A universal method for adaptive merging and reorganization of multi-lattice structures

Abstract

The present invention discloses a universal method for adaptive merging and reorganization of polylattice structures, wherein the object to be 3D printed is scanned to obtain point cloud data of the object, and then the three-dimensional structure of the object is reconstructed using the point cloud data; the internal area of ​​the three-dimensional structure is divided according to the physical field, and the lightweight support structures of different sizes that need to be filled in the divided area are determined, and the lightweight support structures of different sizes are spliced; the spliced ​​lightweight support structures are subjected to Boolean operations, and then the lightweight supports subjected to Boolean operations are reorganized through point cloud information and triangular facets, and the three-dimensional structure is merged with the lightweight support structure to obtain a structure that can be used for 3D printing. The method can fill lightweight support structures of different sizes in the internal area of ​​the model to effectively save printing materials, and can achieve the goal of using different lightweight support structures in different positions.

Description

Technical Field

[0001] The present invention belongs to the technical field of lightweight support structure design for digital models in additive manufacturing, and specifically relates to a universal method for adaptive merging and reorganization of multi-lattice structures. Background Art

[0002] The research on porous supports has attracted widespread attention from the scientific and industrial communities. It is of great significance in many engineering fields and plays a key role in heat transfer of foam metals, manufacturing of high permeability structures, and optimization of membrane distillation structures. For CAD models of different shapes, there is a significant correlation between the lightweight structure and the CAD model. The lightweight support structure can provide better support and controllable porosity density, providing more robust mechanical properties for the required design model. At the same time, for different CAD models, it is necessary to fill the lightweight support structure when printing with a 3D printer, and there are multiple options for filling according to different physical properties.

[0003] Qureshi et al. showed that TPMS-based phase change materials have good application potential in latent heat energy storage systems. The results showed that when a TPMS-type unit structure was used, the thermal conductivity was significantly improved, with the main influencing factors being the unit lattice type, structure and porosity.

[0004] Rueger et al. proposed a three-dimensional lattice structure composed of rotating cubes with flexible links at the corners. It has an anisotropic negative Poisson's ratio, and size effects appear when twisting and bending. These effects are consistent with Cosserat elasticity. At the same time, Cosserat elastic structures show great sensitivity to strain gradients.

[0005] Ozdemir et al. conducted Hopkinson pressure bar (HPB) tests and showed that cubic and rhombus lattice structures can disperse the impact load in time and reduce the peak impact stress, identifying a significant rate dependence of the load-deformation characteristics.

[0006] Liang et al. fabricated porous Fe-based metallic glass (MG)-based composites with rhombic dodecahedron microstructures by selective laser melting (SLM), which were then applied to catalytic degradation of cibacron brilliant red 3B-A (BR3B-A) dye solution. The SLM-fabricated porous Fe-based MG-based composites exhibited excellent overall catalytic ability, high reaction rate constants, and low activation energy.

[0007] The technique of incorporating designs based on various lattice structures into numerical algorithms is heuristic and has attracted widespread attention to different categories of integrated optimization algorithms. Garner et al. proposed a multiscale method based on microstructural topology optimization. Their method includes functionally gradient materials and multiscale structures, and is extended to comprehensively consider the characteristics of length gradient and isotropy. In a series of extensive studies, they studied the maximum bulk modulus and showed that the bulk modulus of a single cell reached the theoretical boundary predicted by the Hashin-Shtrikman model, which means that the optimization of compatibility will not affect the performance of a single cell.

[0008] Li et al. proposed a method that combines periodic minimal surface-based algorithms and random strategies to control the average pore size in hexahedral meshes. By mapping the pores of periodic minimal surfaces into hexahedral meshes, porous scaffolds have greater degrees of freedom and flexibility, and then additive manufacturing technology is used to manufacture porous scaffolds.

[0009] Gao et al. proposed a parallel topology optimization algorithm for multi-scale design of honeycomb composite structures, adopted an improved SIMP method to achieve concurrent topology design, and evaluated the macroscopic effective performance of the microstructure using an energy-based homogenization method (EBHM).

[0010] Yu et al. proposed an efficient and systematic multi-scale topology optimization method based on PFM. This method uses a composite design algorithm of multi-region microstructures to optimize the objective function under the condition of satisfying various constraints. In order to combine the computational characteristics of lattice structure with structural size and reduce the design complexity of porous scaffolds in the design process, an effective numerical method was constructed to address the limitations of the theoretical design space.

