JP4803809B2 - Structural optimization method - Google Patents

Description

  The present invention relates to a structure optimization method, and more particularly to a structure optimization method capable of greatly reducing the number of man-hours required for analysis processing when applying a phase optimization method for weight reduction to a target structure. .

  In recent years, various optimization methods for giving an optimum shape to a target structure have been used in the field of component design using a computer.

Patent Document 1 discloses a phase optimization method for obtaining an optimal material arrangement within a predetermined design region under preset load conditions and constraint conditions. According to this phase optimization method, it is possible to know where the material can be concentrated in the target structure to obtain the maximum rigidity. It is possible to proceed with the designed work. As a result, it is possible to expect a reduction in the number of repetitive examination steps until the target is achieved, as compared with the conventional design work in which iterative tests and CAE (Computer Aided Engineering) analysis are performed based on past experience.
JP 2004-348691 A

  However, applying the phase optimization method disclosed in Patent Document 1 for the purpose of reducing the weight of the target structure has the following problems.

  First, when the phase optimization method is applied to the initial shape of the target structure, the analysis result shows the optimal material placement in the design region in the initial shape, so the shape based on this analysis result is shown. The change cannot provide more rigidity than can be obtained from the initial shape. Therefore, a method is generally used in which the phase optimization method is applied after virtually expanding the design area, thereby maintaining the rigidity of the initial shape and considering the weight reduction. In the case of a part having a complicated shape, the number of man-hours for modeling work for expanding the design area becomes very large, and there is a problem that the time required for analysis increases. In addition, when the target structure is a two-wheeled vehicle frame, the engine and auxiliary equipment are arranged close to the frame members, so it is difficult to greatly expand the design area itself, and the phase optimization method is applied. There is also a problem that the effect of achieving this cannot be obtained sufficiently.

  The object of the present invention is to solve the above-mentioned problems of the prior art, and to optimize the structure that can greatly reduce the man-hours required for the analysis process when applying the phase optimization method for weight reduction to the target structure. It is to provide a conversion method.

  In order to achieve the above-mentioned object, the present invention provides a structure optimization method for obtaining an optimal structure form that satisfies an objective function under a predetermined constraint condition using a phase optimization method, The first feature is that, when the performance of the structure is lower than the constraint condition, the phase optimization method is applied after the performance of the structure is set virtually high.

  The performance of the structure is the magnitude of rigidity with respect to external force or the like. When the magnitude of the rigidity is smaller than the constraint condition, the structure is assumed to have a large longitudinal elastic modulus. There is a second feature in that the rigidity of is set virtually large.

  According to the first invention, when the performance of the structure is lower than the constraint condition, the phase optimization method is applied after setting the performance of the structure virtually high with a margin for the performance of the structure. Thus, the phase optimization method can be applied without expanding the design area of the target structure, and the optimum form of the structure can be obtained. Thereby, the modeling work for expanding the design area of the target structure can be reduced, the number of analysis steps can be greatly reduced, and the time required to obtain the final shape can be shortened.

  According to the second invention, the performance of the structure is the magnitude of rigidity with respect to external force or the like, and when the magnitude of rigidity is smaller than the constraint condition, the longitudinal elastic modulus of the structure is assumed to be large, thereby Since the rigidity of the structure is set virtually large, a simple method of setting the virtual longitudinal elastic modulus can give a margin to the rigidity value of the structure and expand the design area of the target structure The optimum form of the structure can be obtained without any problem. Thereby, the modeling work for expanding the design area of the target structure can be reduced, the number of analysis steps can be greatly reduced, and the time required to obtain the final shape can be shortened.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing an example of a hardware configuration for executing the structure optimization method according to the present invention. Connected to the computer 1 are input means 2 for inputting design information of a target structure by CAD (Computer Aided Design) or the like and output means 3 such as a display for outputting and displaying the result of arithmetic processing by the structure optimization method. ing. The computer 1 includes a storage unit 101 that stores design information input from the input unit 2 and a calculation processing result executed inside the computer 1, and a finite element method based on the design information stored in the storage unit 101. a structure analysis processing unit 102 that performs a structural analysis of the target structure by the virtual rigid setting unit 103 for setting a virtual modulus for setting the virtual stiffness value in the structural analysis model created by the structure analysis unit 102 And a phase optimization processing unit 104 that applies the virtual longitudinal elastic modulus to execute a phase optimization method on the structural analysis model. Details of the structure optimization method according to the present embodiment will be described below.

