python code, and the generated fiber coordinates and fiber radius information can be automatically stored. And by writing code in Python to generate the three-dimensional structure of Abaqus, the three-dimensional structure of RVE can be quickly generated by reading the program through Abaqus, preparing for the prediction of the elastic constants of composite materials or the multi-scale analysis of composite materials in the later stage, and having great engineering application value. As Figure 7 shown, taking the Abaqus three-dimensional solid model diagram of an RVE with 60% (δ = 25, (l min ,l max ) = (0.2, 0.3), r f = 3.3 μm, h = 5 μm, where h is the thickness of the RVE) as an example, where (a) represents the three-dimensional solid model of the RVE, (b) represents the three-dimensional model diagram of the resin matrix of the RVE, and (c) is the three-dimensional model diagram of the fibers of the RVE. By inputting the material parameters of the fibers and the resin matrix of the composite material in Abaqus, and the mesh types of both the fibers and the resin matrix are C3D8R, the prediction of the elastic constants of the composite material can be carried out through the easypbc plug-in or the self-written Python program.

[0113] Figure 5 The following shows the step flow chart of the present invention. In this example, taking r f = 3.3 μm, δ = 50 (δ is the ratio of the length of the fiber to the radius of the fiber), and the length and width of the RVE are a = b = 165 μm as examples, the random distribution structures of fibers with 50% ((l min ,l max ) = (1.5, 2)), 60% (0, 1.45), and 65% (0, 0.2) are respectively generated to introduce the above-mentioned fiber random distribution generation method in detail:

[0114] Step (1): The window area is x, y ∈ (0, 165). In the given window area x1, y1 ∈ (a2 - r f , a2 + r f ), the center coordinates of the first fiber are randomly generated.

[0115] Step (2): Taking the center coordinates of the first fiber as the center, using the artificial fish swarm algorithm, successively generate the center coordinates of the second, third, fourth... fibers (generally 4 - 5 fibers) around the first fiber until no new fiber center coordinates can be generated around the first fiber. Taking the center coordinates of the second fiber as the center, successively generate new fiber center coordinates around the second fiber,..., until the required number of fibers is finally generated, and the artificial fish swarm algorithm stage ends. However, at this time, there are large blank areas in the corners of the RVE window, as shown in Figure 6 (a), Figure 6 (b), and Figure 6 (c) as shown in S1. S1 represents the initial fiber cluster generated by the artificial fish swarm algorithm.

[0116] Step (3): Establish a rectangular search box with a size of 15×15μm, conduct a blank search on the entire window area. If the fiber volume fraction within the search box is less than the preset volume fraction, perform hardcore scattering. For each step movement, conduct 700 attempts of hardcore scattering, randomly generate new center coordinates in the area where the search box is located. The generated fibers should not interfere with other fibers and should also meet the periodic boundary conditions. When the entire window search ends, there are no longer large blank areas, as shown in Figure 6 (a), Figure 6 (b), and Figure 6 (c) as shown in S2. S2 represents the blank area search and hardcore scattering (where the hollow circles represent the fibers added by the hardcore model).

[0117] Step (4): Divide the window into regions according to the periodic boundary conditions, randomly delete points in the window, as shown in S3 in Figures 6(a), (b), and (c). S3 represents random deletion (the black solid circles represent the randomly deleted fibers), until the preset volume fraction is met, and the entire algorithm process ends. The final RVEs with three fiber volume fractions are as shown in Figure 6 (a), Figures (b), and (c) as shown in S4. S4 represents the final RVE.

[0118] For the above method, an embodiment of the present invention provides a fiber random distribution generation system, including:

[0119] The first fiber generation module is used to randomly generate the first fiber within the fiber generation window.

[0120] The initial fiber cluster generation module is used to take the first fiber as the center and, based on the periodic boundary conditions, repeatedly use the artificial fish swarm algorithm to generate an initial fiber cluster within the fiber generation window.