[0011] Since the digital model in the physical scene consists of a large amount of data consisting of triangles connected by vertices, files with unified specifications should be used for complex operating conditions. In the past decade, how to combine the multi-level design of digital models with limited storage space and simple computing operations has been a hot research direction.

[0012] Navangul et al. developed an algorithm for selectively modifying STL files to achieve the required tolerances specified on part features. In their design process, defective parts were eliminated and material expenditures in the additive manufacturing framework were reduced. Stroud et al. proposed an extended STL format to overcome the defects of STL file format redundancy and rough approximation methods. This results in a smaller final model and allows faster geometry processing, and is able to reconstruct the exact object geometry, thereby avoiding volume distortion, to consistently control slicing and layer approximation, and to provide a mechanism that allows additional manufacturing information.

[0013] Zha et al. proposed a new STL modification algorithm, called surface-based modification algorithm (SMA). Based on the chord error, contour error, and cylindricity error of a single surface of a part, SMA selectively and adaptively modifies certain STL surfaces in an STL model to reduce the cusp height of the model without increasing the file size unnecessarily.

[0014] Although the above methods have made certain improvements in the storage space and processing speed of STL files, there are still some shortcomings. When the current 3D printer prints a given STL file, it fills the interior of the three-dimensional structure with lightweight support structures based on the existing algorithm. Usually, the relatively basic units are filled indiscriminately. The filled lightweight support structures are of the same size and single, and the purpose is only to support the STL. Such an operation will produce the following three disadvantages: 1) The printed model does not have a good physical property, such as stiffness or bending resistance requirements; 2) The filled lightweight support structures are of the same size, which will increase the cost of printing materials; 3) The filled lightweight support structure is single, and the goal of filling different lightweight support structures in different areas of the model cannot be achieved. Summary of the invention

[0015] In view of the problems in the prior art, the purpose of the present invention is to provide a universal method for adaptive merging and reorganization of polylattice structures, which can fill the internal area of ​​the model with lightweight support structures of different sizes to effectively save printing materials and achieve the goal of using different lightweight support structures in different positions.

[0016] A universal method for adaptive merging and reorganization of multi-lattice structures includes the following steps:

[0017] 1) Scan the object to be 3D printed to obtain the point cloud data of the object, and then use the point cloud data to reconstruct the three-dimensional structure of the object;

[0018] 2) Divide the internal area of ​​the three-dimensional structure according to the physical field, determine the lightweight support structures of different sizes that need to be filled in the divided areas, and splice the lightweight support structures of different sizes;

[0019] 3) Performing Boolean operations on the assembled lightweight support structure, and then placing the lightweight support structure after the Boolean operation into the interior of the three-dimensional structure; merging the three-dimensional structure with the lightweight support structure to obtain a structure that can be used for 3D printing.

[0020] Furthermore, a 3D scanner is used to scan the 3D printed object.

[0021] Furthermore, the physical field is a stress field, a signed distance field or a temperature field.

[0022] Furthermore, lightweight support structures of different sizes are spliced ​​according to the interpolation method.

[0023] Furthermore, the interpolation formula used by the interpolation method is:

[0024] φ scaffold (x,y,z)=f(x,y,z)φ1(x,y,z)+(1-f(x,y,z))φ2(x,y,z)

[0025] Where f represents a continuous weighting function varying between 0 and 1, φ1 and φ2 represent two lightweight support structures of different sizes or types;

[0026] f(x,y,z)=1 / (1+expkG(x,y,z))

[0027] Where G is the function describing the transition position, and k is the parameter used to control the transition gradient.

[0028] Furthermore, the three-dimensional structure is combined with the lightweight support structure to obtain a structure that can be used for 3D printing, including the following steps:

[0029] The point cloud coordinate matrix of the three-dimensional structure is combined and reorganized with the point cloud coordinate matrix of the lightweight support structure to obtain a point cloud combined coordinate matrix;

[0030] The triangulation matrix of the three-dimensional structure and the triangulation matrix of the lightweight support structure are reorganized to obtain a combined triangulation matrix;

[0031] According to the point cloud combined coordinate matrix V and the triangulation matrix T, the three-dimensional structure and the lightweight support structure are merged to obtain a three-dimensional structure that can be used for 3D printing.

[0032] Furthermore, the point cloud combined coordinate matrix V is as follows:

[0033] V={v1,v2,...,v p1 ,v p1+1 ,...,v p1+p2},

[0034]

[0035] Where p1 is the number of points on the three-dimensional structure, and p2 is the number of points on the lightweight support structure;

[0036] The combined triangulation matrix T is as follows:

[0037]

[0038]

[0039] Where q1 is the number of triangular meshes in the three-dimensional structure, and q2 is the number of triangular meshes in the lightweight support structure.

