JP2009122999A - Three-dimensional shape optimization apparatus and three-dimensional shape optimization method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To shorten the time required to correct a three-dimensional shape model for meeting design specifications, by automatically generating from a schematic shape a three-dimensional CAD model for shape optimization in which local dimensions and members can be changed when necessary and applying the generated three-dimensional CAD model as an initial shape to shape optimization calculations. <P>SOLUTION: A three-dimensional shape optimization apparatus for determining types and dimensions of parts constructing a three-dimensional CAD model includes a means for inputting a schematic shape, a means for generating and displaying a three-dimensional CAD model from the schematic shape and a means for specifying members that need reinforcing. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a three-dimensional shape optimization apparatus and a three-dimensional shape optimization method.

  Usually, in the design of a machine structure, in order to satisfy a given design specification, parameters such as the shape and material conditions of the machine structure are used by using CAD (Computer Aided Design) and CAE (Computer Aided Engineering). Various parameter surveys are carried out, and design solutions that satisfy the design specifications are derived from the analysis results obtained there. At this time, the optimization technique is used to automate the parameter survey, and by performing the shape optimization calculation for predicting the design solution, the derivation can be performed in a short time.

  Usually, when a new design is performed, an outline shape has been found by using an optimization technique represented by a homogenization method. This schematic shape clearly shows only a portion necessary for strength in a region called a voxel mesh. For this reason, it is close to the image diagram without considering whether it can be manufactured. For this reason, when actually applying to the design, a new 3D CAD model is generated by interactive work with reference to the output approximate shape, and shape optimization calculation is performed using this as an initial shape. The design solution was calculated by

  Conventionally, Patent Document 1 discloses a three-dimensional shape automatic generation technique for generating a three-dimensional model from a shape such as an image diagram. In Patent Document 1, in order to stably extract a three-dimensional shape of a subject from an arbitrary subject image, an image input unit that inputs a subject image including an arbitrary background, and an image cutout unit that extracts the subject shape from the subject shape And a method using shape correction means for correcting the shape of a standard three-dimensional shape model similar to the subject based on the cut subject shape.

Japanese Patent Laid-Open No. 2000-194859

  In the conventional three-dimensional shape automatic generation technology, a corresponding three-dimensional shape model is extracted from the input shape by a similarity search of the input shape. However, generally, a three-dimensional shape model used for machine design is created by a system called three-dimensional CAD. This three-dimensional CAD generates a three-dimensional shape model by combining basic shapes such as rectangular solids and cylinders called shape features. Furthermore, components such as beams and girders can be handled as shape features, and the shape features can be expressed as parts such as H-shaped steel and U-shaped steel and expressed as an assembly of those shape features. is there.

  Now, when a 3D shape model is extracted by a conventional similarity search, a 3D shape model with a similar input overall shape is obtained, and individual shape features such as beams constituting the input shape are obtained. There is no guarantee that the shape the designer wants. That is, the housing structure of an assembly composed of beams and girders is originally searched from the similarity of the whole structure in the conventional method, so that the beams and girders are connected to form a single structure. there is a possibility. For this reason, in the case of a single structure, it is very difficult to perform shape optimization calculation using the dimension value of only the beam portion as a design variable. In addition, optimization calculation including the replacement of members with the members themselves as design variables is difficult to replace only the beams with other types of beams when the beams and girders are a single part. .

  When a 3D shape model is generated from a rough shape using such a conventional 3D shape automatic generation technology, in order to execute the shape optimization calculation due to the above-mentioned problem, beams and girders constituting the 3D shape model are used. Therefore, it takes a very long time to edit the three-dimensional CAD.

  Also in Patent Document 1, since only similar shapes are searched for and displayed with respect to the input approximate shape, it is referred to as generation of a three-dimensional shape model that takes into account the shapes and connection states of individual members constituting the mechanical structure. The point is not fully considered.

  The present invention automatically generates a three-dimensional CAD model for shape optimization that can change the size of a necessary part from a rough shape and change a necessary member, and optimizes the shape using the generated three-dimensional CAD model as an initial shape. The purpose of this is to reduce the time required to correct the 3D shape model to satisfy the design specifications.

  The three-dimensional shape optimization apparatus of the present invention is a three-dimensional shape optimization apparatus that determines the types and dimensions of parts constituting a three-dimensional CAD model, and includes means for inputting a rough shape, and the third order from the rough shape. It includes means for generating and displaying an original CAD model, and means for clearly indicating a member that needs reinforcement.

  According to the present invention, a three-dimensional CAD model for shape optimization that can change the size of a necessary portion and change a necessary member from a rough shape that is an output result of an optimization technique typified by a homogenization method. Is automatically generated, and the generated three-dimensional CAD model is used as the initial shape and applied to the shape optimization calculation, so that a design solution that satisfies the design specifications can be obtained in a short time.

