CN117763927B - Automatic updating method for simulation model driven by geometry-grid twinning - Google Patents
Automatic updating method for simulation model driven by geometry-grid twinning Download PDFInfo
- Publication number
- CN117763927B CN117763927B CN202410195563.4A CN202410195563A CN117763927B CN 117763927 B CN117763927 B CN 117763927B CN 202410195563 A CN202410195563 A CN 202410195563A CN 117763927 B CN117763927 B CN 117763927B
- Authority
- CN
- China
- Prior art keywords
- model
- grid
- geometric model
- control point
- plane
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 102
- 238000000034 method Methods 0.000 title claims abstract description 98
- 238000013507 mapping Methods 0.000 claims abstract description 45
- 238000012986 modification Methods 0.000 claims abstract description 27
- 230000004048 modification Effects 0.000 claims abstract description 27
- 238000013461 design Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims abstract description 12
- 238000012545 processing Methods 0.000 claims abstract description 10
- 230000001131 transforming effect Effects 0.000 claims abstract description 3
- 238000005070 sampling Methods 0.000 claims description 57
- 238000004458 analytical method Methods 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 11
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 239000000463 material Substances 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 4
- 230000008676 import Effects 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000010801 machine learning Methods 0.000 claims description 2
- 239000013598 vector Substances 0.000 claims description 2
- 230000007547 defect Effects 0.000 abstract description 2
- 238000005457 optimization Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 11
- 230000003993 interaction Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 125000000174 L-prolyl group Chemical group [H]N1C([H])([H])C([H])([H])C([H])([H])[C@@]1([H])C(*)=O 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000002715 modification method Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
An automatic updating method of a simulation model driven by geometry-grid twins belongs to the field of digital twins, and comprises the following steps: firstly, processing a reference geometric model to obtain a reference simulation model and a modified geometric model; secondly, establishing two corresponding curved surface grid models based on the reference geometric model and the modified geometric model, and mapping to obtain two planar grid models; thirdly, acquiring control point sets of two curved surface grid models according to the corresponding relation of the node numbers; fourthly, training the coordinate relation of the two control point sets to obtain the mapping relation between the two control point sets; and finally, transforming coordinates of the finite element nodes in the reference simulation model based on the mapping relation to realize twin of the modified grid model and complete automatic updating of the simulation model. The invention solves the defects of complicated steps and slow iteration speed when the simulation model is updated due to the modification of partial geometry in the simulation model, has high efficiency and robustness, and is convenient for developing subsequent structural optimization design.
Description
Technical Field
The invention belongs to the field of digital twinning, and provides an automatic updating method of a simulation model driven by geometric-grid twinning.
Background
In order to improve the strength-to-weight ratio of structures such as aircrafts and spacecrafts and increase the reliability of the structures, various types of thin-wall parts such as integral wallboards, frames, beams, ribs and the like are widely applied to structural designs. The thin-wall parts generally have the characteristics of complex structure, various characteristics, high design precision requirement and the like. In the design stage, multiple configuration modifications are often required to be carried out by considering various factors such as structural mechanical properties, part installation dimensions, space avoidance and the like.
The existing simulation model modification method is that a local geometric model is firstly built in CAD software (such as Pro/E, UG), the local geometric model is stored into a CAD file with a certain format, then the model is imported, gridding is conducted, material properties are set, assembling, loading, boundary conditions and other preprocessing operations are conducted, and then subsequent mechanical analysis is conducted. Once the geometry changes again, the model needs to be reintroduced and then subjected to finite element pre-processing to obtain a simulation model corresponding to the model. The method requires analysts to perform corresponding finite element pretreatment work on the changed geometric model each time, modeling steps are complicated, design analysis efficiency is severely restricted, and accordingly development period is prolonged and cost is increased.
Therefore, it is needed to invent a method for automatically updating a simulation model driven by geometry-grid twinning, which does not need complex and complicated finite element pretreatment on the geometry model, and automatically obtains a modified simulation model by fusing the information of the geometry model before and after modification and the grid information of a reference simulation model.
Xu W et al [Xu W, Neumann I. Finite element analysis based on a parametric model by approximating point clouds[J]. Remote Sensing, 2020, 12(3): 518], by combining the actually measured geometric model with the CAD modeling model, improves the accuracy of the geometric model, and directly calculates the point cloud by a finite element method to realize the continuity of CAD and CAE, but the essence of the method is that the existing CAD model is slightly changed and is difficult to be fused into the existing finite element analysis framework.
Disclosure of Invention
Aiming at the problems existing in the prior art, the invention provides an automatic updating method of a simulation model driven by geometry-grid twinning, which mainly solves the problems of complicated simulation model modification steps and long iteration period.
