JPH07262171A - Area input method and simulation program generating method using the same - Google Patents

Abstract

PURPOSE:To provide a method to input a graphic or the model of numerical simulation to a computer with less labor and time. CONSTITUTION:The shape of plural areas and the name of substance in respective area, etc., are inputted by using a mouse, etc., and they are held as user definition information 901 by a shape defining part 1012, etc. A shape processing program 106 checks the superposition of shape of the plural areas, and a partial area of area shape located at the highest position for a superimposed area is employed, and a hidden part is deleted. The information of a material constant and a source item supplied to an original area via the mouse, etc., by a user are added on each partial area. An analysis area is decided as a set of such partial areas. The external boundary of the analysis area is sampled. The information of a boundary condition supplied previously to the side of the external boundary of the analysis area is attached on each side of the external boundary of a sampled analysis area. Physical model information 102 consisting of such analysis area and information annexed to the area can be generated. A simulation code 104 can be generated by a simulation code conversion program 103 by using such information.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an input method for drawing a drawing on a computer, and in particular, an input method capable of reducing the number of simulation steps by simply inputting a model for numerical simulation. Involved in

[0002]

2. Description of the Related Art Generally, when solving a so-called continuous system problem such as thermal diffusion or fluid analysis using a numerical simulation on a computer, the problem to be solved is modeled in a form represented by a partial differential equation. To do. This is called a simulation model.

The simulation model consists of:

・ Area shape of the object to be analyzed ・ Partial differential equation governing the system ・ Analysis conditions (material constants, source terms of equations, boundary conditions) Generally, in the numerical simulation of a continuous system, the finite element method or the like is used. The region shape is divided into meshes using the discretization method, and the equations that are discretized are solved for each of the divided nodes. In executing the simulation, it is necessary to give the analysis conditions of the equation to each node.
Material constants and source terms are given to each node inside the area shape, and boundary conditions are given to each node belonging to the outer boundary of the area shape (outermost contour of the area shape to be simulated).

In general, in order to perform a good (accurate and low cost) analysis, it is necessary to construct an appropriate simulation model. The obstacle at this time is the time and effort required to input the simulation model into the computer. Especially, the work of inputting a simulation model having a complicated region shape is complicated and very heavy for the user.

Conventionally, as a method for interactively inputting a simulation model, an accurate area shape is first input,
Next, many methods have been adopted in which mesh division is performed and then analysis conditions are given to individual nodes.

For example, Japanese Patent Laid-Open No. 4-117573 discloses a method of interactively inputting an analysis simulation. In this, the process of inputting and creating a geometric model, an analytical model, and analytical condition (material condition and boundary condition) data is described as a separate step.

Further, Japanese Patent Laid-Open No. 3-256172 discloses a method of interactively setting boundary conditions. Described here is the setting method when the shape and mesh have already been created.

Further, Japanese Patent Laid-Open No. 4-257074 also discloses a method of interactively setting boundary conditions. When the shape and the mesh have already been created, the boundary condition can be set for a plurality of mesh nodes with a small number of operating procedures.

[0010]

The conventional method has the following problems.

That is, when inputting a simulation model having a complicated shape, the number of steps is remarkably increased. There are three main factors.

One is that the user has to input the area shape information along a data structure which is an internal representation of the area shape computer. Generally, since the internal structure of a computer is redundant, the man-hours required for input tend to increase. In particular, the correction and correction of the simulation model often occur in the scene where the simulation is performed, but the amount of work is particularly large.

The other one arises from the method of inputting the region shape and the analysis conditions as completely different steps. The user sets the analysis conditions after inputting the region shape once, and must know which analysis condition should be set for which region shape. This causes an input error due to a memory error or the like.

Another reason arises from the fact that the user must know the outer boundary of the area shape. Due to the characteristics of the numerical simulation, it is not possible to specify the external boundary when the user inputs the area shape. For example, in a simulation model in which liquid and solid coexist, the liquid region is not the target of structural analysis, and the solid region is not the target of fluid analysis. Only when the material constant is set in addition to the area shape, the external boundary is determined. Boundary conditions must be set on all external boundaries and not on external boundaries. In the conventional method, since the area shape is input in a separate step, the user has to know the external boundary, which also causes an input error.

