JPH0363880A - Object modeling system - Google Patents

Abstract

PURPOSE:To fractionize each element of a model for analysis further at need and to improve analytic accuracy by a finite element method by providing an object model generating part and a model generating part for analysis. CONSTITUTION:When an operator inputs data via input means 12 and 13, the object model generating part 21 and the model generating part 23 for analysis are operated, and the data of an object model is generated and taken out. And the data 33 of an input model for finite element analysis and the data 34 of an output model for finite element analysis are generated sequentially, and they are inputted to a data base 30, respectively. The data of a fractional model which fractionizes a prescribed element targeted to be fractionzed out of the elements of the input model for finite element analysis is generated by using generated data, and the generated data of the fractional model is inputted to the data base 30 of the input model for finite element analysis. In such a way, the analytic accuracy can be improved.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is directed to CA E (Computer Aided
Engineer-ing), CAD (Co+n
putter Aided Design), CAM(
Computer Aided Manufacturer
The present invention relates to an object modeling system that is the core of the construction of an automation system for various tasks in design and production using computers, such as a computer-aided computer-based system, and particularly relates to a method for subdividing a model for finite element analysis.

[Conventional technology]

Conventionally, when a model for finite element analysis is generated and a part of the model is subdivided, the subdivided model includes many triangular elements.

The above subdivisions are described in ``Automatic Design Methodology J Noriobu Okino (Yokendo)''.

[Problem to be solved by the invention]

In the above subdivision method, the model after subdivision will contain many subdivision elements of trigonal or tetrahedral elements, and the analysis data generated from it will be a low-order approximation. When performing element analysis, it was often the case that sufficient analysis accuracy could not be obtained.

Further, even when editing the analysis results and outputting them to a CRT or the like, there is a problem in that for models including trigonal or tetrahedral elements, the results at the model position intended by the user cannot be output with sufficient accuracy.

An object of the present invention is to provide an object modeling system that can generate a model that improves analysis accuracy when subdividing an object model and converting it into a model suitable for analysis.

[Means to solve the problem]

In order to achieve the above object, the present invention includes an object model generation section and an analytical model generation section, the object model generation section generates data of an object model for an analysis target, and the object model generation section generates object model data for an analysis target. In the object modeling system, the model generation unit generates data of an analysis model for finite element analysis using the object model data, and the analysis model generation unit generates data of an analysis model of the generated analysis model. The present invention is characterized in that data of an analysis model is generated by further subdividing a predetermined subdivision target element.

The analysis model generation unit may include means for evaluating the analysis accuracy of the generated analysis model and determining the predetermined subdivision target element based on the evaluation result.

In this case, the means for determining the subdivision target elements determines whether or not the analysis error or stress exceeds a predetermined range for each element of the generated analysis model, and designates the elements that exceed the predetermined range to be subdivided among the subdivision target elements. It can be an element.

Further, the predetermined subdivision target element may be given by an operator6.Furthermore, the subdivision target element determining means may include an element adjacent to the subdivision specified element group in the subdivision target element. be.

It is desirable that the subdivision target element be subdivided using a predetermined subdivision pattern.

In this case, it is desirable that the predetermined subdivision pattern has a minimum subdivision pattern of quadrilateral or hexahedral elements.

[Effect]

When an operator inputs data via an input means,
The object model generation unit operates and generates object model data. Next, when the operator inputs data through the input means, the analysis model generation unit operates, and the data of the object model is taken out, the data of the input model for finite element analysis, the data of the output model for finite element analysis. are generated in sequence and input into the database. Using the data of the generated object model, the data of the input model for finite element analysis, the data of the output model for finite element analysis, etc., the predetermined subdivision target elements among the elements of the input model for finite element analysis are further subdivided. Since the data of the subdivision model is generated and the data of the generated subdivision model is input into the database of the input model for finite element analysis, the accuracy of analysis is improved.

In addition, when subdividing, the element to be subdivided is subdivided using a subdivision pattern consisting of quadrilateral and hexahedral elements, and the elements surrounding the subdivision element are divided using a subdivision pattern of quadrilateral and hexahedral elements, so the user can It is possible to obtain highly accurate finite element analysis output results.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings.

A. Basic System Configuration FIG. 1 shows the system configuration of an embodiment of the present invention, in which input/output display means 10. It includes a calculation result 20° database 30. The input/output display means 10 includes, for example, a graphic display 11°, a stylus 12, and a keyboard 13. In this example, stylus 12
is used, but it is also possible to change this to a mouse. Also.