  The present invention obtains the shape of a plane portion parallel to the XY plane, the XZ plane, and the YZ plane with respect to the input approximate shape by inputting an approximate shape that is an optimum calculation result typified by a homogenization method. Means for generating a two-dimensional plane area for the acquired plane partial shape, acquiring coordinate value data of the generated area, acquiring the generated plane area information, and acquiring the area of the acquired plane area; Calculate the ratio with the area of the rectangle that includes the plane area, and the plane area that exceeds the specified ratio will recognize all areas of the rectangle that are included as a plane area. Generate and acquire the coordinate value data of the neutral line, and further, as width information, obtain the means for calculating the diameter data of the inscribed circle created when generating the neutral line, and the created plane area information, Cuboid from database Obtain a three-dimensional CAD model, change the rectangular parallelepiped according to the coordinate value and thickness information of the plane information, obtain the neutral line information, obtain the three-dimensional CAD model of the H-section steel from the database, Change the H-section steel according to the coordinate value and width information, display the 3D CAD model generated with the rough shape to the operator, and if the same neutral line and width information are different, highlight the target member , Means for notifying that reinforcement is necessary, and the maximum number of generations, the number of populations for performing optimization calculation using an objective function, constraint conditions, and a genetic algorithm necessary for shape optimization calculation, A means for inputting design variables such as member changes and dimension changes, a means for inputting material conditions, constraint conditions, load conditions, and mesh conditions necessary for structural analysis, and creation Set the material conditions, restraint conditions, and load conditions entered in the analysis condition input section for the 3D shape model that has been created, and generate an analysis model by generating a mesh area according to the mesh conditions, and the created analysis model A means for performing structural analysis, evaluating the results of the analysis performed, and using genetic algorithms to calculate the best design variables that satisfy the constraints with minimum or maximum objective functions, and optimal Obtain the best 3D CAD model, the result, the approximate shape input in the approximate shape input unit, and the analysis result at that time, and display the difference so that the operator can recognize the difference. A three-dimensional shape optimizing apparatus including a method and a method thereof are proposed.

  This makes it possible to automatically generate a three-dimensional CAD model from the approximate shape and perform shape optimization calculation.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing a system configuration of an embodiment of a three-dimensional shape optimization apparatus according to the present invention. 1 includes a schematic shape input unit 101, a planar shape acquisition unit 102, a two-dimensional shape generation unit 103, a neutral line / plane generation unit 104, a three-dimensional CAD shape generation unit 105, and an optimization condition input unit 106. Analysis condition input unit 107, analysis model creation unit 108, analysis unit 109, optimum calculation control unit 110, optimum calculation display unit 111, database 112, and computer 113.

  The approximate shape input unit 101 inputs an approximate shape which is an optimum calculation result typified by a homogenization method. The planar shape acquisition unit 102 acquires the shape of the plane portion parallel to the XY plane, the XZ plane, and the YZ plane with respect to the approximate shape input by the approximate shape input unit 101. The two-dimensional shape generation unit 103 generates a two-dimensional planar region for the planar partial shape acquired by the planar shape acquisition unit 102, and acquires coordinate value data of the generated region.

  The neutral line / plane generation unit 104 acquires the plane area information generated by the two-dimensional shape generation unit 103, calculates the ratio between the area of the acquired plane area and the area of the rectangle that includes the plane area, and is designated For a plane area that is greater than or equal to the ratio, the entire rectangular area that it contains is recognized as a plane area. When the ratio is equal to or less than the ratio, a neutral line is generated by the line segment, and the coordinate value data of the neutral line is acquired. Furthermore, as the width information, diameter data of the inscribed circle created when the neutral line is generated is calculated.

  In the three-dimensional CAD shape creation unit 105, the plane area information created by the neutral line / plane generation unit 104 is obtained, a three-dimensional CAD model of the rectangular parallelepiped is obtained from the database, and the rectangular parallelepiped is obtained according to the coordinate value and thickness information of the plane information. To change. In addition, the neutral line information is acquired, the three-dimensional CAD model of the H-section steel is acquired from the database, the H-section steel is changed according to the coordinate value and width information of the neutral line information, and the tertiary generated along with the rough shape by the operator The original CAD model is displayed. Further, when the width information is different in the same neutral line, the target member is highlighted and a notice is given that reinforcement is necessary.

  In the shape optimization condition input unit 106, the objective function necessary for the shape optimization calculation, the constraint conditions, the maximum generation number for performing the optimization calculation using the genetic algorithm, the number of populations, the change of members, the size change, etc. Enter the initial value of the design variable. The analysis condition input unit 107 inputs material conditions, constraint conditions, load conditions, and mesh conditions necessary for the structural analysis.

  The analysis model creation unit 108 sets the material conditions, constraint conditions, and load conditions input by the analysis condition input unit 107 for the 3D shape model created by the 3D CAD shape creation unit 105, and sets the region according to the mesh conditions. Generate an analysis model by generating a mesh. The analysis unit 109 acquires the analysis model created by the analysis model creation unit 108 and performs structural analysis.

  The optimal calculation control unit 110 evaluates the analysis result executed by the analysis unit 109, and calculates the best design variable that satisfies the constraint condition with the objective function being the minimum or maximum using a genetic algorithm. In the optimum result display unit 111, the best three-dimensional CAD model obtained by executing the optimum calculation in the optimum calculation control unit 110, the result at that time, the approximate shape input by the approximate shape input unit 101, and the analysis result at that time Are displayed so that the operator can recognize each difference.

