A method for automatically modeling heterogeneous concrete based on a topologically disordered hexagonal connection grid
By adopting an automatic modeling method for heterogeneous concrete based on topologically disordered six-way connected meshes, the problem of the inability to automatically identify heterogeneous features in existing technologies is solved, achieving efficient and accurate concrete modeling and simulation, which is suitable for batch analysis of specimens of arbitrary shapes.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- TAIYUAN UNIVERSITY OF TECHNOLOGY
- Filing Date
- 2025-06-27
- Publication Date
- 2026-06-23
AI Technical Summary
Existing numerical modeling methods for concrete cannot automatically identify and simulate heterogeneous features, have high modeling complexity, limited applicability, and are difficult to achieve efficient batch simulation analysis.
An automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh is adopted. By dividing the concrete specimen into thin-sheet unit cell meshes, generating random nodes and connecting them in six directions, a topologically disordered three-dimensional lattice element network is constructed. Combined with two-dimensional numerical images, a material property state matrix is generated to achieve automatic modeling.
It achieves efficient simulation of the heterogeneous characteristics of concrete, improves modeling and calculation efficiency, is applicable to efficient batch simulation analysis of specimens of arbitrary shapes, and improves the accuracy and universality of simulation results.
Smart Images
Figure CN120764019B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of concrete modeling technology, and in particular to an automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh. Background Technology
[0002] With the development of computer technology, numerical simulation methods have been widely applied to the analysis of concrete damage and failure processes. Given the significant heterogeneity of concrete, a reasonable modeling approach is crucial for effectively analyzing the damage and failure process. Coarse macroscopic models often fail to accurately represent the heterogeneous characteristics of concrete, leading to simulation results that significantly deviate from reality. Conversely, fine and microscopic models that can accurately represent the heterogeneous distribution of concrete often imply a significant increase in modeling complexity, and the simulation results are influenced by individual minute details, failing to represent the overall general situation of the simulated object. Therefore, balancing the accurate representation of the overall heterogeneity of concrete with modeling complexity has become an urgent technical problem to be solved in order to better simulate the concrete damage and failure process.
[0003] Chinese patent application CN202311061350.4 discloses a modeling method for numerical simulation of concrete cracking characteristics in tunnels. This method establishes a tunnel geometric model, determines the location of the cracked area within the model, and obtains the corresponding geometric set. The tunnel geometric model is then meshed to obtain mesh models for each region. Based on the geometric set corresponding to the cracked area, a set of concrete damage elements in the mesh model is determined. Based on the tunnel geometric model, the tunnel mesh model, and the set of concrete damage elements, the running parameters of the element deletion script in the numerical simulation software are set, and the corresponding elements are deleted after reaching a set mechanical level. However, this prior art has the following shortcomings: it requires pre-determining the location of the cracked area manually, and cannot automatically identify and simulate the heterogeneous distribution of concrete; the element deletion method is relatively crude, which may affect the stability and accuracy of numerical calculations; it is mainly applicable to specific structural forms such as tunnels, and its applicability to modeling concrete specimens of arbitrary shapes is limited; the modeling process requires manual setting of many parameters, resulting in low automation and difficulty in achieving efficient batch simulation analysis. Summary of the Invention
[0004] In view of this, and to address the shortcomings of existing numerical modeling methods for concrete, this invention proposes an automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh. This method can establish a macroscopic concrete model that quantitatively simulates the heterogeneous characteristics of concrete from the perspective of macroscopic topological disorder. At the same time, it can also meet the requirements of different structural features of specimens and perform efficient batch simulation analysis of the damage and failure development process of concrete specimens of arbitrary shapes.
[0005] The technical solution of this invention is implemented as follows: This invention provides an automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh, comprising the following steps:
[0006] S1. Divide the concrete specimen into several thin slices, and establish a unit cell grid in each thin slice to form a three-dimensional unit cell network structure.
