A method for optimizing and visualizing a CAD model

Abstract

The application relates to a CAD model optimization method and a visualization method, which comprises the following steps: reading in CAD design data and CAE simulation data; constructing a multi-disciplinary data format parser to convert the read data; discretizing the CAD model into a CAD grid; inserting CAE grid vertices as feature points into the CAD grid to perform local grid subdivision on the CAD grid; processing the CAD grid according to a grid model automatic lightweighting criterion and retaining the feature points; and the grid model automatic lightweighting criterion is any one of the following: a maximum error lightweighting criterion and a compression percentage lightweighting criterion. Based on a simulation body data visualization engine as a basic platform, the existing CAE grid data is used to guide the optimization process of the CAD model, the CAD model is automatically compressed under the premise of ensuring the display quality of the CAD model, the rendering efficiency is improved, and a feasible technical path is provided for the multi-disciplinary fusion visualization of civil aircraft complex products.

Description

Technical Field

[0001] This invention relates to the field of fusion technology of CAD design data and CAE simulation data, specifically an optimization method and visualization method for CAD numerical models. Background Technology

[0002] Multidisciplinary collaborative evaluation is an important means to shorten the development cycle of complex products and improve the quality of manufacturing. Complex products refer to modern industrial complex product systems, including aircraft, spacecraft, high-speed trains, and large ships. Multidisciplinary collaborative evaluation refers to using advanced visualization techniques (VR / AR / MR visualization) to display the calculation results of multiple disciplines within a single application. In aircraft design, CAD and CAE designs influence each other, and their tools are completely different. Therefore, visualizing the CAD and CAE design results within a single application is crucial for facilitating communication and collaborative evaluation between different disciplines.

[0003] Currently, designers use CAD (Computer-Aided Design) systems to design 3D digital models of products, while simulation engineers use CAE (Computer-Aided Engineering) systems to perform simulation calculations in fields such as finite element analysis and fluid mechanics. CAD design data generated by CAD systems is surface data, while CAE simulation data generated by CAE systems is volume data. These two types of data are typically displayed using different visualization systems. Generally, CAD design data is also referred to as CAD digital model or CAD model.

[0004] To achieve the fusion and visualization of CAD design data and CAE simulation data, the common approach is to first preprocess the CAE simulation data, converting volumetric data into surface data, and then import it into a general surface data rendering engine (such as Unity3D or Unreal Engine). Whether done manually or through custom-developed automated or semi-automated software tools, this process is not only inefficient, but more importantly, general surface data rendering engines are often incompatible with volumetric data processing, making it impossible to perform interactive visualization and analysis of CAE simulation data and failing to meet the practical needs of multidisciplinary collaborative evaluation of complex products.

[0005] In practical engineering applications, CAE simulation data (volume data) often constitutes a large proportion. A more reasonable solution would be to integrate CAD design data (surface data) with a simulation volume data visualization engine (such as VTK). However, the burden of interactive visualization processing of volume data is already heavy. To achieve CAE / CAD multidisciplinary data fusion visualization and ensure the real-time nature of interactive visualization analysis, it is not advisable to directly render the original CAD model. After the CAD model is read into the simulation volume data visualization engine, it must be discretized (tessellated) into a triangular facet model before it can be processed by the simulation volume data visualization engine. The discretization of CAD models is mature, but the discretized triangular facet models are often very fragmented, with a large data volume and poor smoothness. Therefore, there are problems with large data volume and difficulty in guaranteeing visualization quality.

[0006] The "CAE / CAD multidisciplinary data fusion visualization" described in this invention includes:

[0007] Multidisciplinary integration refers to the integration and visualization of CAD (3D numerical modeling) and CAE (simulation calculation in the fields of finite element method and fluid mechanics) in civil aircraft design.

[0008] Data fusion visualization refers to the fusion of data from different disciplines. The final presentation can be understood as the visualization and overlay of CAD and CAE models within a single visualization system (such as a computer desktop, VR headset, or CAVE system). For example, fuselage visualization is based on CAD design data, while wing visualization is based on CAE simulation data.

[0009] Through this fusion, the spatial states of CAD design data and CAE simulation data within the 3D scene of the volumetric data visualization platform are coordinated and consistent. They can not only be treated as a whole and interactive behaviors can be applied to them, updating the data state synchronously, but also be selected individually from the internal data structure for analysis operations.