  In the database 112, the approximate shape input unit 101, the plane shape acquisition unit 102, the two-dimensional shape generation unit 103, the neutral line / plane generation unit 104, the three-dimensional CAD shape generation unit 105, the optimization condition input unit 106, and the analysis condition input unit 107, the analysis model creation unit 108, the analysis unit 109, the optimum calculation control unit 110, and the results calculated by the optimum calculation display unit 111 are accumulated. The computer 113 controls 101 to 112 constituting this apparatus.

  A processing procedure of the embodiment configured as described above will be described with reference to FIGS. 2, 3, 4, 5, and 6 are flowcharts illustrating a use procedure in the three-dimensional shape optimization apparatus illustrated in FIG. 1. The utilization procedure of the present invention is roughly divided into three phases. The first is phase 1 in which a rough shape is input and a three-dimensional CAD model that is an initial shape for optimization calculation is generated. The second phase is a phase 2 in which conditions necessary for shape optimization and conditions necessary for execution of structural analysis are input. The third is to perform the structural analysis according to the shape optimization conditions and analysis conditions input in the second phase with the first created 3D CAD model as the initial shape, and the objective function is maximized or In phase 3, the optimum calculation is executed so as to satisfy the minimum and constraint conditions, and the best result is output.

  First, S1 of phase 1 will be described.

  In S1, when performing the optimization calculation, a rough shape from which a three-dimensional CAD model serving as an initial shape is created is input. FIG. 7 shows an example of a schematic shape to be input and an input screen. The approximate shape is an output result optimized by the homogenization method, and is a computer case. The homogenization method divides a three-dimensional optimization target region into a difference grid called a voxel mesh, and executes optimization calculation to clearly show cells composed of each part of the grid that is necessary for strength. .

  FIG. 7 shows that in the computer case, the hatched portion and the solid line portion indicate areas that are required in terms of strength, and the other portions are not required. That is, in the phase 1, a three-dimensional CAD model of the housing composed of the hatched area and the solid line area is created. The operator presses the “reference” button to display the data list, and selects the target rough shape data. The schematic shape input unit 101 reads the selected data and displays it on the screen. If the approximate shape of the displayed screen is appropriate, the operator presses the “input” button, and the approximate shape input unit 101 stores the input approximate shape in the database.

  In S2, the planar partial shape of the approximate shape input in S1 is acquired. The planar shape acquisition unit 102 obtains a rectangular parallelepiped shape including the approximate shape stored by the approximate shape input unit 101 by calculation. Here, a method for obtaining a planar partial shape parallel to the XY plane will be described. The planar shape acquisition unit 102 calculates the number of cells on the XY plane in the first stage of the voxel cell from the minimum value in the Z-axis direction of the rectangular parallelepiped shape obtained by calculation, and XY one by one up to the maximum value in the Z-axis direction. The number of planar cells is calculated and stored in the database. FIG. 8 shows the distribution of the number of cells in the XY plane from the stored minimum value to maximum value in the Z-axis direction. As is apparent from the schematic shape, it can be seen that there are many cells in the floor portion and the ceiling portion of the schematic shape.

  Next, the planar shape acquisition unit 102 calculates the Z-axis value at the specified threshold value and stores each value in the database. The threshold value can be arbitrarily specified by the operator, but is assumed to be 300 here. The Z-axis values stored in the database are referred to as A, B, C, and D for the sake of explanation, and are indicated by black circles in FIG. Between A and B is the floor portion, and between C and D is the ceiling portion. Further, FIG. 8 shows the positional relationship between points A, B, C, and D having a schematic shape.

  Next, the planar shape acquisition unit 102 acquires schematic shape data, acquires a shape parallel to the XY plane and the shape from point A to point B on the Z axis, parallel to the XY plane and shape from point C to point D. And store each in the database. This process can be realized by acquiring voxel cells in a specified area. Thus, a plane part shape parallel to the XY plane can be acquired, and parts other than the floor part and the ceiling part can be excluded.

  Next, the planar shape acquisition unit 102 also acquires planar partial shapes parallel to the XZ plane and the YZ plane by the same method, and stores these shapes in the database. An example of the planar partial shape parallel to the acquired XZ plane is shown in FIG. The planar shape acquisition unit 102 acquires a total of six shapes, including two planar partial shapes parallel to the XY plane, two planar partial shapes parallel to the XZ plane, and two planar partial shapes parallel to the YZ plane. Is stored in the database.

  In S3, the two-dimensionalization of the planar shape acquired in S2 is performed. The two-dimensional shape generation unit 103 acquires the planar partial shape data stored in S2, and obtains a rectangular parallelepiped shape including the acquired planar partial shape data by calculation. Here, processing of a planar shape portion parallel to the XY plane will be described.

  Next, the two-dimensional shape generation unit 103 converts the planar partial shape into two dimensions. That is, the shape projected onto the XY plane is obtained by calculation and stored in the database. This can be done by calculating the union of voxels parallel to the XY plane. Next, the coordinate value of the two-dimensional shape is obtained by calculation. This can be calculated from the coordinate value of the vertex of the rectangular parallelepiped shape. For the X and Y coordinates, the coordinate values obtained from the vertices of the rectangular parallelepiped are used as they are, and for the Z coordinates, the values of intermediate points at two points in the Z-axis direction in the rectangular parallelepiped are used and stored in the database. That is, the planar partial shape has rectangular coordinate value data including the planar shape at an intermediate position in the thickness direction of the rectangular parallelepiped. Next, the two-dimensional shape generation unit 103 calculates coordinate values for the remaining planar partial shapes by the same method, performs two-dimensionalization processing, and stores each in the database.