[0007] S2. Based on the set randomness parameters, generate random nodes in each unit cell and perform minimum distance checks on the distance between nodes to form a set of random nodes;
[0008] S3. Connect the random nodes in each unit cell with the nodes in the adjacent unit cells in six directions to construct a topologically disordered three-dimensional lattice cell network.
[0009] S4. Generate a material property state matrix based on the two-dimensional numerical images of each thin sheet, assign material property values to the three-dimensional lattice unit network, and complete the automatic modeling of the concrete model.
[0010] Based on the above technical solution, preferably, in step S1, the unit length x of the unit cell grid in the X, Y, and Z directions is selected according to the structural characteristics of the concrete specimen in the X, Y, and Z directions. e y e z e The unit length of the unit cell mesh is chosen to satisfy the following conditions:
[0011] z between each layer e x and different plane positions within the same layer e y e They may not be exactly the same, but the z-shapes of different regions within the same layer... e The x values must be the same across all layers in the same region. e y e They must be the same.
[0012] Based on the above technical solutions, preferably, step S2 specifically includes:
[0013] S21. Generate a sub-cell within each unit cell, establish a local rectangular coordinate system for each sub-cell, set the randomness, and generate the local coordinates of random nodes within the sub-cell using a random function.
[0014] S22. Based on the sequential index values of each subcell within the current thin layer, transform the coordinates of random nodes in the local coordinate system into coordinate values in the global coordinate system.
[0015] S23. During the node generation process, the minimum distance between the newly generated node and the old node in the X, Y, and Z directions is checked. If the minimum distance requirement is not met, the node coordinates are regenerated.
[0016] S24. Store the coordinates of each generated random node in a coordinate matrix according to the order in which the nodes are generated, and finally form a node set containing all random nodes.
[0017] Based on the above technical solutions, the preferred method for calculating the nodal coordinate transformation between the local coordinate system and the global coordinate system is as follows:
[0018]
[0019]
[0020] In the formula, x e y e z e ∑y represents the unit length of the unit cell mesh in the X, Y, and Z directions, respectively; i is the row index, j is the column index, and k is the lamellae index; x0(i,j,k), y0(i,j,k), and z0(i,j,k) are the X, Y, and Z coordinates of the random node corresponding to the current index (i,j,k) in the local coordinate system within the sub-unit cell, respectively; x(i,j,k), y(i,j,k), and z(i,j,k) are the X, Y, and Z coordinates of the random node corresponding to the current index (i,j,k) in the global coordinate system, respectively; e (i) represents the cumulative length of the unit cell in the Y direction from row 1 to row i, ∑x e (j) represents the cumulative length of the unit cell in the X direction from column 1 to column j, ∑z e (k) is the cumulative length of the unit cell in the Z direction from the 1st layer to the kth layer; rand(1) generates a random number between 0 and 1, and random is the degree of randomness.
[0021] Based on the above technical solutions, preferably, in step S23, the minimum distance check includes: based on the divided three-dimensional cubic unit cell network, performing a distance check between the new node generated in each unit cell and the old node in the neighboring area of the unit cell, wherein the neighboring area consists of all unit cells that share edges or corners with the unit cell in the X, Y, and Z directions.
[0022] Based on the above technical solutions, preferably, step S3 specifically includes:
[0023] Each node in a single cell is connected to its adjacent nodes in the three main directions of the three principal planes to form a six-directional lattice unit local network of a single node, forming a topologically disordered three-dimensional lattice unit network. The coordinates of the midpoint of each unit in each direction are stored in the matrix according to the connection order.
[0024] The three main planes include the XY plane, XZ plane, and YZ plane; the three main directions include horizontal, vertical, and diagonal; and the six-directional lattice unit local area network of a single node includes six directional units: XX, YY, ZZ, XY, XZ, and YZ.
