Shape data generation method and shape data generation program

Abstract

To obtain shape data for a component that maintains a specified external shape while keeping generated stress values within an acceptable range.SOLUTION: A shape data generation method includes: step S21 of obtaining a stress function D21 with discrete stress data D11 generated in a target-shaped object as a dependent variable; step S22 of obtaining internally generated point cloud data D22, which indicates a stress distribution generated within the target-shaped object by the density of multiple internally generated fixed points D22a, using the stress function D21; step S23 of obtaining surface-generated point cloud data D23, which includes multiple surface-generated fixed points D23a for defining the external shape of the target-shaped object, without relying on the stress distribution generated within the target-shaped object; and step S30 of obtaining truss information D30, which includes a fixed truss D32a, using the internally generated point cloud data D22 and the surface-generated point cloud data D23.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a shape data generation method and a shape data generation program. [Background technology]

[0002] The use of three-dimensional modeling devices, commonly known as 3D printers, is increasing. To create a model using a three-dimensional modeling device, data in an STL format that defines the object is required. This data is prepared by designers using three-dimensional CAD. Designers sometimes determine the final shape based on their own knowledge and experience. They also sometimes determine the shape based on the results of simulations using analysis software used in conjunction with three-dimensional CAD.

[0003] One method for determining the shape of a part through simulation is the topology optimization method. Patent Documents 1 and 2 disclose techniques related to the topology optimization method. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2020-66104 [Patent Document 2] Patent Publication No. 2021-182201 Summary of the Invention [Problem to be solved by the invention]

[0005] A part is designed to satisfy a design goal under several design conditions. One example of a design condition is the external shape of the part. One example of a design goal is the stress value generated in the part. In other words, the external shape of the part must be a predetermined shape while the stress value generated in the part must be within an allowable range. However, as the name suggests, the topology optimization methods disclosed in Patent Documents 1 and 2 optimize the external shape of the part to satisfy the design goal. In other words, the techniques disclosed in Patent Documents 1 and 2 change the external shape of the part to satisfy the design goal.

[0006] An object of the present invention is to provide a shape data generation method and a shape data generation program for obtaining a part that has a predetermined external shape while keeping the generated stress value within an allowable range. [Means for solving the problem]

[0007] A shape data generation method that is one form of the present invention includes the steps of: obtaining a stress function in which coordinate values ​​defined for a target-shape object are independent variables and stress values ​​generated in a target-shape object for which predetermined constraint conditions and load conditions are set are dependent variables; using the stress function to obtain internally generated point cloud data that indicates the stress distribution generated inside the target-shape object by the density of multiple internally generated points; obtaining surface generated point cloud data that includes multiple surface generated points for defining the outer shape of the target-shape object without being based on the stress distribution generated inside the target-shape object; and using the internally generated point cloud data and the surface generated point cloud data to obtain truss information that includes truss elements that connect the internally generated points and / or surface generated points to each other.

[0008] This shape data generation method obtains internally generated point cloud data that indicates the stress distribution occurring inside a target-shape object by the density of multiple internally generated points. A truss structure obtained using this internally generated point cloud data can increase the strength of the target-shape object simulated by the shape data. Furthermore, this shape data generation method obtains surface generation point cloud data that includes multiple surface generation points for defining the outer shape of the target-shape object, without being based on the stress distribution occurring inside the target-shape object. The truss structure obtained using this surface generation point cloud data does not change the appearance of the target-shape object simulated by the shape data, because it is not based on the stress distribution occurring inside the target-shape object. Therefore, it is possible to obtain shape data for a part that can keep the generated stress value within an allowable range while maintaining a predetermined outer shape.

[0009] In the above-described shape data generation method, the step of obtaining internally generated point cloud data may include the steps of: using a stress function to obtain multiple candidate-point stress values ​​occurring at multiple internally generated candidate points included in the target-shape object; obtaining an inter-point evaluation distance for treating a first candidate-point stress value occurring at a first internally generated candidate point selected from the multiple internally generated candidate points as virtual distance information from the first internally generated candidate point; and selecting the first internally generated candidate point as an internally generated determined point, which is an internally generated point, when a second internally generated candidate point different from the first internally generated candidate point does not exist within the range indicated by the inter-point evaluation distance, which is virtual distance information set for the first internally generated candidate point. This step allows for obtaining internally generated point cloud data corresponding to the stress distribution.

[0010] In the above-described shape data generation method, the shape of the target-shape object is defined by a plurality of meshes each defined by a plurality of nodal elements and connected elements connecting the plurality of nodal elements, and the step of obtaining surface generation point cloud data may include the steps of: setting the nodal elements defining both ends of the connected elements and additional points overlapping the connected elements as a plurality of surface generation candidate points; and determining the second surface generation candidate point as the surface generation point when the distance from the first surface generation candidate point to the second surface generation candidate point is greater than the minimum spatial distance between the first and second surface generation candidate points. This step allows for the obtaining of surface generation point cloud data with a low computational load.