  In S4, the two-dimensional shape calculated in S3 is neutralized and planarized. The neutral line / plane generation unit 104 acquires the two-dimensional shape data calculated in S3. Next, the neutral line / plane generation unit 104 calculates the area of the region. Furthermore, the coordinate value is acquired, the area of the rectangle that includes the area is calculated, and the ratio of the area of the area to the area of the rectangle is calculated. If the ratio is equal to or greater than the specified value, the neutral line / plane generation unit 104 stores the target region in the database as a plane. The ratio can be arbitrarily specified by the operator, but here it is set to 0.7. If the ratio is equal to or less than the specified value, the neutral line / plane generation unit 104 generates a neutral line of the region.

  First, the boundary line of the plane part region is extracted. Next, the neutral line / plane generation unit 104 creates a neutral line for the internal region of the extracted boundary line. The neutral line can be created by using a known technique such as a MAT (Medium Axis Transform) method. Next, the neutral line / plane generation unit 104 connects the end points on the boundary of the generated neutral line with straight lines, calculates the coordinate values of the end points, and stores them in the database. Next, the neutral line / plane generation unit 104 also stores the diameter of the circle used when creating the neutral line using the MAT method in the database. This circle is inscribed in the area and means the width of the area.

  FIG. 10 shows an example of plane creation when the process of S4 on the XY plane is performed. FIG. 11 shows an example of creating a neutral line when the process of S4 on the XZ plane is performed. In addition, an inscribed circle is also indicated, indicating that the inscribed circle corresponds to the width. That is, when the ratio is 0.7 or more, the entire region is filled, and when it is 0.7 or less, the region is made neutral. In this way, if a three-dimensional CAD model is faithfully created in a portion necessary for strength, the shape of a hole or the like becomes more complicated than necessary, and actual production becomes difficult. However, since the neutral line / plane generation unit 104 simplifies flattening and neutral lines, a three-dimensional CAD model can be configured with simple members, and standard parts can be used.

  In S5, a three-dimensional CAD model is automatically generated according to the plane and neutral line information obtained up to S4. The three-dimensional CAD shape creation unit 105 acquires the planarization information and neutralization information obtained in S4. First, generation of a three-dimensional CAD model from a neutral line will be described. The three-dimensional CAD shape creation unit 105 acquires a three-dimensional CAD model of the H-shaped steel from the database 112, and changes the length, width, and position of the H-shaped steel from the coordinate value and width information of the neutral line information. Here, since the diameter of the inscribed circle is different at each point of the neutral line, the average value of the width information is used.

  Next, generation of a three-dimensional CAD model from a plane will be described. The three-dimensional CAD shape creation unit 105 acquires a three-dimensional CAD model of a rectangular parallelepiped from the database 112, and changes the vertical, horizontal, thickness, and position of the rectangular parallelepiped from the coordinate values of the plane information. A similar process is performed on all neutral plane and plane information to create a three-dimensional CAD model and store it in the database. An example of the created three-dimensional CAD model is shown in FIG. In the figure, the schematic shape input by the schematic shape input unit 101 is displayed in the upper stage, and the three-dimensional CAD model generated in S5 is displayed in the lower stage. Here, as shown in the lower part of the figure, a member having an appropriate shape is clearly shown at a place where reinforcement is required. This member is called a “member that needs reinforcement”.

  Also, the three-dimensional CAD shape creation unit 105 acquires the width information of the neutral line information, and when the width at each point of the neutral line exceeds a value specified by the operator, such as 20% or more, the neutral line information is obtained. The H-section steel generated from is clearly indicated to the operator, and the fact that reinforcement is necessary is notified. This is because when the reinforcement of the base of the member is necessary, the reinforcement shape is calculated in the general shape, but all of the reinforcement information is replaced with a straight H-section steel in S5. This is to prevent omission. In addition, the operator can compare the input approximate shape with the generated three-dimensional CAD model and interactively correct it in view of restrictions in production.

  Next, phase 2 will be described. In phase 2, in order to perform optimization calculation of the three-dimensional CAD model generated in phase 1, an operator inputs various conditions necessary for the optimization calculation.

  In S6, input necessary for the shape optimization calculation is performed. An example of an input screen for shape optimization conditions is shown in FIG. The shape optimization condition input unit 106 acquires the three-dimensional CAD model generated in phase 1 and displays it to clearly indicate to the operator that it is a shape optimization target. The operator inputs conditions necessary to satisfy the design specifications for the optimization calculation. In order to obtain a rough shape as a result of the optimal calculation by the homogenization method and a three-dimensional CAD shape at the same level, a condition obtained by calculating the rough calculation is set. Here, the minimum mass is input as the objective function, the maximum displacement is 0.5 mm or less, and the maximum generation number 50 and the population number 40 for performing the optimization calculation using the genetic algorithm are input as the constraint conditions.