[0025] Based on the above technical solutions, preferably, storing the midpoint coordinates of each unit in each direction in the matrix according to the connection order specifically includes: obtaining the index of the other end node according to the current node index and its positional relationship with the other end node; obtaining the node coordinates from the random node coordinate matrix according to the index; obtaining the unit midpoint coordinates by taking the average of the two node coordinates; and storing the midpoint coordinates of the units in each direction in the matrix according to the unit generation order.
[0026] Based on the above technical solutions, preferably, in step S4, generating the material property state matrix based on the two-dimensional numerical images of each thin sheet specifically includes: obtaining a two-dimensional numerical image representing the structural characteristics of each thin sheet, generating the corresponding node network material property state matrix based on the two-dimensional numerical image, and then generating the three-dimensional lattice unit network material property state matrix.
[0027] Based on the above technical solutions, the preferred method for generating the node network material property state matrix includes:
[0028] RGB image processing is performed on two-dimensional numerical images to make the color thresholds distinct to correspond to different materials;
[0029] Set the image pixel size to equal the number of unit cells in the current layer, extract all pixel RGB value arrays, and assemble them into a matrix according to the unit cell generation order index to represent the RGB value corresponding to each unit cell in the current thin layer's unit cell network;
[0030] The material type corresponding to each node is determined based on the RGB values, and a node network material property state matrix is generated.
[0031] Based on the above technical solutions, the preferred method for generating the material property state matrix of a three-dimensional lattice element network includes:
[0032] For each node in the node network material property state matrix, the material property value of its neighboring nodes in the node network state matrix is searched according to the index value corresponding to the node's positional relationship. Based on the material property state and positional relationship between the current node and its neighboring nodes, the material property category corresponding to the lattice unit formed by the connection is determined and stored in matrix form.
[0033] The automatic modeling method for heterogeneous concrete based on topologically disordered six-directional connected meshes of the present invention has the following advantages over existing technologies:
[0034] This invention, based on discrete element method (DEM) simulation technology and concrete materials science, considers the modeling and computational efficiency of concrete heterogeneity, the universality of simulation results in practical applications, the requirements for irregularly shaped specimens, and the requirements for composite material specimens (such as reinforced concrete). It designs an integrated macroscopic heterogeneous concrete modeling method with arbitrary shapes. This method enables the automatic generation of macroscopic heterogeneous concrete models, using topological disorder determined by randomness to simulate the internal heterogeneity of macroscopic concrete while improving modeling and computational efficiency. This method can establish more realistic and efficient macroscopic computational models and also facilitates batch numerical simulations. The macroscopic concrete models established using this method are very suitable for simulating the mechanical response of large concrete specimens. Attached Figure Description
[0035] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0036] Figure 1 This is a flowchart of the automatic modeling method for heterogeneous concrete based on topologically disordered six-way connected meshes according to the present invention;
[0037] Figure 2 A schematic diagram of establishing a unit cell grid for the plate-shaped concrete specimen of the present invention;
[0038] Figure 3 A schematic diagram of the relevant parameters for defining the randomness of this invention;
[0039] Figure 4 This is a flowchart of the minimum distance inspection process of the present invention;
[0040] Figure 5 This is a schematic diagram of the random node set and its index according to the present invention;
[0041] Figure 6 This is a schematic diagram illustrating the construction of the topologically disordered three-dimensional lattice cell network of the present invention;
[0042] Figure 7 A macroscopic lattice model of concrete constructed for the technical solution of this invention. Detailed Implementation
[0043] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.
[0044] like Figure 1 As shown, this invention provides an automatic modeling method for heterogeneous concrete based on a topologically disordered six-directional connected mesh, comprising the following steps:
[0045] S1. Divide the concrete specimen into several thin slices, and establish a unit cell grid in each thin slice to form a three-dimensional unit cell network structure.
