JP2005222178A - Optimum design device for structure, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To optimally design a structure including a frame element and a panel element. <P>SOLUTION: A prescribed design area is divided by a node, while the frame element, the panel element and a spring element are disposed between the nodes. Initial information thereof is stored in an initial information storage part 12. Analysis conditions including volume restriction of the frame element and the panel element, and cost restriction of the spring element are stored in an analysis condition information storage part. An optimizing arithmetic part 16 optimizes a thickness or a spring constant about each element such that a configuration of a structure of the design area becomes maximum, on the basis of the analysis conditions. The optimized initial information (information as an updated arithmetic result) is outputted from an output part 18. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to optimum design support for a structure, and more particularly to design support for generating a shape in an initial design stage.

  Conventionally, methods for designing various structures have been proposed. For example, in Patent Document 1, a design model in which a design area of a structure is divided by a plurality of nodes, adjacent nodes and non-adjacent nodes are coupled with frame elements is created, and the design model is optimized. A design support apparatus is disclosed.

JP 2001-195444 A

  In the above prior art, a method based on a frame element has been proposed. However, in a more general structural design, there is a limit in evaluating performance only with a frame element. In the body suspension of an automobile, the structure is composed of a frame structure and a panel structure. Therefore, an optimum design based on such more general frame elements and panel elements is desired.

  The present invention relates to node position information indicating the position of a node that divides a design area of a structure, a frame element that is arranged by connecting the nodes, and a panel element that is arranged in a divided area defined by the nodes. An initial information storage unit that stores initial information including element information that is information related to each other and element arrangement information that is information arranged so that the elements are connected to each other at node positions, and the elements are connected An analysis condition information storage unit for storing analysis condition information, which is information indicating an objective function to be optimized with respect to a design region of a structure formed by the method, and a constraint condition at the time of optimization of the objective function, Based on the analysis condition information, unnecessary elements are deleted from the frame elements and panel elements defined in the initial information, and the element information is changed in necessary elements. And optimization calculation section for optimized, computed by the optimizing operation unit, characterized by comprising an output unit for outputting the changed initial information as the operation result information.

  The analysis condition information preferably includes a boundary condition including an external force applied to the structure and a fixed condition, and a volume condition of the frame element and the panel element.

  In addition, it is preferable that the initial information storage unit further includes a spring element that connects the nodes.

  Further, it is preferable that the analysis condition information includes a cost condition of the spring element.

  Further, it is preferable that the panel element is modeled on the assumption of stress, and the rigidity of the panel element is limited to the in-plane direction.

  It is also preferred that the panel element has rigidity only with respect to shear stress.

  In addition, it is preferable that the optimization calculation unit determines a necessary part of the frame element and a necessary part of the panel element in the structure by causing substantial disappearance of unnecessary frame elements or panel elements. is there.

  In addition, it is preferable that the optimization calculation unit determines a position where the structure is to be joined by causing the unnecessary disappearance of the spring element.

  In addition, the optimization calculation unit performs a quantitative comparison of the optimal structure obtained by using the panel element having rigidity only for stress in a specific direction and the panel element having rigidity for stress in all directions. It is preferable to extract a mechanical factor that gives

  Moreover, it is preferable that the analysis condition information of the frame element and the panel element includes volume constraints set separately.

  Moreover, it is preferable that the analysis condition information of the frame element and the panel element includes a volume constraint set for the sum of the frame element and the panel element.

  Further, the optimum design support program according to the present invention includes a node position information indicating the position of a node that divides a design area of a structure, a frame element arranged by connecting the nodes, and a division defined by the nodes. Initial information including element information, which is information related to panel elements arranged in a region, and element arrangement information, which is information arranged so that the elements are connected to each other at node positions, is stored in the initial information storage unit Analysis condition information which is information indicating a step, an objective function to be optimized with respect to a design region of a structure formed by connecting the elements, and constraint conditions for optimization of the objective function An unnecessary condition is deleted from the frame element and the panel element defined in the initial information based on the analysis condition information storage step and the analysis condition information. And a step of causing the computer to execute a step of changing and optimizing the element information of necessary elements and a step of outputting the changed initial information as operation result information by the optimization. To do.