The database 30 contains object model data that is data modeling an object as an analysis object, analysis input model data obtained by converting this object model data 31 into a model suitable for analysis, and information for outputting and displaying analysis results. Databases 31 and 33 each store analysis output model data converted into a suitable model.
It is equipped with 34. The computing device 20 includes an object model generation unit 21 that generates an object model to be analyzed using object model data.

Analysis input model generation unit 2 that generates an analysis input model using object model data and analysis input model data
3, and an analytical output model generation section 24 that generates an analytical output model from the object I-model data, analytical human model data, and analytical output model data.

A1. Analysis Model Generation The analysis model generation operation in the analysis system of this embodiment will be described below.

(1) As shown in FIGS. 1 and 2, first, the graphic display 11. The arithmetic device 20 via input means such as a tablet and a stylus 12 or a keyboard 13.
Activate. As a result, the object model generator 2
1, generate the object model data and input it into the object model data database 31 (
Step 201).

Next, the object model data generated in step 201 is retrieved from the database 31, and using this data, the analysis input model generation unit 23 performs
Generate analysis input model data and input it into the analysis input model data database 33 (step 20
2).

Next, the analysis input model data generated in step 202 is retrieved from the database 33, and the analysis output model generation unit 24 generates analysis output model data using this data. (Step 203).

According to this embodiment, the analysis input model generation unit 23 uses object model data in the database 30 to generate an analysis human model of the analysis target suitable for the analysis content,
Since the results are output to the database 33, the analysis can be performed with high precision and the analysis processing work becomes easy.

(2) FIG. 3 shows another embodiment related to the generation of an analytical model. The finite element analysis model is automatically generated by automatically dividing the image using the shape function.

Specifically, as shown in FIGS. 3 and 4, first, object model data is extracted from the database 31, and
The divided object model generation unit 22 converts the divided object model data and inputs it into the database 32 (step 2021). Here, the split object model is a model for automatically splitting the object model (generating a model for finite element analysis), and this model cannot describe hole shapes, etc.

Next, the divided object model data is extracted from the database 32, and the analysis input model generation unit 22 generates analysis input model data, which is input into the analysis input model data database 33 (step 2022).
.

According to this embodiment, a divided object model is generated using object model data, and this divided object model is automatically divided by the mapping method, so that automatic division of the object model is facilitated.

(3) FIG. 5 shows yet another embodiment regarding analysis model generation. In this embodiment, the process (step 203) of generating analysis output child model data from the analysis input model data in FIG. This is how it was done.

That is, as shown in FIGS. 5 and 6, input model data for analysis is first retrieved from the database 33, and input card image model data for each analysis program is generated in the input card image model generation section 25. is input into the input card image model data database 35 (step 2031).

Next, the input card image data is extracted from the database 35, and the analysis output model generation unit 24 generates the analysis output model data using each analysis program, and inputs this into the analysis output model database 34 (step 2032). ).

According to this embodiment, since the analysis input model data is once converted into input card image model data of each analysis program, connection with various analysis programs is facilitated.

(4) The data for each model in Figures 1, 2, and 5 are as follows:
As shown in FIG. 7, each generation processing section generates a geometric model and attribute data of the constituent elements of the model, and stores them separately in corresponding areas of the database 30.

However, when the components of a geometric model and their attribute data are related, and the data of each model includes the data of that component, the attributes of the component of the model generated in the next step are The data is automatically generated from the attribute data of the model and its components generated in the previous step.

Further, as shown in FIG. 8, the attribute database shown in FIG. 7 is distributed and stored in two, an input attribute database and an output attribute database. However, the attribute data stored in each attribute database is associated with the constituent elements of the geometric model.

By doing this, as shown in Fig. 9, even in various analyzes such as structural/vibration analysis, if data on the attributes of the geometric model (attributes of the constituent elements) is prepared,
Geometric model data for finite element analysis and boundary element analysis can be used in common in each analysis process. In addition, with this configuration, as shown in Figure 10, for example, output model data (geometric model and its input attribute data) for fluid analysis is generated, and this data is used as input attribute data for structural/vibration analysis. Various combinations can be analyzed, such as inputting data into a database and performing structural/vibration analysis using data in the generated input attribute database.

Note that the attributes of the geometric model in FIGS. 7 to 10 include the analysis program name (input attribute), data specific to the analysis program, and the like.

A2. Incorporation into the CAE Database Here, an example of the description of the basic configuration of CAE and the geometric object model used therein (consisting of geometric model data and attribute data of its constituent elements) will be shown. Also,
It shows which object model of the CAE system the data of each model in the database 30 of FIGS. 1, 3, and 5 corresponds to, and which object model data is generated from.