[0046] S2. Based on the set randomness parameters, generate random nodes in each unit cell and perform minimum distance checks on the distance between nodes to form a set of random nodes;
[0047] S3. Connect the random nodes in each unit cell with the nodes in the adjacent unit cells in six directions to construct a topologically disordered three-dimensional lattice cell network.
[0048] S4. Generate a material property state matrix based on the two-dimensional numerical images of each thin sheet, assign material property values to the three-dimensional lattice unit network, and complete the automatic modeling of the concrete model.
[0049] This invention achieves an automated modeling method for quantitatively simulating the heterogeneous characteristics of concrete from macroscopic topological disorder by dividing concrete specimens into thin-layer unit cell meshes, generating random nodes within each unit cell based on a randomness parameter, and connecting these random nodes along six principal directions to form a topologically disordered three-dimensional lattice element network. Then, by automatically identifying and assigning material properties to elements through two-dimensional numerical images of each layer of the specimen, the method realizes this automated modeling method. The entire modeling process utilizes the sequential indexing of node and element generation to achieve efficient geometric positioning and material property assignment, avoiding the complex search and judgment processes of traditional methods. Simultaneously, the topological disorder controlled by the randomness parameter can simulate the internal heterogeneity of concrete while avoiding the influence of individual characteristics in the microscopic model on the overall result. This significantly improves modeling and computational efficiency while ensuring modeling accuracy, enabling the automatic generation of macroscopic concrete models. It is particularly suitable for efficient batch simulation analysis of the damage and failure development process of concrete specimens of arbitrary shapes, providing an effective technical solution for simulating the mechanical response of large concrete specimens.
[0050] Furthermore, in step S1, the unit length z of the unit cell mesh in the Z direction is selected based on the structural characteristics of the specimen in the Z direction. e The specimen was divided into h thin layers along the Z direction. Based on the structural characteristics of the specimen in the X and Y directions, the unit cell mesh length x in the X and Y directions was selected.e y e Within each thin slice, a three-dimensional cubic unit cell network of m×n is divided according to the unit cell grid length of that layer, such as... Figure 2 As shown, where Figure 2 The left and middle images show a slab-shaped concrete specimen and its z-shape. e The diagram on the right shows a detailed schematic of the unit cell mesh of the k0th thin film.
[0051] To accommodate potentially irregular specimen shapes, the unit length of the unit cell mesh is selected based on the following condition: the z-axis distance between each layer... e x and different plane positions within the same layer e y e They may not be exactly the same, but the z-shapes of different regions within the same layer... e The x values must be the same across all layers in the same region. e y e They must be identical, meaning the unit cell grids of each layer must correspond vertically. Specifically, for unit cells at index (i,j) within adjacent layers k0 and k1, it is allowed that... But the following conditions must be met:
[0052] Furthermore, step S2 specifically includes:
[0053] S21. Generate a sub-cell within each unit cell, establish a local Cartesian coordinate system for each sub-cell, set the randomness, and generate the local coordinates of random nodes within the sub-cell using a random function.
[0054] S22. Based on the sequential index values of each subcell within the current thin layer, transform the coordinates of random nodes in the local coordinate system into coordinate values in the global coordinate system.
[0055] Specifically, such as Figure 3 As shown, Figure 3 The diagram illustrates the relevant parameters for defining randomness. Randomness is the ratio of the subcell side length to the cell side length in the X, Y, and Z directions, representing the maximum range of random distribution of nodes within the cell. The method for calculating randomness is as follows:
[0056]
[0057] In the formula, random represents the degree of randomness, and L sub-cell,x L sub-cell,y L sub-cell,z These are the X, Y, and Z side lengths of the subunit, respectively, and L. cell,x L cell,y L cell,z These are the X, Y, and Z side lengths of the unit cell, respectively.