  In the present invention, the design area of the structure is divided at a plurality of nodes. Then, adjacent or non-adjacent nodes are connected by frame elements and / or panel elements. Also, the nodes can be connected by a spring element. As a method for optimization based on node position information and element arrangement information composed of nodes and frame elements, panel elements, or spring elements, for example, given boundary conditions are given, and frame elements, panel elements, or spring elements By correcting the rigidity, a shape capable of obtaining the maximum rigidity can be generated. As boundary conditions, external force and fixed conditions are added, and it is conceivable to perform optimization that minimizes displacement under a certain volume constraint for frame elements and panel elements and a cost constraint for spring elements. The optimization is performed by changing the rigidity of the frame element, the panel element, or the spring element, but for the frame element, the panel element, or the spring element that does not substantially contribute to the rigidity maximization, the optimization is performed. The volume condition and the cost condition can be satisfied. Here, “substantially disappearing” refers to the case where the frame element, the panel element, or the spring element is directly deleted from the data, but also exists as data, but the presence of the frame element, the panel element, or the spring element. This includes the case where the existence of the parameter is indirectly eliminated by setting the affecting parameter (for example, cross-sectional area for the frame element, plate thickness for the panel element, spring constant for the spring element, etc.) to zero. When the data itself is erased, the number of data contributing to the calculation varies and the calculation becomes complicated. However, the complexity of the calculation can be suppressed by substantially erasing the data.

  In addition, it is suitable to extract the mechanical factors that can obtain the optimum structure by changing the formulation of the panel element and performing the optimal design using the panel element that has rigidity only for a specific stress. It is. By comparing the optimal solution obtained using this element with the optimal solution obtained using the panel element having rigidity with respect to stress in all directions, the mechanical mechanism of the structure becomes clear. That is, by comparing the optimum structure obtained from both from the mechanical viewpoint, the dominant stress of the structure can be judged, and thereby the mechanical factor for obtaining the optimum structure can be extracted.

  Furthermore, by optimizing the frame element and the panel element as design elements at the same time, it is possible to determine a necessary part of the frame element and a necessary part of the panel element in the structure. As a result, the location of the frame element that supports the axial stress and the location of the panel element that supports the in-plane stress, particularly the shear stress, can be determined. it can.

  Then, by optimizing the frame element or the panel element and the spring element simultaneously as design elements and eliminating the spring element at unnecessary portions of the spring element, the optimum structure and the optimum joint location can be determined.

  The present invention also provides a program that supports optimal design of a structure executed by a computer. When the computer executes this program, parameters that directly affect the presence of a frame element, panel element, or spring element (for example, a cross-sectional area for a frame element or a panel element for a panel element) based on the input boundary conditions. Optimization can be performed by modifying the rigidity of the frame element, the panel element, and the spring element indirectly.

  As described above, according to the present invention, an optimum design can be performed for a structure including both a frame structure and a panel structure. Therefore, the structural design can be automatically performed in a more general structural design such as an automobile body suspension. The efficiency of design evaluation of the actual structure can be improved.

  Hereinafter, embodiments of the present invention will be described.

  FIG. 1 shows a schematic configuration of an optimum design support apparatus according to the embodiment. As described above, the optimum design support apparatus 10 includes the initial information storage unit 12, the analysis condition information storage unit 14, the optimization calculation unit 16, and the output unit 18.

  In the initial information storage unit 12, the node position information indicating the position of the node determined in the desired design area of the structure, the element information that is information regarding the panel element, the frame element, and the spring element to be arranged in the design area, Initial information including element arrangement information, which is information arranged so that elements are connected to each other at node positions, is received and stored. The initial information may be directly input by an input unit such as a keyboard, or data generated by another computer or the like may be received, and the input unit is not limited. An example of typical initial information is stored for each design object, and it is also preferable to create initial information by selecting or slightly modifying it.

  The analysis condition information storage unit 14 receives analysis condition information that is information indicating an objective function to be optimized and a constraint condition at the time of optimization for a design region of a structure formed by connecting the elements. . This analysis condition information is also input in the same manner as the initial information.