In order to achieve the above purpose, the invention adopts the following technical scheme:
An automatic updating method of a simulation model driven by geometry-grid twins comprises the following steps: firstly, obtaining a reference geometric model through CAD design, obtaining a corresponding reference simulation model through finite element processing, and obtaining a geometric model modified by the reference geometric model through CAD modification; secondly, establishing two corresponding curved surface grid models based on the reference geometric model and the modified geometric model, and mapping the two curved surface grid models respectively by a quasi-conformal mapping method to obtain two mapped planar grid models; thirdly, acquiring control point sets of two plane grid models based on a point cloud sampling method, and acquiring the control point sets of the two curved surface grid models according to the corresponding relation of node numbers; training the coordinate relationship of the two curved surface control point sets in the third step by using a fitting method to obtain a mapping relationship between the two curved surface control point sets; and fifthly, transforming coordinates of the finite element nodes in the reference simulation model based on the mapping relation, realizing twin of the modified grid model, and automatically inheriting finite element setting of the reference simulation model in the first step to complete automatic updating of the simulation model. The method specifically comprises the following steps:
The first step, obtaining a reference geometric model through CAD design, obtaining a corresponding reference simulation model through finite element processing, and obtaining a geometric model modified by the reference geometric model through CAD modification. Specific:
Step 1.1: and establishing a reference geometric model A through CAD three-dimensional design software according to the design requirement of the structure, and outputting the reference geometric model A in a format (IGES, STP, SAT) of the geometric model.
Step 1.2: and performing geometric import, grid division, material attribute setting, assembly and connection relation setting in finite element analysis software for the obtained reference geometric model A to obtain a reference simulation model A Fem, wherein the reference simulation model comprises finite element grid model information (unit type, node number and node set) and finite element setting information (material attribute, connection relation, load, boundary condition and analysis type).
Step 1.3: and modifying the reference geometric model A according to the requirements of engineering design, obtaining a modified geometric model B by modifying the shape, changing the size and adjusting the angle position of the geometric model, and outputting the modified geometric model B in a format (IGES, STP, SAT) of the geometric model.
Secondly, adopting unit types of triangles to carry out grid division on the reference geometric model A and the geometric model B, and establishing curved surface grid models corresponding to the two geometric models; respectively carrying out plane parameterization on the two curved surface grid models by a quasi-conformal mapping method to obtain two mapped plane grid models; specific:
Step 2.1: the method comprises the steps of performing grid division on a reference geometric model A by adopting triangular grid units to obtain a curved surface grid model A tra of the geometric model A, selecting 4 reference points from outline nodes of the curved surface grid model A tra, performing plane parameterization by adopting a quasi-conformal mapping method to obtain a rectangular plane grid model A tra-Plane, wherein the node numbers and unit types of the curved surface grid model A tra before plane parameterization and the grid model A tra-Plane after plane parameterization are unchanged and correspond to each other.
Step 2.2: the modified geometric model B is subjected to grid division by adopting triangular grid units to obtain a curved surface grid model B tra of the geometric model B, 4 reference points are selected from outline nodes of the curved surface grid model B tra, planar parameterization is performed by adopting a quasi-conformal mapping method to obtain a rectangular planar grid model B tra-Plane, the node numbers and unit types of the curved surface grid model B tra before planar parameterization and the planar grid model B tra-Plane after planar parameterization are unchanged and correspond to each other.
Thirdly, grid node sets of the two plane grid models A tra-Plane and B tra-Plane obtained in the second step are obtained based on a point cloud sampling method, node sets of the curved surface grid models A tra and B tra are obtained according to the corresponding relation of node numbers, and the node sets are used as control point sets of a reference geometric model A and a geometric model B; specific:
step 3.1: sampling the nodes of the planar grid model A tra-Plane obtained in the step 2.1 by adopting a point cloud sampling method, setting the number of sampling points in the length direction as m and the number of sampling points in the width direction as n, obtaining m x n sampling point coordinates, calculating Euclidean distances between all grid nodes and the sampling points, and selecting the grid node with the smallest Euclidean distance from the sampling points; further obtaining nodes of m x n planar rectangular grids; the node numbers and the unit types of the plane parameterized front surface mesh model A tra and the plane parameterized front and back surface mesh model A tra-Plane are not changed; therefore, according to the correspondence of the node numbers, m×n nodes of the surface mesh model a tra corresponding to the reference geometric model a can be obtained as the control point set of the reference geometric model a.