[0121] A final fiber generation module is used to move a search box within a fiber generation window for generating the initial fiber clusters. When the fiber volume fraction within the search box after each movement is less than the preset volume fraction, new fibers are scattered within the search box after each movement based on a hardcore model until the search box has moved through the entire fiber generation window. The fibers within the fiber generation window are pruned according to the preset volume fraction to obtain the finally generated fibers.

[0122] In practical applications, the initial fiber cluster generation module specifically includes:

[0123] A fiber set generation unit for the first iteration number is used to generate a fiber set for the first iteration number within the fiber generation window based on the periodic boundary condition with the first fiber as the center by using an artificial fish swarm algorithm.

[0124] A first judgment unit is used to judge whether the fiber volume fraction within the fiber generation window is greater than or equal to the preset volume fraction to obtain a first judgment result.

[0125] The fiber set generation unit is used, if the first judgment result is negative, to generate a fiber set for the next iteration number within the fiber generation window based on the periodic boundary condition with each fiber within the fiber set for the current iteration number as the center by using an artificial fish swarm algorithm, update the iteration number, and return "judge whether the fiber volume fraction within the fiber generation window is greater than or equal to the preset volume fraction to obtain a first judgment result".

[0126] A stop iteration unit is used, if the first judgment result is positive, to stop the iteration.

[0127] In practical applications, the final fiber generation module specifically includes:

[0128] A setting unit is used to set the length, width, and moving step size of the search box.

[0129] A fiber volume fraction calculation unit is used to move the search box within the fiber generation window for generating the initial fiber clusters and calculate the fiber volume fraction within the search box after the movement.

[0130] A second judgment unit is used to judge whether the fiber volume fraction within the search box after the movement is less than the preset volume fraction to obtain a second judgment result.

[0131] A sprinkling unit, which is used to, if the second judgment result is yes, sprinkle points within the search box based on the hardcore model and periodic boundary conditions, and return "move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the moved search box" until the search box finishes moving within the fiber generation window.

[0132] A repeating unit, which is used to, if the second judgment result is no, return "move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the moved search box" until the search box finishes moving within the fiber generation window.

[0133] A deleting unit, which is used to, based on the periodic boundary conditions, delete the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers.

[0134] The present invention also provides an electronic device, including:

[0135] A memory and a processor, where the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the fiber random distribution generation method described above.

[0136] The present invention also provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the fiber random distribution generation method described above is implemented.

[0137] The present invention has the following beneficial effects:

[0138] Many algorithms such as the Random Sequential Expansion (RSE) algorithm and the Random Distribution Particle Swarm Optimization (RDPSO) algorithm are prone to generating fiber cluster phenomena when generating fibers. The occurrence of this phenomenon will result in poor randomness in the spatial distribution of the generated fibers in the fiber cross-section, which is far from the real distribution of fibers in production and manufacturing. Therefore, the generated Representative Volume Element (RVE) cannot represent the real fiber distribution. The present invention can effectively generate a random distribution structure of fibers and provide a new solution idea for the cluster problems that occur in most algorithms when generating random fibers.

[0139] In this specification, the various embodiments are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple. For the relevant parts, reference can be made to the description in the method section.

[0140] In this article, specific examples are used to elaborate on the principles and implementation manners of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those of ordinary skill in the art, based on the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation on the present invention.

Claims

1. A method for generating a random distribution of fibers, characterized in that, Including: Randomly generate the first fiber within the fiber generation window; Taking the first fiber as the center, based on the periodic boundary condition, use the artificial fish swarm algorithm multiple times to generate an initial fiber cluster within the fiber generation window; Specifically including: Based on the periodic boundary condition, use the artificial fish swarm algorithm with the first fiber as the center to generate a fiber set at the first iteration number within the fiber generation window; Judge whether the fiber volume fraction within the fiber generation window is greater than or equal to the preset volume fraction to obtain a first judgment result; If the first judgment result is no, then based on each fiber in the fiber set at the current iteration number as the center, use the artificial fish swarm algorithm based on the periodic boundary condition to generate a fiber set at the next iteration number within the fiber generation window, update the iteration number, and return "Judge whether the fiber volume fraction within the fiber generation window is greater than or equal to the preset volume fraction to obtain a first judgment result"; If the first judgment result is yes, stop the iteration; Move the search box within the fiber generation window where the initial fiber cluster is generated. When the fiber volume fraction within the search box after each movement is less than the preset volume fraction, then scatter points within the search box after each movement based on the hardcore model to obtain new fibers until the search box finishes moving within the fiber generation window. Trim the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers. Specifically including: Set the length, width, and movement step size of the search box; Move the search box within the fiber generation window where the initial fiber cluster is generated, and calculate the fiber volume fraction within the search box after the movement; Judge whether the fiber volume fraction within the search box after the movement is less than the preset volume fraction to obtain a second judgment result; If the second judgment result is yes, then scatter points within the search box based on the hardcore model and the periodic boundary condition, and return "Move the search box within the fiber generation window where the initial fiber cluster is generated, and calculate the fiber volume fraction within the search box after the movement" until the search box finishes moving within the fiber generation window; If the second judgment result is no, then return "Move the search box within the fiber generation window where the initial fiber cluster is generated, and calculate the fiber volume fraction within the search box after the movement" until the search box finishes moving within the fiber generation window; Based on the periodic boundary condition, trim the fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers.

2. A fiber random distribution generation system, characterized in that, Including: The first fiber generation module is used to randomly generate the first fiber within the fiber generation window; The initial fiber cluster generation module is used to take the first fiber as the center, based on the periodic boundary condition, use the artificial fish swarm algorithm multiple times to generate an initial fiber cluster within the fiber generation window. Specifically including: The fiber set generation unit at the first iteration number is used to generate a fiber set at the first iteration number within the fiber generation window by using the artificial fish swarm algorithm with the first fiber as the center based on the periodic boundary condition; A first judgment unit, configured to judge whether the fiber volume fraction within the fiber generation window is greater than or equal to a preset volume fraction, and obtain a first judgment result; A fiber set generation unit, configured to, if the first judgment result is negative, generate a fiber set for the next iteration within the fiber generation window based on each fiber in the fiber set at the current iteration number, using an artificial fish swarm algorithm based on periodic boundary conditions, update the iteration number, and return "judge whether the fiber volume fraction within the fiber generation window is greater than or equal to a preset volume fraction, and obtain a first judgment result"; A stop iteration unit, configured to stop iteration if the first judgment result is positive; A final fiber generation module, configured to move a search box within the fiber generation window for generating the initial fiber cluster, and when the fiber volume fraction within the search box after each movement is less than the preset volume fraction, scatter points within the search box after each movement based on a hardcore model to obtain new fibers until the search box finishes moving within the fiber generation window, and delete fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers, specifically including: A setting unit, configured to set the length, width, and movement step size of the search box; A fiber volume fraction calculation unit, configured to move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the search box after the movement; A second judgment unit, configured to judge whether the fiber volume fraction within the search box after the movement is less than the preset volume fraction, and obtain a second judgment result; A scatter point unit, configured to, if the second judgment result is positive, scatter points within the search box based on the hardcore model and periodic boundary conditions, and return "move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the search box after the movement" until the search box finishes moving within the fiber generation window; A repetition unit, configured to, if the second judgment result is negative, return "move the search box within the fiber generation window for generating the initial fiber cluster, and calculate the fiber volume fraction within the search box after the movement" until the search box finishes moving within the fiber generation window; A deletion unit, configured to, based on the periodic boundary conditions, delete fibers within the fiber generation window according to the preset volume fraction to obtain the finally generated fibers.

3. An electronic device, characterized in that, Including: A memory and a processor, where the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the fiber random distribution generation method according to any one of claims 1.

4. A computer-readable storage medium, characterized in that, It stores a computer program, and when the computer program is executed by the processor, it implements the fiber random distribution generation method according to any one of claims 1.

Images