  Based on the analysis condition information, the optimization calculation unit 16 deletes unnecessary elements from the panel elements, the frame elements, and the spring elements defined by the initial information, and the element information of necessary elements. To change. That is, the optimum configuration for obtaining the target rigidity is determined by the optimization process.

  The output unit 18 outputs the initial information that has been calculated and modified (optimized) by the optimization calculation unit 16 as calculation result information.

  The optimum design support device can be executed on a computer as a processing program in a computer such as a personal computer or a workstation. That is, in the storage unit of the computer, the initial information storage unit 12 and the analysis condition information storage unit 14 are provided, and a program for executing the optimization calculation is stored. Then, by using the stored information, the CPU executes the program while using the RAM or the like, so that unnecessary elements are substantially erased and optimized initial information (information as a calculation result) is obtained. Can be obtained.

  Here, since the design model of this embodiment uses elements based on structural mechanics such as frame elements, panel elements, and spring elements, a large-capacity memory is unnecessary. For example, the output unit is on a personal computer. It is also possible to use spreadsheet software as a front end. The processing program can be recorded on any medium that can hold information electromagnetically, optically, or chemically, such as a CD-ROM, DVD, hard disk, or semiconductor memory. For example, the processing program can be installed by recording the processing program on a CD-ROM and supplying the processing program from the CD-ROM to the hard disk of the computer. Of course, it is also possible to record a processing program on the hard disk or ROM of a computer from the beginning and use it as a computer dedicated to structural model design.

  Hereinafter, an optimum design method in the optimum design support apparatus will be described.

<Formulation of optimal design method>
First, setting of the objective function will be described. As shown in FIG. 2, a static deformation problem of a structure in which nodes are connected by discrete elastic elements will be considered. The nodes are connected by frame elements, and panel elements are arranged in a region defined by the nodes. Also, the hatched lines in the figure are fixed ends, and the nodes existing here are fixed.

Each node of the structure has three degrees of freedom of x, y, and z in the translation direction and three degrees of freedom of θ _x , θ _y , and θ _{z in} the rotation direction. Then, it is assumed that u is the deformation vector of the structure when the boundary Γ _d of the elastic structure shown in FIG. 2 is completely fixed and the load vector f is applied to the node P _i in the structure. Here, f and u are expressed by the following equations. Here, f and u are vectors, which are shown in bold in the figures and formulas, but are shown in normal letters for convenience in the text of the specification. The same applies to other vectors and matrices.

  Here, for simplification, a case where a load is applied to one node of the structure is considered, but a plurality of loads may be simultaneously applied to a plurality of nodes.

  At this time, an average compliance l (el), which is a product of the load vector f and the deformation vector u shown in the following equation, is a measure of rigidity when the load vector f is applied to a node.

  That is, the rigidity can be maximized by minimizing this l.

<Disappearance method of elements in optimization calculation unit>
The basic idea of the disappearance of elements is to set a fixed design area in topology optimization and introduce a characteristic function shown in the following equation. That is, a fixed design area D that covers the original design area to be the optimum structure is first provided, and the optimization problem is replaced with a material distribution problem or an element arrangement problem using the fixed design area and the characteristic function.

  By using χΩ in the above equation, the optimum structure can be determined by determining whether or not the point at the coordinate x in the fixed design region D is inside. When this characteristic function is applied to a discretized element, there arises a problem that there is a finite number of discontinuities regarding whether or not the element exists. Therefore, here, the design space is replaced by replacing the characteristic function with a continuous function by the following formula based on the density method concept.

Where ρ _v is a normalized design variable,
Takes the value Further, P is a parameter that gives a penalty to ρ _v , whereby the difference in the calculated volume density can be emphasized more than the actual difference. If Expression (5) is used, the optimal design problem can be replaced with a continuous distribution problem of structural elements in the fixed design region D.

<Formulation of design elements>
As described above, in this embodiment, a frame element, a panel element, and a spring element are assumed as design elements. The details will be described below.

  The frame element has 6 degrees of freedom of rigidity with respect to axial force, bending and torsion. In the present embodiment, the simplest case is assumed, the cross section is assumed to be a solid circle, and the normalized cross sectional area of the following equation is used as a design variable.