Step 3.2: sampling the nodes of the planar grid model B tra-Plane obtained in the step 2.2 by adopting a point cloud sampling method, setting the number of sampling points in the length direction as m and the number of sampling points in the width direction as n, obtaining m x n sampling point coordinates, calculating Euclidean distances between all grid nodes and the sampling points, and selecting the grid node with the nearest Euclidean distance; further obtaining nodes of m x n planar rectangular grids; the node numbers and the unit types of the plane parameterized front surface mesh model B tra and the plane parameterized rear surface mesh model B tra-Plane are not changed; therefore, according to the correspondence of the node numbers, m×n nodes of the curved surface mesh model B tra corresponding to the geometric model B can be obtained as the control point set of the geometric model B.
Further, the control point sets in step 3.1 and step 3.2 may be considered as descriptions of geometric model contours.
Further, the point cloud sampling method in step 3.1 includes, but is not limited to, a grid sampling method, a uniform sampling method, and a set sampling method, and the point cloud sampling method in step 3.2 includes, but is not limited to, a grid sampling method, a uniform sampling method, and a set sampling method, where the point cloud sampling methods in step 3.1 and step 3.2 are the same.
Further, the control point set of the geometric model a in the step 3.1 is a reference geometric model control point set. And 3.2, the control point set of the geometric model B in the step is a modified geometric model control point set.
And fourthly, training the position coordinate relation of the two geometric model control point sets in the third step by using a fitting method to obtain the mapping relation between the two geometric model control point sets. Specific:
And (3) training the fitting relation of the two geometric model control point sets in the third step by adopting a fitting method, wherein the input of the training process is the coordinate value of the reference geometric model control point set obtained in the step (3.1), and the output is the coordinate value of the modified geometric model control point set obtained in the step (3.2). For a three-dimensional structure, the input coordinate values should include coordinates in three directions of x, y and z, that is, m×n three-dimensional vectors. After training, the mapping relation between the control point sets of the reference geometric model A before modification and the geometric model B after modification is obtained. The mapping relation refers to a fitting relation between coordinate values of the reference geometric model control point set and coordinate values of the modified geometric model control point set.
Further, the fitting method comprises a radial basis function neural network, a machine learning method of a back propagation neural network or a proxy model method based on a radial basis function method and a kriging method.
And fifthly, based on the mapping relation obtained in the fourth step, twinning the reference simulation model A Fem in the first step to finish automatic updating of the simulation model for subsequent simulation analysis and structural performance evaluation. Specific:
And (3) carrying out coordinate transformation on grid information in the reference simulation model A Fem established in the first step based on the mapping relation between the control point set of the reference geometric model A before modification and the control point set of the geometric model B after modification, which is obtained in the fourth step, wherein the input of the mapping relation is all grid node coordinates of the reference simulation model A Fem, and the output is grid node coordinates corresponding to the geometric model B after modification. On the basis, the finite element setting information (connection relation, load, boundary condition and analysis type) in the first step is integrated, the simulation model is automatically updated, and the simulation model B Fem corresponding to the geometric model B is obtained and used for subsequent simulation analysis and structural performance evaluation.
The beneficial effects of the invention are as follows:
(1) Aiming at the problems that the traditional geometric model is difficult to parameterize and read and uniform control point distribution cannot be obtained, the invention carries out triangular mesh processing on the geometric model of the curved surface, maps the geometric model into a plane based on quasi-conformal mapping, obtains the control points of the curved surface through obtaining the control points of the plane according to the node numbers, and realizes the obtaining of the control points of the geometric model. The method is different from the traditional point cloud generation method, and the same and uniformly distributed control point sets can be obtained aiming at different geometric models.
(2) Aiming at the problems of complicated modification steps and long modification period of the traditional simulation model, the invention does not need to carry out complicated and complicated finite element pretreatment on the geometric model, realizes the generation of the corresponding grid model by fusing the information of the geometric model before and after modification, can inherit the grid information of the reference simulation model, and automatically obtains the modified simulation model. Therefore, the method is simple to operate and high in robustness, is used for various structural forms such as entities, shells and opening-containing features, can quickly update the simulation model without complex interaction, can solve the defects of complicated updating steps and long iteration period of the simulation model, has high efficiency and robustness, is convenient for subsequent structural design iteration in the practical application process, and shortens the design period.