[0010] The current technical approach still relies on manual visualization and optimization of CAD models. This manual process is not only time-consuming and labor-intensive, but also requires a high level of technical expertise from personnel, making it one of the bottleneck technical problems hindering the multidisciplinary integration and visualization of complex products.

[0011] Typically, CAD models generated by CAD systems (such as Catia) have precise mathematical expressions, such as Non-Uniform Rational B-Splines (NURBS). These need to be discretized into mesh models (mesh models composed of triangular faces) before they can be processed by simulation data visualization engines. Since directly discretized mesh models from CAD models are often not optimized, with fragmented triangular faces, direct use for rendering not only results in a heavy rendering burden but also poor display quality. Existing mesh optimization algorithms can improve the smoothness of the mesh model, but they struggle to effectively reduce the number of triangular faces, thus failing to achieve lightweight mesh models. Lightweighting of CAD models still requires significant manual intervention, which is not only time-consuming but also demands a high level of technical expertise from personnel.

[0012] The information disclosed in this background section is intended only to enhance the understanding of the general background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0013] To address the shortcomings of existing technologies, the present invention aims to provide an optimization and visualization method for CAD models. Based on a simulation data visualization engine, it utilizes existing CAE mesh data to guide the optimization process of CAD models. While ensuring the display quality of CAD models, it automatically compresses the CAD models, improves rendering efficiency, and provides a feasible technical path for the multidisciplinary fusion visualization of complex civil aircraft products.

[0014] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0015] A method for optimizing CAD digital models, characterized by comprising the following steps:

[0016] Step 1: In the volume data visualization platform, import CAD design data from the CAD system; in the volume data visualization platform, import CAE simulation data from the CAE system.

[0017] Step 2: Construct a multidisciplinary data format parser. Use the multidisciplinary data format parser to read CAD design data and CAE simulation data, and convert the read data into a data structure that can be processed by the pipeline and rendering engine of the volumetric data visualization platform.

[0018] Step 3: Discretize the CAD model to obtain a discretized CAD mesh;

[0019] Step 4: Using CAE mesh data as an optimization template for lightweight CAD modeling, insert CAE mesh vertices in the CAE mesh data as feature points into the discretized CAD mesh to perform local mesh subdivision of the CAD mesh.

[0020] Step 5: Process the CAD mesh according to the automatic lightweighting criteria of the mesh model. During the lightweighting process, the feature points obtained in Step 4 are retained, thereby achieving the purpose of optimizing the CAD mesh.

[0021] The automatic lightweighting criterion for the mesh model is any one of the following: maximum error lightweighting criterion, compression percentage lightweighting criterion.

[0022] Based on the above technical solution, in step 1, the CAD design data includes: geometric information and material texture information; the CAD design data is CAD design data for complex products.

[0023] The CAE simulation data includes: a mesh model and calculation results, wherein the calculation results are tensor calculation results or vector calculation results.

[0024] Based on the above technical solution, in step 1, the import of CAD design data and CAE simulation data can be achieved by building a data interface. The import of CAD design data and CAE simulation data can also be converted using open data file formats such as STEP and 3dxml.

[0025] Based on the above technical solution, in step 2, the pipeline and rendering engine adopt a simulation volume data visualization engine;

[0026] The data structure includes: CAD digital models corresponding to CAD design data; and CAE mesh data corresponding to CAE simulation data.

[0027] Based on the above technical solutions, in step 5, the strategy for lightweighting the CAD mesh is to reduce the mesh resolution and remove only geometric details that do not significantly affect the shape, thereby significantly reducing the number of vertices and polygons.

[0028] When the QEM method is used as a lightweight CAD meshing strategy, constraints are added to the selection of vertices and edges:

[0029] Constraint a: Valid edges must be edges that exist in the mesh;

[0030] Constraint b, V c ∈{V i}, V c Refers to the lightweight vertex, {V i} refers to the set of vertices before lightweighting, meaning the new vertices are the original vertices; the constraint here is to prevent the generation of new vertices after lightweighting, thus avoiding significant lightweighting deformation problems.

[0031] Constraint c: The inserted CAE mesh vertices are marked as feature points and are retained during lightweight processing.

[0032] Based on the above technical solution, the specific processing steps of step 4 are as follows:

[0033] In the simulation data visualization engine, based on the CAE mesh vertex list, starting from the first CAE mesh vertex, each vertex P is selected sequentially. After each vertex P is selected, the following processing is performed:

[0034] Traverse the triangular faces in the CAD mesh and calculate the distance between each vertex V1, V2, V3 and vertex P of the triangular face;

[0035] If the distance between a vertex of a triangle and vertex P is less than a preset threshold δ, then the vertex of the triangle is marked as a feature point.