[0040] Furthermore, the point cloud coordinate matrix V1 of the three-dimensional structure is as follows:

[0041]

[0042] The triangulation matrix T1 of the three-dimensional structure is as follows:

[0043]

[0044] The point cloud coordinate matrix V2 of the lightweight support structure is as follows:

[0045]

[0046] The triangulation matrix T2 of the lightweight support structure is as follows:

[0047]

[0048] In the formula, v i,j A 3×1 vector representing the coordinates of a point; Δ i,j A 3-by-1 vector representing the triangulated surface.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The present invention can effectively save printing materials by filling the internal area of ​​the model with lightweight support structures of different sizes. At the same time, by using different lightweight support structures in different areas, a smooth transition between different lightweight support structures can be achieved without any breakage. The present invention can construct a porous support based on any lattice structure and design a lightweight support structure for any model, and at the same time, the printed three-dimensional model satisfies the relevant physical properties. The present invention is independent of the complex shape of the target model, has low computational complexity, and has a small memory for storing surface mesh reconstruction. The designed multi-scale three-dimensional structure maintains a smooth surface and can be directly applied to additive manufacturing. The present invention can simply generate large-scale three-dimensional porous bodies, which can be used for the design and filling of lightweight support structures in mechanical processing processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a schematic diagram of merging two Schwarz P surfaces of different sizes in 2D space. (a) is the result before merging; (b) is the symbol The result of linear interpolation and merging of the area between the and symbols.

[0052] Figure 2 It is a schematic diagram of merging two Schwarz P surfaces of different sizes in 3D space. Among them, (a) is the result of discretizing the pyramid part of the larger Schwarz P surface in 3D space; (b) is the enlarged view of the interpolation area; (c) is the result after interpolation; (d) is the result of modifying another direction part; (e) is the result of modifying the other two directions parts.

[0053] Figure 3 It is a schematic diagram of two basic tetrahedral structures of different sizes being merged into a target structure. Among them, (a) is the two basic tetrahedral structures before merging; (b) is the result after merging the two basic tetrahedral structures; (c) is the 3D model printed using a 3D printer.

[0054] Figure 4 This is a demonstration diagram of the filling design of a lightweight porous loose support structure for a hemispherical shell. Among them, (a) is the first step, which is to divide the multi-sized Schwarz P structure into units and linearly interpolate the corresponding interfaces under the guidance of the signed distance field; (b) is the second step, which is to perform Boolean operations on the designed lightweight support structure, eliminate points that exceed the region boundary, and thicken the spherical shell; (c) is a summary of the two processes of Figure (a) and Figure (b), that is, the result of filling the hemispherical shell with a lightweight support structure from different perspectives. DETAILED DESCRIPTION

[0055] The method of the present invention is described in detail below with reference to the accompanying drawings.

[0056] Aiming at the problem that the physical properties of the model in the prior art are poor and the size of the filled lightweight support structure is the same and single, the present invention provides a universal method for adaptive merging and reorganization of multiple lattice structures, which aims to design a lightweight bracket based on an arbitrary lattice structure according to the existing model. First, the physical field in the actual model is obtained by external input or solving partial differential equations; then, a plurality of lightweight support structures are merged, and they can be spliced ​​regardless of whether the size, porosity or unit type of the lightweight support structure is the same. Finally, two different types of structures are merged, and the merged model can be directly used for 3D printing. The method directly merges from the point cloud, thereby reducing the amount of information storage. The method can be used to connect two lattice structures with different unit sizes or types in the radial direction, wherein the transition region is obtained by smooth interpolation between the two lattice structures with different unit sizes or types. Therefore, based on this method, large-scale three-dimensional porous bodies can be simply generated, which can be used for the design and filling of lightweight support structures in the machining process.

[0057] The present invention provides a universal method for adaptive merging and reorganizing multiple lattice structures. The method is universal and can design a lightweight bracket based on any lattice structure and splice multiple lightweight support structures. The surface of the prepared multi-scale composite bracket is smooth and free of fractures, and can be directly applied to additive manufacturing. The method comprises the following steps:

[0058] 1) Use a 3D scanner to perform a 360-degree scan of any object to be 3D printed in reality. After scanning, the point cloud data of the object can be obtained, and then the three-dimensional structure of the object can be reconstructed through software and using the point cloud data.