  Next, input initial values of design variables. Here, whether to change to another U-shaped steel or L-shaped steel is specified for individual members such as an H-shaped steel constituting the three-dimensional CAD model. In FIG. 13, a list of member names constituting the three-dimensional CAD model such as member 1 and member 2 is displayed, and whether or not to change the member is designated for each. If the member is clicked with a mouse on the display screen of the three-dimensional CAD model, the operator can know the member name by displaying the member names such as member 1 and member 2. Since the floor material does not need to be changed in the optimization calculation, the member 3 corresponding to the floor material is designated OFF in FIG. 13 (change of member 3) and is not changed. In addition, by pressing a detail button, design variables can be designated for each member from among members that can be changed, such as H-shaped steel, U-shaped steel, and L-shaped steel.

  Here, members such as H-shaped steel, U-shaped steel, and L-shaped steel, which are elements for generating a three-dimensional CAD model, are referred to as basic members. Data relating to the basic member is recorded in advance on a computer-readable recording medium, that is, a database. The three-dimensional CAD model can be regarded as an assembly including basic members.

  FIG. 14 shows an example of a detailed setting screen. H-shaped steel, U-shaped steel, L-shaped steel, and T-shaped steel can be changed for the specified member, but here the members are changed from H-shaped steel, U-shaped steel, and L-shaped steel. Is done. Next, design variables are also input regarding the dimensions of the members. In FIG. 13, it is specified whether or not to change the dimensions of the members 1 and 2. Here, if a detailed button is pressed, detailed design variables such as dimensional tolerances such as the height and width of the H-section steel can be input. FIG. 15 shows an example of a detailed setting screen. For the specified member, the range in which the dimension can be changed in the optimization calculation and the initial state value are specified. The shape optimization condition input unit 106 acquires the input information and stores it in the database.

In S7, analysis conditions necessary for the structural analysis are input. An example of the analysis condition input screen is shown in FIG. The analysis condition input unit 107 displays an analysis model for the three-dimensional CAD model generated in phase 1. The operator inputs conditions necessary for analysis. As a load condition, an evenly distributed load of 1 N / mm ³ is input to the upper surface of the casing. Further, as a constraint condition, a constraint in the XYZ directions is input to a surface where the casing and the floor are in contact. The operator can input conditions by clicking a mouse on a surface, a line, a point, or the like for inputting a load condition or constraint condition on the analysis model display screen. Moreover, steel is input as a material. As a result, the analysis condition input unit 107 can acquire member-specific attributes such as steel Young's modulus and Poisson's ratio from the database. In addition, it is also possible to input those materials for each member constituting the analysis model. In addition, the size for mesh generation necessary for executing structural analysis is input. As described above, various conditions necessary for the analysis are input. The analysis condition input unit 107 acquires the input information and stores it in the database.

  Next, phase 3 will be described. In phase 3, the 3D CAD model generated in phase 1 is subjected to structural analysis in accordance with the initial state, optimization calculation conditions and analysis conditions input in phase 2, and the results are represented by genetic algorithms (Genetic Algorithm, in the figure). Optimization calculation is executed by iterative calculation using GA), and the best three-dimensional CAD model is calculated.

  In S8, a three-dimensional CAD model is generated. The three-dimensional CAD shape generation unit 105 acquires iteration number information from the database. When the number of iterations is 1, the 3D CAD model generated in phase 1 is acquired. If the number of iterations is 2 or later, obtain design variable information from the database and change the height and width dimension values of the H-section steel of each member according to the information described in the design variable information. A new three-dimensional CAD model is generated and stored in a database by replacing it with L-shaped steel or T-shaped steel.

  In S9, an analysis model is created. The analysis model creation unit 108 acquires the three-dimensional CAD model generated in S8 from the database, and creates an analysis model. The analysis model creation unit 108 acquires the analysis condition information input in S7, generates a calculation region according to the input constraint conditions, load conditions, and material condition settings, and the input mesh size, and stores them in the database. .

  In S10, structural analysis is executed. The analysis unit 109 acquires an analysis model from the database, executes structural analysis, and stores the analysis result in the database.

  In S11, the analysis result is evaluated, and optimization calculation is performed using a genetic algorithm (GA). The optimal calculation control unit 110 acquires the result of the structural analysis executed in S10 from the database. Further, the information input in the optimization problem setting input in S6 is acquired from the database. The optimum calculation control unit 110 repeats if the mass that is the objective function is the minimum and the maximum displacement that is the constraint condition is within 0.5 mm or exceeds the maximum generation number 50 input in S6. The calculation is terminated and the process proceeds to S12. Otherwise, crossover / mutation is performed using genetic algorithm (GA), design variables are determined to generate 40 new populations of the next generation, design variable information and number of iterations are stored in the database And return to S8.

  In S12, the optimum calculation result is displayed. In the optimum result display unit 111, the best three-dimensional CAD model obtained up to S11 is acquired from the database, and is output by the homogenization method input in S1, but the approximate shape is acquired from the database. Display the difference so that the operator can recognize it. FIG. 17 shows an example of a result comparison screen displaying the approximate shape input in S1 and the best three-dimensional CAD model obtained up to S11. As shown in this figure, the operator can recognize the difference in each structure, and the performance can also be grasped by displaying each structural analysis result.

  In this embodiment, when generating a three-dimensional CAD model from a schematic shape, an H-shaped steel is used. However, other shapes such as an L-shaped steel can be used. When generating a three-dimensional CAD model from neutral line information, an L-section steel may be acquired and created instead of an H-section steel from a database.

  In this embodiment, the members are changed from H-shaped steel to L-shaped steel in each member, but it is possible to include the presence or absence of members in the optimization calculation, such as deleting the H-shaped steel at any location. It is. In the present embodiment, the optimization calculation result by the homogenization method is input as the approximate shape, but the voxel shape generated from the X-ray CT scanner can also be set as the approximate shape. In the present embodiment, the optimization calculation result by the homogenization method is input as the approximate shape, but the optimization calculation result by the density method can be set to the approximate shape. In this embodiment, the optimization calculation result by the homogenization method is input as the approximate shape, but the voxel image can be approximated.

  The above-described procedure (means) is usually implemented as a program for causing a computer to function. A computer-readable recording medium in which this program is recorded is also included in the three-dimensional shape optimization apparatus according to the present invention.

It is a whole block diagram of the three-dimensional shape optimization apparatus by this invention. It is a figure which shows the utilization procedure (phase 1) of this invention. It is a figure which shows the utilization procedure (phase 1 continuation) of this invention. It is a figure which shows the utilization procedure (phase 2) of this invention. It is a figure which shows the utilization procedure (phase 3) of this invention. It is a figure which shows the utilization procedure (phase 3 continuation) of this invention. It is a figure which shows an example which displayed on screen the schematic shape to input. It is a figure which shows distribution of the number of the cells parallel to XY plane. It is a figure which shows an example of the shape of the plane part parallel to a XZ plane. It is a figure which shows an example of the plane creation in XY plane. It is a figure which shows an example of the neutral line creation in XZ plane. It is a figure which shows an example of the produced | generated three-dimensional CAD model. It is a figure which shows an example of the input screen of an optimization problem. It is a figure which shows an example of the detailed setting screen of member change. It is a figure which shows an example of the detailed setting screen of a member dimension. It is a figure which shows an example of the input screen of analysis conditions. It is a figure which shows an example of the screen which compares a rough shape and the result of the best three-dimensional CAD model.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 101 ... Outline shape input part, 102 ... Planar shape acquisition part, 103 ... Two-dimensional shape generation part, 104 ... Neutral line / plane generation part, 105 ... Three-dimensional CAD shape creation part, 106 ... Shape optimization condition input part, 107 Analyzing condition input unit 108 Analyzing model creation unit 109 Analyzing unit 110 Optimal calculation control unit 111 Optimal result display unit 112 Database 113 Computer

Claims (20)