Here P _A is Pena metallization parameter, A is the cross-sectional area, A _max is the cross-sectional area at the maximum. In addition, each physical quantity required when creating the stiffness matrix of the frame element is obtained as follows. If the diameter of the frame is d,
Since the following relationship holds, the sectional moments I _y and I _z are
The cross-sectional secondary pole moment J _x is
It becomes. An element stiffness matrix is created using the above relationship.

  Next, the panel elements are formulated. In the thin panel constituting the mechanical structure, the balance in the in-plane direction is dominant, and the rigidity required for the panel is mainly in the in-plane direction. On the other hand, since the rigidity in the out-of-plane direction is low and this rigidity is rarely required as the performance of the mechanical structure, the out-of-plane direction rigidity is considered negligible in the evaluation at the time of conceptual design. Therefore, in the present embodiment, only the in-plane direction balance is considered in the panel element formulation.

The design variable in the panel element is the normalized panel thickness ρ _{t given} by

Here, P _t is the penalization parameter, and t _max is the maximum plate thickness.

  Next, the element stiffness matrix is formulated based on the stress assumption. In the stress assumption, relatively accurate results can be obtained when in-plane bending is dominant, and the interpolation function of stress s and the interpolation function of displacement u can be defined independently. It has characteristics that can be assumed in advance. First, if the correspondence between the local coordinate system and the global coordinate system is formulated, the following formula is obtained.

Here, X _C is an element center vector, and e _x , e _y , and _ez are unit vectors in the local coordinate system. Based on the local coordinates of the above equation, the stress distribution s in the element is formulated with the following two assumptions. In the first case, true stress and shear stress are interpolated by the following equations.

  In this case, true stress and shear stress can be supported in the same manner as a general panel. In the second case, only the shear stress is interpolated by the following equation.

  The reason why the two panel elements are formulated here is that the mechanical mechanism of the structure is clarified by comparing the optimal solutions obtained using these two elements. That is, if the optimum structure obtained from both is compared from the mechanical point of view, the dominant stress of the structure can be judged, and the formation or structure of the structure can be understood. Hereinafter, a normal panel indicates a panel formulated by Equation (13), and a shear panel indicates a panel formulated by Equation (14).

  The displacement u is expressed by the following equation using a general bilinear interpolation function N.

Here, d is a displacement vector of four nodes, and if this is used, the minute strain e is expressed as follows using the displacement vector u.

On the other hand, the stress-strain relational expression is obtained as follows.

However, since the stress and displacement are assumed independently in this element, the strain obtained from Equation (16) does not match the strain obtained from Equation (17). here,
In other words, energy balance is ensured by relaxing Equation (17) so that the energy residual is minimized. Here, D is an elastic matrix, and Ω _e is an element region. Therefore, the following equation is obtained from the equations (17) and (18).

Substituting equation (19) into equation (16),
Therefore, the element stiffness matrix is as follows.

Finally, the spring element is formulated. The design variable in the spring element is a normalized spring constant ρ _k . However, the object is not to obtain the optimum welding strength but to search for the optimum welding position based on the obtained result. At this time, the spring constant k _t for translation and the spring constant k _r for rotation are shown by the following equations.

Is the P _k penalization parameter, and k _tmax and k _rmax are the maximum spring constants for translation and rotation, respectively. Incidentally, k _tmax, by way of setting k _rmax, it is possible to express various welding methods.

<Formulation of optimization problem>
When there are m multiloading problems in which loads are separately applied to the design area D composed of n _A frame elements, n _t panel elements, and n _k spring elements, Maximize rigidity. That is, when the load represented by the load vector f ^j (j = 1,..., M) is applied, the rigidity is maximized under the constraint conditions described later. Note that the displacement when the boundary Γd is completely fixed and a load is applied is u ^j (j = 1,..., M), and the average compliance is l ^j (j = 1,..., M). And As a constraint condition for optimization, the frame and panel elements are set to the total volume, and the spring elements are set to the total cost proportional to the design variables. At this time, the optimization problem is formulated as follows.