[0011] In the above-described shape data generation method, the step of obtaining truss information may include the steps of obtaining candidate truss data using the internally generated point cloud data and the surface generated point cloud data, and obtaining definite truss data including definite truss elements that constitute the target-shape object by removing, from the candidate truss elements included in the candidate truss data, distorted truss elements that are not included in the target-shape object. This step can prevent changes in the appearance of the target-shape object simulated by the shape data.

[0012] Another form of the present invention is a shape data generation program that causes a computer to function as a first calculation unit that obtains a stress function in which coordinate values ​​defined in a target-shape object are independent variables and stress values ​​generated in a target-shape object for which predetermined constraint conditions and load conditions are set are dependent variables; a second calculation unit that uses the stress function to obtain internally generated point cloud data that indicates the stress distribution generated inside the target-shape object by the density of multiple internally generated points; a third calculation unit that obtains surface generated point cloud data including multiple surface generated points for defining the outer shape of the target-shape object without being based on the stress distribution generated inside the target-shape object; and a fourth calculation unit that uses the internally generated point cloud data and the surface generated point cloud data to obtain truss information including truss elements that connect the internally generated points and / or surface generated points to each other.

[0013] This shape data program makes it possible to obtain shape data for a part that can keep the generated stress value within an allowable range while maintaining the external shape of the part as specified. [Effects of the Invention]

[0014] According to the present invention, a shape data generation method and a shape data generation program are provided for obtaining a part that has a predetermined external shape while keeping the generated stress value within an allowable range. [Brief explanation of the drawings]

[0015] [Figure 1] FIG. 1 is a flowchart of the shape data generation method according to this embodiment. [Figure 2] FIG. 2 is a graph for explaining the RBF interpolation method. [Figure 3] FIG. 3 is a detailed flowchart of the steps for obtaining internally generated point cloud data. [Figure 4] FIG. 4 is a diagram for explaining the steps of obtaining internally generated point cloud data. [Figure 5] FIG. 5 is a detailed flowchart of the steps for obtaining surface generated point cloud data. [Figure 6] 6(a), 6(b), and 6(c) are diagrams for explaining the steps of generating surface generation candidate points. [Figure 7] FIG. 7 is a diagram illustrating the steps of generating surface generation candidate points, following FIG. [Figure 8] FIG. 8 is a diagram showing the results of executing the step of obtaining point cloud information. [Figure 9] Fig. 9(a) is a diagram showing the original shape of the target shape model, and Fig. 9(b) is a diagram for explaining a distortion candidate truss. [Figure 10] FIG. 10 is a detailed flowchart of the steps for obtaining a deterministic truss. [Figure 11]FIG. 11 is a diagram for explaining the inside / outside determination regarding the step of obtaining a definite truss. [Figure 12] FIG. 12 is a diagram illustrating a computer that executes a shape data generating program. [Figure 13] FIG. 13 is a diagram illustrating functions realized on a processor by executing a shape data generation program on a computer. [Figure 14] Figures 14(a) and 14(b) show the truss information model before the shape data generation method is applied. Figure 14(c) shows the analysis results of the stress distribution generated in the truss information model shown in Figure 14(a). [Figure 15] Figures 15(a) and 15(b) show the truss information model after applying the shape data generation method. Figure 15(c) shows the analysis result of the stress distribution generated in the truss information model shown in Figure 15(a). [Figure 16] 16(a) and 16(b) are diagrams for examining the results of the calculation example. DETAILED DESCRIPTION OF THE INVENTION

[0016] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are designated by the same reference numerals, and duplicated explanations will be omitted.

[0017] Fig. 1 is a flowchart showing the main steps of the shape data generation method, and Fig. 13 shows some internal parameters used when executing the shape data generation method.

[0018] Discrete stress data D11 is obtained (S10). The discrete stress data D11 includes coordinate values ​​(x, y, z) and stress values ​​(σ) associated with the coordinate values. First, in step S10, a stress analysis process is executed (S11). In step S11, the discrete stress data D11 is obtained using model shape data D101, model material data D102, constraint condition data D103, and load condition data D104. The process in step S11 is performed by matrix structural analysis of a plane truss. It may also be performed by finite element analysis. For example, if the target-shaped object has a two-dimensional shape, the model shape data D101 represents a two-dimensional truss structure. Then, using the above data, a two-dimensional truss analysis is performed, including the following steps (1) to (4). Step (1) Derive element stiffness equations for the members that make up the truss. Step (2) Derive the overall stiffness equation from the derived element stiffness equations. Step (3) After inputting the boundary conditions (constraint conditions, load conditions) into the global stiffness equation, solve the stiffness equation with the nodal displacements as unknowns. Step (4) Calculate the stress within the element using the nodal displacement.

[0019] In this embodiment, step S10 of obtaining the discrete stress data D11 is described as being executed as part of the shape data generation method. Furthermore, the discrete stress data D11 is the result of finite element analysis, as described above. Therefore, the shape data generation method may execute a step of reading the discrete stress data D11 instead of step S10 of obtaining the discrete stress data D11. In this case, the discrete stress data D11 can be prepared by using software for finite element analysis, for example.

[0020] Point cloud information D20 is obtained (S20). In step S20, a point cloud is generated inside and on the surface of the target-shape object. The method for generating a point cloud inside the target-shape object is different from the method for generating a point cloud on the surface of the target-shape object. The method for generating a point cloud inside the target-shape object uses the discrete stress data D11 obtained in step S11. On the other hand, the method for generating a point cloud on the surface of the target-shape object does not use the discrete stress data D11 obtained in step S11. The method for generating a point cloud on the surface of the target-shape object uses model shape data D101 that defines the target-shape object.

[0021] A stress function D21 is obtained (S21). One example of a process for obtaining the stress function D21 is the Radial Basis Function interpolation method (hereinafter referred to as "RBF interpolation"). RBF interpolation obtains a line segment that passes through all of a plurality of points set in a coordinate system. For example, as shown in FIG. 2, RBF interpolation makes it possible to obtain, for given points P2a to P2i, a curve G2 (function s: see formula (1)) that passes through all of these points P2a to P2i.

number

[0022] In step S21 to obtain the stress function D21, RBF interpolation is performed on a group of points defined in a three-dimensional space. The group of points here refers to a plurality of nodes defined in the stress analysis process (step S11). Each of the plurality of nodes is associated with a coordinate value (x, y, z) and a stress value (σ) corresponding to the coordinate value. The coordinate values ​​(x, y, z) and the stress value (σ) are the discrete stress data D11 described above.

[0023] Internally generated point cloud data D22 is obtained (S22). Figure 3 is a detailed flowchart of step S22 for obtaining the internally generated point cloud data D22. Step S22 for obtaining the internally generated point cloud data D22 generates a point cloud inside the target-shaped object. That is, step S22 obtains the internally generated point cloud data D22 using the discrete stress data D11 obtained in step S11.

[0024] One example of a method for generating an internally generated point cloud is the Poisson Disk Sampling method (hereinafter referred to as the "PDS method"). The PDS method generates points in space and generates a point cloud in which the points are spaced apart by a certain distance or more. The PDS method includes the following steps (1) to (4). Step (1) Randomly generate one point and set that point as the fixed point. Step (2) Generate a new point and set it as a candidate point. Step (3) A circle with the minimum inter-point distance (minimum inter-point density distance D22e, which will be described later) is generated for the candidate points. Step (4): If there are no other points inside the circle, the candidate point is determined as the fixed point. If there are other points inside the circle, the candidate point P is deleted.

[0025] Then, steps (2) to (4) are repeated until the number of generated points exceeds N or until candidate points are eliminated m times in a row. Note that the minimum inter-point distance does not need to be constant. In other words, the PDS method may set the minimum inter-point distance to any value for each generated point.

[0026] The algorithm for the internally generated point cloud data D22 using the PDS method executed in step S22 will be described. Step S22 for obtaining the internally generated point cloud data D22 includes five steps S221, S222, S223, S224, and S225. In step S221, internally generated candidate points D22b are generated. In step S222, candidate point stress values ​​D22c occurring at the internally generated candidate points D22b are obtained. In step S223, a candidate point density D22d is derived from the candidate point stress values ​​D22c. In step S224, a lowest inter-point density distance D22e is derived from the candidate point density D22d. In step S225, it is determined whether an internally generated determined point D22a exists within the range of a virtual sphere having a radius equal to the lowest inter-point density distance D22e.

[0027] An internal generation candidate point D22b is generated (S221). One internal generation candidate point D22b is generated inside the object of target shape. The internal generation candidate point D22b can be defined as coordinate values ​​(x, y, z).

[0028] A candidate point stress value D22c occurring at the internal generation candidate point D22b is obtained (S222). By substituting the coordinate values ​​(x, y, z) that define the internal generation candidate point D22b into the stress function D21 obtained in step S21, the candidate point stress value D22c occurring at the internal generation candidate point D22b indicated by the coordinate values ​​(x, y, z) can be obtained.