[0059] 2) For any physical field (such as stress field, signed distance field or temperature field, etc.), the internal area of ​​the three-dimensional structure is divided, and the lightweight support structures of different sizes that need to be filled in the divided area are determined, and the lightweight support structures of different sizes are spliced ​​according to the interpolation method.

[0060] Specifically, for the signed distance field, a three-dimensional structure is taken as a three-dimensional hemispherical shell as an example. Figure 4 As shown in (a), the internal area of ​​the three-dimensional hemispherical shell is divided according to the size of the signed distance field. According to the size of the signed distance of the divided area, the lightweight support structures of different sizes required to fill different areas are determined, and the lightweight support structures of different sizes are spliced ​​according to the interpolation method. The specific splicing process is as follows: Figure 1 , the operation of merging two Schwarz P surfaces of different sizes in two-dimensional space is performed in the same way in three-dimensional space, as Figure 2 shown.

[0061] Given two Schwarz P surfaces, let the two-dimensional space Ω=(0,L x )×(0,L y ), hL x / N x -1)=L y / N y -1 is a uniform grid size, where L x is the length in the x direction, L y is the length in the y direction, N x is the number of grids in the x direction, N y is the number of grids in the y direction, N x and N y is a positive integer. i,j is an approximation of ψ((i-1)h,(j-1h). ψ i,j is the corresponding value in the grid, i is the grid number in the x direction, h is the grid size, and j is the grid number in the y direction.

[0062] like Figure 1 As shown, using the symbol The points are placed on the diagonal line of the lower quarter of the larger Schwarz P surface, and the symbol · represents the average of the right end of the larger Schwarz P surface and the left end of the smaller Schwarz P surface. Then, the point values ​​of the larger Schwarz P surface are evenly discretized (with the symbol The area between the point value of the Schwarz P surface (denoted by the symbol ·) and the smaller size is linearly interpolated to obtain the interpolation result of the lower quarter of the larger Schwarz P surface. The corresponding interpolation empirical formula is as follows:

[0063] φ scaffold (x,y,z)=f(x,y,z)φ1(x,y,z)+(1-f(x,y,z))φ2(x,y,z)

[0064] In the formula, f represents a continuous weighting function varying between 0 and 1, φ1 and φ2 represent two lightweight support structures of different sizes or types, and the sigmoid function is used as the weighting function in the above formula. The specific sigmoid function is as follows:

[0065] f(x,y,z)=1 / (1+expkG(x,y,z))

[0066] Where G is the function describing the transition position, and k is the parameter used to control the transition gradient.

[0067] When connecting two lightweight support structures of different sizes or types, the above interpolation empirical formula can be used to change the gradient type between different structures. For the corresponding physical properties of each lattice, the corresponding gradient type in the interpolation empirical formula can be changed, and then mapped to the calculation space to achieve different physical properties for lightweight support structures of different sizes or types.

[0068] 3) Perform Boolean operations on the assembled lightweight support structure to eliminate points that exceed the region boundary, such as Figure 4 As shown in (b), the lightweight support structure after Boolean operation is then placed inside the three-dimensional structure (i.e., the three-dimensional hemispherical shell).

[0069] 4) After the lightweight support structure is placed inside the hemispherical shell, the structure cannot be used directly for printing. Next, the point cloud between the three-dimensional structure (i.e., the three-dimensional hemispherical shell) and the lightweight support structure needs to be reorganized, that is, the point cloud is connected according to the connection relationship to form a triangular mesh, and each triangular mesh is connected according to the connection relationship, so as to merge the three-dimensional structure (i.e., the three-dimensional hemispherical shell) and the lightweight support structure to obtain a merged structure, which can be directly used for 3D printing.

[0070] The specific process of merging the three-dimensional structure (i.e., the three-dimensional hemispherical shell) with the lightweight support structure to obtain the merged structure is as follows:

[0071] The point cloud and triangulation matrix between the three-dimensional structure (i.e., the three-dimensional hemispherical shell) and the lightweight support structure are reorganized. The specific process is as follows:

[0072] Assume that V1 and T1 represent the point cloud coordinate matrix and triangulation matrix of the three-dimensional hemispherical shell, and V2 and T2 represent the point cloud coordinate matrix and triangulation matrix of the lightweight support structure, specifically:

[0073]

[0074]

[0075] In the formula, v i,j 3×1 vector representing the coordinates of a point, p1 is the number of points on the hemispherical shell, p2 is the number of points in the lightweight support structure, q1 is the number of triangular meshes in the hemispherical shell, and q2 is the number of triangular meshes in the lightweight support structure. i,j A 3-by-1 vector representing the triangulated surface.