Constraints
Or

Here [rho _{A, iA} is i _A-th frame element design variables, ρ _{t, it} is design variables i _t th panel element, ρ _{k, ik} is i _k-th spring element design variables, L _iA is i _a-th frame element length of, V _B is the sum of the volume of the frame element, V _B ^U is the upper limit of the volume constraints of the frame element, a _{t, it} is the area of i _t th panel element, V _P is the panel Sum of volume of elements, V _P ^U is the upper limit value of the volume constraint of the panel element, V _U is an upper limit value of the volume constraint that limits the total volume of the frame elements and panel elements, and C _max is the cost per spring element the maximum value of, TC ^U is the sum of the cost of the spring element, TC is the upper limit value of the cost constraints of the spring element, K is a stiffness matrix of the overall design area.

  As a method for solving the multi-objective problem of the above equation, an objective function f of the following equation is used in the present embodiment.

  If the above formula is used, the one with the maximum average compliance is minimized, so that the rigidity is improved when the rigidity is the lowest. Therefore, as a result, an optimum structure having an average rigidity is obtained in all cases.

FIG. 3 shows an optimization flowchart. First, a stiffness matrix K is obtained (S1), and the sensitivity of this stiffness matrix K is calculated (S2). Next, the balance equation is solved to obtain the displacement vector u ^j (S3). Then, the average compliance is obtained using the value of the displacement vector u ^j , the volume constraint of the frame element and the panel element, and the cost constraint of the spring element are calculated (S4), and the convergence determination is performed (S5).

If the objective function has converged, the optimization is terminated. If the objective function has not converged, the sensitivity with respect to the design variables ρ _A , _iA , ρ _t , _it , ρ _k , _ik of the average compliance l (el) and the constraints (S6), and the design variables ρ _A , _iA , ρ _t , _it , ρ _k , and _ik are updated using the calculated values (S7). Thereafter, the process returns to S1, the rigidity matrix K is recreated, and the same calculation is repeated.

CONLIN (Convex Linearization), which is a sequential convex function approximation method, is used as an optimization method for updating design variables. CONLIN has a characteristic that high convergence can be obtained without a heuristic method for a rigidity problem. The basic concept of CONLIN is to set a function h that is a target function or a constraint condition,
Approximation using the following parametric variable. The target function used here has the same dimension as the strain energy and its sensitivity is negative, so the second approximation of the above equation is applied. On the other hand, the strain energy is inversely proportional to the stiffness value of each element in the load regulation problem, that is, the variables ρ _A , _iA , ρ _t , _it , ρ _k , and _{ik in} this formulation. This indicates that the above equation is approximated in a form close to the original target function, and high convergence can be expected.

<Numerical example>
The validity of this design support device is verified by some numerical examples. In all cases, steel was assumed as the material of the structure, Young's modulus was 209 GPa, and Poisson's ratio was 0.3. The parameter P _A penalize for each design element, P _t, P _k was respectively 1 to take the physical integrity of the relationship between design variables and the volume or cost.

  First, a two-dimensional model is used to show that different optimal solutions can be obtained by different panel elements formulated by Equation (13) and Equation (14).

  FIG. 4 shows the design region D. In this example, the design area D is divided into four parts in the x direction and the y direction, and a panel element is arranged in each divided area. A frame element is arranged on the circumference of each arranged panel element. Furthermore, the optimal solution is obtained when the left side of the design region D is completely fixed (the fixed end is indicated by diagonal lines in the figure) and a load of 10.0 N is applied to the node at the right boundary center in the -y direction.

Here, the only design elements panel element, the frame element is a non-design elements, the influence of the frame element is made sufficiently small, the cross-sectional area of all the frame elements was 1.26 × 10- ⁵ mm ^2. In addition, the maximum panel thickness t _max was set to 3.0 mm, and the volume constraint V _P ^U was set to 30% of the entire design area.

FIG. 5 shows the optimum structure obtained. If the case of FIG. 5 (a) using an ordinary panel of formula (13), in case the case shown in FIG. 5 (b) using a shear panel of formula (14), the normalized thickness [rho _t The distribution of is shown. The light and dark bars in the figure were assigned between the maximum value and the minimum value of ρ _t in order to clarify the difference in plate pressure of each optimum structure. A dark-colored panel element requires a large value as a plate thickness, and a white panel element is an unnecessary panel element.