  A three-dimensional shape optimization method for determining the types and dimensions of parts constituting a three-dimensional CAD model, the step of inputting a rough shape, and the step of generating and displaying the three-dimensional CAD model from the rough shape A method for optimizing a three-dimensional shape including a step of clearly indicating a member that needs reinforcement.   Obtaining a two-dimensional region from the schematic shape; creating a model including a neutral line and the two-dimensional region; and generating the three-dimensional CAD model from the neutral line and the two-dimensional region. The three-dimensional shape optimization method according to claim 1, further comprising:   Obtaining a region of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape; creating the two-dimensional region from the shape of the plane portion; and obtaining coordinate data of the two-dimensional region. Acquiring, generating a model including the neutral line and the two-dimensional area from the two-dimensional area, and generating and displaying the three-dimensional CAD model from the neutral line and the two-dimensional area A process of inputting an objective function, a constraint condition, and a design variable for optimization calculation, a process of inputting a material condition, a constraint condition, and a load condition for a structural analysis, and a process of generating an analysis model The structural analysis using the analytical model, and the design variable is determined from the result of the structural analysis so that the objective function is minimized or maximized while satisfying the constraint condition. With features Three-dimensional shape optimization method according to claim 2, wherein that.   A step of obtaining a region of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape, and creating a two-dimensional region from the shape of the plane portion to obtain coordinate data of the two-dimensional region A step of creating a model including a neutral line and the two-dimensionalized region from the two-dimensionalized region, a step of inputting an objective function, a constraint condition, and a design variable for optimization calculation, and a structural analysis Inputting the material conditions, constraint conditions, and load conditions for the process, generating the analysis model, performing the structural analysis using the analysis model, and determining the constraint conditions from the results of the structural analysis. The design variable is determined so that the objective function is minimized or maximized while satisfying, and the three-dimensional CAD model that is an assembly including a basic member is generated from the neutral line and the two-dimensionalized region. And a step of optimizing the design variables using a genetic algorithm so that the objective function is minimized or maximized while satisfying the constraint condition from the result of the structural analysis. The three-dimensional shape optimization method according to claim 2, wherein the method is optimized.   Obtaining a region of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape, creating a two-dimensional region from the shape of the plane portion, and obtaining coordinate data of the two-dimensional region A step, a step of creating a model including a neutral line and the two-dimensionalized region from the two-dimensionalized region, a step of inputting an objective function, a constraint condition, a design variable for optimization calculation, and a material for structural analysis The step of inputting conditions, constraint conditions and load conditions, the step of generating an analysis model, the step of performing the structural analysis using the analysis model, and the objective function is minimized while satisfying the constraint conditions from the result of the structural analysis Or a step of determining the design variable so as to be maximized, a step of generating and displaying a three-dimensional CAD model that is an assembly including a basic member from the neutral line and the two-dimensionalized region; From the results of the structural analysis, the step of optimizing the design variable using a genetic algorithm so as to minimize or maximize the objective function while satisfying the constraint conditions, the approximate shape, and the performance in the approximate shape And a step of displaying the three-dimensional CAD model obtained by the optimization calculation so that the performance when using the three-dimensional CAD model can be compared, and a step of clearly indicating a member that needs reinforcement. The three-dimensional shape optimization method according to claim 1, wherein:   A three-dimensional shape optimizing apparatus for determining the type and size of a part constituting a three-dimensional CAD model, comprising: means for inputting a rough shape; means for generating and displaying the three-dimensional CAD model from the rough shape; The three-dimensional shape optimizing device for carrying out the three-dimensional shape optimizing method according to claim 1, further comprising: means for clearly indicating a member that needs reinforcement.   Means for acquiring a two-dimensional region from the schematic shape; means for creating a model including a neutral line and the two-dimensional region; and generating the three-dimensional CAD model from the neutral line and the two-dimensional region. The three-dimensional shape optimization apparatus according to claim 6, further comprising:   Means for acquiring an area of a plane part parallel to the XY plane, XZ plane, and YZ plane of the approximate shape, and creating the two-dimensional area from the shape of the plane part, and obtaining the coordinate data of the two-dimensional area Means for obtaining, means for creating a model including the neutral line and the two-dimensionalized region from the two-dimensionalized region, and generating and displaying the three-dimensional CAD model from the neutral line and the two-dimensionalized region Means for inputting objective functions, constraint conditions, and design variables for optimization calculation, means for inputting material conditions, constraint conditions, and load conditions for structural analysis, means for generating an analysis model, And means for analyzing the structure using the analysis model, and means for determining the design variable so that the objective function is minimized or maximized while satisfying the constraint condition from the result of the structure analysis. With features Three-dimensional shape optimization apparatus of claim 7, wherein that.   Means for acquiring an area of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape, creating a two-dimensional area from the shape of the plane section, and acquiring coordinate data of the two-dimensional area Means for creating a model including a neutral line and the two-dimensionalized region from the two-dimensionalized region, a means for inputting an objective function for optimization calculation, constraints, design variables, and a material for structural analysis Means for inputting conditions, constraint conditions, load conditions, means for generating an analysis model, means for performing the structural analysis using the analysis model, and the objective function is minimized while satisfying the constraint conditions from the result of the structural analysis Or means for determining the design variable so as to be maximized, means for generating and displaying a three-dimensional CAD model, which is an assembly including basic members, from the neutral line and the two-dimensionalized region; A means for optimizing the design variable using a genetic algorithm so that the objective function is minimized or maximized while satisfying the constraint condition from a result of structural analysis, and means for clearly indicating a member that needs reinforcement The three-dimensional shape optimization apparatus according to claim 7, comprising:   Means for acquiring an area of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape, creating a two-dimensional area from the shape of the plane section, and acquiring coordinate data of the two-dimensional area Means for creating a model including a neutral line and the two-dimensionalized region from the two-dimensionalized region, a means for inputting an objective function for optimization calculation, constraints, design variables, and a material for structural analysis Means for inputting conditions, constraint conditions, load conditions, means for generating an analysis model, means for performing the structural analysis using the analysis model, and the objective function is minimized while satisfying the constraint conditions from the result of the structural analysis Or means for determining the design variable so as to be maximized, means for generating and displaying a three-dimensional CAD model, which is an assembly including basic members, from the neutral line and the two-dimensionalized region; The means for optimizing the design variable using a genetic algorithm so that the objective function is minimized or maximized while satisfying the constraint condition from the result of the structural analysis, the approximate shape, and the performance in the approximate shape And means for displaying the three-dimensional CAD model obtained by the optimization calculation so as to be able to compare the performance when using the three-dimensional CAD model, and means for clearly indicating a member that needs reinforcement. The three-dimensional shape optimization apparatus according to claim 6, wherein   Means for inputting a rough shape; means for acquiring a two-dimensional region from the rough shape; means for creating a model including a neutral line and the two-dimensional region; the neutral line and the two-dimensional region; For generating and displaying a three-dimensional CAD model from the program and a program for functioning as means for clearly indicating a member that needs reinforcement.   