[0076] In order to achieve the purpose of merging the hemispherical shell and the lightweight support structure, the point cloud coordinate matrix V1 of the hemispherical shell and the point cloud coordinate matrix V2 of the lightweight support structure are first merged and reorganized according to certain rules to obtain the point cloud combined coordinate matrix V:

[0077]

[0078]

[0079] Then define Δ ′ ={p1,p1,p1} T , where p1 is the number of points on the hemispherical shell. By reorganizing T1 and T2 in a certain rule and using Δ ′ Replan the connection relationship of T2 and obtain the combined triangulation matrix T as follows:

[0080]

[0081]

[0082] According to the above method, after the point cloud coordinate matrix and triangulation matrix of the hemispherical shell and the lightweight support structure are merged and reorganized according to certain rules, the hemispherical shell and the lightweight support structure can be merged, that is, Figure 4 As shown in (c), the merged three-dimensional structure can be directly used for 3D printing.

[0083] The universal method for adaptive merging and reorganization of polycrystalline structures of the present invention has the following advantages: 1) The method can select a lightweight support structure with corresponding performance to fill the CAD model according to the desired physical properties (such as bending resistance, stiffness and Poisson's ratio, etc.); 2) The method can fill the internal area of ​​the model with lightweight support structures of different sizes to effectively save printing materials. For example, the model was originally filled with rhombuses with a side length of 1 cm. The method can now use rhombuses with a side length of 2 cm to fill the inside of the model, thereby saving a certain proportion of printing materials. 3) The method can achieve the goal of using different lightweight support structures in different positions. For example, hexagonal units are used on the left side of the model, and square units are used on the right side of the model. The method can achieve the transition from hexagon to square, and can ensure that there is a smooth surface without fractures when transitioning from hexagon to square.

Claims

1. A universal method for adaptive merging and reorganization of multi-lattice structures, characterized in that: The following steps are involved: 1) Scan the object to be 3D printed to obtain the point cloud data of the object, and then use the point cloud data to reconstruct the three-dimensional structure of the object; 2) Divide the internal area of ​​the three-dimensional structure according to the physical field, determine the lightweight support structures of different sizes that need to be filled in the divided areas, and splice the lightweight support structures of different sizes; 3) Perform Boolean operation on the spliced ​​lightweight support structure, and then reorganize the lightweight support structure after Boolean operation through point cloud information and triangular facets, merge the three-dimensional structure with the lightweight support structure, and obtain a structure that can be used for 3D printing; Lightweight support structures of different sizes are spliced ​​together according to the interpolation method; The interpolation formula used by the interpolation method is: In the formula, represents a continuous weighting function varying between 0 and 1, and Indicates two lightweight support structures of different sizes or types; In the formula, is the function describing the transition position, is a parameter used to control the transition gradient; Combining a three-dimensional structure with a lightweight support structure to obtain a structure that can be used for 3D printing includes the following steps: The point cloud coordinate matrix of the three-dimensional structure is combined and reorganized with the point cloud coordinate matrix of the lightweight support structure to obtain a point cloud combined coordinate matrix; The triangulation matrix of the three-dimensional structure and the triangulation matrix of the lightweight support structure are reorganized to obtain a combined triangulation matrix; Combined coordinate matrix based on point cloud With the triangulation matrix , combining the three-dimensional structure with the lightweight support structure to obtain a three-dimensional structure that can be used for 3D printing; Point cloud combined coordinate matrix as follows: In the formula, is the number of points on the three-dimensional structure, is the number of points of the lightweight support structure; Combined triangulation matrix as follows: In the formula, is the number of triangular meshes in the three-dimensional structure, is the number of triangular meshes in the lightweight support structure; Point cloud coordinate matrix of the 3D structure as follows: Triangulation matrix of a 3D structure as follows: Point cloud coordinate matrix of lightweight support structure as follows: Triangulation matrix of lightweight support structure as follows: In the formula, A 3×1 vector representing the coordinates of a point; A 3-by-1 vector representing the triangulated surface.

2. A universal method for adaptive merging and reorganization of multi-lattice structures according to claim 1, characterized in that: Use a 3D scanner to scan the 3D printed objects.

3. A universal method for adaptive merging and reorganization of multi-lattice structures according to claim 1, characterized in that: The physical field is a stress field, a signed distance field, or a temperature field.

Images