Next, it will be shown that different optimum solutions can be obtained by setting the two volume constraints shown in the two equations (24) and (25) with the frame element and the panel element as design elements at the same time. The two-dimensional model shown in FIG. 4 was used as the optimization model, and a normal panel was used as the panel element. The maximum cross-sectional area of the frame was 1.13 × 10 ² mm ² , the maximum panel thickness was 3.0 mm, and the volume constraint was optimized as 30% of the total design area in any case.

  FIG. 6 shows the optimum structure obtained. In the figure, (a) shows the result when separate volume constraints are set for the panel element and the frame element in Expression (24), and (b) in the figure shows a common volume restriction for both elements in Expression (25). This is the result. Thus, it can be seen that the optimum structure is greatly different by making the volume constraints different.

Next, the result when a three-dimensional model is used is shown. As shown in FIG. 7, five two-dimensional design areas are prepared, and each plane of the rectangular parallelepiped is set as a design area. The setting of the design area on each surface is the same as in the case of FIG. An optimal solution is obtained when the left side of the design region D is completely fixed and a torque of 10.0 N · m is applied around the x axis at the center of the right boundary surface. The maximum cross-sectional area of the frame was 2.83 × 10 mm ² , the maximum thickness of the panel was 3.0 mm, and the volume constraint was 50% of the total design area based on Equation (25).

  8 and 9 show the optimum structure. FIG. 8 shows the result when a normal panel is used, and FIG. 9 shows the result when a shear panel is used. Thus, it can be seen that even in the three-dimensional model, an optimum structure is obtained, and the optimum structure varies depending on the type of panel element.

Next, an optimal solution when a spring element is used as a design element in addition to a frame element and a panel element is shown. FIG. 10 shows the design area. As shown in the figure, five spring elements are arranged from the top in the center of FIG. 9 used in the two-dimensional model. The constraint conditions for the spring elements were set as C _max = 1.0 and TC ^U = 0.40. C _max is the maximum value of the cost per one spring element, TC ^U is the upper limit value of the cost constraints of the spring element. That is, in this example, it is an object to search for an optimum welding position on the assumption that welding of two places out of five places is permitted. The maximum value of the spring constant is set as k _tmax = 10.0 N / m and k _rmax = 10.0 N · m. At the same time, the frame element and panel element were also used as design elements.

  The results are shown in Table 1 and FIG. That is, Table 1 shows the optimum value of the spring constant when the frame element, the panel element, and the spring element are simultaneously used as the design elements in the design region of FIG. FIG. 11 shows the necessary degree of the frame element and the panel element at that time by color coding. A dark color is a necessary element.

In Table 1, the value of ρ _k corresponds to 1 to 5 in FIG. 10 from the top. The values at the top and bottom are large, indicating that it is appropriate to weld the top and bottom.

  Subsequently, an optimum reinforcing shape of the T-shaped joint portion of the automobile body is derived using the panel element. The design area is shown in FIG. In this example, the frame and the external panel are non-design elements, and the internal reinforcing panel is a design element. As for the load condition, a torsion of 1.0 N · m and a vertical load of 10 N are applied at the position shown in FIG.

  The results are shown in FIG. In the figure, (a) shows the result using a normal panel, and (b) shows the result using a shear panel. It can be seen from the results that a reinforcing structure as shown in FIG. 14 is suitable for such a T-shaped joint.

  Thus, according to the present embodiment, it is possible to perform an optimization design for a structure using a panel element such as an automobile body part. Therefore, this result can be used for design, trial manufacture, manufacture, etc. of various members.