[0036] When the distance between all vertices of the triangular facet and vertex P is greater than or equal to the preset threshold δ, then:

[0037] Traverse the triangular faces in the CAD mesh and calculate the projection distance from vertex P to each side of the triangular face;

[0038] When a certain projection distance is less than a preset threshold ε, the projection point P' of vertex P on the edge is inserted into the CAD mesh, and the projection point P' is marked as a feature point.

[0039] When the projected distance from vertex P to each side of the triangle is greater than or equal to the preset threshold ε, then:

[0040] Insert vertex P and mark vertex P as a feature point.

[0041] Based on the above technical solution, in step 4, the local meshing of the CAD mesh refers to triangulation of the CAD model through feature points, forming new triangles through retopology, and then adding the feature points to the vertex list of the CAD model.

[0042] A visualization method for CAD digital models, characterized in that texture material mapping is performed on the vertex list of the CAD model obtained by any one of the above-described optimization methods for CAD digital models;

[0043] After texture mapping, a lightweight CAD mesh was obtained, which includes vertex positions, vertex colors, and texture information;

[0044] The lightweight CAD mesh is visualized using a simulation data visualization engine.

[0045] Based on the above technical solution, for the color attribute of the mesh, let triangle T = (V1, V2, V3), P be the vertex coordinates, S be the color (rgb), and the error metric of triangle T be expressed as:

[0046]

[0047] in It is the square of the distance from vertex P to projection point P'. It is the distance deviation from S to S', where S' is derived from the color interpolation of the projection point P';

[0048] The distance from vertex P to projection point P' is usually very small. It can be approximated that point P is located inside triangle T. Its appearance attribute value can be obtained by directly interpolating the attribute values ​​at the three vertices, thereby greatly reducing the amount of calculation.

[0049] The CAD digital model optimization method and visualization method described in this invention have the following advantages:

[0050] Beneficial effects:

[0051] 1. Based on the simulation data visualization engine, it integrates CAD design data (surface data) to provide a feasible technical path for the multidisciplinary fusion visualization of complex products such as civil aircraft. While ensuring the display quality of CAD digital models, it automatically compresses CAD digital models and improves rendering efficiency.

[0052] 2. Make full use of existing CAE mesh data to guide the optimization process of CAD model, and optimize and compress the CAD model while ensuring the display quality of the CAD model.

[0053] 3. By greatly compressing the number of faces in the CAD digital model (the number of triangles in the CAD visualization model), the rendering burden of the simulation data visualization engine is reduced, ensuring interactive operation and analysis of multidisciplinary data fusion visualization of complex products, and also making the fusion display of CAD and CAE more seamless.

[0054] CAE mesh data is usually optimized, which not only preserves the shape of the CAD design, but also has fewer facets. If existing CAE mesh data is fully utilized, the number of facets in the CAD model can be greatly reduced while ensuring the display quality of the CAD model.

[0055] The CAD digital model optimization and visualization method described in this invention can be used in fields such as collaborative design and evaluation of civil aircraft digital prototypes, multidisciplinary collaboration, virtual simulation, and training. It can realize real-time rendering of large-scale engineering data, which helps researchers discover potential technical problems, reduce rework, and improve the efficiency of civil aircraft development.

[0056] The CAD numerical model optimization and visualization method described in this invention is characterized by its versatility and fundamental nature. It can also be applied to multidisciplinary collaborative visualization of complex products such as ships, engines, and automobiles, and has market promotion value. Attached Figure Description

[0057] The present invention includes the following figures:

[0058] The accompanying drawings are provided to better understand the invention and are not intended to unduly limit the scope of the invention. Wherein:

[0059] Figure 1 The flowchart of the CAD digital model optimization method and visualization method described in this invention.