Drawings
FIG. 1 is a flow chart of an implementation of a method for automatically updating a simulation model driven by geometry-grid twinning;
FIG. 2 is a schematic diagram of various models; (a) is a reference geometric model schematic; (b) a reference simulation model schematic; (c) a modified geometric model schematic;
FIG. 3 is a schematic diagram of the model and reference points before and after the quasi-conformal mapping of the different models; (a) Modeling conformal mapping of a front model and a rear model and a reference point schematic diagram for a reference geometric model; (b) Simulating conformal mapping of the front model and the rear model and a reference point schematic diagram for the modified geometric model;
FIG. 4 is a graph of control point distribution for different models; (a) A control point profile for modifying the pre-reference geometric model; (b) a control point profile for the modified geometric model;
FIG. 5 is a schematic diagram of various simulation models; (a) is a truncated semi-cylindrical simulation model; (b) a curved surface simulation model corresponding to the modification; (c) final assembling the updated simulation model;
FIG. 6 is a first order frequency cloud obtained by different methods; (a) Modifying a first-order frequency cloud image of the method for a general simulation model; (b) Automatically updating a first-order frequency cloud picture of the method for the simulation model;
FIG. 7 is a schematic diagram of a typical surface structure simulation automatic update flow; (a) is a reference geometric model; (b) a set of control points for a reference geometric model; (c) a reference simulation model; (d) is a modified geometric model; (e) a set of control points for the modified geometric model; (f) a simulation model modified by the geometric model;
FIG. 8 is a schematic diagram of a simulated automatic update flow for a curved surface structure with openings; (a) is a reference geometric model; (b) a set of control points for a reference geometric model; (c) a reference simulation model; (d) is a modified geometric model; (e) a set of control points for the modified geometric model; (f) a simulation model modified by the geometric model;
FIG. 9 is a schematic diagram of a simulation automatic update flow of a physical structure; (a) is a reference geometric model; (b) a set of control points for a reference geometric model; (c) a reference simulation model; (d) is a modified geometric model; (e) a set of control points for the modified geometric model; (f) a simulation model modified by the geometric model;
Detailed Description
In order to make the solution to the problems of the method, the method scheme adopted and the effect of the method achieved by the invention more clear, the invention is further described in detail below with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting thereof. It should be further noted that, for convenience of description, only some, but not all of the matters related to the present invention are shown in the accompanying drawings.
Fig. 1 is a flowchart of an implementation of a method for automatically updating a simulation model driven by geometry-grid twinning, which is provided by an embodiment of the present invention, and includes the following steps:
The first step, obtaining a reference geometric model through CAD design, obtaining a corresponding reference simulation model through finite element processing, and obtaining a geometric model modified by the reference geometric model through CAD modification. Specific:
Step 1.1: the bottom of the geometric model used in this embodiment is a solid disc with a diameter of 350mm and a height of 15mm, the upper part of the geometric model is a semi-cylindrical surface with a diameter of 300mm and a height of 450mm, and the geometric model is modeled in CAD software and exported in the form of IGES.
Step 1.2: the geometric model is imported into ABAQUS finite element software, finite element setting is carried out, finite element setting information comprises grid division, boundary setting and material attribute application, an overall simulation model is built, a semi-cylindrical surface is selected as a structure needing modification and is used as a reference geometric model A, the number of grids of the simulation model A Fem,AFem, which is a model corresponding to the semi-cylindrical surface structure, in the simulation model is 2115, and the simulation model A is shown in fig. 2 (b).
Step 1.3: and according to engineering requirements, shape change is carried out in the range of the semi-cylindrical surface, and a modified curved surface model B is established in CAD software, as shown in fig. 2 (c).
Secondly, adopting triangle unit types to carry out grid division on the reference geometric model A before modification and the geometric model B after modification, and establishing curved surface grid models corresponding to the two geometric models, wherein the number of the curved surface grid models is two; and respectively carrying out plane parameterization on the two curved surface grid models by a quasi-conformal mapping method to obtain two mapped plane grid models. Specific:
step 2.1: and carrying out grid division on the semi-cylindrical surface by adopting triangular grid units to obtain a curved surface grid model of the semi-cylindrical surface, wherein the number of the triangular units is 16920, 4 reference points are selected from outer contour nodes of the curved surface grid model, the reference points are positioned on the curved surface grid model nodes, and carrying out plane parameterization by adopting a quasi-conformal mapping method to obtain a rectangular plane grid model. The mapping process is shown in fig. 3 (a).
Step 2.2: and carrying out grid division on the modified geometric model B by adopting triangular grid units to obtain a curved surface grid model of the modified geometric model, wherein the number of the triangular grid units is 11092, 4 reference points are selected from outline nodes of the curved surface grid model, and plane parameterization is carried out by adopting a quasi-conformal mapping method to obtain a rectangular plane grid model. The mapping process is shown in fig. 3 (b).
Thirdly, grid node sets of the two plane grid models obtained in the second step are obtained based on a uniform sampling method, the node sets of the curved surface grid model are obtained according to the corresponding relation of the numbers, and the node sets are used as control point sets of the geometric models before and after modification. Specific:
Step 3.1: for the rectangular plane grid model (the number of units is 16920) obtained in the second step 2.1, sampling nodes of the plane grid model by adopting a uniform sampling method, setting the number of sampling points in the length direction to be 30, setting the number of sampling points in the width direction to be 30, obtaining 900 sampling point coordinates, calculating Euclidean distances between all grid nodes and the sampling points, and selecting the grid node with the smallest Euclidean distance from the sampling points. And then 900 nodes of the plane rectangular grid are obtained. Therefore, according to the corresponding relation of the node numbers, 900 nodes of the initial grid model can be obtained and used as a control point set of the semi-cylindrical surface. Fig. 4 (a) is a schematic view of a curved control point obtained in this embodiment.