It is a block diagram of the optimal design support apparatus. It is a figure of the structure which fixed the edge and applied the load. It is a processing flowchart of an embodiment. It is a figure which shows a two-dimensional design area | region. FIG. 5 is a diagram showing an optimal solution using a panel element as a design element in the design region of FIG. FIG. 5 is a diagram illustrating an optimal solution in which a panel element and a frame element are simultaneously used as design elements in the design region of FIG. 4. It is a figure which shows a three-dimensional design area | region. FIG. 8 is a diagram illustrating an optimal solution in which the panel element and the frame element are simultaneously used as design elements in the design region of FIG. 7. FIG. 8 is a diagram illustrating an optimal solution in which the panel element (shear panel) and the frame element are simultaneously used as design elements in the design region of FIG. 7. It is a figure which shows the design area | region in the case of including a spring element as a design element. FIG. 11 is a diagram illustrating an optimal solution when a frame element, a panel element, and a spring element are simultaneously used as design elements in the design region of FIG. 10. This is a design area simulating a T-shaped joint of an automobile body. It is the optimal solution in the design area | region of FIG. It is the reinforcement shape of the T-shaped joint part devised from the optimal solution of FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Optimal design support apparatus, 12 Initial information storage part, 14 Analysis condition storage part, 16 Optimization calculation part, 18 Output part.

Claims (12)

This is information on node position information indicating the position of a node that divides a design area of a structure, frame elements that are arranged by connecting the nodes, and panel elements that are arranged in a divided region defined by the nodes. An initial information storage unit that stores initial information including element information and element arrangement information that is information arranged so that the elements are connected to each other at node positions;
Analysis for storing analysis condition information, which is information indicating an objective function to be optimized for a design area of a structure formed by connecting the elements, and constraint conditions for optimization of the objective function A condition information storage unit;
Based on the analysis condition information, an optimization operation that deletes unnecessary elements from the frame elements and panel elements defined in the initial information, and changes and optimizes the element information for necessary elements. And
An output unit that outputs the initial information that is calculated and changed by the optimization calculation unit as calculation result information;
An optimum design support device for a structure, comprising: The apparatus of claim 1.
The structure optimum design support apparatus, wherein the analysis condition information includes boundary conditions including an external force applied to the structure and a fixed condition, and volume conditions of the frame element and the panel element. The apparatus according to claim 1 or 2,
The initial design information storage unit further includes a spring element for connecting the nodes. The apparatus of claim 3.
The structure optimum design support apparatus, wherein the analysis condition information includes a cost condition of the spring element. The apparatus according to claim 1.
A structure characterized in that the optimization calculation unit determines a necessary position of the frame element and a necessary position of the panel element in the structure by causing substantial disappearance of the unnecessary frame element or panel element. Optimal design support device for things. The apparatus of claim 3.
An optimization design support device for a structure, wherein the optimization calculation unit determines a portion to which the structure is to be joined by causing an unnecessary disappearance of an unnecessary spring element. The apparatus of claim 1.
The optimization calculation unit performs optimal comparison by quantitatively comparing the optimal structure obtained using at least one of panel elements that have rigidity only for stress in a specific direction or panel elements that have rigidity for stress in all directions. An optimum design support device for a structure, characterized in that a mechanical factor for obtaining the structure is extracted. The apparatus of claim 7.
The panel element having rigidity with respect to the stress in all directions is modeled based on the stress assumption, and the rigidity of the panel element is limited to the in-plane direction. Optimal design support device. The apparatus of claim 7.
An optimum design support device for a structure, wherein the panel element having rigidity only in the specific direction has rigidity only in terms of shear stress. The device according to any one of claims 1 to 8,
The structure design optimum design support apparatus, wherein the analysis condition information of the frame element and the panel element includes volume constraints set separately. The device according to any one of claims 1 to 8,
The structure optimum design support apparatus, wherein the analysis condition information of the frame element and the panel element includes a volume constraint set for the sum of the frame element and the panel element. This is information on node position information indicating the position of a node that divides a design area of a structure, frame elements that are arranged by connecting the nodes, and panel elements that are arranged in a divided region defined by the nodes. Storing initial information including element information and element arrangement information, which is information arranged such that the elements are connected to each other at node positions, in an initial information storage unit;
Analysis condition information is analysis condition information, which is information indicating an objective function to be optimized for a design region of a structure formed by connecting the elements, and constraint conditions for optimization of the objective function. Storing in the storage unit;
Based on the analysis condition information, the unnecessary elements are deleted from the frame elements and panel elements defined in the initial information, and the necessary elements are changed and optimized for the elements, and
By this optimization, a step of outputting the changed initial information as operation result information;
An optimal design support program for a structure, which causes a computer to execute