[0060] Figure 2. (a) Schematic diagram of the QEM edge contraction algorithm proposed by Garland and Heckbert Figure 1 ;

[0061] Figure 2.(b) Schematic diagram of the QEM edge contraction algorithm proposed by Garland and Heckbert;

[0062] Figure 3.(a) Schematic diagram of the distance between vertices of the triangular facet and vertex P. Figure 1 ;

[0063] Figure 3.(b) Schematic diagram of the distance between vertices of the triangular facet and vertex P;

[0064] Figure 3.(c) Schematic diagram of the projection point P' of vertex P onto the edge;

[0065] Figure 3.(d) Schematic diagram of triangulation;

[0066] Figure 3.(e) Schematic diagram of vertex P directly used as a feature point Figure 1 ;

[0067] Figure 3.(f) Schematic diagram of vertex P directly used as a feature point;

[0068] Figure 4 A schematic diagram illustrating the methods for handling color and texture in model lightweighting; Detailed Implementation

[0069] The present invention will be further described in detail below with reference to the accompanying drawings. This detailed description is an illustration in conjunction with exemplary embodiments of the invention, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0070] like Figure 1 As shown, this invention provides an optimization method for CAD digital models, comprising the following steps:

[0071] Step 1: In the volume data visualization platform, import CAD design data from the CAD system; in the volume data visualization platform, import CAE simulation data from the CAE system.

[0072] The CAD design data includes: geometric information and material texture information; the CAD design data is for complex products.

[0073] The CAE simulation data includes: a mesh model and calculation results, wherein the calculation results are tensor calculation results or vector calculation results;

[0074] CAD design data and CAE simulation data can be imported by building a data interface. Taking the commercial CAD system Catia as an example, a 3DK data interface and a CADExchanger data interface can be built to import CAD design data and CAE simulation data.

[0075] CAD design data and CAE simulation data can also be converted using open data file formats such as STEP and 3dxml;

[0076] Step 2: Construct a multidisciplinary data format parser. The multidisciplinary data format parser reads CAD design data and CAE simulation data, and converts the read data into a data structure that can be processed by the pipeline and rendering engine of the volumetric data visualization platform. The pipeline and rendering engine adopts a volumetric data visualization engine (such as any existing volumetric data visualization engine such as VTK).

[0077] The data structure includes: CAD digital model corresponding to CAD design data; CAE mesh data corresponding to CAE simulation data;

[0078] CAE mesh data is usually optimized to not only preserve the shape of the CAD design, but also have fewer facets, typically about 20% of the number of facets in the CAD discrete mesh.

[0079] CAE mesh data is obtained by discretizing the CAD design model. In order to ensure the accuracy of subsequent CAE calculations, the mesh needs to be optimized. For complex-shaped products, the optimization process often requires the participation of experienced technical personnel.

[0080] Although the optimization criteria for CAE mesh data are proposed for subsequent simulation calculations, they are highly consistent with the visualization processing criteria for CAD models. That is, fewer mesh vertices and larger meshes are arranged in relatively flat areas, and denser meshes are arranged in areas with drastic curvature changes, while trying to maintain the shape characteristics of the original CAD product.

[0081] Therefore, CAE mesh data can be regarded as a high-quality simplified version of CAD discrete model, thereby guiding the lightweight processing of the original CAD discrete mesh model;

[0082] Step 3: Discretize the CAD model to obtain a discretized CAD mesh;

[0083] Step 4: Using CAE mesh data as an optimization template for lightweight CAD modeling, insert CAE mesh vertices in the CAE mesh data as feature points into the discretized CAD mesh to perform local mesh subdivision of the CAD mesh.

[0084] Step 5: Process the CAD mesh according to the automatic lightweighting criteria of the mesh model. During the lightweighting process, the feature points obtained in Step 4 are retained, thereby achieving the purpose of optimizing the CAD mesh.

[0085] The automatic lightweighting criterion for the mesh model is any one of the following: maximum error lightweighting criterion, compression percentage lightweighting criterion; the above lightweighting criteria are all conventional lightweighting criteria, and will not be described in detail in this invention.

[0086] The strategy for lightweighting CAD meshes is to reduce the mesh resolution and remove only geometric details that do not significantly affect the shape, thereby significantly reducing the number of vertices and polygons.