Step 3.2: for the planar rectangular grid model (the number of units is 11092) obtained in the second step 2.2, sampling nodes of the grid model by adopting a uniform sampling method, setting the number of sampling points in the length direction to be 30, setting the number of sampling points in the width direction to be 30, obtaining 900 sampling point coordinates, calculating Euclidean distances between all grid nodes and the sampling points, and selecting the grid node with the smallest Euclidean distance from the sampling points. And then 900 nodes of the plane rectangular grid are obtained. Therefore, according to the corresponding relation of the node numbers, 900 nodes of the initial grid model can be obtained and used as a control point set of the modified geometric model. Fig. 4 (b) is a schematic view of the curved control point obtained in this embodiment.
In this embodiment, the control point set in step 3.1 and step 3.2 is a description of the geometric model contour.
And fourthly, training the third step by using a radial basis function neural network to obtain two control point sets with the number of 900, and obtaining the mapping relation between the two control point sets. The input of the radial basis function neural network is coordinate values X i,Yi,Zi (i=1, 2,3, …, 900) of 900 control point sets corresponding to the reference geometric model, the output of the radial basis function neural network is coordinate values X j,Yj,Zj (j=1, 2,3, …, 900) of 900 control point sets corresponding to the changed geometric model, each node has coordinates in the X, y and z directions, and the data are fitted through the radial basis function to obtain a mapping relation.
And fifthly, carrying out coordinate transformation on the grid information (the number of grids is 2115) in the reference simulation model established in the first step based on the mapping relation in the fourth step, wherein the input of the mapping relation is all grid node coordinates of the semi-cylindrical model, as shown in fig. 5 (a), and the output is grid node coordinates corresponding to the modified geometric model, as shown in fig. 5 (b). On the basis, the finite element setting information in the first step is integrated, the automatic updating of the simulation model is completed, the simulation model corresponding to the modified geometric model is obtained, the automatic updating of the simulation model is completed, and the simulation model is used for subsequent simulation analysis and structural performance evaluation, as shown in fig. 5 (c).
Results and effects of this embodiment verify:
The general simulation model updating method needs to import the modified geometric model into finite element analysis software for section attribute giving, assembly, connection relation setting, load setting, grid division and submitting calculation. The number of user interactions was 28, and the whole process required 10 minutes. According to the result of the modal analysis, the first-order frequency is 180.20Hz, and the cloud chart is shown in FIG. 6 (a); for the method, the simulation model can be directly updated and submitted only by manually setting the control point, the user interaction times are 2 times, and the process consumes 1 minute. From the results of the modal analysis, the first-order frequency was 180.45Hz, and the resulting cloud was as shown in FIG. 6 (b). Compared with the traditional geometric model simulation modeling method, the simulation result error is (180.45-180.20)/180.20 =0.13%. In order to further verify the universality of the method, a typical curved surface, a curved surface containing holes and a solid structure are taken as examples, and the demonstration is carried out by the proposed method.
Fig. 7 is a schematic diagram of an automatic simulation update flow of a typical curved surface structure, for the simulation model update of the typical curved surface structure, a geometric model is built in CAD software to be used as a reference geometric model, as shown in fig. 7 (a), finite element processing is performed to obtain a reference simulation model, as shown in fig. 7 (c), the reference geometric model is modified in CAD software to obtain a modified geometric model, as shown in fig. 7 (d), two groups of control points with the same number are obtained by the method, control points corresponding to the reference geometric model are shown in fig. 7 (b), control points corresponding to the modified geometric model are shown in fig. 7 (e), the mapping relation between the two control points is trained by a radial basis function, and the simulation update of the modified geometric model is realized by the mapping relation, and the simulation model after automatic update is shown in fig. 7 (f).
Fig. 8 is a schematic diagram of an automatic simulation update flow of a structure with an open-cell curved surface, for the update of the simulation model with the open-cell curved surface, a geometric model is built in CAD software to be used as a reference geometric model, as shown in fig. 8 (a), finite element processing is performed to obtain a reference simulation model, as shown in fig. 8 (c), the modified geometric model is obtained by modifying the model in CAD software, as shown in fig. 8 (d), two groups of control points with the same number are obtained by the method, the control points corresponding to the reference geometric model are shown in fig. 8 (b), the control points corresponding to the modified geometric model are shown in fig. 8 (e), the mapping relation between the two control points is trained by a radial basis function, and the simulation update of the modified geometric model is realized by the mapping relation, and the simulation model after automatic update is shown in fig. 8 (f).