As described above, the conventional method compels the user to perform an unnecessarily complicated procedure particularly when inputting a complicated simulation model, which easily induces an input error by the user and a decrease in work efficiency. There is a problem that it is very inefficient.

An object of the present invention is to provide a shape inputting method which allows the area shapes of a complicated simulation model to be input with a small labor by allowing the area shapes to overlap each other.

Another object of the present invention is to provide a simulation model input method capable of setting the area shape of a simulation model and analysis conditions in parallel.

Still another object of the present invention is to provide a simulation model input method capable of automatically extracting the outer boundary of the area shape and setting the analysis condition.

[0019]

According to the present invention, the area shape information that can be overlapped is automatically removed from the area shape information that is covered by another area shape to obtain consistent area shape information as a simulation model. To generate.

Further, information on analysis conditions such as material constants and source terms of equations can be always set for the region shape.

Further, the information on the outer boundary is extracted from the information on the area shape.

[0022]

Since the area shapes can be overlapped with each other, the number of man-hours required for correcting the area shapes can be suppressed.

Further, since it is possible to always set the analysis conditions for the area shape, it is possible to reduce input mistakes.

Further, by automatically extracting the external boundary, the boundary conditions can be set without contradiction, and the input error can be reduced.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the processing procedure when the present invention is applied to the front end of a numerical simulation system will be described in detail below.

A procedure for expressing a physical phenomenon under the system according to the present invention will be described in detail with reference to FIGS. 1 and 2.

FIG. 2 shows the configuration of the entire apparatus. The central processing unit 201 controls user operations via an input device 202 such as a mouse.
In addition, the result of the user's operation and the processing accompanying it is displayed on the display device 203 such as a CRT display. Of these, the programs that realize the main functions of the system are stored in the storage device 204.

This system comprises a physical model creation support program 101, a shape processing program 106, and a simulation code conversion program 103.

The physical model creation support program 101 receives the user's operation via the input device 202 and processes it to create the user definition information 901. The shape processing program 106 creates from this information 901 information called physical model information 102 that defines a numerical simulation problem.

Simulation code conversion program 10
3 is a simulation code 104 which is information for performing a simulation by inputting the physical model information 102.
And the calculation grid data 105 is output. The output simulation code 104 and calculation grid data 105 are compiled into a form executable by the compiler 205, and a simulation is executed by the simulation execution program 206 so that the user can obtain the result of the simulation.

The outline of each program will be described with reference to FIG.

The user operation is obtained via the user interface. The input processing unit 1011 activates any one of the shape defining unit 1012, the material defining unit 1013, the source term defining unit 1014, and the boundary condition defining unit 1015 depending on which part of the user interface receives the input. Each definition part is
Of the information for defining the physical phenomenon to be simulated, called the user-defined information, the part related to the shape, the part related to the material, the part related to the source term of the partial differential equation that governs the physical phenomenon, and the part related to boundary conditions. Defines the information for.

Under the present system, the user-defined information expresses information on the physical phenomenon to be simulated. In the present embodiment, the physical phenomenon is represented by a set of a region shape and its associated physical attribute. The area shape is
It can be defined like drawing a picture through the user interface and can be superimposed on each other. Details of the contents of the user-defined information will be described later.

The physical phenomenon cannot be accurately expressed only by the information expressed by the user-defined information. The main problem is that overlapping area shapes are physically incorrect. A shape processing program 106 is prepared to eliminate this contradiction.

The external boundary extraction unit 120 in this program 106
1 uses the information on the part related to the shape and the information on the part related to the material of the user-defined information, and extracts the part corresponding to the external boundary from which the boundary condition of the simulation should be set, and stores it in the physical model information 102.

The direct sum decomposition unit 1501 uses the information of the part relating to the shape and the information of the part relating to the material in the user-defined information, to determine the part of the overlapped region shape which is covered by another region shape. By deleting the contradiction, the contradiction is resolved, and the information of the area shape in which the contradiction is resolved is stored in the physical model information 102.