[0058] The method for generating random node coordinates is as follows: Establish a local Cartesian coordinate system O'x'y'z' for each sub-cell, such as... Figure 2 As shown in the right figure, under a specific degree of randomness, the coordinates of the random nodes in the subcell of the i-th row and j-th column of the k-th thin layer are generated by the rand function in MATLAB software. The coordinates of the random nodes in the local coordinate system are:
[0059]
[0060] Establish a global rectangular coordinate system Oxyz at the lower left corner of the specimen, as follows: Figure 2 As shown in the left figure, the coordinates of random nodes in the local coordinate system are transformed into coordinates in the global coordinate system based on their sequential index values within the current thin layer:
[0061]
[0062] In the formula, x e y e z e ∑y represents the unit length of the cell mesh in the X, Y, and Z directions, respectively; i is the row index, j is the column index, and k is the lamellae index; x0(i,j,k), y0(i,j,k), and z0(i,j,k) are the X, Y, and Z coordinates of the random node corresponding to the current index (i,j,k) in the local coordinate system within the sub-cell; x(i,j,k), y(i,j,k), and z(i,j,k) are the X, Y, and Z coordinates of the random node corresponding to the current index (i,j,k) in the global coordinate system; ∑y e (i) represents the cumulative length of the unit cell in the Y direction from row 1 to row i, ∑x e (j) represents the cumulative length of the unit cell in the X direction from column 1 to column j, ∑z e (k) is the cumulative length of the unit cell in the Z direction from the 1st layer to the kth layer; rand(1) generates a random number between 0 and 1, and random is the degree of randomness.
[0063] The coordinates of each random node are stored in a matrix according to the order in which the nodes are generated. Each matrix element is the three-dimensional coordinates (x, y, z) of a node.
[0064] S23. During the node generation process, the minimum distance between the newly generated node and the old node in the X, Y, and Z directions is checked. If the minimum distance requirement is not met, the node coordinates are regenerated.
[0065] For each new node with index (i,j,k) generated, the neighborhood index value is obtained based on the positional relationship, the coordinates of the old node in the location matrix are located, and a distance check is performed on the new and old nodes. If the minimum distance requirement is not met, the new node is regenerated; if it is met, the next new node is generated.
[0066] The minimum distance check includes: based on the divided 3D cubic unit cell network, performing a distance check between the new node generated in each unit cell and the old nodes in the neighboring unit cell, wherein the neighboring unit cell consists of all units that share edges or corners with the unit cell in the X, Y, and Z directions.
[0067] The minimum distance check method can determine the number and coordinate information of old neighboring nodes based on the node index value. If the new node does not meet the requirements, only the coordinates of the new node need to be adjusted, saving a lot of repetitive search and check work. For example, for the cell in the i-th row and j-th column of the k-th layer, the old nodes in its neighboring cells can be located based on the sequential index value (i,j,k) of the cell. Therefore, for each new node, the number of searches for minimum distance check is determined by its index value. Based on the generated sequential index to locate the neighboring node information, the maximum complexity of this algorithm is O(2^6), which avoids the old domain search with higher search cost (complexity O(i×j×k-1)), that is, searching and minimum distance checking all old nodes generated before the new node, and further avoids the global search with higher search and adjustment cost (complexity O(m×n×h-1)), that is, performing minimum distance check on all remaining nodes.
[0068] Specifically, such as Figure 4As shown, the minimum distance checking process in this embodiment adopts a three-level nested loop structure. The checking order is j=1:n (number of columns), i=1:m (number of rows), k=1:h (number of slice layers), processing each cell position one by one. For the current index (i,j,k) position, a new cell and the corresponding random node coordinates (x,y,z) are first generated. Next, it is determined whether the current cell is located at a boundary position, i.e., i=1 or m, j=1 or n, k=1 or h. Based on the determination result, the search range of adjacent cells is determined: if it is a boundary cell, the number of searches S is determined according to its specific position; if it is an internal cell, all 26 neighboring cells around it are searched (i.e., S=26). Then, the adjacent search loop is entered. The adjacent index is calculated one by one according to the correspondence between the cell position and the index. The corresponding node coordinates are obtained from the coordinate matrix. The distance between the newly generated node and each adjacent node is calculated, and it is determined whether the distance is greater than or equal to the set minimum distance. If yes, continue checking the next neighboring node; if no, regenerate the random node corresponding to the current index and restart the minimum distance check process for that position. Once the distance check of all neighboring nodes has passed, store the coordinates of the current random node in the coordinate matrix according to the index, and proceed to the processing of the next cell position, until the minimum distance check of all cells is completed.