Fig. 9 is a schematic diagram of an automatic updating flow of a physical structure simulation, for updating a physical structure simulation model, by establishing a geometric model in CAD software as a reference geometric model, as shown in fig. 9 (a), performing finite element processing to obtain a reference simulation model, as shown in fig. 9 (c), modifying the reference geometric model in CAD software to obtain a modified geometric model, as shown in fig. 9 (d), obtaining two groups of control points with the same number by the method, wherein the control points corresponding to the reference geometric model are as shown in fig. 9 (b), the control points corresponding to the modified geometric model are as shown in fig. 9 (e), training a mapping relation between the two control points by a radial basis function, and realizing simulation updating of the modified geometric model by the mapping relation, and the simulation model after automatic updating is as shown in fig. 9 (f).
The examples described above represent only embodiments of the invention and are not to be understood as limiting the scope of the patent of the invention, it being pointed out that several variants and modifications may be made by those skilled in the art without departing from the concept of the invention, which fall within the scope of protection of the invention.
Claims (7)
1. The automatic updating method of the simulation model driven by the geometry-grid twins is characterized by comprising the following steps of:
The method comprises the steps that firstly, finite element processing is conducted on a reference geometric model A to obtain a corresponding reference simulation model, and then CAD modification is conducted to obtain a geometric model B modified by the reference geometric model;
Secondly, adopting unit types of triangles to carry out grid division on the reference geometric model A and the geometric model B, and establishing curved surface grid models corresponding to the two geometric models; respectively carrying out plane parameterization on the two curved surface grid models by a quasi-conformal mapping method to obtain two mapped plane grid models; the method comprises the following steps:
Step 2.1: the method comprises the steps of performing grid division on a reference geometric model A by adopting triangular grid units to obtain a curved surface grid model A tra of the geometric model A, selecting 4 reference points from outline nodes of the curved surface grid model A tra, performing plane parameterization by adopting a quasi-conformal mapping method to obtain a rectangular plane grid model A tra-Plane, wherein the node numbers and unit types of the curved surface grid model A tra before plane parameterization and the grid model A tra-Plane after plane parameterization are unchanged and correspond to each other;
Step 2.2: grid division is carried out on the modified geometric model B by adopting triangular grid units to obtain a curved surface grid model B tra of the geometric model B, 4 references are selected from outline nodes of the curved surface grid model B tra, planar parameterization is carried out by adopting a quasi-conformal mapping method to obtain a rectangular planar grid model B tra-Plane, the node numbers and unit types of the curved surface grid model B tra before planar parameterization and the planar grid model B tra-Plane after planar parameterization are unchanged and correspond to each other;
Thirdly, acquiring control point sets of two plane grid models based on a point cloud sampling method, and acquiring the control point sets of the two curved surface grid models according to the corresponding relation of node numbers; the method comprises the following steps:
Step 3.1: sampling the nodes of the planar grid model A tra-Plane obtained in the step 2.1 by adopting a point cloud sampling method, setting the number of sampling points in the length direction as m and the number of sampling points in the width direction as n, obtaining m x n sampling point coordinates, calculating Euclidean distances between all grid nodes and the sampling points, and selecting the grid node with the smallest Euclidean distance from the sampling points; further obtaining nodes of m x n planar rectangular grids; the node numbers and the unit types of the plane parameterized front surface mesh model A tra and the plane parameterized front and back surface mesh model A tra-Plane are not changed; therefore, according to the correspondence of the node numbers, m×n nodes of the curved surface mesh model a tra corresponding to the reference geometric model a can be obtained as a control point set of the reference geometric model a;
step 3.2: sampling the nodes of the planar grid model B tra-Plane obtained in the step 2.2 by adopting a point cloud sampling method, setting the number of sampling points in the length direction as m and the number of sampling points in the width direction as n, obtaining m x n sampling point coordinates, calculating Euclidean distances between all grid nodes and the sampling points, and selecting the grid node with the nearest Euclidean distance; further obtaining nodes of m x n planar rectangular grids; the node numbers and the unit types of the plane parameterized front surface mesh model B tra and the plane parameterized rear surface mesh model B tra-Plane are not changed; therefore, according to the corresponding relation of the node numbers, m×n nodes of the curved surface mesh model B tra corresponding to the geometric model B can be obtained and used as a control point set of the geometric model B;
Training the coordinate relationship of the two curved surface control point sets in the third step by using a fitting method to obtain a mapping relationship between the two curved surface control point sets;
Fifthly, transforming coordinates of the finite element nodes in the reference simulation model based on the mapping relation, realizing twin of the modified grid model, and automatically inheriting finite element setting of the reference simulation model in the first step to complete automatic updating of the simulation model; the method comprises the following steps:
based on the mapping relation between the control point sets of the reference geometric model A and the geometric model B before modification, which are obtained in the fourth step, carrying out coordinate transformation on the grid information in the reference simulation model A Fem established in the first step, wherein the input of the mapping relation is all grid node coordinates of the reference simulation model A Fem, and the output is the grid node coordinates corresponding to the geometric model B after modification; on the basis, the finite element setting information in the first step is integrated, automatic updating of the simulation model is completed, and the simulation model B Fem corresponding to the geometric model B is obtained and used for subsequent simulation analysis and structural performance evaluation.