Details of the contents of the physical model information 102 and details of the processing of the external boundary extraction unit 1201 and the direct sum decomposition unit 1501 will be described later.

The simulation code generation unit 2401 extracts the information on the material-related portion from the physical model information 102, performs matching with the material database 2501, obtains a material constant, and reflects it in the simulation code 104. Further, the simulation code generation unit 2401 extracts the information of the source term and the part related to the boundary condition from the physical model information 102, performs matching with the numerical calculation knowledge base 2601, and the equation and the equation for solving the target physical phenomenon. The source term and the numerical solution are obtained and reflected in the simulation code 104.

The calculation grid generation unit 2491 uses the physical model information 10
The information on the part related to the shape is taken out from 2, and this is divided into a finite element mesh and output as calculation grid data 105.

Details of the contents of the simulation code 104, details of the calculation grid data 105, details of the processing of the simulation code generation unit 2401, and details of the contents of the material database 2501 and the numerical calculation knowledge base 2601 will be described later.

FIG. 3 shows a coupled problem of thermal analysis and fluid analysis taken as an example.

Steel pillars 3 are provided in the flow path divided by the wall 302.
04 is standing. Water 303 flows in from the left side 305 of the flow path,
It flows to the right 306. Further, the iron pillar 304 is generating heat. For reference, the dimensions of each part are given in the form of x and y coordinates in the figure.

FIG. 4 is a typical example of the screen of the display device 203 (FIG. 2).

The display screen includes a window 401, a drawing area 402 in the window 401, a graphic operation menu 404, and an analysis field menu 40.
8, material menu 413, source term menu 417, boundary condition menu 419 and simulation execution start button 423
Consists of.

A general procedure for a user to define a physical phenomenon will be described. The physical phenomenon defined by the user is stored in the user-defined information 101 in FIG.

(1) Select the target physical phenomenon from the analysis field menu 408. The physical phenomenon to be analyzed and the physical phenomenon to be excluded are distinguished by the color of the button.

(2) Select the physical phenomenon to be edited from the analysis field menu 408. The reason why this selection operation is necessary is that the information on the source term and the boundary condition differs depending on the physical phenomenon. This selection operation does not affect the created region shape and material. This selection operation changes the display of source terms and boundary conditions. This selecting operation can always be performed throughout the editing work. This selection operation is
It can be performed only from the physical phenomenon selected for analysis in (1).

(3) A polygon creation button 405 or a circle creation button 406 which means creation of a figure in the figure operation menu 404
Select the shape to create from the list. The selected figure is processed by the shape definition unit 1012 (FIG. 1).

(4) Select a material to be added to the figure from the material menu 413. The selected material is processed by the material definition unit 1013 (FIG. 1).

(5) Select a source term to be added to the figure from the source term menu 417. The selected source term is processed by the source term definition unit 1014 (FIG. 1).

(6) From the boundary condition menu 419, select a boundary condition to be added to the figure. The selected boundary condition is processed by the boundary condition definition unit 1015 (FIG. 1).

(7) The shape defining section 1012 creates and displays the figure selected in (3) in the drawing area 402. The material, source term and boundary condition selected in (4), (5) and (6) are added to the created figure.

(8) Modify button 407 of the graphic operation menu 404
Press to correct the created figure. The shape defining unit 1012 also performs this correction.

(9) A graphic is created by repeating (3) to (8). Figures can be overlaid. The superposition order is such that the graphic created later is on the top, but it can also be changed later by the user's operation. Information such as a figure selected by the above is stored as user-defined information 901.

(10) By pressing the simulation execution button 423, the editing work is terminated, and the physical model information 102 is generated from the user-defined information 901 by the external boundary extraction unit 1201 and the direct sum decomposition unit 1501.

When the editing work is completed, the system generates the simulation code 104 and the calculation grid data 105 from the physical model information 102 by the simulation code conversion program 103, creates the execution format by the compiler 205, and executes the simulation by the simulation execution program 206. And give results.

FIG. 5 shows the procedure for defining the physical phenomenon shown in FIG.
Starting from Fig.8. Since the phenomenon to be analyzed is a combination of heat and fluid, select heat and fluid from the analysis field menu.

FIG. 5 shows a rectangle 501 representing a flow path. First, select “heat (502)” from the analysis field menu as the analysis field to edit, select polygon from the figure operation menu, “water” from the material menu, “none” from the source item menu, and “heat” from the boundary condition. Select "Conduction" to create a rectangle in the drawing area.

FIG. 6 is a diagram in which the source term and the boundary condition in the fluid analysis are added to the rectangle representing the flow channel created in FIG. First, the analysis field to be edited is changed to "fluid (602)" by the analysis field menu. There is no change in the display of the material of the rectangle 601 and the material menu, but there is a change in the source term menu 603 and the boundary condition menu 604. Select None from the Source section menu to set it to a rectangle. "Inflow / Outflow" was selected from the boundary condition menu and set on the left and right sides of the rectangle 601. I selected "stationary wall" from the boundary condition menu and set it on the upper and lower sides of the rectangle.

FIG. 7 shows a triangle which represents an iron pillar. Following the same procedure as in Fig. 5, the analysis field to be edited is returned to "heat", the material is "iron", and the source is "heat".
, And select “none” as the boundary condition (note that the side of this triangle will not become an external boundary in the thermal analysis due to the boundary extraction process that follows, so it is not necessary to set the boundary condition strictly). A triangle was created in the drawing area.

FIG. 8 is a diagram in which the source term and the boundary condition in the fluid analysis are added to the triangle representing the iron column created in FIG. Switch the analysis field to be edited to "Fluid", select "None" for the source term (Note that the inside of this triangle will not be the analysis target area in the fluid analysis due to the processing of the direct sum decomposition section later, (It is not strictly necessary to set the source term) Set it to a triangle, select "stationary wall" as the boundary condition (the side of this triangle will become an external boundary in the fluid analysis by the boundary extraction processing later) Set on the side of.

FIG. 9 shows a general format of the user definition information 901.

Reference numeral 902 is information on the entire problem, which includes a problem name, whether or not the problem is time-dependent (a stationary problem or a non-stationary problem), and a field (one or more) to be analyzed.

Reference numeral 903 is information on vertices in the defined shape, and each vertex is composed of a name and coordinates.

Reference numeral 904 is side information in the defined shape.
Each side is composed of a name, points at both ends, and a set of boundary conditions set on the side (the number of which corresponds to the number of analysis target fields).

Reference numeral 905 is surface information in the defined shape.
Each face consists of a name, a set of edges that make up the face, and a set of materials and source terms set on the face (the number of which corresponds to the number of analysis target fields).

FIG. 10 is a visual representation of the user-defined information 901 created in FIGS. The appearance changes depending on the analysis field. Reference numeral 1001 represents a physical phenomenon from the viewpoint of thermal analysis, and 1002 represents a physical phenomenon from the viewpoint of fluid analysis. Each vertex, each side, and each face are given names that start with v, e, and s and are identified by serial numbers.

FIG. 11 is a specific example of the user-defined information 901 of FIG. 1102 is the problem name, time dependency, analysis field (two in this case). 1103 is vertex information, and each of the seven vertices is represented by coordinates. Reference numeral 1104 denotes side information, which has seven sides including a start point, an end point, and boundary conditions (each of which records boundary conditions corresponding to two analysis fields). 1105 is surface information, one triangle and one rectangle, material (not depending on the number of analysis fields) and source term (there are only the number of analysis fields)
Information has been added.

When the definition of the physical phenomenon by the user is completed, the shape processing program 106 is started and the user definition information 901 is converted into the physical model information 102.

FIG. 14A shows an example of a user-defined area. Three rectangles are stacked in the order 1404, 1403, 1402. Rectangle 1404 and 1403 are made of the same material,
The material of 1402 shall be different. The thick solid line (1405) is the visible outline of the rectangle, and the thin dashed line (1406) is hidden by other rectangles.

FIG. 12 is a processing flow of the boundary extraction unit.

The external boundary extraction unit 1201 first divides the contour of the polygon into minimum units called subdivision sides, that is, line segments that do not intersect with other sides (step 1502). The details are shown in FIG. 16 (a). Here, the region after the generation of subdivided edges is shown, and it is judged whether or not all edges have intersections with other edges (step 1
603), divide each side into line segments that do not intersect with other sides. Figure 14
In (b), each thin solid line is a subdivision side. Subdivision edges are subdivision edges that have nothing on one side, such as 1408, with one side between them, subdivision edges that have the same material on both sides, such as 1409, and subdivision edges that have different materials on both sides, such as 1410. There is a side. The outer boundary is a subdivision with nothing on one side.

Therefore, inside each subdivision side
(The inside of the surface to which the subdivision side belongs is referred to as the inside) and the material on the outside is referred to, and if one is valid and the other is invalid, the subdivision is defined as the external boundary (steps 1203, 1204, 1).
301, 1205, 1301, 1206-1208). FIG. 14 (c) shows the area after extraction of the outer boundary. The thick solid line (1412) is the extracted outer boundary. The valid / invalid of the material is a typical example of the presence / absence of the material as shown in Fig. 14, but other things such as liquid or gas / solid in fluid analysis and solid / liquid or gas in structural analysis Including that. In other words, whether or not each subdivision side becomes an external boundary depends not only on the material but also on the analysis field. FIG. 13 (1301) is a processing flow of the material effectiveness determination processing 1301. Utilizing that the condition that the material becomes invalid is either "cavity figure" (step 1302), "solid in fluid analysis" (step 1305), or "liquid or gas in structural analysis" (step 1307). , The given material is valid (step 1
309) It is determined whether it is invalid (steps 1303, 1306, 1308).

FIG. 15 shows a processing flow of the direct sum decomposition unit 1501. The direct sum decomposition is a process of dividing and integrating a plurality of area shapes into area shapes having no overlapping portions.

FIG. 18A shows an example of a region before the direct sum decomposition is applied. Three rectangles are stacked in the order of 1804, 1803, 1802. The rectangles 1804 and 1803 are made of the same material. The material of rectangle 1802 is different. Thick solid line (for example, 1
805) is the visible outline of the rectangle, the thin dashed line (eg 1806) is obscured by another rectangle.

The direct sum decomposition unit 1501 first divides the polygonal contour into subdivided sides similarly to the boundary extraction unit (step 1507).
This process is the same as the process shown in FIG. Figure 18 (b)
Indicates the area after the subdivision edge generation. Each thin solid line is a subdivision. Subdivision edges are subdivision edges that have nothing on one side, such as 1808, with one side in between, subdivision edges that have the same material on both sides, such as 1809, and subdivision edges that have different materials on both sides, such as 1810. There is a side. In step 1503, the region is divided based on the generated subdivision side (step 1503). Divide into subdivided parts along the subdivision side.
Here, as shown in FIG. 16B, the overlapping area is divided for each surface (steps 1607 and 1608). FIG. 19A shows the partial area after such area division. 1902, 1903, 190
4, 1905, 1906 are hidden under other areas. In FIG. 15, the invalid partial area is subsequently erased (step 150
Four). This deletion, as shown in FIG. 17 (a), for each partial area (step 1702), it is determined whether or not it is hidden under another partial area (step 1703), and if so, that partial area to erase. FIG. 19 (b) shows the partial area after this processing. At this stage, the overlapping of figures disappears. Here, as shown in FIG. 17 (b), for each partial region (step 1706), if the material and source term of the adjacent partial regions match, these partial regions are integrated (steps 1707, 1708, 1709). Finally, the partial areas having the same attribute are integrated (step 1505). Finally, as shown in Fig. 20 (a), the defined polygons are 2002, 2003, and 2004.
Is directly decomposed into two regions. This obtained area is shown in Figure 20.
Displayed as (b).

The physical model information is generated by the above processing.

FIG. 21 shows a general format of physical model information (2101). 2102 is information about the entire model, which includes the model name, time dependence, and analysis field (one or more). twenty one
02 is point information, and the information of each point consists of its name and coordinates.
Reference numeral 2104 is line information, and each line includes a name of a start point and an end point, whether or not it is an external boundary (this differs for each analysis field), and boundary conditions (this also differs for each analysis field). 2105 is a region name, and each region is a set of lines that make up it, material (this does not depend on the analysis field), whether it is an analysis area (this is different for each analysis field), source item
(This also differs for each analysis field).

FIG. 22 is a visual representation of physical model information obtained by converting the user-defined information of FIG. 10 by the boundary extraction processing and the direct sum decomposition processing. The appearance changes depending on the analysis field. 2201 represents a physical model from the viewpoint of thermal analysis, and 2202 represents a physical model from the viewpoint of fluid analysis.
Each point, each line, and each area is given a name beginning with p, l, and r and identified by serial numbers.

FIG. 23 is a specific example of the physical model information of FIG. 2302 is the problem name, time dependency, analysis field (two in this case). 2303 is information of points, and each of the 7 points is represented by coordinates. Reference numeral 2304 denotes line information, which has seven sides including a start point, an end point, and boundary conditions (each of which records boundary conditions corresponding to two analysis fields). 1105 is surface information, one triangle, one rectangle with a triangular hole, whether it is a material (regardless of the number of analysis fields) and analysis target
Information is added (corresponding to the effectiveness of the materials described above, which differs for each analysis field), and source term (which differs for each analysis field). Especially r002 (rectangle with triangular holes)
Note the composition line of. The addition of the three lines forming the hole is a remarkable difference from the rectangle s002 when compared with the user-defined information in FIG.

As shown in FIG. 1, the physical model information 102
Is converted into a simulation code 104 and calculation grid data 105 by the simulation code conversion program 103. Simulation code conversion program 103
Is composed of a simulation code generation unit 2401 that uses a material database 2501 and a numerical calculation knowledge base 2601, and a calculation grid generation unit 2491 that generates a finite element mesh.

As a method for generating a calculation grid, many methods have been proposed in the past, including those manually created and automatically generated, and since it is irrelevant to the essence of the present invention, a specific processing I will omit the contents.

The simulation code generation method uses the same method as the invention of Japanese Patent Application No. 3-338165.
A brief explanation will be given here as well.

FIG. 24 shows a simulation code generator 2401.
Is a processing flow of.

First, the model name, the analysis field, and the time dependency are obtained from the physical model information 102 (step 2402), and further the calculation grid data 105 generated in advance is obtained (step 2403).

Next, variables are obtained from the numerical calculation knowledge base 2601 for each analysis field (step 2404) (step 2405).
Furthermore, we collect the necessary information about the area to be analyzed and the external boundary. Whether or not a certain area is an analysis target (step 2407) differs depending on the analysis field and is recorded in the physical model information 102. For the area determined to be the analysis target, the finite element from the computational grid data, the material constant from the material database 2501 with the material name attached to that area as the key, and the source term name attached to that area as the key Numerical calculation knowledge base
Obtain source term information from 2601 respectively (step 2408
~ 2410). Also, for each external boundary, the boundary condition name attached thereto is used as a key to obtain boundary condition information from the numerical calculation knowledge base 2601. It should be noted that the fact that a certain boundary is an external boundary indicates that it is an analysis target in the analysis field, and therefore it is not necessary to check again whether it is the analysis target.

Next, a rule for selecting a solution method is obtained from the numerical calculation knowledge base 2601 (step 2413), and a numerical solution method suitable for a given physical model is selected from among several candidates (step 2413). Step 2414). A simulation program for executing each of the selected numerical solution method and boundary conditions is generated (steps 2415, 2416), and these are combined into one simulation program. In this way, the simulation code generation process is completed.

FIG. 25 shows a general format of the material database 2501. 2502 is information on the entire material database, listing the types of material constants recorded in the database. 2503 is information about each material, which is the name and phase of the material.
(Discrimination between solid, liquid and gas), and the value of each material constant.

FIG. 26 shows a general format of the numerical calculation knowledge base 2601. Build for each analysis field. 2602 is information about one analysis field as a whole.
Describe the variables, equations, and material constants used. 2603 is information for each numerical solution candidate, which consists of a numerical solution name, conditions for selecting the numerical solution, and a program for generating a simulation code.

FIG. 27 shows a concrete example of the concrete material database 2501. The type of material constant recorded in the database is recorded in 2702, and the value of the material constant for each material is recorded in 2703.

FIG. 28 is a concrete example of a concrete numerical calculation knowledge base 2601. 2802 is a knowledge base for thermal analysis, in which variables such as temperature, equations and material constants are described. 2803 is a candidate solution for thermal analysis. 2804 is a knowledge base for fluid analysis.

FIG. 29 is a general format of simulation code. The specific simulation program includes the information listed here. The first is information about the entire simulation, which is composed of 2903 simulation names and time dependence, analysis fields, and 2903 computational grid data names. After 2904, information for each analysis field, analysis field name and variable name, material constant and source term and boundary condition,
It consists of a solution and a simulation program. It is not always necessary to describe in the form of the information listed here, and it may be described without dividing it into analysis fields, or information common to the entire area may be specified collectively.

FIG. 30 shows a general format of calculation grid data for storing information on finite elements. 3002 is information regarding the entire calculation grid data, which is composed of the calculation grid name. 3003 is node information, and the information of each node consists of a node name and coordinates. 30
04 is finite element information, and each finite element consists of a finite element name and the node names that make up it. 3005 is boundary element information,
Each boundary element is composed of a boundary element name and the node names that compose it. Note that the boundary element is, of the constituent edges of the finite element,
It is included in the outer boundary.

FIG. 31 shows a concrete example of the simulation code. Here, the system according to the present invention outputs the source code of the programming language DEQSOL for partial differential equations, but it is also possible to output a general programming language such as FORTRAN. Although the format is different from that in Fig. 29, it contains the information necessary for performing the simulation.

FIG. 32 shows a concrete example of the calculation grid data. 32
02 is node data, and the information of each node consists of a name and coordinates identified by serial numbers starting with n. 3203 is finite element data, and the information of each finite element includes a name that starts with m and is identified by a serial number, the type of finite element (triangular element in this case), and the names of the nodes that make up the finite element. Reference numeral 3204 denotes area information, which includes the number and dimension of finite elements belonging to each shape area (here, the area is a plane and thus is two-dimensional), and the names of finite elements constituting the area. The external boundary information 3205 includes the number and dimension of the nodes belonging to each external boundary (one-dimensional because the boundary is a line element), and the names of the nodes forming the boundary.

[0096]

By allowing the overlapping of the area shapes, it becomes easy to input and correct a complicated area shape.

Next, since the area shape and the analysis condition can be set in parallel, it is possible to describe the problem in a more natural form for the user.

Further, by automatically extracting the external boundary, it is possible to eliminate the complexity of managing the boundary condition for the user.

As described above, according to the present invention, the problem input at the time of simulating the physical phenomenon can be simplified, the input error can be reduced, and the man-hour can be greatly reduced.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of each processing module according to the present invention, a flow of processing, a flow of information inside a system, and a configuration of a processing module.

FIG. 2 is a diagram showing a configuration of the entire apparatus.

FIG. 3 is a diagram showing a coupled problem of thermal analysis and fluid analysis taken as an example.

FIG. 4 is a diagram showing an appearance of a user interface.

FIG. 5 is a diagram showing a procedure for defining a problem.

FIG. 6 is a diagram showing a procedure for defining a problem.

FIG. 7 is a diagram showing a procedure for defining a problem.

FIG. 8 is a diagram showing a procedure for defining a problem.

FIG. 9 is a diagram showing a general format of user-defined information.

FIG. 10 is a diagram visualizing user-defined information corresponding to an example.

FIG. 11 is a specific example of user-defined information.

FIG. 12 is a flow of boundary extraction processing.

FIG. 13 is a flowchart of a material effectiveness determination process.

FIG. 14 is a conceptual diagram of boundary extraction processing.

FIG. 15 is a flow of a direct sum decomposition process.

FIG. 16 is a flowchart of subdivision side generation processing and area division processing.

FIG. 17 is a flowchart of a process of deleting an invalid partial region and a process of integrating partial regions having the same attribute.

FIG. 18 is a conceptual diagram of a direct sum decomposition process.

FIG. 19 is a conceptual diagram of a direct sum decomposition process.

FIG. 20 is a conceptual diagram of a direct sum decomposition process.

FIG. 21 is a general format of a physical model.

FIG. 22 is a diagram visualizing physical model information corresponding to an example.

FIG. 23 is a specific example of physical model information.

FIG. 24 is a flow of a simulation code generation process.

FIG. 25 is a general format for a material database.

FIG. 26 is a general form of a Numerical Computation Knowledge Base.

FIG. 27 is a specific example of a material database.

FIG. 28 is a specific example of a numerical calculation knowledge base.

FIG. 29 is a general form of simulation code.

FIG. 30 is a general format of computational grid data.

FIG. 31 is a specific example of simulation code.

FIG. 32 is a specific example of calculation grid data.

[Explanation of symbols]

901 ... General format of user-defined information, 1001 ... User-defined information regarding thermal analysis, 1002 ... User-defined information regarding fluid analysis, 1101 ... Specific example of user-defined information, 2101 ... General format of physical model information, 2201 ... Regarding thermal analysis Physical model information, 2202 ... Physical model information regarding fluid analysis, 2301 ... Specific example of physical model information, 2501 ... General format of material database, 2601 ... General format of numerical calculation knowledge base, 2701 ... Specific example of material database, 2801 ... Specific example of numerical calculation knowledge base, 2901 ... General format of simulation code, 30
01 ... General format of calculation grid data, 3101 ... Specific example of simulation code, 3201 ... Specific example of calculation grid data.

Claims (6)

[Claims] 1. An area to be input in a system having a processing device, an input device for inputting information to the processing device, and a display device for displaying the input information and the information generated by the processing device. The operator inputs area information including information about a figure representing the area and information identifying a substance occupying the area, and based on the input area information, the area is The substance is displayed on the display device so that it can be identified, and the input and display are repeated for a plurality of other regions and substances occupying the respective regions, and in the repetition, the region information input later is displayed. When one area to be designated has a portion which overlaps with another area designated by at least one other area information previously input,
When the substance occupying the one region and the substance occupying the other region are different from each other, with respect to the overlapping portion, the region input later is preferentially displayed, and the substance occupying the one region. And the material occupying the other region is the same, and if the one region has a portion that overlaps the other region, or if the one region is adjacent to the other region,
The other areas are displayed as areas connected to the one area, and after the repetition, a plurality of effective areas configured by a plurality of areas designated by a plurality of area information input during the repetition Region information including graphic information representing each of the plurality of effective regions and information identifying the substance occupying each of the plurality of effective regions is generated, and in the determination, the region information specified later is designated. When one area has a portion overlapping with another area specified by at least one other area information previously input, when the substance occupying the one region and the substance occupying the other region are different, ,
Of the one area, the overlapping portion is invalidated, the material occupying the one area is the same as the material occupying the other area, and the one area is overlapped with the other area. Or the one region is adjacent to the other region, the one region and the other region are integrated into one continuous region, and the invalidity and the integration are different from each other. An area input method that repeats for a set of areas. 2. The area information representing the plurality of valid areas includes information about an internal boundary representing a boundary between each valid area and another valid area, and a set of the plurality of valid areas and a set other than the set. The area input method according to claim 1, further comprising information on an outer boundary representing a boundary with the area. 3. The area information entered by the operator,
2. The area input method according to claim 1, further comprising the step of adding information specifying a predetermined substance to the input area information when the area information does not include information about the material occupying the specified area. . 4. The area input method according to claim 1, wherein one of the plurality of area information input by the operator includes area information designating a cavity indicating that there is no area to be input. 5. The area input method according to claim 1, wherein the determination of the plurality of effective areas includes a step of performing a direct sum decomposition of overlapping area shapes and converting the overlapping area shapes into a set of non-overlapping figures. 6. A plurality of valid areas are analyzed by further inputting information describing a problem to be simulated, and based on the area information designating the plurality of judged valid areas and the inputted problem description information. The method further comprises the step of automatically generating a simulation program for obtaining a solution to the problem.
The described simulation program generation method.