Means for acquiring an area of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape; creating a two-dimensional area from the shape of the plane portion; and coordinate data of the two-dimensional area Means for generating a model including a neutral line and the two-dimensional area from the two-dimensional area, means for generating and displaying a three-dimensional CAD model from the neutral line and the two-dimensional area, Means for inputting objective functions, constraint conditions, design variables for optimization calculation, means for inputting material conditions, constraint conditions, load conditions for structural analysis, means for generating analysis models, using the analysis model Means for analyzing the structure, means for determining the design variable so that the objective function is minimized or maximized while satisfying the constraint condition from the result of the structure analysis, and reinforcement is required According to claim 11 of a program to function as demonstrating means member.   Means for acquiring an area of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape; creating a two-dimensional area from the shape of the plane portion; and coordinate data of the two-dimensional area Means for obtaining a model including a neutral line and the two-dimensional area from the two-dimensional area, an objective function for optimization calculation, a constraint condition, a means for inputting design variables, a structure analysis Means for inputting material conditions, constraint conditions, load conditions, means for generating an analysis model, means for performing the structural analysis using the analysis model, and satisfying the constraints from the result of the structural analysis A three-dimensional CAD model, which is an assembly including basic members, is generated from the means for determining the design variable so that the function is minimized or maximized, the neutral line, and the two-dimensionalized region. Means for displaying, means for optimizing the design variable using a genetic algorithm so that the objective function is minimized or maximized while satisfying the constraints from the result of the structural analysis, and a member that needs reinforcement 12. The program according to claim 11, which functions as a means for specifying.   Means for acquiring an area of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape; creating a two-dimensional area from the shape of the plane portion; and coordinate data of the two-dimensional area Means for obtaining a model including a neutral line and the two-dimensional area from the two-dimensional area, an objective function for optimization calculation, a constraint condition, a means for inputting design variables, a structure analysis Means for inputting material conditions, constraint conditions, load conditions, means for generating an analysis model, means for performing the structural analysis using the analysis model, and satisfying the constraints from the result of the structural analysis A three-dimensional CAD model, which is an assembly including basic members, is generated from the means for determining the design variable so that the function is minimized or maximized, the neutral line, and the two-dimensionalized region. Means for displaying, means for optimizing the design variable using a genetic algorithm such that the objective function is minimized or maximized while satisfying the constraints from the result of the structural analysis, the schematic shape, and the schematic Means for displaying so that the performance in shape, the three-dimensional CAD model obtained by the optimization calculation, and the performance when using the three-dimensional CAD model can be compared, and means for clearly indicating the member that needs reinforcement The program of Claim 11 for functioning as.   A computer-readable recording medium on which a program for causing a computer to function as means of the three-dimensional shape optimization apparatus according to claim 6 is recorded.   Means for acquiring a two-dimensional area from the schematic shape, means for creating a model including a neutral line and the two-dimensional area, and generating a three-dimensional CAD model from the neutral line and the two-dimensional area 16. A recording medium according to claim 15, wherein a computer-readable recording medium storing a program for functioning as a display means and a means for clearly indicating a member that needs reinforcement is recorded.   Means for acquiring a region of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape; creating a two-dimensional region from the shape of the plane portion; and coordinate data of the two-dimensional region Means for generating a model including a neutral line and the two-dimensional area from the two-dimensional area, means for generating and displaying a three-dimensional CAD model from the neutral line and the two-dimensional area, Means for inputting objective functions, constraint conditions, and design variables for optimization calculation, means for inputting material conditions, constraint conditions, and load conditions for structural analysis, means for generating an analysis model, and using the analysis model Means for performing the structural analysis, means for determining the design variable so that the objective function is minimized or maximized while satisfying the constraint condition from the result of the structural analysis, and means for clearly indicating the member that needs reinforcement. Recording medium recording a computer-readable claim 16, wherein the program for functioning.   Means for acquiring a region of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape; creating a two-dimensional region from the shape of the plane portion; and coordinate data of the two-dimensional region , Means for creating a model including a neutral line and the two-dimensional area from the two-dimensional area, an objective function for optimization calculation, a constraint condition, a means for inputting design variables, a structure analysis Means for inputting material conditions, constraint conditions, load conditions, means for generating an analysis model, means for performing the structural analysis using the analysis model, and satisfying the constraints from the results of the structural analysis Means for determining the design variable such that a function is minimized or maximized, means for generating and displaying a three-dimensional CAD model that is an assembly including a basic member from the neutral line and the two-dimensionalized region, and the structure Solution In order to function as a means for optimizing the design variable using a genetic algorithm so that the objective function is minimized or maximized while satisfying the constraint condition, and as a means for clearly specifying a member that needs reinforcement The computer-readable recording medium according to claim 16, wherein the program is recorded.   Means for acquiring a region of a plane portion parallel to the XY plane, XZ plane, and YZ plane of the approximate shape; creating a two-dimensional region from the shape of the plane portion; and coordinate data of the two-dimensional region , Means for creating a model including a neutral line and the two-dimensional area from the two-dimensional area, an objective function for optimization calculation, a constraint condition, a means for inputting design variables, a structure analysis Means for inputting material conditions, constraint conditions, load conditions, means for generating an analysis model, means for performing the structural analysis using the analysis model, and satisfying the constraints from the results of the structural analysis Means for determining the design variable such that a function is minimized or maximized, means for generating and displaying a three-dimensional CAD model that is an assembly including a basic member from the neutral line and the two-dimensionalized region, and the structure Solution Means for optimizing the design variable using a genetic algorithm so that the objective function is minimized or maximized while satisfying the constraint condition, the rough shape, the performance in the rough shape, and the optimization A program for functioning as a means for displaying the three-dimensional CAD model obtained by calculation so as to be able to compare the performance when using the three-dimensional CAD model, and a means for clearly indicating a member that needs reinforcement The recorded computer-readable recording medium according to claim 15.   In order to determine the types and dimensions of the parts constituting the three-dimensional CAD model, means for inputting a rough shape, means for generating and displaying the three-dimensional CAD model from the rough shape, and a member that needs reinforcement And a computer-readable recording medium on which data relating to a basic member serving as an element for generating the three-dimensional CAD model from the schematic shape is recorded.

Images