  2A and 2B are conceptual diagrams of a structural analysis model used when the phase optimization processing unit executes the phase optimization method. FIG. 2A shows a conventional method, and FIG. 2B shows a structure optimization method according to the present invention. Indicates what applies. As described above, the phase optimization method and the phase optimization process refer to a method that can obtain an optimum material arrangement within a predetermined design region and a calculation process thereof under a preset constraint condition. The phase optimization method includes a homogenization method and a density method. In the present invention, the density method is applied.

In the structural analysis, it is assumed that Hooke's law “ε = σ / E” is established, with ε: strain, σ: stress, and E: longitudinal elastic modulus. The density method is a method for obtaining an optimal phase using the density of each mesh as a design variable on the assumption that the longitudinal elastic modulus E of the material is proportional to the power of the mesh density set by the finite element method. . That is, it is assumed that “E = ρ ^N E ₀ ” is established for the longitudinal elastic modulus E as E: longitudinal elastic modulus used in calculation, ρ: density, N: multiplier, E ₀ : set longitudinal elastic modulus. Is. Thus, longitudinal elasticity coefficient E used in the calculation, the density of each mesh fluctuates 0.0～E ₀ by varying the 0.0 to 1.0, the derivation of the optimal solution of the phase optimization process, The density of each mesh is used as a design variable, and this is performed by iterative calculation for maximizing or minimizing the objective function while satisfying the constraint conditions.

  However, in the density method as described above, assuming that the mesh density in the initial shape of the target structure is 1.0, the magnitude of rigidity with respect to external force and the like after the phase optimization process, that is, the rigidity value is greater than the initial shape. I can't do it. Therefore, in the conventional method, as shown in FIG. 2A, a method of executing the phase optimization process after expanding the design area of the target structure has been used. Since the target structure 11 in FIG. 2A has a margin in rigidity by expanding the design area with respect to the target structure 10 having the initial shape, the phase optimization method can be applied. Become.

  In contrast to the conventional method, in the conceptual diagram of the structural analysis model according to the present invention shown in FIG. 2B, in order to apply the phase optimization method without expanding the design area, the target structure 10 having an initial shape is used. It is shown that the object structure 12 having the same shape as the initial shape and having twice the rigidity value can be obtained by virtually setting the longitudinal elastic modulus E to 2 times. Thereby, for example, when the longitudinal elastic modulus is set to double, the strain with respect to the stress due to the external force or the like is halved, and a margin is generated in the rigidity value, so that the phase optimization process can be executed. Note that the magnitude of the virtual longitudinal elastic modulus can be arbitrarily set according to the purpose.

  FIG. 3 is a flowchart showing the procedure of the structure optimization method according to the present invention. A series of processing shown in this flowchart is realized by using the computer 1. In step S1, the target stiffness value of the target structure is set according to the designer's intention. This target stiffness value is input by the input unit 2 and stored in the storage unit 101 of the computer 1. In subsequent step S2, the structural analysis processing unit 102 executes structural analysis of the target structure by the finite element method. A structural analysis model composed of a plurality of meshes is created by this structural analysis processing.

  In step S3, it is determined whether or not the current stiffness value of the target structure subjected to the structural analysis is smaller than the target stiffness value set in step S1. If an affirmative determination is made in step S3, it is determined that there is no margin of rigidity necessary for the phase optimization process, and the process proceeds to step S4. In step S4, the ratio between the target stiffness value and the current stiffness value is calculated. Next, in step S5, a virtual longitudinal elastic coefficient for virtually increasing the rigidity of the target structure is set based on the calculated ratio. This virtual longitudinal elastic modulus can be set to, for example, 2 times based on the calculated ratio.

  In step S6, the virtual longitudinal elastic modulus set in step S5 is applied to the structural analysis model of the target structure. By the processing in step S6, the target structure is provided with a marginal rigidity that allows the shape adjustment, and therefore, it is possible to proceed to step S7 and execute the phase optimization processing. In step S7, the phase optimization processing unit 104 executes phase optimization processing, and a series of processing ends. If a negative determination is made in step S3, it is assumed that there is sufficient rigidity necessary for phase optimization, and the process proceeds directly to step S7 to execute phase optimization processing.

  According to the structure optimization method as described above, the phase optimization method can be executed without expanding the design area of the target structure, so the modeling work for expanding the design area can be reduced and the analysis man-hours can be greatly reduced. The time until the final shape is obtained can be reduced.