[0087] In the strategy of lightweighting CAD meshes, the choice of difference metric is crucial. Theoretically, it should reflect the degree of visual change in the model; however, this is difficult to measure in practice. Therefore, researchers have proposed many specific difference metrics, such as geometric distance in object space (e.g., Hausdorff distance), pixel distance in screen space, or difference in attribute space (e.g., normals, colors, or textures). A widely used one is the edge shrinkage algorithm proposed by Garland and Heckbert, which uses the quadratic error matrix (QEM) as the metric cost, as shown in Figures 2.(a) and 2.(b). It has a fast computation speed and high simplification quality. When selecting a suitable edge for iterative shrinkage, the QEM edge shrinkage algorithm defines a variable Δ to describe the edge shrinkage cost, as follows:

[0088] For each vertex V in the mesh, a 4×4 symmetric error matrix Q is predefined, then vertex V = [VxVyVz 1] T The error is in the form of its quadratic term Δ(V) = V T QV;

[0089] Suppose that for a contracting edge (V1, V2), its vertex becomes Vc after contraction, and its error matrix Qc is Qc = Q1 + Q2;

[0090] There are two common strategies for calculating the position of vertex Vc: one simple strategy is to choose a position among V1, V2, and (V1+V2) / 2 that minimizes the contraction cost Δ(Vc); the other strategy is to numerically calculate the position of vertex Vc that minimizes Δ(Vc). Since the expression for Δ is a quadratic term, setting the first derivative to 0 makes the expression equivalent to solving:

[0091]

[0092] Where q ij This is the corresponding element in matrix Qc;

[0093] If the coefficient matrix is ​​invertible, the position of the new vertex Vc can be obtained by solving the above equation. If the coefficient matrix is ​​not invertible, the position of the new vertex Vc can be obtained by the first strategy.

[0094] Based on the above description, the steps of the QEM algorithm are as follows ([1] Garland M, Heckbert P S. Simplifying surfaces with color and texture using quadric error metrics[C] / / Visualization.IEEE, 1998.):

[0095] (1) Calculate the Q matrix for all initial vertices;

[0096] (2) Select all valid edges (here we select connected edges, or we can classify edges whose distance is less than a threshold as valid edges);

[0097] (3) For each valid edge (V1, V2), calculate the optimal extraction target. error (Q1+Q2) It is the cost of extracting this edge;

[0098] (4) Put all the edges into a heap according to their cost weights;

[0099] (5) Remove the edge with the lowest cost each time and update the cost of all valid edges containing V1;

[0100] However, the conventional QEM method described above will change the topology of the original mesh, and there is a risk of changing the edges, holes and concave shapes in the original mesh. Furthermore, the vertices in the original mesh are directly discretized from the CAD design model, meaning that these vertices are interpolation points with accurate spatial positions, and their positions should not be changed.

[0101] This invention modifies the conventional QEM method by adding constraints to the selection of vertices and edges when the QEM method is used for lightweight CAD meshes.

[0102] Constraint a: Valid edges must be edges that exist in the mesh;

[0103] Constraint b, V c ∈{V i}, V c Refers to the lightweight vertex, {V i} refers to the set of vertices before lightweighting, meaning the new vertices are the original vertices; the constraint here is to prevent the generation of new vertices after lightweighting, thus avoiding significant lightweighting deformation problems.

[0104] Constraint c: The inserted CAE mesh vertices are marked as feature points and are retained during lightweight processing;

[0105] Adding constraints not only effectively preserves the topology of the CAD mesh, but also greatly improves the calculation speed. To ensure the smoothness of the final mesh, the mesh processed above can be optimized as a whole.

[0106] Based on the above technical solution, the specific processing steps of step 4 are as follows:

[0107] In the simulation data visualization engine, based on the CAE mesh vertex list, starting from the first CAE mesh vertex, each vertex P is selected sequentially. After each vertex P is selected, the following processing is performed:

[0108] Traverse the triangular faces in the CAD mesh and calculate the distance between each vertex V1, V2, V3 and vertex P of the triangular face;

[0109] If the distance between a vertex of a triangle and vertex P is less than a preset threshold δ, then the vertex of the triangle is marked as a feature point.

[0110] As shown in Figures 3.(a) and 3.(b), if the distance between vertex V1 and vertex P of the triangular facet is less than the preset threshold δ, then vertex V1 of the triangular facet is marked as a feature point and denoted as V1(P).

[0111] When the distance between all vertices of the triangular facet and vertex P is greater than or equal to the preset threshold δ, then:

[0112] Traverse the triangular faces in the CAD mesh and calculate the projection distance from vertex P to each side of the triangular face;

[0113] When a certain projection distance is less than a preset threshold ε, the projection point P' of vertex P on the edge is inserted into the CAD mesh, and the projection point P' is marked as a feature point.