2. The method for automatically updating a simulation model driven by geometry-grid twinning according to claim 1, wherein: the first step, obtaining a reference geometric model through CAD design, obtaining a corresponding reference simulation model through finite element processing, and obtaining a geometric model modified by the reference geometric model through CAD modification; specific:
Step 1.1: establishing a reference geometric model A through CAD three-dimensional design software according to the design requirement of the structure, and outputting the reference geometric model A in a geometric model format;
Step 1.2: performing geometric import, grid division, material attribute setting, assembly and connection relation setting in finite element analysis software aiming at the obtained reference geometric model A to obtain a reference simulation model A Fem, wherein the reference simulation model comprises finite element grid model information and finite element setting information;
Step 1.3: modifying the reference geometric model A according to the requirements of engineering design, obtaining a modified geometric model B by modifying the shape, changing the size and adjusting the angle position of the geometric model, and outputting the modified geometric model B in the format of the geometric model;
The fourth step is to train the position coordinate relation of the two geometric model control point sets in the third step by using a fitting method to obtain the mapping relation between the two geometric model control point sets; specific:
Training the fitting relation of the two geometric model control point sets in the third step by adopting a fitting method, wherein the input of the training process is the coordinate value of the reference geometric model control point set obtained in the step 3.1, and the output is the coordinate value of the modified geometric model control point set obtained in the step 3.2; for a three-dimensional structure, the input coordinate values should include coordinates in three directions of x, y and z, namely m×n three-dimensional vectors;
After training, obtaining a mapping relation between a control point set of the reference geometric model A before modification and the geometric model B after modification; the mapping relation refers to a fitting relation between coordinate values of the reference geometric model control point set and coordinate values of the modified geometric model control point set.
3. The method for automatically updating a simulation model driven by a geometry-grid twinning according to claim 2, wherein in the step 1.2, the finite element grid model information comprises a unit type, a node number and a node set, and the finite element setting information comprises a material property, a connection relationship, a load, a boundary condition and an analysis type.
4. The method for automatically updating a simulation model driven by geometry-grid twinning according to claim 1, wherein the control point set in the step 3.1 and the step 3.2 is a description of the contour of the geometric model.
5. The method for automatically updating a simulation model driven by geometry-grid twinning according to claim 1, wherein the point cloud sampling methods in the step 3.1 and the step 3.2 are the same, and the method comprises a grid sampling method, a uniform sampling method or a set sampling method.
6. The method for automatically updating a simulation model driven by geometry-grid twinning according to claim 1, wherein the control point set of the geometric model a in the step 3.1 is a reference geometric model control point set; and 3.2, the control point set of the geometric model B in the step is a modified geometric model control point set.