(1) Basic configuration FIG. 11 is an example of an internal model of data of a geometric object model used in CAE, in which each data is distributed and stored in a database (abstract complex chain model). In other words, it is a distributed database. The data of each model can also be used for data conversion between family systems, and can be used for calculations on geometric object models (basic operations and their combinations) including figures, cell complex chains, and abstract cell complex chain operations. Can perform data conversion. The arrow → in the figure indicates a conversion operation from one data to another, and indicates that chain conversion is also possible between these data.

(2) Data description of target models Among the target models shown in Figure 1, part model 3 is the target model that becomes the basis for their configuration. Therefore,
First, the data description of part model 3 will be explained, and then based on this, the data descriptions of other models will be explained.

(a) Part model 3 The geometric model of part model 3 is cell complex chain A (Er)
(r=o ~ 3), and the data is the drafting model, part model 1, part model 29, part model 4, which will be described later.
can be generated from the data.

FIG. 12 is an example in which data of part model 3 is described by one internal model (represented in table format) of an abstract cell complex chain (database model). Each table name in Figure 12 (
For example, 3 internal chains), symbols (C, etc.) are the names of the components of the oil cell complex chain, and each table shows the cell complex chain A (E
r) component data (realized values) are stored.

The cell complex chain table manages each table that stores cell complex chain data, and stores data such as a pointer representing the address of each table and the number of elements stored in the table.

4 Each row (record) of the internal chain (connected component) table contains O~
Stores data such as start and end pointers to a three-boundary cycle chain (connected component) table in which data of elements of three external chains are arranged two-dimensionally. 4. The internal chain (connected component) table stores data for different parts. That is, the data stored in the three boundary cycle chain (connected component) tables corresponding to each row (record) of the four internal chain (connected component) tables is different. The 3-boundary cycle chain (connected component) table stores data such as pointers to the O-3 external chain table, which is a two-dimensional arrangement of the data of the O-3 external chain elements of the cell complex chain. .

Other tables include other cell complex chain components (external chains, internal chains,
Boundary cycle chain) data is stored. In particular, the cell body εk
When the r space rat' (see Fig. 12) that defines r is a subspace, for example, a free-form surface patch 221 as shown in Fig. I3, the free-form surface patch ε2' is defined as A(c2' )
Describe with data. In this case, A(
Add a table of the internal model of another abstract cell complex chain that stores the data of A (station'), and add the data of the internal chain of A (station') to the row of the internal chain table in Figure I2, as shown in Figure 14. Stores data such as pointers to tables that store .

Figures I5 to 18 are examples of cell complex chains expanded so that the shapes can have holes.

Fig. 15 shows an example of cell bodies 1c and l with holes, Fig. 16 shows an example of the cell body with holes of Fig. 15, and Fig. 17 shows an example of cell bodies 1c and 1 with holes.
This is a cell complex chain produced from cell complex K (lcl I) in Figure 6. In addition, in describing the data of the cell complex chain expanded in this way, as shown in FIG. 18, the r-1 internal cycle chain table is added to the table in FIG. In (row), the data of the pointer to the internal chain table in r-1 and the data of its attributes are collectively stored. Each element (row) stores pointer data to the r-1 internal chain table and its attribute data.

FIGS. 19 to 23 are examples in which the cell complex chain is expanded so that a combination model of wire frame, surface, and solid can be easily created.

It is described by the wire frame shown in Fig. 19, the surface shown in Fig. 20, and the IC shown in the second engineering drawing.

Describe. In addition, to describe the data of the cell complex chain expanded in this way, as shown in FIG. 22, add the combined cell table to the table of FIG. 12, and write r- 1,・
...0 pointer data to the binding cell table, and r-1
, . . . In each element (row) of the O-bond cell table, data of a pointer to an element of the internal chain table of the same dimension is stored.

FIG. 23 shows an example of a finite element or boundary element model (described later) generated by subdividing the cell complex chains in FIGS. It can be connected to the upper node Jel.

Furthermore, the above-mentioned conjugate cell is also used to describe boundary conditions such as geometric constraints of an object model. In this case, the boundary condition data is made to correspond to the data of the internal chain elements in FIG. 12, which correspond to the data of the bound cell stored in the bound cell table of FIG. 22.

In this part model 3, a complex structure such as a fluid machine is constructed using a so-called wire frame.

You can model surfaces, solids, and their combination models. In addition to objects, it is also possible to model fluids outside the objects. It is also possible to describe combination models of structures and fluids. Furthermore, it is possible to describe (topological) data on the coupling relationships between these models and their components.

Therefore, it is particularly suitable for modeling complex structures such as fluid machines.

(b) Finite element (boundary element) model The finite element (boundary element) model is a typical analysis input/output model, and its data is stored in the database 33 in Figure 1.
．． The information is stored in the database 34.

Here, the finite element model is a subdivision of the part model 3, and can be described by a cell complex chain A (E') (r = o ~ 3),
The data can also be stored in each table in FIG. However, the number of elements of the boundary cycle of part model 3 and the type of cells are limited. For example, the planes of solid elements are usually tetra- to hexahedral, and the planes of shell elements are 3°4
Restricted to sides. In addition, the shape of the constituent surfaces is a 1st to 3rd order curved surface. The shape of the component is limited to linear to cubic curves. In other words, the number of data items in the 1-2 boundary cycle table in FIG. 12 is limited. Also, the constituent surfaces and constituent solids do not have holes. In other words, the inner cycle data shown in FIG.
It does not have data on the connective cell shown in FIG.

Due to these restrictions, the data of A (E') is
In Figure 2, the data of l~3 internal internal data and 1~3 boundary cycle chain correspond to the elements of the outer chain, and the elements are ○
It can also be described by replacing the data with a set of ordered node sequences corresponding to the elements of the internal chain. For example, the data of these node columns are described by pointers to rows of the O internal chain table. In addition, the data of 2-3 internal chains and 2-3
It can also be described by replacing the data of the 3-boundary cycle chain with data of an ordered set of edge sequences corresponding to the elements of the outer chain, whose elements correspond to the elements of the O inner chain. For example, the data of these edge line columns are described by pointers to rows of the l-internal chain table.

The data of the finite element model is usually generated using the data of the part models 2 to 4. In this case, in each row inside O~3 that stores the data of the finite element model, number data of the corresponding component model component (O~3 internal element) is stored. This number data indicates which component (O~3 internal internal element) of each finite element is generated from the subdivision of which component (O~3 internal internal element) of the part model (for example, part model 3). This is data showing.

In this way, by associating each component of the finite element model with the component of the part model, it is possible to extract, for example, data on the component surface of the part model 3 that includes a certain surface element of the finite element model. In addition, the component model 3 and the finite element model can be treated as one, making it possible to easily extract the elements of the finite element model that correspond to the components of the component model 3, and for example, to process their CRT display at high speed. can.

Note that the divided object model (cell complex chain A (Er) (r=0 to 3) stored in the database 32 in FIG.
) and the above finite element (boundary element) model have the same data structure, except for the fineness of the subdivision.

That is, a finite element (boundary element) model is generated by subdividing the divided object model.

('C) Others The above-mentioned component model 3 (cell complex chain) includes a plurality of target models shown below. Further, these basic operations are also included in the basic operations of the component model 3. These target models can be abstract models (semantic models of data) of parts models and product models, respectively (however, their frameworks are different).

(a) Subset of cell complexes (b) Partial cell complexes (c) External chains of cell complex chains (d) External chains of cell complex chains and their internal chains (e) Internal chains of cell complex chains (f) In the set (a) of the internal chain of the cell complex chain and its boundary cycle chain, the data (realized value) is stored in the O~3 external chain table, the O~3 internal, and the 0~3 internal chain The rows (records) of the table store data for O to 3 cells. Also, 0 to 3
Integer coefficients of the outer chain table.

Ignore coupling coefficients.

In (b), the same data (realized values) as in the case of the cell complex chain are stored in the O~3 internal and O~2 boundary cell chain tables,
O~3 Ignore integral coefficients and coupling coefficients in the external chain table.

(c) is the O~3 external chain table, the 0~3 internal chain table, and the data (
actual value), and the records of the 0 to 3 internal chain table contain O to
Stores data of 3 cells.

(d) to (f) store the same data (realized values) as the pectoral collar body chain in each component, and in (e) and (f), O
~3 Ignore the integral coefficients and coupling coefficients of the external chain table.

Each of the above models is based on its characteristics, such as CAE analysis system represented by finite element analysis (see Figure 1), process design, NC tape creation, robotics, etc.
Connected to the AM system.

Note that the above data model can also be used for data description of intermediate files for data conversion between different systems.

B. High-precision analysis system The configuration of a high-precision analysis system based on an object model will be described below with reference to FIGS. 24 and 25. However, for the sake of convenience, the object model in the figure is the highly versatile component model 3 (three chest-neck body chains) in FIG. 11.

B1. Subdivision of the analysis model As shown in FIGS. 24 and 25, first, object model data is generated in the object model generation unit 21, and this is stored in the object model data database 3↓
(Step 201).

Next, the object model data generated in step 201 is extracted, and using this data, the analytical human model generation unit 23 generates analytical human model data, which is input into the analytical input model data database 33. (Step 202).

Next, the object model data generated in step 201 and the analytical human model data generated in step 202 are extracted, and using these data, the analytical output model generation unit 24 generates data for the analytical output model. , this is input into the analysis output model data database 34 (step 203).

Here, detailed processing contents of step 203 will be explained. First, the analysis input model data generated in step 2027 or step 2024A described later is retrieved from the database 33, and based on this, the analysis output model generation processing unit 24 generates analysis output model data, and this is converted into analysis output model data. database 34 (step 2031).

Next, the data of the analytical input model (geometric model and input attributes) generated in step 202 or step 2034A described later is obtained from the database 33, and the data of the analytical output model generated in step 2031 ( Using these data, the analysis input model generation processing unit 23 creates analysis error data for each element of the analysis output model (analysis input model) (step 2032').

Next, it is determined whether the analysis error of each element generated in step 2032 is smaller than a predetermined value (step 2033). If the analysis error of each element is the same as or smaller than the predetermined value, the process is terminated, and if the analysis error of each element is greater than the predetermined value, proceed to the next step 20.
Proceed to step 34.

Next, the object model data (geometric model data) generated in step 201 is retrieved from the database 31, and the object model data and the analytical input model (geometric model and input data of the object model generated in step 201, step 20
Using the analysis error data of each element generated in 32,
The analytical input model generation processing unit 23 subdivides the analytical input model data, generates new analytical input model data that is a subdivision of the object model, and inputs this data into the analytical input model database 33 (step 203
4A).

Proceed to the next step 2031.

In the above configuration, the data of the component of the object model and the data of the component of the already created analysis input model are associated, so new analysis input that becomes a subdivision of the object model by subdivision of the already created analysis input model Models can be easily generated.

In addition, in the above, for example, extracting the constituent elements of the high stress value part of the structural analysis in the object model,
By applying these methods to those components, the processing speed can be increased.

Subdivision C1 of C1 finite element (boundary element) model, basic processing configuration As shown in the processing procedure of FIG. 26, first, the graphic display 11. Arithmetic device 20 via input means such as a tablet and stylus 12 or keyboard 13
is activated, the object model generation unit 21 generates object model data, and this object model data is input into the database 31 (step 20
1).

Next, the data of the object model generated in step 201 is extracted from the database 31 and the data of the analysis input model is extracted from the database 33, and using these data, the analysis hole model generation unit 23 generates the data of the analysis input model. This is then input into the analysis input model database 33 (step 301).

In this step, first, data of the object model generated in step 201 is retrieved from the database 31, and using this data, the analysis hole model generation unit 23 generates data of the analysis input model, and this data is used as the analysis input model data. Input into database 33 (step 2
02).

Next, it is determined whether the analysis input model generated in step 202 is appropriate, and if appropriate, proceed to step 203, and if inappropriate, proceed to the next step 303 (step 302).
, this determination can be made by an operator by outputting and displaying the generated model on the graphic display 11, and by performing steps 2031 to 2033 in FIG. 25 described above.
A method using error judgment similar to that described above can be applied.

In step 303, from the database 31 in step 2
The data of the object model generated in step 01 is extracted, and the analysis input model is subdivided in the analysis hole model generation unit 23 using the data of the analysis input model, and the data of the analysis input model that is the subdivision of the object model is generated. This is input into the analysis input model database 33 (step 303).

In step 203, data of the analysis input model is extracted from the database 33, and using this data, the analysis output model generation unit 24 generates data of the analysis output model, and inputs this into the database 34 of analysis output model data ( Step 203).

Next, the method of generating the analytical input model or divided object model (subdivision operation) in step 303 of FIG. 26 will be explained with reference to FIG.

C2. Subdivision by designation of subdivision element group As shown in FIG. (for example, the data generated in step 202 in FIG. 26) and the data of the analysis output model generated in step 203 from the database 34, and using these data, the analysis hole model generation unit 23 generates an analysis input model. Generate subdivision target elements for (
Step 401). This subdivision target element group includes a subdivision specified element group and surrounding element groups, which will be described later.

In this step 401, as shown in FIG. 28, first, a subdivision specified element group of the analysis input model is automatically generated using data of the analysis output model (for example, stress data or analysis error data of structural analysis). do. Alternatively, the operator specifies a region including a group of subdivision designation elements, and generates a group of subdivision designation elements included therein (step 4011).

Next, the elements surrounding the subdivision specified element group generated in step 4011 are edited (see step 4022 described below).
, a subdivision target element group (subdivision designated element group and surrounding element group) is generated (step 4012).

Next, the analysis input model generation unit 23 subdivides the subdivision target element group of the analysis human model generated in step 401,
Generate a new analytical input model and input it into the analytical input model data database 33 (step 401
2).

In this step, as shown in FIG. 28, each element of the subdivision specified element group of the analysis input model generated in step 4011 is first subdivided using the basic subdivision pattern shown in FIGS. 29 and 30 (step 4021). ). At this time, as shown in FIG. 31, the generated model generally does not satisfy the configuration conditions of the analytical model.

That is, for example, the number of constituent sides of the elements surrounding the subdivision designated element group is not 3 or 4.

Therefore, next, elements around the subdivision specified element group are subdivided using an appropriate basic subdivision pattern shown in FIG. 32, for example, to create a subdivision model (data model) that satisfies the configuration conditions of the analysis input model, as shown in FIG. 33. (step 4022)
.

C3. In order to improve the accuracy of subdivision finite element analysis by selecting a subdivision pattern or the output accuracy of the results, the subdivision specified element group and the surrounding element group should be
It is desirable to subdivide a subdivision basic pattern consisting of quadrilateral elements that can be approximated to a higher order. Below, we will explain a specific method of sequentially subdividing each element around the subdivision specified element group in the pattern shown in Fig. 32(b) along the boundary of the subdivision specified element group as shown in Fig. 35. The flow will be explained in order.

(1) Selection of subdivision target element groups and their division patterns (step 401).

First, step 4011) of generating data of a plurality of subdivided elements (subdivision designated element group) (e=) shown in FIGS. 36 and 37 by specifying a subdivision area.

Next, the entire element group (whole element group) of the subdivision target model shown in Figs. 36 and 38, and the subdivision designated element group (
At ), first generate data for the boundary (bi ) of the entire element group (B1) shown in FIGS. The data is subdivided into the patterns shown in FIG. 29 or 30, and subdivision data of (eI) is generated (step 4021).

Next, (et) (j=t, 2.-, Nb) is sequentially subdivided along the boundary in the pattern shown in FIG. 32 (step 40).
1). In this case, as shown in FIG. 41, the following pattern is created (step 4012A).

Tube 401 of the entire element group (es) with Nb) as the boundary
2B). These are shown in Figures 36 and 40 (e
J).

(2) Subdividing the subdivision target element group and subdividing the subdivision model into raw patterns and matching the whole.

At this time, (i) As shown in Fig. 42, when eI is a trigonal element, first subdivide the triangular element eJ in the pattern shown in Fig. 29, and then divide e J+1 into the next part of ea-1. Think of it as an element and subdivide it.

(ii) As shown in FIG. 43(a), including eJ, when e J-1 has the pattern in FIG. 43(b), subdivide eJ into the pattern in FIG. 43(b).

- When e4-1 has the pattern shown in FIG. 43(c), subdivide eJ into the pattern shown in FIG. 43(c).

Moreover, as shown in FIG. 44, the ei , use eJ as a subdivision specification element.

field data and newly subdivide from e J-1. Further, 8J is subdivided into the pattern shown in FIG. 32(C).

When the number of elements in (el) including 5 nodes (bl) is three or more, in addition to (ej) with j of these eJ, data of a new boundary element group is generated, and the new generated e
Subdivision is performed starting from J+1.

(iv) When the total number Nb of (eJ) is odd, the final requirement eNb
The subdivision makes it impossible to maintain consistency as a whole. Therefore, in this case, subdivision is performed using the pattern shown in FIG. 32(a).

(V) The newly generated node is the boundary of the entire element group 1. b
i) and the boundary (bt) of the subdivision specified element group
, these nodes are moved to the boundary of the original target model.

If so, the above processing is performed independently for each.

FIG. 46 shows a subdivision model generated by the above method when the target model shown in FIG. 36 and the data of the subdivision target element group are given.

Although the above is a two-dimensional case, this idea can be similarly applied to shell element models and solid element models in three-dimensional space.

FIG. 47 is an example of subdivision of a shell element model in three-dimensional space.

FIG. 48 is a typical example of a subdivision pattern of a subdivision target element group of solid elements.

FIG. 43 is an example in which a solid object model is subdivided into corners where three surfaces intersect using the pattern shown in FIG. 48.

FIG. 50 is an example in which a solid object model is subdivided into a corner where two surfaces intersect with a subdivision pattern consisting of the set of patterns shown in FIG. 48.

The fifth engineering drawing is an example of a solid object model in which one boundary surface is subdivided using a subdivision pattern consisting of the set of patterns shown in FIG.

In this way, the idea of subdividing existing elements using a predetermined subdivision pattern by combining these subdivision patterns can also be applied to a solid target model.

D. Subdivision of Object Models Object models are different from basic object models and models for finite element analysis, and the above-mentioned subdivision method cannot generally be applied to them. However, it can be used when generating a basic object model from an object model (cutting the model). FIG. 52 is an example of an object model, and FIG. 53 is an example of subdivisions generated by instructions of points ij...op on the boundary of the object model in FIG. 4 in model subdivision
It consists of a set of two basic object elements (hexahedrons).

〔Effect of the invention〕

As described above, according to the present invention, each element of the analytical model is further subdivided as necessary, so that the accuracy of analysis by the finite element method is improved.

In addition, since the subdivision pattern is set to a quadrilateral or hexahedral element as the minimum pattern, and the subdivision target elements (111 designated elements and/or boundary elements) are subdivided using this pattern, high-order approximation is possible and analysis accuracy is improved. Further improvement.

[Brief explanation of drawings]

Figure 1 is a basic configuration diagram of an analysis model generation system according to an embodiment of the present invention, Figure 2 is its processing flow diagram, and Figure 3 is a diagram of the first embodiment of the analysis model generation system.
Figure 4 is a basic configuration diagram of analysis input model generation using a split object model in the embodiment, Figure 4 is its processing flow diagram, and Figure 5 is a basic configuration diagram of the analysis input model generation using the card image model in the embodiment shown in Figure 1. The basic configuration diagram, Figure 6 is its processing flow diagram, and Figure 7 shows the data of each model in Figures 1, 3, and 5 as the geometric model data and the attribute data of its constituent elements. Figure 8 shows that the data can be stored separately in the model database.
Figure 9 shows that the attribute data of each model consists of input model attribute data and output model attribute data, and can be stored separately in the attribute database. Figure IO shows that a combination of models can be analyzed by inputting the attribute data of each analysis output model into the attribute database of other analysis input models, and Figure 11 shows that CAE Figure 12 is a diagram explaining the basic internal model of the typical geometric target model in Figure 11, and Figure 13 is a diagram explaining the data conversion between each geometric target model used in A diagram showing a foam whose defining space is a subspace; Figure I4 is a diagram explaining the internal model of the data of the geometric object model whose constituent elements include the foam shown in Figure 13; The figure shows an example of a foam with holes, Figure I6 shows an example of a chest collar produced from the foam of Figure I5, and Figure 17 shows a chest collar produced from the chest collar of Figure 16. FIG. 18 is a diagram showing an example of a body chain, and is a diagram for explaining an internal model of a chest-collar body chain whose shape has a hole, as shown in FIG. 17. Figure 19 shows a pectoral collar chain generated from an external chain with one cell body and two two cell bodies;
Figure 21 shows that the pectoral chain in Figure 20 can be expressed as a pectoral chain generated from an external chain that has a mechanical foam and one bicellular body as its constituent elements. Figure 22 is a diagram to explain the internal model of the pectoral collar chain whose constituent elements are foams with connective cells. Figure 23 is a diagram showing the foam with connective cells and its structure. Figure 24 shows that a finite element model (chest chain) can be generated by subdividing the chest chain generated from external chains in elements. Figure 24 shows how data for the analysis input model is generated by re-subdividing the object model. Basic configuration diagram for
Figure 25 is a processing flow diagram showing that the analysis input model is subdivided using object model data, analysis input model data, and analysis output model data to generate analysis input model data that is the subdivision of the object model. , FIG. 26 is a processing flow diagram for subdividing a finite element (boundary element) analysis input model and generating data of the analysis input model that becomes the subdivision of the object model, and generating a group of elements to be subdivided in FIG. 27. A processing flow diagram for subdividing a finite element (boundary element) analysis input model and generating data for the analysis input model that becomes the subdivision of the object model. Processing flow diagram for subdividing a finite element (boundary element) analysis input model and generating data of an analysis input model that becomes subdivision of an object model, Part 2
Figures 9 and 30 are diagrams of the subdivision pattern of the subdivision designation element.
1 is a diagram showing a subdivision example of a subdivision designated element group, FIG. 32 is a diagram of a subdivision pattern of boundary elements of a subdivision designated element group, and FIG. 33 is a diagram showing a subdivision example of each boundary element of a subdivision designated element group, 34th
The figure is a diagram explaining the pattern and subdivision order of the boundary elements of the subdivision specified element group, and Fig. 35 shows the subdivision target element group and subdivides the subdivision target element group with a predetermined subdivision pattern. (Boundary elements) A processing flow diagram for generating subdivision data of the analysis input model. Figure 36 is a diagram explaining the constituent elements of the target finite element model. Figure 37 is a diagram explaining the subdivision specified element group and its boundaries. Figure 38 is a diagram explaining the entire element group of the target finite element model and its boundaries, Figure 39 is a diagram explaining the boundaries of the subdivision specified element group but not the boundaries of the entire element group, and Figure 40 is a diagram explaining the subdivision FIG. 41 is a diagram explaining the boundary elements of the specified element group. FIG. Figure 43 is a diagram explaining the subdivision method when the two sides of the boundary element of the subdivision specified element group are the boundaries in Figure 38. Figure 44 is a diagram explaining the subdivision method when the two sides of the boundary element of the subdivision specified element group are the boundaries The third side is the 38th
Figure 45 is a diagram explaining the subdivision method when there are three or more elements of the subdivision designation element pattern including the node of the boundary element to be subdivided.
Figure 6 is a diagram showing an example of subdivision of the target finite element analysis model, Figure 47 is a diagram showing a subdivision specification element group of a shell element model in three-dimensional space and an example of its subdivision, and Figure 48 is a diagram showing subdivision specification of a solid element model. Figure 49 is a diagram showing an example of an element pattern. Figure 49 is a diagram explaining a subdivision pattern of an element where three faces intersect in a solid element model. Figure 50 is a diagram illustrating a subdivision pattern of an element where two faces intersect in a solid element model. A diagram explaining the pattern, the fifth construction diagram, is a diagram showing an example of subdividing one boundary surface with a subdivision pattern consisting of the set of patterns in Figure 48 in a solid element model, and Figure 52 is a diagram of the object (solid) model. For example, FIG. 53 is a diagram showing an example of subdivisions generated by creating points on the boundaries of the object model of FIG. 52. DESCRIPTION OF SYMBOLS 10... Input/output display means, 1... Graphic display, 12... Stylus, 13... Keyboard, 20... Arithmetic unit, 21... Object model generation processing unit, 23... - Human force model generation unit for finite element analysis, 24... Output model generation unit for finite element analysis,
30... Database, 31... Database of object model data, 33... Database of input model data for finite element analysis, 34... Number (2) 3 for finite element analysis Figure 7 No. (2) 7-technical dike, matakaya/θ, 4, Shitsuzo, Sakamo, and island data, 11, hard leaves, 3, Atsushi, 4-, inner town, seasonal table rod, 15, Kokutaku, 6th ------"°--ゞA'Q ne-end K (lc?l) non-B figure (-1r@% page table 0 size ↑ to ζ inter whisper/7 clear 2nd figure 3 ward) t 24 Kuni No. 6 No. 35 Fig. 3 Fig. 4 - Fig. 1 Pressure Pressure - Fig. 6 Fig. 37 z Fig. 3 Fig. (b) all eD r 4th θ Fig. L ＋++−−−−−−−J r Rain 1 Fig. L---+++--J Product 44 Fig. (no straw 5 Trouble 41st 7th Fig. 9 Grid force 5θ ■