[0069] S24. Store the coordinates of each generated random node in a coordinate matrix according to the order in which the nodes were generated, ultimately forming a node set and its coordinate matrix containing all random nodes (i.e., m×n×h points). The matrix index corresponds to the node position, such as... Figure 5 As shown.
[0070] Furthermore, step S3 specifically includes:
[0071] Each node in a single cell is connected to its adjacent nodes in the three principal directions on the three principal planes to form a six-directional lattice unit local network of a single node, thereby forming a topologically disordered three-dimensional lattice unit network. The coordinates of the midpoint of each unit in each direction are stored in the matrix according to the connection order.
[0072] The three main planes include the XY plane, XZ plane, and YZ plane; the three main directions include horizontal, vertical, and diagonal; and the six-directional lattice unit local area network of a single node includes six directional units: XX, YY, ZZ, XY, XZ, and YZ.
[0073] Specifically, the midpoint coordinates of each element in each direction are stored in the matrix according to the connection order. This includes: obtaining the index of the other end node based on the current node index and its positional relationship with the other end node; obtaining the node coordinates from the random node coordinate matrix based on the index; obtaining the midpoint coordinates of the element by taking the average of the two node coordinates; and storing the midpoint coordinates of the elements in each direction in the matrix according to the element generation order.
[0074] Following the node generation order from left to right and bottom to top, each node is connected to a maximum of six adjacent nodes to form a unit, including XX-direction units, XY-direction units, XZ-direction units, YY-direction units, YZ-direction units, and ZZ-direction units, forming a six-directional lattice unit local area network of nodes, such as... Figure 6 As shown in (a). Figure 6 (a) is a six-way lattice cell local area network with node index (i,j,k0); once all nodes are connected, a topologically disordered three-dimensional lattice cell network is formed, such as Figure 6 As shown in (b). Figure 6 (b) is a topologically disordered three-dimensional lattice cell network composed of multiple connected nodes.
[0075] Further, in step S4, generating the material property state matrix based on the two-dimensional numerical images of each thin sheet specifically includes: first obtaining a two-dimensional numerical image representing the structural characteristics of each thin sheet, generating the corresponding node network material property state matrix based on the two-dimensional numerical image, and then generating the three-dimensional lattice unit network material property state matrix.
[0076] The method for generating the material property state matrix of the node network includes: acquiring a two-dimensional numerical image that can intuitively represent the structural characteristics of each thin sheet layer through color; performing RGB image processing on the image to make the color thresholds distinct to correspond to different materials, even if each color represents only one material; setting the image pixel size to m×n, corresponding to the unit cell grid size of each layer, so that each pixel corresponds to one unit cell; and performing RGB image processing on the image to make the color thresholds distinct, ensuring that each color in the image corresponds to a different type of material.
[0077] Using MATLAB software, extract the RGB value arrays of all pixels and assemble them into a matrix according to the generation order of the unit cells. This matrix represents the RGB value of each unit cell in the current thin-layer unit cell network. Based on the RGB values, determine the material type of each node and generate the node network material property state matrix.
[0078] The method for generating a three-dimensional lattice element network material property state matrix includes: for each node, searching for the material property values of its neighboring nodes in the node network state matrix based on the index value corresponding to the positional relationship; determining the material property category of the lattice element formed by the connection based on the material property state and positional relationship of the current node and its neighboring nodes, and storing it in the form of a state matrix.