7. The method for automatically updating a simulation model driven by geometry-grid twinning according to claim 2, wherein in the fourth step, the fitting method comprises a radial basis function neural network, a machine learning method of a back propagation neural network or a proxy model method based on a radial basis function method and a kriging method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410195563.4A CN117763927B (en) | 2024-02-22 | 2024-02-22 | Automatic updating method for simulation model driven by geometry-grid twinning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410195563.4A CN117763927B (en) | 2024-02-22 | 2024-02-22 | Automatic updating method for simulation model driven by geometry-grid twinning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117763927A CN117763927A (en) | 2024-03-26 |
| CN117763927B true CN117763927B (en) | 2024-05-17 |
Family
ID=90314831
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410195563.4A Active CN117763927B (en) | 2024-02-22 | 2024-02-22 | Automatic updating method for simulation model driven by geometry-grid twinning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117763927B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111027250A (en) * | 2019-12-11 | 2020-04-17 | 大连理工大学 | Special-shaped curved surface reinforced shell modeling method based on grid deformation technology |
| US10657301B1 (en) * | 2016-02-17 | 2020-05-19 | National Technology & Engineering Solutions Of Sandia, Llc | Systems, methods and computer program products for constructing complex geometries using layered and linked hexahedral element meshes |
| CN115238392A (en) * | 2022-08-09 | 2022-10-25 | 大连理工大学人工智能大连研究院 | Free deformation method-based curved surface reinforcement shell modeling method and system |
| CN116579040A (en) * | 2023-06-16 | 2023-08-11 | 国网甘肃省电力公司 | Method for optimizing geometry and material parameters of insulator based on finite element simulation |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111783241B (en) * | 2020-06-16 | 2022-09-30 | 大连理工大学 | Prediction method for thin-wall micro-milling deformation |
-
2024
- 2024-02-22 CN CN202410195563.4A patent/CN117763927B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10657301B1 (en) * | 2016-02-17 | 2020-05-19 | National Technology & Engineering Solutions Of Sandia, Llc | Systems, methods and computer program products for constructing complex geometries using layered and linked hexahedral element meshes |
| CN111027250A (en) * | 2019-12-11 | 2020-04-17 | 大连理工大学 | Special-shaped curved surface reinforced shell modeling method based on grid deformation technology |
| CN115238392A (en) * | 2022-08-09 | 2022-10-25 | 大连理工大学人工智能大连研究院 | Free deformation method-based curved surface reinforcement shell modeling method and system |
| CN116579040A (en) * | 2023-06-16 | 2023-08-11 | 国网甘肃省电力公司 | Method for optimizing geometry and material parameters of insulator based on finite element simulation |
Non-Patent Citations (6)
| Title |
|---|
| Automatic and high-quality surface mesh generation for CAD models;Jianwei Guo et al.;Computer-Aided Design;20191231;全文 * |
| 参数化网格划分方法研究及其***实现;谢世坤;黄菊花;桂国庆;郑慧玲;;中国机械工程;20070215(第03期);全文 * |
| 基于SolidWorks-MATLAB-ANSYS集成框架的结构设计优化;李珺;杨扬;;计算机应用与软件;20200312(第03期);全文 * |
| 水工结构数值仿真等几何分析方法初探;李明超;张梦溪;;水利学报;20180129(第03期);全文 * |
| 航空航天结构轻量化设计与实验方法研究进展;王博等;宇航学报;20230430;全文 * |
| 面向注塑模CAE的有限元网格自动生成技术;朱晓靖;卢炎麟;王玉槐;付飞均;;轻工机械;20090815(第04期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117763927A (en) | 2024-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10740509B2 (en) | Machine learning three-dimensional fluid flows for interactive aerodynamic design | |
| US20080275677A1 (en) | System, methods, and computer readable media, for product design using coupled computer aided engineering models | |
| CN108229085A (en) | computer simulation analysis system and method | |
| EP2869226A1 (en) | Updating a CAD model to reflect global or local shape changes | |
| KR20160074427A (en) | Interactive 3d experiences on the basis of data | |
| CN107229794B (en) | Model construction system based on CAE and VR and management method thereof | |
| CN115688276A (en) | Aircraft appearance automatic optimization method, system, equipment and medium based on discrete companion method | |
| Hofmeyer et al. | Automated design studies: topology versus one-step evolutionary structural optimisation | |
| US9965574B2 (en) | CAD-based initial surface geometry correction | |
| CN111310344B (en) | Method for considering coupling effect of wind field and fire field in forest fire spreading simulation | |
| CN115774968A (en) | Curved surface mesh generation method and system based on recursive decomposition and computer equipment | |
| CN110334450B (en) | Method for repairing object plane projection error in multi-block structure grid generation | |
| US20210141869A1 (en) | Automated Analysis of Mechanical Designs | |
| CN113486429B (en) | Automatic construction method of space intersection structure based on interpolation algorithm | |
| Pernot et al. | A shape deformation tool to model character lines in the early design phases | |
| Wong et al. | Graph neural network based surrogate model of physics simulations for geometry design | |
| CN117763927B (en) | Automatic updating method for simulation model driven by geometry-grid twinning | |
| JP3221137B2 (en) | Analysis device | |
| CN114676615B (en) | Tool compensation profile construction error assessment method, device, equipment and medium | |
| CN108197368B (en) | Method for simply and conveniently calculating geometric constraint and weight function of complex aerodynamic shape of aircraft | |
| CN115130340A (en) | Pipeline modeling method based on fractional Brownian motion | |
| WO2021222697A1 (en) | Method and apparatus for structural design | |
| CN110555267A (en) | Parameterized horizontal set structure topology optimization method based on implicit B-spline | |
| CN117576339B (en) | Unstructured grid generation method with highest resolution of 1km | |
| CN109684675B (en) | Method for analyzing strength of aviation thin-wall support based on abaqus plate coupling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |