JP3218569B2 - Mesh generator - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mesh generating apparatus for generating a model mesh.

[0002]

2. Description of the Related Art In recent years, the role of simulation in product development has grown. Structural analysis, magnetic field analysis, etc. are such. It is absolutely necessary to perform mesh division as a pre-analysis step. The mesh division is to divide a model to be analyzed into small elements. In the analysis, a predetermined formula is applied to each of the divided elements (mesh) to calculate. Depending on the method of dividing the mesh, the analysis result may differ greatly, and the work of dividing the mesh in the analysis occupies a very important position. For this reason, it is required that the analyst is skilled in mesh division.

In addition, even if a model can be divided into meshes,
If this mesh cannot be input to the computer system, it cannot be analyzed. To input a mesh to a computer system, the coordinates of points (nodes) are input, and further, the arrangement of the node numbers is input. The arrangement of the node numbers represents a mesh, which is formed by connecting the previously input nodes into a form such as a hexahedron. For example, FIG.
The meshes 1 and 2 of FIG. Here, the meshes 1 and 2 in (a) and (b) of FIG. 24 represent the XZ plane among the three dimensions as shown in the figure.

Now, as two models, for example, a mesh 1 shown in FIG. 24A and a mesh 2 shown in FIG.
Take for example. These two models have substantially the same shape, but have slightly different dimensions in the portions shown in the figure. Even if the dimensions are slightly different in this part, the way of dividing the mesh is the same, but the coordinates of the nodes are different, so the analyst can input the meshes of the two models individually,
They copied and modified different parts by hand.

[0005]

As described above, conventionally, the analyst has manually input the mesh data (node coordinates, arrangement of the node numbers, etc.) of the model to be simulated sequentially. When creating a mesh of a model that has almost the same shape but slightly changed dimensions, there was a problem that it was necessary to input individually from the beginning or to copy and correct it . At this time, the analyst must divide the mesh in consideration of how to divide the model to be analyzed to obtain an effective analysis result, and input the data into a computer system. If you are not familiar with both the concept of mesh division and the method of input to the computer system, you cannot input and simulate, or even if you can, it will require a lot of effort, and you will not be able to simulate quickly. As a result, there has been a problem that development of a new product is delayed.

[0006] The present invention solves these problems,
A model mesh and a change designation table are provided, and a mesh data structure is generated according to the change designation table for each combination of variable items (values) specified for the selected model mesh, and a mesh of a desired model is automatically generated. It is intended to be.

[0007]

Means for solving the problem will be described with reference to FIG. In FIG. 1, a model mesh file 12 stores a data structure of a model mesh of a model in advance.

The change designation table 13 stores a procedure for calculating a value of a variable item of the template in the template mesh file 12 and generating a data structure of the mesh.

[0009] The mesh generation function 4 is used for a model mesh selected from the model mesh file 12.
According to the change designation table 13, the value of the variable item designated by the domain is calculated to generate a data structure of the mesh, or a combination of a plurality of variable items (values) designated by the domain is generated, and For example, a value is calculated to generate a mesh data structure.

The mesh editing function 5 determines the aspect ratio of each element in the data structure of the generated mesh, and if the aspect ratio does not fall within the allowable value, sets a node on the side of the long element that does not fit and divides it. When the total number of meshes after division exceeds a predetermined number, the data structure of the mesh before division is output.

[0011]

According to the present invention, as shown in FIG. 1, a predetermined variable item of the template mesh selected from the template mesh file 12 and its default value are displayed. Based on the value of the designated variable item and the default value (or the designated value) for the other items, the value of the variable item is calculated according to the change designation table 13 to generate a mesh data structure. I have.

At this time, a combination of variable items is generated in accordance with the specification of the range of the variable item and the number of divisions,
A change designation table 13 for each of these generated combinations
Is calculated according to the above formula, and a mesh data structure is generated.

When the mesh editing function 5 finds the aspect ratio of each element and does not fall within an allowable value for the data structure of the generated mesh, a node is set on the side of the long element that does not fit. I try to split it.

When the total number of meshes after division so that the aspect ratio of each element falls within an allowable value exceeds a predetermined number, the data structure of the mesh before the division is output.

Accordingly, a model mesh file 12 and a change specification table 13 are provided, and a mesh data structure is generated according to the change specification table 13 for each combination of variable items specified for the selected model mesh. Further, by generating a data structure of the divided mesh when the aspect ratio does not fall within the allowable value, or by outputting a data structure of the mesh before division when the number of meshes after division exceeds a predetermined number,
Automatic generation of mesh of desired model, automatic generation of data structure of all combinations of meshes according to domain specification, division of automatically generated mesh so that aspect ratio is within allowable value In addition, the total number of meshes can be reduced to a predetermined number or less.

[0016]

Next, the configuration and operation of an embodiment of the present invention will be described in detail with reference to FIGS.

FIG. 1 is a block diagram showing one embodiment of the present invention.
In FIG. 1, an analysis condition input 1 is a mesh condition (eg, material properties, boundary conditions, etc.) of a model to be analyzed, and is input to the solver 14.

The mesh generator 2 automatically generates a mesh for simulating a model. In this case, the mesh generator 2 generates a mesh data structure that can be input to a computer system. The mesh generator 2 includes an input unit 3, a mesh generation function 4,
It comprises a mesh editing function 5, an input data file 6, a data structure 7, and the like.

The input unit 3 is for inputting various data. Here, the input unit 3 selects and inputs from screens (screens 1, 2, 3, etc., which will be described later), and inputs data to input fields on the screen. It is for inputting data.

The mesh generation function 4 is used for a model mesh selected from the model mesh file 12.
A mesh data structure is generated by calculating the value of a variable item designated as a domain according to the change designation table 13, or a combination of a plurality of variable items designated as a domain is calculated and the value is calculated for each combination. To generate a mesh data structure (described later with reference to the flowchart of FIG. 7).

The mesh editing function 5 determines the aspect ratio of each element in the data structure of the generated mesh and determines that the element does not fall within the allowable value. When the total number of meshes after division exceeds a predetermined number, the data structure of the mesh before division is output (described later with reference to the flowchart of FIG. 15).

The input data file 6 stores data input from a screen or the like, which is the input unit 3,
It stores data such as shape and dimensions, shape member names, dimension values of variable items, and the like.

The data structure 7 is a data structure of the mesh generated by the mesh generation function 4, and is a data describing a method of mesh division of a shape having a size to be analyzed (see FIG. 14). This data structure 7 is stored in a file.

The shape database 8 is a database relating to the shape of the model. In this case, the shape classification file 9, the variable item file 10, the graphic file 1
1, a model mesh file 12, a change designation table 13, and the like.

The shape classification file 9 stores the shapes registered in the model mesh file 12, and here, as shown in FIG. 2A, a shape code, a shape name, and a shape member. The name is stored. When the analyst selects a model to be analyzed by looking at the shape name list and inputs the code, the corresponding shape member name is recorded. In the processing after this recording, a file such as the template mesh file 12 is referred to by the recorded member name.

The variable item file 10 is for storing a variable item by giving a name to each side of the shape. For example, the height of a certain shape is named variable item name H and width W (definition)
Then, it is clarified that the inputted dimension is a dimension of the shape of the model. The variable item file 10 stores typical dimension values (default values) of the model shape to assist in inputting the dimensions of the model to be analyzed (see FIG. 6). (A)).

The graphic file 11 stores data necessary for graphically displaying the shapes registered in the model mesh file 12. The graphic file 11 has nothing to do with the mesh division, but stores data necessary for confirming the shape selected from the shape classification file 9. Further, together with the model shape, the variable items also store data necessary for display, for example, as shown in FIGS. 6 (a) and 6 (b).

The model mesh file 12 stores a data structure of a mesh serving as a model for each shape in advance, and stores a data structure of a model mesh when an expert actually divides the mesh. It was done.

The change designation file 13 stores how to rewrite the template mesh file 12 (procedure). When a specific dimension is given, the data structure of the template mesh is rewritten according to the dimension. FIG. 13 stores a location and a procedure for calculating a numerical value to be rewritten.
reference).

The solver 14 includes the mesh generator 2
Under the condition given by the analysis condition input 1, the data structure of the mesh automatically generated by
According to the flowchart of FIG. 20, the value (for example, magnetic flux density) of each part of the mesh is calculated. The calculated analysis contents are printed on a sheet of paper or displayed so as to be easily understood as two-dimensional or three-dimensional graphics (FIG. 2).
2, see FIG. 23).

FIG. 2 shows an example of a file according to the present invention. FIG.
(A) shows an example of a shape classification file. This is shown in FIG.
This is an example of a shape classification file 9 constituting the shape database 8 of the present embodiment. Here, a shape code, a shape name, and a shape member name of those previously registered in the model mesh file 12 are registered as a pair. Therefore, the analyst:
As described above, when a list of shape names is extracted from the shape classification file 9 and displayed, and a corresponding shape code is input, a shape member name corresponding to the shape code is recorded, and the recorded shape member name is recorded in the subsequent processing. A file such as the template mesh file 12 is referred to by the shape member name.

FIG. 2B shows an example of a variable item file. This is an example of the variable item file 13 constituting the shape database 8 in FIG. 1, and corresponds to the shape name in the shape classification file 9 in FIG. In addition, typical dimension values (expressed by basic dimension code, basic dimension name, parameter name, parameter dimension) and the like are registered in advance (see FIG. 6).

Next, a procedure for inputting model shape data according to the order shown in the flowchart of FIG. 3 will be described in detail. In FIG. 3, S1 inputs a shape code. In this case, the contents read from the shape classification file 9 of FIG. 2A are displayed as shown in a screen 1 of FIG. 4, and “1” of the “core shape to be analyzed” is input from this screen 1 here. Fill in the fields. As a result, the core shape of the model to be analyzed becomes “EI center chip gap”.
Will be entered.

In step S2, a basic dimension code is input. This means that the contents read from the variable item file 10 of FIG. 2B are displayed as shown in a screen 2 of FIG.
Therefore, here, "11" of "basic dimension code" is entered in the input field. As a result, the basic dimension code of the core shape of the model to be analyzed is input as “11”.

In step S3, the default value of the variable item is determined using the basic dimension code as a key. In this case, the default values (default values, initial values) of the variable items of the basic dimension code “11”, which are input in S1 and S2, respectively, are extracted here with the core shape “EI center gap”, for example, as shown in FIG. a),
It is displayed as shown in FIG.

In step S4, the dimensions of the analysis model are input. For example, among the variable items displayed as shown in the screen 3 of FIG. 6 in S3, the initial values in the figure are rewritten to the dimensions of the model to be analyzed, and the MIN (minimum value), The MAX (maximum value) and the number of times (the number of divisions) are arbitrarily input (for the range, see the description of FIG. 6). These input data are stored in the input data file 6.

As described above, the default values (initial values, default values) of the variable items corresponding to the shape code and the basic dimension code are displayed on the screen 3 of FIG. MAX, MI in the domain
By inputting N and the number of divisions, data necessary for creating the mesh of the model can be input by a simple operation, and the variable items correspond to any of the model shapes, as shown in FIG. It is displayed graphically as shown in FIG.

FIG. 4 shows an example of a shape code selection screen (screen 1) of the present invention. This is a list of the shape names of the shape classification file 9 shown in FIG. Here, a list of core shapes is displayed. From this list, enter the shape code “1” in the input field and enter “E”
"I center gap".

FIG. 5 shows an example of a basic dimension code selection screen (screen 2) of the present invention. This is "1" on screen 1 in FIG.
In response to the selection of ".", A list of the basic dimension codes of the variable item file 10 shown in FIG. 2B is displayed. Here, the basic dimension code “11” is input from this list, and “EI30 / 26K” is selected.

FIG. 6 shows a default value (initial value) of a variable item of the model of the present invention and a screen example (screen 3) of a variable area.
This corresponds to the selection of the shape code “1” on the screen 1 in FIG. 4 and the selection of the basic dimension code “11” on the screen 2 in FIG. Screen 3 shows the initial value (default value) of the basic item dimensions
It is shown above.

FIG. 6A shows variable items and variable areas.
This is, as shown in the parameter input table shown,
Parameters (variable items) W1, W2, W3, WG, H1,
H2, H3, H, D, EGP, BG1, BH1, BL1
Are displayed as a total of 13 initial values (default values). The positions of these variable items in the model are shown in FIG.

The domain sets the MIN (minimum value) and MAX (maximum value) to be divided, and divides the interval by the number of times. For example, the parameter Initial value MIN of the domain MAX number of times of the domain WG 5.000 4.000 6.000 3 indicates that the value of MAX is divided into three from the MIN of the WG, and in this example, WG = 3.000 5.000 6.0000 The same applies to other variable items for which a variable is set.

Next, according to the order shown in the flowchart of FIG. 7, a procedure for generating a data structure of a model mesh based on the input data inputted in the description of FIGS. 3 to 6 will be specifically described.

In FIG. 7, S11 reads input data (shape member name, variable item). this is,
The input data input in the description of FIGS. 3 to 6 is read, for example, as shown in FIG.

In step S12, a model mesh is selected using the shape member name as a key. In this method, a model mesh is selected and extracted from the model mesh file 12 using the shape member name as a key, for example, as shown in FIG.

In step S13, the number N of sets of variable items is calculated. This is, for example, the input data example of FIG. 8 read in S11, and calculates the number N of sets of variable items whose division number is 1 or more. In the case of FIG. 8, the variable items whose division number is 1 or more are The two. Since the values of these WG and H are WG45.6H11.52, respectively, it is clear that there are a total of 9 combinations of both as shown in FIG. Therefore, N = 9.

In step S14, I = 0 is initialized. S15
Is I = I + 1. In step S16, a set of variable items is determined.

In step S17, variables are defined. These S16,
In S17, for example, it is determined to be the first set of the variable item set in FIG. 9, and the value of the variable (variable item) WG = 4, H = 1 at that time
Is defined.

In step S18, two records are read from the changed item. This means that two records are sequentially read from the change item record of the change instruction table 13 in FIG. Here, the first two records GRID 2 8.25 0.0 0.0 ← representing a location record... GRID 2 <G2> 0.0 0.0 ← representing a changed record. Read.

In step S19, a model mesh record matching the location record is found. This means that, for example, the above-mentioned location record extracted in S18 is found as a matching record from, for example, FIG. 12, which is the template mesh file 12 having the same shape member name.

At S20, it is determined whether or not it has been found. Here, since it is found as shown in FIG. 12, the template mesh record is rewritten with the change record in S21 and written in the data structure. In this example, the record of FIG. 12 is rewritten with the change record of FIG. 13, and is written in the data structure of the mesh of FIG. At this time, <G2
> Substitutes the values W1 = 30 and W2 = 20 of the variable items in FIG. 8 into SET G2 = (W1−W2) /2.0 of the variable definition record in FIG. 13, and G2 = (30−20) / 2 = 5, and this value is set at the position of <G2>. Then, the process proceeds to S22. On the other hand, if NO in S20, it is not found, so an error (unchangeable) is made in S24, and the location record is output as a mesh data structure in S25.

In S22, it is determined whether or not there is a record of the change item. In some cases, S18 and the subsequent steps are repeated until they disappear. On the other hand, if no, the process proceeds to S23.

In S23, it is determined whether N = I, that is, whether or not the number N of variable item sets determined in S13 has been repeated. In the case of YES, the process ends because mesh data structures have been generated for the number N of sets of variable items. On the other hand, in the case of NO,
In S15, I = I + 1, and the next set of variable items is S1.
Step 6 and subsequent steps are repeated.

As described above, the data structure of the mesh shown in FIG. 14 for the number N of the sets of variable items calculated in S13 is automatically generated. Here, the number of variable item sets and the data structure of 9 sets of meshes in FIG. 9 are automatically generated.

FIG. 8 shows an example of input data according to the present invention. This is an example of input data input in accordance with the flowchart of FIG. 3. Here, there are 13 variable items, and the variable areas are set as shown in the figure.

FIG. 9 shows an example of a set of variable items according to the present invention. This is based on the input data example of the variable items in FIG.
For two variable items WG and H having two or more divisions, nine combinations are arranged as a combination thereof. Since these nine sets have variable items WG = 4, 5, 6H = 1, 1.5, 2, these are combined without overlapping.

FIG. 10 shows a specific example of the data structure of the template mesh according to the present invention. Here, as for the record having the GRID of the node designation record, the node number 1 is created at the coordinates (0, 0, 0) as described on the right side. Also,
As shown on the right side, the record having the GCIR of the node designation record has a radius of 1 around the node number 11.
0, an angle of 45 ° with the X axis, and a Z coordinate of 0. As described on the right side, the record having the node designation record MOVE is obtained by adding 10 to the Z coordinate of the node numbers 1 to 22, and adding the node numbers 101 to 122.
Create with turn.

The element designation record indicates designation of a hexahedral element or designation of a pentahedral element as shown. As described in the lower part, the identifier at the head of the node designation record is GRID: coordinate value input GCIR: intersection of a circle and a straight line passing through the center of the circle MOVE: movement and copying of a defined node The identifier at the head of the designated record indicates an ELE8: hexahedral element.

ELE6: Represents a pentahedral element. FIG. 11 shows a mesh diagram of the present invention. This is shown in FIG.
5 schematically shows the data structure of the model mesh of the above. For example, 1 represents the node “1” and corresponds to the first record GRID 1 of the node designation record in FIG. Similarly, 2, 11, 12, and 22 correspond to FIG.
Correspond to the second to fifth records of the node designation record.

The hexahedron of element number 1 corresponds to the first record ELE81 of the element designation record in FIG.
The pentahedron with the element number 2 is the element designation record 2 in FIG.
This corresponds to the second record ELE62.

FIG. 12 shows an example of a template mesh file according to the present invention. This is extracted from the model mesh file 12 using the shape member name as a key (FIG. 7).
S12). This template mesh file 12
An expert registers the data structure of a model mesh having a shape and dimensions preset for the shape of the model.

FIG. 13 shows an example of a change designation table according to the present invention. The change designation table 13 includes a variable definition record,
And change item records. The variable definition record defines a formula for calculating the value of a variable (variable item) in the variable item record.

The change item record is a pair of the location record and the change record, and replaces the corresponding record in the data structure of the template mesh and changes its value. Specifically, when a record that matches the location record is found in the data structure of the template mesh, the record of the template mesh is replaced with the change record, and the data structure of the desired mesh is automatically generated. At this time, the variable (variable item) in the change record
For example, when <G2> is found, the corresponding calculation formula SET G2 = (W1-W2) /2.0 of the variable definition record is calculated and replaced with this value.

FIG. 14 shows an example of the data structure of a mesh according to the present invention. This is a data structure of a mesh automatically generated by changing the template mesh file of FIG. 12 according to the change designation table 13 of FIG. The data structure of the mesh in FIG. 14 is automatically generated in accordance with the flowchart in FIG. 7 by the number of sets of variable items in FIG.

FIG. 15 shows a mesh editing flowchart according to the present invention. This is based on the data structure of the mesh of FIG. 14 automatically generated by the description of FIGS.
The aspect ratio of each element is obtained and divided so as to be within an allowable value, thereby automatically improving analysis accuracy. At this time, if the total number of meshes after the division exceeds the calculation capability of the computer system, the total number of meshes within the limit is automatically suppressed, and the analysis by simulation is performed with the maximum accuracy with limited resources. . This will be described below.

In FIG. 15, S31 rewrites the data structure to a mesh editing format. In step S32, the aspect ratio of each element is calculated. Here, the aspect ratio is
For example, taking the element number 606 of the data structure (mesh editing format) in FIG. 16 as an example, this one element number 606
, And one of them is as follows: aspect ratio = [side (606-706)] / [side (606-607)] = 0.1 / 0.335 ((FIG. 18 ( a) Mesh diagram 1). Similarly, length ratios are determined for other combinations of sides.

In S33, it is determined whether or not all the aspect ratios fall within the allowable values. This is because, for example, when the allowable value of the aspect ratio is set to 1/2 ≦ aspect ratio ≦ 2, the side (606-607) = 0.335 with respect to the side (606-706) = 0.1 in FIG. The side (706-707) = 0.335, which indicates that the length is too long and the ratio between the two does not fall within an allowable value (an allowable value between 1/2 and 2) (see mesh diagram 1 in FIG. 18A). ). If it does not fit, the process proceeds to the division from S34. On the other hand, if it fits, the mesh diagram is graphically output in S40, and the process proceeds to S39.

In step S34, a node is set on the longest side of the element that does not fall within the allowable value, and is also set on the side related to this side. In this case, for the element determined to have an aspect ratio not within the allowable value in S33, a node is set so that long sides are equally divided, for example, so that the aspect ratio is within the allowable value. Specifically, FIG.
As shown in FIG. 8B, the node 1001 on the side (606-627) on the longest side (606-607) that does not fit in the allowable value, the node 1002 on the side (706-707) The node 1004 is set on the 1003 side (726-727) (see FIG. 18B). here,
The element is also changed from the old 506 to the new 506 and the 1001 element. And these new 1001 to 1004
As the number of nodes increases, nodes are also set (1005 and 1006) on the side related thereto (FIG. 18).
(B) points indicated by x on the mesh diagram 2), and the elements are also newly changed (new 606 and 1002).

In step S35, a mesh is added and deleted using the new node. By setting and changing the above nodes, the final stage is reached as shown in the mesh diagram 3 of FIG. FIG. 17 shows the edited data structure (mesh editing format) at this time. In FIG. 17, among GRID 606 1.365 0 2.03 GRID 1001 1.5325 0 2.03 GRID 607 1.700 0 2.03, GRID 1001 is edge (606-607).
Is a node that is set by dividing the space equally. Similarly, GR
The values described above are similarly set for IDs 1002 to GRID 1006.

In step S36, the total number of meshes is counted. S37
Determines whether the total number of meshes is within the capacity of the solver. If it fits, S32 and subsequent steps are repeated. On the other hand, if it does not fit, the mesh before re-division is output graphically in S38. Then, the process proceeds to S39.

In step S39, the analyst determines whether this mesh is sufficient. If it is good, the series of mesh division editing processing ends. If not, use another template mesh or cut a new mesh.

As described above, for example, based on the data structure of the mesh of FIG. 16 (see the mesh diagram 1 of FIG. 18A), the aspect ratio of each element is obtained, The nodes are set so as to fit in the mesh and the mesh is divided, and the data structure of the mesh in FIG.
9 is automatically generated. This makes it possible to automatically generate an appropriate mesh data structure by setting nodes on sides where the aspect ratio does not fall within the allowable value, and dividing the data structure of meshes with slightly different dimensions automatically generated from the model mesh. It becomes possible.

FIG. 16 shows an example of the data structure (mesh editing format) of the present invention. This is a data structure example in which the data structure of the automatically generated mesh is converted into a mesh editing format according to the flowchart of FIG. 7 and the description of FIGS. Here, for example, in the first record GRID 506 1.365 0.0 1.775, the element number of the node is 506, and the position is the coordinate value of X, Y, Z that follows. Others are the same. FIG. 18A is a mesh diagram 1 schematically showing these.

FIG. 17 shows an example of the edited data structure (mesh editing format) according to the present invention. This is a data structure after nodes are set to be equally divided for sides where the aspect ratio of each element of the data structure of the mesh in FIG. 16 does not fall within the allowable value. Here, the nodes 1001 to 1006 of the nodes indicated as divided by the arrows are newly set. These are schematically shown in a mesh diagram 3 in FIG.

FIG. 18 shows an example of mesh editing according to the present invention. FIG. 18A shows a state before editing. This is a schematic display (graphic output) of the data structure of FIG. 16 that is the object of mesh editing. The node number “506” described in the mesh FIG.
“507” “606”, “607”, “706”, “7”
07 ″ is the GRID 506, GRID 50 in FIG.
7, GRID 606, GRID 607, GRID
706 and GRID 707 are displayed based on the coordinate values.

FIG. 18B shows a state in the middle of editing. This corresponds to the node numbers “1001”, “1003”, and “100” in the mesh diagram 1 of FIG.
5 "shows a state in which a node is set at an associated position of" X "after 5" is equally divided and set.

FIG. 19 shows an example of mesh editing (final stage of editing) according to the present invention. As shown in FIG. 18B, this is the final step after mesh division is performed by repeatedly setting a node to a long side and a related side so that the aspect ratio falls within an allowable value. Is schematically displayed. In this final stage, mesh division is automatically performed so that the aspect ratio when focusing on each element is equal to or less than an allowable value.

FIG. 20 shows a solver flowchart of the present invention. This means that the solver 14 in FIG. 1 gives analysis conditions (material properties, boundary conditions, etc.) for the data structure of the mesh of the model automatically generated by the mesh generator 2, for example, the data structure of the mesh in FIG.
This is an operation when a predetermined equation is calculated on the mesh and the result is obtained. This will be described below.

In FIG. 20, S51 reads data. This is the condition for the analysis (boundary conditions,
The data structure of the mesh after editing shown in FIG. 17 (see FIG. 3 in FIG. 19) is read.

In step S52, a coefficient matrix K is created. This creates, for example, a coefficient matrix K shown in FIG. 21 for solving simultaneous linear equations in S54. S5
3 creates a right-hand side vector f. This creates, for example, the right-hand side vector f shown in FIG. 21 for solving the simultaneous linear equations in S54.

In step S54, a simultaneous linear equation is solved (K · U
= F). This solves the simultaneous linear equation KU = f shown in FIG. 21, for example, when performing a static magnetic field analysis. S55 outputs the result. As a result, as shown in FIG. 22, a magnetic flux density distribution diagram is graphically output. Further, as shown in FIG. 23, the magnetic flux distribution density is graphically output in a perspective view.

As described above, the template mesh file 1
Based on the domain specified by the analyst as a variable item for the template mesh selected from the sample mesh 2, the mesh data structure of all combinations is automatically generated according to the previously specified change specification table 13, and the aspect ratio of each element Automatically generates a mesh data structure such that the value falls within an allowable range, and analyzes the generated mesh data structure (for example, the data structure in FIG. 17) with analysis conditions (boundary conditions, material characteristics, etc.). ) To automatically analyze the magnetic flux density of each mesh. Then, it is possible to graphically display the result and display the analysis result in an easy-to-view manner for an analyst.

FIG. 21 shows an example of simultaneous linear equations. This is an example of a system of linear equations used when performing static magnetic field analysis, applied to the data structure of an automatically generated mesh. Here, K is a coefficient matrix, which is determined by the shape of the magnetic material, boundary conditions (symmetry), and material characteristics.

F is a magnetic field matrix, which is determined by the magnetic material, the shape of the current element, the current value, and the like. U is a magnetization matrix, which is unknown;
This is for obtaining this and calculating the magnetic flux density and the like from the value of U.

FIG. 22 shows an analysis example of the present invention. This is because the solver 14 solves a system of linear equations in accordance with the flowchart of FIG. 20 for the data structure of the mesh of FIG. 17, obtains the magnetic flux density for each mesh, and sets a vector (magnitude and direction of the magnetic flux density ). The figure shows the state of the XZ plane.

FIG. 23 shows an analysis example of the present invention. This is a graphic representation of the same magnetic flux density distribution as in FIG. 22 with a three-dimensional effect in a perspective view.

[0087]

As described above, according to the present invention,
Model mesh file 12 and change designation table 1
3, a data structure of the selected model mesh is generated according to the change designation table 13 for each combination of variable items, and a data structure divided into meshes such that the aspect ratio is within an allowable value is generated. Or when the number of meshes after division exceeds a predetermined value, the data structure of the mesh before division is output. (1) The data structure of the mesh of the desired model is automatically generated. (2) Automatically generate the mesh data structure of all combinations according to the specification of the domain and the number of divisions. (3) Mesh division so that the aspect ratio of the automatically generated mesh is within the allowable value. (4) Further, the total number of meshes must be kept within a predetermined value so that analysis can be performed within limited resources of the computer system. Can. As a result, even an inexperienced analyst can easily create a mesh at the level of an expert with a simple operation, and perform analysis quickly, efficiently, and accurately.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of one embodiment of the present invention.

FIG. 2 is an example of a file according to the present invention.

FIG. 3 is an input flowchart of the present invention.

FIG. 4 is an example (screen 1) of a shape code selection screen of the present invention.

FIG. 5 is an example (screen 2) of a basic dimension code selection screen of the present invention.

FIG. 6 is a screen example (screen 3) of default values and variable areas of variable items of the model of the present invention.

FIG. 7 is a mesh generation flowchart of the present invention.

FIG. 8 is an example of input data according to the present invention.

FIG. 9 is an example of a set of variable items according to the present invention.

FIG. 10 is a specific example of a data structure of a model mesh of the present invention.

FIG. 11 is an example of a mesh diagram of the present invention.

FIG. 12 is an example of a template mesh file of the present invention.

FIG. 13 is an example of a change designation table according to the present invention.

FIG. 14 is an example of a data structure of a mesh according to the present invention.

FIG. 15 is a mesh editing flowchart of the present invention.

FIG. 16 is a data structure (mesh editing format) of the present invention.
It is an example.

FIG. 17 is an example of a data structure (mesh editing format) after editing according to the present invention.

FIG. 18 is an example of mesh editing according to the present invention.

FIG. 19 is an example of mesh editing (final stage of editing) according to the present invention.

FIG. 20 is an example of a solver flowchart of the present invention.

FIG. 21 is an example of simultaneous linear equations according to the present invention.

FIG. 22 is an example of an analysis result of the present invention.

FIG. 23 is an example of an analysis result according to the present invention.

FIG. 24 is an explanatory diagram of a conventional technique.

[Explanation of symbols]

 1: Analysis condition input 2: Mesh generator 3: Input unit 4: Mesh generation function 5: Mesh editing function 6: Input data file 7: Mesh data structure 8: Shape database 9: Shape classification file 10: Variable item file 11: Graphic file 12: Model mesh file 13: Change designation table 14: Solver

Claims (4)

(57) [Claims] 1. A mesh generating apparatus for generating a model mesh, comprising: a model mesh file (12) for storing a data structure of a model mesh of a model in advance; and a mesh item of the model mesh file (12). Among them, a change designation table (13) for calculating a value of a predetermined variable item to generate a mesh data structure and storing a procedure, and an arbitrary model are selected from the model mesh file (12). In response to this, a method of displaying a predetermined variable item and its default value from the template
In response to the change of the default value of the displayed variable item or the designation of the area, the change designation table (13)
Means for calculating the values of the variable items and replacing the item values of the template mesh according to
Mesh generator with. 2. The method according to claim 1, wherein the generating means generates a combination of the variable items in accordance with the specification of the range of the variable item and the number of divisions, and for each of the generated combinations, the change designation table (13). 2. The mesh generating apparatus according to claim 1, wherein the values of the variable items are calculated according to the following formula, and the data structures of the meshes in which the item values of the template mesh are replaced are generated. 3. The generating means calculates an aspect ratio of each element in the data structure of the generated mesh, and when the aspect ratio does not fall within an allowable value, sets a node on a side of the long element that does not fit. there the claim 1 is described for dividing
Mesh generation apparatus of claim 2 wherein. 4. When the total number of meshes after division so that the aspect ratio of each element falls within an allowable value exceeds a predetermined number, the generating means outputs a data structure of the mesh before the division. The mesh generation device according to claim 3.