Claims (1)

[Claims] 1. An object model generation section and an analysis model generation section, the object model generation section generates object model data regarding an analysis target, and the analysis model generation section is an object modeling system that generates data for an analysis model for finite element analysis using the object model data, wherein the analysis model generation unit divides the elements of the generated analysis model into predetermined subdivisions. An object modeling system characterized by generating data for an analysis model in which target elements are further subdivided. 2. The analysis model generation unit includes means for evaluating the analytical accuracy of the generated analysis model and determining the predetermined subdivision target element based on the evaluation result. 1. The object modeling system according to 1. 3. The object modeling system according to claim 1, wherein the predetermined subdivision target element is given by an operator. 4. The subdivision target element determining means determines whether the analysis error for each element of the generated analysis model exceeds a predetermined range, and designates the elements that exceed the predetermined range as subdivision designation elements among the subdivision target elements. The object modeling system according to claim 2, characterized in that: 5. The subdivision target element determining means determines whether the analytical stress for each element of the generated analytical model exceeds a predetermined range, and designates the elements that exceed the specified range as subdivision designation elements among the subdivision target elements. 3. The object modeling system according to claim 2, characterized in that: 6. The object modeling system according to claim 4 or 5, wherein the subdivision target element determining means includes elements adjacent to the subdivision specified element group as subdivision target elements. 7. The object modeling system according to claim 4, wherein the object modeling system generates data of a boundary element group that does not include the boundary of the analytical model at the boundary of the subdivision designated element group. 8. The object modeling system according to any one of claims 4, 5, 6, and 7, wherein data of a group of elements whose boundaries are boundary elements is generated. 9. Claims 1, 2, 3, and 4, wherein the subdivision of the subdivision target element is performed in a predetermined subdivision pattern.
9. The object modeling system according to any one of 5, 6, 7, and 8. 10. The object modeling system according to claim 9, wherein the predetermined subdivision pattern has a minimum subdivision pattern of quadrilateral or hexahedral elements. 11. The object modeling system according to claim 1, wherein the analysis model includes an analysis input model used for analysis calculation processing and an analysis output model representing the analysis calculation processing result.