[0114] As shown in Figure 3.(c), if the projection distance from vertex P to the edge formed by the vertex V1 and the vertex V2 of the triangular facet is less than the preset threshold ε, then the projection point P' of vertex P on the edge is inserted into the CAD mesh and the projection point P' is marked as a feature point.

[0115] When the projected distance from vertex P to each side of the triangle is greater than or equal to the preset threshold ε, then:

[0116] Insert vertex P and mark vertex P as a feature point;

[0117] The feature point is derived from vertex P. As mentioned above, when vertex P is very close to the vertex of the triangular facet, the vertex of the triangular facet is the feature point; when vertex P is very close to one of the edges of the triangular facet, the projection point P' from vertex P to that edge is the feature point; when vertex P is neither close to the vertex of the triangular facet nor close to one of the edges of the triangular facet, then vertex P is directly used as the feature point.

[0118] The CAD model is triangulated using feature points, new triangles are formed through retopology, and the feature points are then added to the vertex list of the CAD model. The specific algorithm for triangulation is implemented using existing technology and will not be described in detail in this invention.

[0119] As shown in Figure 3.(d), when vertex P is very close to one side of a triangular facet, connect the vertices of the two triangular faces, including the side containing the projection point P', to the projection point P'. Then, the original two triangles △V1V2V3 and △V1V2V4 are split into four triangles △V1P. ’ V3, △V2P ’ V3、△V1P ’ V4, △V2P ’ V4, after completing triangulation, update the CAD mesh data structure linked list: add new feature points P ’ Add to the vertex list, delete the vertex indices of the original two triangles, and add the vertex indices of the four triangles after triangulation to the list;

[0120] As shown in Figures 3.(e) and 3.(f), when vertex P is neither close to a vertex of a triangle nor close to an edge of a triangle, vertex P is connected to vertices V1 and V2V3 of the triangle respectively, splitting the original triangle △V1V2V3 into three triangles △V1PV2, △V2PV3, and △V1PV3. After completing the triangulation, the CAD mesh data structure linked list is updated: the newly added feature point P is added to the vertex linked list, the vertex index of the original triangle is deleted, and the vertex indices of the three triangles after triangulation are added to the linked list.

[0121] Based on the above-mentioned optimization method for CAD digital models, this invention also provides a visualization method for CAD digital models, which performs texture and material mapping based on the vertex list of the CAD model obtained by the aforementioned optimization method.

[0122] To maintain the original material, texture, and color attributes on the CAD design model, it is necessary to assign corresponding attribute values ​​to the newly added feature points. Referring to the color and texture processing methods in Garland's ([1] Garland M, Heckbert P S. Simplifying surfaces with color and texture using quadric error metrics [C] / / Visualization.IEEE, 1998.) model lightweighting, these values ​​are obtained through linear interpolation, such as... Figure 4 As shown:

[0123] For the color attribute of the mesh, let triangle T = (V1, V2, V3), P be the vertex coordinates, and S be the color (rgb). The error metric for triangle T is expressed as:

[0124]

[0125] in It is the square of the distance from vertex P to projection point P'. It is the distance deviation from S to S', where S' is derived from the color interpolation of the projection point P';

[0126] For the application scenario in this invention, the distance from vertex P to projection point P' is usually very small. It can be approximated that point P is located inside triangle T, and its appearance attribute value can be obtained by directly interpolating the attribute values ​​at the three vertices, thereby greatly reducing the amount of calculation.

[0127] After texture mapping, a lightweight CAD mesh was obtained, which includes vertex positions, vertex colors, and texture information;

[0128] The lightweight CAD mesh is visualized using a simulation data visualization engine.

[0129] The present invention uses an improved QEM algorithm, which only removes vertices during the lightweighting process, resulting in high efficiency in lightweighting CAD meshes with textured materials.

[0130] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

[0131] The above description is only a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. Any equivalent modifications or changes made by those skilled in the art based on the content disclosed in the present invention should be included in the scope of protection set forth in the claims.