  FIG. 4 shows an example of an output result by the phase optimization process according to the present invention. In the present embodiment in which the phase optimization process is performed on the vehicle body frame 20 as the target structure, the vehicle body frame 20 expressed entirely as a mesh is displayed on the output means 2 including a color display or the like. Each mesh is colored, and the density of the mesh is expressed by this color arrangement state. On the actual display, from the highest density part to the lowest density part can be displayed in 8 levels such as red → purple → yellow → green → light blue → blue → navy → white. The portion with the highest density (for example, A portion) is shown in black, the portion with the lowest density (for example, B portion) is shown in white, and the mesh display is also omitted in the intermediate portion for explanation. The virtual longitudinal elastic modulus is virtually set in order to execute the phase optimization process and obtain the above output result, and therefore has no meaning as an absolute value. The target structure can be not only a body frame of a motorcycle but also various parts.

  A vehicle body frame 20 as a target structure according to the present embodiment includes a left and right main pipe 21, a seat rail base 24 extending rearward and upward of the main pipe 21, and a connecting pipe that connects the left and right main pipes 21 at a substantially central portion of the vehicle body. 25, a head pipe 22 for rotatably supporting a stem shaft (not shown), and an engine hanger 23 integrally coupled below the main pipe 21. According to the output result of the phase optimization process for the vehicle body frame 20, the color is dark and the density is high, for example, the A portion or the C portion is “reinforcement required portion”, and the color is thin and the density is low. It becomes possible to determine the portion, the D portion, etc. as “weakening required portions”. As described above, in the phase optimization processing according to the present embodiment, the shade of the color of the mesh can be regarded as the level of sensitivity with respect to the target rigidity, so such an analysis method is called “phase sensitivity analysis”. To do. And, by this phase sensitivity analysis, it is possible to clarify the shape change policy of strengthening the high-sensitivity portion and weakening the low-sensitivity portion, whereby the weight reduction of the body frame 20 is relatively easy. Can be achieved.

  In the example of the vehicle body frame 20, a portion A between the main pipe 21 and the engine hanger 23: the thickness is increased and a rib is added, and a portion B below the head pipe 22 and the side surface in the vehicle width direction is thinned. C part of the connecting pipe 25: It is possible to easily consider countermeasures such as increasing the pipe diameter. As described above, according to the phase sensitivity analysis, the aim of the strengthening required portion and the weakening required portion can be visually grasped, so it is easy to obtain an idea of the shape change, and the weight reduction can be achieved from the examination stage of the basic skeleton of the component. In addition, it is possible to proceed with design work that also takes into account the above, and a great effect is exhibited in reducing the total number of analysis steps. Also, in the past, shape change locations and methods were studied based on the stress diagram of the structural analysis result and the deformation diagram for the stress, etc., so advanced technology and know-how were required for the judgment, and the number of work steps increased. However, in the phase sensitivity analysis as described above, the changed part can be directly determined from the density distribution displayed on the display or the like, so that the determination error can be reduced and the analysis time can be shortened.

  Note that the phase sensitivity analysis as described above is applied to the analysis aiming at improving the rigidity while minimizing the increase in weight in addition to the purpose of reducing the weight while maintaining the rigidity of the initial shape. Is possible. In addition, the calculation conditions for the phase optimization process are the design variables: the whole model, the objective function: weight minimization, and the constraints are set to “stiffness equivalent”, “improvement of torsional rigidity and lateral rigidity is equivalent”, etc. Is possible.

  FIG. 5 is a conceptual diagram showing the process and results of the phase optimization method according to the present invention. When the virtual stiffness value G2 is set for the current stiffness value G1 of the target structure, a margin width of the stiffness value between G1 and G2 is created, and a shape change is performed for the purpose of reducing the weight within this margin width. It becomes possible to do. This conceptual diagram shows that if the rigidity of the target structure is unchanged, the weight of the target structure can be reduced from W1 to W2 by changing the shape based on the phase optimization processing result. . In the actual design work, the structural analysis for the target structure after the shape change and the examination of the further shape change are repeated until the final shape is determined.

  Hereinafter, with reference to FIGS. 6 to 8, verification results of the effectiveness of the structure optimization method according to the present invention will be described. FIG. 6 is a perspective view of a test piece used for verification of the structure optimization method according to the present invention. A thin plate-like test piece 30 having a notch 31 at the lower center is given a restraint condition by a restraint part 32 at one end and a load condition by a load L given to the other end. In addition, each calculation condition in this verification is a design variable: the whole model, a constraint condition: rigidity equivalent, and an objective function: weight minimization.

  FIGS. 7A and 7B show an output result when the Mises stress analysis is performed on the test piece 30 and an output when the phase optimization processing according to the present invention is performed. A result (b) is shown. In both figures, the difference in sensitivity is expressed by the shade of color, and in the Mises stress analysis result of FIG. 7A, the black H portion indicating that the stress generated in the test piece 30 is the largest, The notch 31 is concentrated in the upper right corner of the figure. On the other hand, in the phase optimization processing result of FIG. 7B, the sensitivity is shown by the phase sensitivity analysis, and the black I portion having the highest sensitivity is constrained by the restraining portion 32. As a result, it appeared in the shape of a substantially right triangle extending from the part to the other end to which the load condition was applied.

  8A and 8B show a state in which a patch as a reinforcing member is applied to the test piece based on the analysis result shown in FIG. For comparison, the weight of the patch was the same. In FIG. 8A, a substantially rectangular patch 40 centered on the H portion is applied so as to reinforce the portion having a large stress based on the Mises stress analysis result. On the other hand, in FIG. 8B, based on the phase sensitivity analysis result, a substantially triangular patch 50 is applied along the I portion indicated as a portion requiring reinforcement. The degree of improvement in the rigidity value verified by applying the load L again to both of them is about 50% for the patch 40 based on the Mises stress analysis result and about 70% for the patch 50 based on the phase optimization processing result. It was verified that a difference of about 20% occurs between the two. Therefore, for example, when the rigidity of both is made equal, a weight reduction corresponding to a rigidity value of about 20% can be achieved theoretically, so that various conditions such as the presence of auxiliary equipment mounting stays are taken into consideration. It is considered that the weight can be reduced by 10 to 20%.

  As described above, according to the structure optimization method of the present invention, when the rigidity value of the target structure is smaller than the target rigidity value, instead of expanding the design area of the target structure, Since the phase optimization process is performed after the elastic coefficient is virtually set large, the modeling work for expanding the design area can be reduced. As a result, the number of analysis steps can be greatly reduced, and the examination time until the final shape is obtained can be shortened. In addition, because phase sensitivity analysis displays the necessary and weakened parts of the target structure directly on the output means, advanced analysis technology and know-how are no longer required, and light weight is ensured while maintaining equivalent rigidity. Can be efficiently performed.

  In addition, the form of information input from the input means and the input means, the form of information output from the output means and the output means, the configuration of each processing unit in the computer, the program used for the structural analysis by the finite element method, the virtual stiffness Of course, the setting value of the value is not limited to the above-described embodiment, and various modifications are possible.

It is the block diagram which showed an example of the hardware constitutions which perform the structure optimization method concerning this invention. It is a conceptual diagram of the structural analysis model used when performing a phase optimization method. It is a flowchart which shows the procedure of the structure optimization method which concerns on this invention. It is an example of the phase optimization process result which concerns on one Embodiment of this invention. It is the conceptual diagram which showed the process and result of the phase optimization method. It is a perspective view of the test piece used for verification of the structure optimization method which concerns on this invention. It is the output result (a) at the time of execution of Mises stress analysis with respect to a test piece, and the output result (b) at the time of execution of phase optimization processing. It is a front view of the test piece which applied the patch.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Computer, 2 ... Input means, 3 ... Output means, 101 ... Memory | storage means, 102 ... Structural analysis process part, 103 ... Virtual rigidity setting part, 104 ... Phase optimization process part, 10 ... Target structure, 20 ... Vehicle body Frame, 30 ... Specimen

Claims (1)

A structure optimization method for obtaining an optimal structure form satisfying an objective function under a predetermined constraint condition using a phase optimization method,
A procedure for inputting design information of the structure (12) to the storage means (101) included in the computer (1) by the input means (2);
A procedure for analyzing the rigidity value of the structure (12) by performing a structural analysis by a finite element method using the design information by the structural analysis processing unit (102) included in the computer (1);
When the rigidity value of the structure (12) is lower than the predetermined constraint, the structure (103) used for calculation of the rigidity value by the virtual rigidity setting unit (103) included in the computer (1). A procedure for setting a virtual longitudinal elastic modulus in which the longitudinal elastic modulus (E) of 12) is virtually increased;
A step of applying the phase optimization method to a structural analysis model to which the virtual longitudinal elastic modulus is applied by a phase optimization processing unit (104) included in the computer (1);
A structure optimization method, comprising: a step of reading out an arithmetic processing result by the structure optimization method from the storage means (101) and outputting and displaying the result on an output means (3) connected to the computer (1). .

Images