[0079] Specifically, the material properties of an element are determined by the material properties of its two endpoints. Based on the index relationship between the two endpoints of elements in different directions (e.g., the endpoint indices of an element in direction XX are (i,j,k) and (i+1,j,k) respectively), the material types corresponding to the two endpoints are located in the node network material property state matrix, and the material property category of the connecting element is determined. If the two nodes have the same material type, the element material property is also that material type; if the two nodes have different material types, the element material property is the interface between them. The material property categories of all elements in the six directions are determined sequentially according to the element connection order, and the determination results are expressed in matrix form (M). xx M xy M xz M yy M yz M zz )storage.
[0080] The implementation method for assigning material properties to elements is as follows: the sequential index values of the material property state matrix of the three-dimensional lattice element network strictly correspond to the element indexes of the coordinate matrix of the element midpoint. The element can be directly located according to the matrix element index values. Then, the matrix index is used to realize the algorithm to efficiently assign material properties to the elements.
[0081] A macroscopic lattice model of heterogeneous concrete with a topologically disordered six-way interconnected network was established using an integrated modeling program, such as... Figure 7 As shown, the degree of topological disorder in the model is determined by the degree of randomness, where, Figure 7 The left-middle figure shows a lattice model with zero randomness. Figure 7 The right-middle figure shows a lattice model with a randomness of 0.5.
[0082] This invention provides an automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh. The method uses a topologically disordered structural model determined by randomness to simulate the internal heterogeneity of concrete. During the modeling process, efficient and automated construction is achieved by generating sequential indexes based on nodes and elements. This solves the problem in the original macroscopic modeling technology of concrete specimens that it is difficult to simultaneously ensure the reproduction of concrete heterogeneity and modeling efficiency.
[0083] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. An automatic modeling method for heterogeneous concrete based on topologically disordered six-directional connected meshes, characterized in that, Includes the following steps: S1. Divide the concrete specimen into several thin slices, and establish a unit cell grid in each thin slice to form a three-dimensional unit cell network structure. S2. Based on the set randomness parameters, generate random nodes in each unit cell and perform minimum distance checks on the distance between nodes to form a set of random nodes; S3. Connect the random nodes in each unit cell with the nodes in the adjacent unit cells in six directions to construct a topologically disordered three-dimensional lattice cell network. S4. Generate a material property state matrix based on the two-dimensional numerical images of each thin sheet, assign material property values to the three-dimensional lattice unit network, and complete the automatic modeling of the concrete model. Step S3 specifically includes: Each node in a single cell is connected to its adjacent nodes in the three main directions of the three principal planes to form a six-directional lattice unit local network of a single node, forming a topologically disordered three-dimensional lattice unit network. The coordinates of the midpoint of each unit in each direction are stored in the matrix according to the connection order. The three main planes include the XY plane, XZ plane, and YZ plane; the three main directions include horizontal, vertical, and diagonal; and the six-directional lattice unit local area network of a single node includes six directional units: XX, YY, ZZ, XY, XZ, and YZ.
2. The automatic modeling method for heterogeneous concrete based on topologically disordered six-way connected mesh as described in claim 1, characterized in that, In step S1, the unit length of the unit cell grid in the X, Y, and Z directions is selected according to the structural characteristics of the concrete specimen in the X, Y, and Z directions. , , The unit length of the unit cell mesh is chosen to satisfy the following conditions: Between each floor and different plane positions within the same floor , They are not entirely the same, but different areas within the same floor Must be identical, within the same area across all floors. , They must be the same.
3. The automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh as described in claim 1, characterized in that, In step S1, the unit length of the unit cell grid in the X, Y, and Z directions is selected according to the structural characteristics of the concrete specimen in the X, Y, and Z directions. , , The unit length of the unit cell mesh is chosen to satisfy the following conditions: the interlayer... and different plane positions within the same floor , They are the same to each other, but different areas within the same floor. Must be identical, within the same area across all floors. , They must be the same.
4. The automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh as described in claim 1, characterized in that, Step S2 specifically includes: S21. Generate a sub-cell within each unit cell, establish a local rectangular coordinate system for each sub-cell, set the randomness, and generate the local coordinates of random nodes within the sub-cell using a random function. S22. Based on the sequential index values of each subcell within the current thin layer, transform the coordinates of random nodes in the local coordinate system into coordinate values in the global coordinate system. S23. During the node generation process, the minimum distance between the newly generated node and the old node in the X, Y, and Z directions is checked. If the minimum distance requirement is not met, the node coordinates are regenerated. S24. Store the coordinates of each generated random node in a coordinate matrix according to the order in which the nodes are generated, and finally form a node set containing all random nodes.
5. The automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh as described in claim 4, characterized in that, The method for calculating the nodal coordinate transformation between the local coordinate system and the global coordinate system is as follows: ， ， In the formula, , , , i, j, and k represent the unit length of the unit cell mesh in the X, Y, and Z directions, respectively; i is the row index; j is the column index; and k is the lamellar layer index. , , Each is the current index The corresponding random node's X, Y, and Z coordinates in the local coordinate system within the sub-unit cell. , , Each is the current index The corresponding random node's X, Y, and Z coordinates in the global coordinate system; This is the cumulative sum of the Y-direction lengths of the unit cells from row 1 to row i. This is the sum of the lengths of the unit cells in the X direction from column 1 to column j. This is the sum of the unit cell lengths in the Z direction from layer 1 to layer k; To generate a random number between 0 and 1, For randomness.
6. The automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh as described in claim 4, characterized in that, In step S23, the minimum distance check includes: based on the divided three-dimensional cubic unit cell network, performing a distance check between the new node generated in each unit cell and the old node in the neighboring unit cell, wherein the neighboring unit cell consists of all unit cells that share edges or corners with the unit cell in the X, Y, and Z directions.
7. The automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh as described in claim 1, characterized in that, The midpoint coordinates of each element in each direction are stored in the matrix according to the connection order. Specifically, this includes: obtaining the index of the other end node based on the current node index and its positional relationship with the other end node; obtaining the node coordinates from the random node coordinate matrix based on the index; obtaining the midpoint coordinates of the element by taking the average of the two node coordinates; and storing the midpoint coordinates of the elements in each direction in the matrix according to the element generation order.
8. The automatic modeling method for heterogeneous concrete based on topologically disordered six-way connected mesh as described in claim 1, characterized in that, In step S4, generating the material property state matrix based on the two-dimensional numerical images of each thin sheet specifically includes: obtaining a two-dimensional numerical image representing the structural characteristics of each thin sheet, generating the corresponding node network material property state matrix based on the two-dimensional numerical image, and then generating the three-dimensional lattice unit network material property state matrix.
9. The automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh as described in claim 8, characterized in that, The generation of the nodal network material property state matrix includes: RGB image processing is performed on two-dimensional numerical images to make the color thresholds distinct to correspond to different materials; Set the image pixel size to equal the number of unit cells in the current layer, extract all pixel RGB value arrays, and assemble them into a matrix according to the unit cell generation order index to represent the RGB value corresponding to each unit cell in the current thin layer's unit cell network; The material type corresponding to each node is determined based on the RGB values, and a node network material property state matrix is generated.
10. The automatic modeling method for heterogeneous concrete based on a topologically disordered six-way connected mesh as described in claim 9, characterized in that, Methods for generating the material property state matrix of a three-dimensional lattice element network include: For each node in the node network material property state matrix, search for the material property values of its neighboring nodes in the node network state matrix according to the index value corresponding to the node's positional relationship. Based on the material property state and positional relationship between the current node and its neighboring nodes, determine the material property category of the lattice unit formed by the connection and store it in matrix form.