Claims

1. A method for optimizing CAD digital models, characterized in that, Includes the following steps: Step 1: In the volume data visualization platform, import CAD design data from the CAD system; in the volume data visualization platform, import CAE simulation data from the CAE system. Step 2: Construct a multidisciplinary data format parser. Use the multidisciplinary data format parser to read CAD design data and CAE simulation data, and convert the read data into a data structure that can be processed by the pipeline and rendering engine of the volumetric data visualization platform. Step 3: Discretize the CAD model to obtain a discretized CAD mesh; Step 4: Using CAE mesh data as an optimization template for lightweight CAD modeling, insert CAE mesh vertices in the CAE mesh data as feature points into the discretized CAD mesh to perform local mesh subdivision of the CAD mesh. Step 5: Process the CAD mesh according to the automatic lightweighting criteria of the mesh model. During the lightweighting process, the feature points obtained in Step 4 are retained, thereby achieving the purpose of optimizing the CAD mesh. The automatic lightweighting criterion for the mesh model is any one of the following: maximum error lightweighting criterion, compression percentage lightweighting criterion.

2. The CAD digital model optimization method as described in claim 1, characterized in that, In step 1, the CAD design data includes: geometric information and material texture information; the CAD design data is CAD design data for complex products. The CAE simulation data includes: a mesh model and calculation results, wherein the calculation results are tensor calculation results or vector calculation results.

3. The CAD digital model optimization method as described in claim 1, characterized in that, In step 1, the CAD design data and CAE simulation data are read in by building a data interface, or by converting the CAD design data and CAE simulation data in an open data file format, namely STEP or 3dxml.

4. The CAD digital model optimization method as described in claim 1, characterized in that, In step 2, the pipeline and rendering engine adopt a simulation volume data visualization engine; The data structure includes: CAD digital models corresponding to CAD design data; and CAE mesh data corresponding to CAE simulation data.

5. The CAD digital model optimization method as described in claim 1, characterized in that, In step 5, the strategy for lightweighting the CAD mesh is to reduce the mesh resolution and remove only geometric details that do not significantly affect the shape, thereby significantly reducing the number of vertices and polygons. When the QEM method is used as a lightweight CAD meshing strategy, constraints are added to the selection of vertices and edges: Constraint a: Valid edges must be edges that exist in the mesh; Constraint b, , Refers to the lightweighted peak. This refers to the set of vertices before lightweighting, meaning the new vertices are the original vertices; the constraint here is to prevent the generation of new vertices after lightweighting, thus avoiding significant lightweighting deformation issues. Constraint c: The inserted CAE mesh vertices are marked as feature points and are retained during lightweight processing.

6. The CAD digital model optimization method as described in claim 1, characterized in that, The specific processing steps in step 4 are as follows: In the simulation data visualization engine, based on the CAE mesh vertex list, starting from the first CAE mesh vertex, each vertex P is selected sequentially. After each vertex P is selected, the following processing is performed: Traverse the triangular faces in the CAD mesh and calculate the distance between each vertex V1, V2, V3 and vertex P of the triangular face; If the distance between a vertex of a triangle and vertex P is less than a preset threshold δ, then the vertex of the triangle is marked as a feature point. When the distance between all vertices of the triangular facet and vertex P is greater than or equal to the preset threshold δ, then: Traverse the triangular faces in the CAD mesh and calculate the projection distance from vertex P to each side of the triangular face; When a certain projection distance is less than a preset threshold ε, the projection point P' of vertex P on the edge is inserted into the CAD mesh, and the projection point P' is marked as a feature point. When the projected distance from vertex P to each side of the triangle is greater than or equal to the preset threshold ε, then: Insert vertex P and mark vertex P as a feature point.

7. The CAD digital model optimization method as described in claim 6, characterized in that, In step 4, the local meshing of the CAD mesh refers to triangulation of the CAD model using feature points, forming new triangles through retopology, and then adding the feature points to the vertex list of the CAD model.

8. A visualization method for CAD digital models, characterized in that, Texture and material mapping is performed on the vertex list of the CAD model obtained by the optimization method of the CAD digital model according to any one of claims 1 to 7. After texture mapping, a lightweight CAD mesh was obtained, which includes vertex positions, vertex colors, and texture information; The lightweight CAD mesh is visualized using a simulation data visualization engine.

9. The visualization method for CAD digital models as described in claim 8, characterized in that, For the color attribute of the mesh, let triangle T = (V1, V2, V3), P be the vertex coordinates, and S be the color (rgb). The error metric for triangle T is expressed as: ， in It is the square of the distance from vertex P to projection point P'. It is the distance deviation from S to S', where S' is derived from the color interpolation of the projection point P'; The distance from vertex P to projection point P' is usually very small. It can be approximated that point P is located inside triangle T. Its appearance attribute value can be obtained by directly interpolating the attribute values ​​at the three vertices, thereby greatly reducing the amount of calculation.

Images