[0029] A candidate point density D22d is derived from the candidate point stress value D22c (S223). The candidate point density D22d is derived from the candidate point stress value D22c using the stress-density relational expression (see Equation (2)).

number

[0030] The lowest inter-point density distance D22e is derived from the candidate point density D22d (S224). Here, it is assumed that the truss structure to be generated later is a tetrahedron. The candidate point density D22d is also assumed to be the ratio between the volume of the tetrahedron and the volume of the truss. Then, the relationship between the candidate point density D22d and the lowest inter-point density distance D22e is expressed as Equation (3).

number

[0031] In other words, this embodiment is characterized in that when applying the PDS method to select an internal generation confirmed point D22a from an internal generation candidate point D22b, the candidate point stress value D22c is converted into a candidate point density D22d based on the allowable stress, and an inter-point distance (lowest inter-point density distance D22e) based on the candidate point density D22d is used.

[0032] It is determined whether or not another internal generation determined point D22a exists within the range of a virtual sphere having a radius equal to the lowest inter-point density distance D22e (S225). If the internal generation determined point D22a exists (S225: YES), the internal generation candidate point D22b is not selected as the internal generation determined point D22a (S226). In other words, the internal generation candidate point D22b is deleted. If the internal generation determined point D22a does not exist (S225: NO), the internal generation candidate point D22b is selected as the internal generation determined point D22a (S227).

[0033] Step S22 repeats steps S221 to S227. The condition for ending the repetition of steps S221 to S227 may be that the number of internal generation determined points D22a is greater than a predetermined number (N). Alternatively, the condition for ending the repetition of steps S221 to S227 may be that the internal generation candidate point D22b has not been selected as the internal generation determined point D22a a predetermined number of times (m times) in succession.

[0034] As a result of executing step S22 for generating the internally generated point cloud data D22, a plurality of internally generated determined points D22a corresponding to the stress distribution are set inside the object T4 (see FIG. 4).

[0035] Surface generation point cloud data D23 is obtained (S23). Figure 5 is a detailed flowchart of step S23 for obtaining surface generation point cloud data D23. In step S23 for obtaining surface generation point cloud data D23, a surface generation point cloud is generated at regular intervals on the surface of the target-shape object. Step S23 for obtaining surface generation point cloud data D23 includes step S231 for generating a surface generation point cloud at regular intervals for the mesh that constitutes the surface of the target-shape object, and step S232 for thinning out the surface generation point cloud data D23 generated on the surface of the target-shape object.

[0036] Surface generation candidate points D23b are generated (S231). The surface of the target shape object is made up of a plurality of meshes of different sizes. Surface generation candidate points D23b are generated based on the meshes. The surface generation candidate points D23b can be obtained by the following steps (1) to (6). In the following explanation, it is assumed that the meshes have a triangular shape.

[0037] In the first step (1), the height of the triangular mesh M6 is set to h (see FIG. 6(a)). This height h is the maximum value that can be defined as the height of the triangular mesh M6. The variable ndiv is defined as in the following equation (3). The constant pitch included in the equation below is a predetermined arbitrary value. For example, when the constant pitch is greater than the height h (pitch>h), the variable ndiv is zero (ndiv=0).

number

[0038] In the second step (2), first dividing points K1a to K1c that equally divide the first side K1, which is not the base side, are set using the variable ndiv (see FIG. 6(b)). Furthermore, second dividing points K2a to K2c that equally divide the second side K2, which is not the base side, are set (see FIG. 6(b)). Note that FIG. 6(b) is an example in which the variable ndiv is 4.

[0039] In the third step (3), line segments Y61 to Y63 are set that connect the first equally divided points K1a to K1c with the second equally divided points K2a to K2c, respectively (see FIG. 6(c)). Furthermore, a line segment Y64 is also set.

[0040] As a fourth procedure (4), the following fifth procedure (5) and sixth procedure (6) are repeatedly executed for each of the line segments Y61 to Y64.

[0041] In the fifth step (5), a new variable ndivx is defined. The new variable ndivx is defined by equation (5). The subscript "x" of the new variable ndivx is a number that identifies the line segments Y61 to Y64. For example, when the constant pitch is longer than the length d1 of the line segment Y61 (pitch>d1), the new variable ndiv1 is zero (ndiv1=0).

number

[0042] In the sixth step (6), a dividing point Y62a that divides the line segment Y62 equally is set using a new variable ndivx (see FIG. 7).

[0043] Then, as a result of performing the fifth step (5) and the sixth step (6) for all line segments Y62, Y63, and Y64, multiple surface generation candidate points D23b (points K0a, K1a, K1b, K1c, K1d, K2a, K2b, K2c, K2d, Y62a, Y63a, Y63b, Y64a, and Y64b) are set on the triangular mesh M6, as shown in FIG. 7.

[0044] Next, a thinning-out process step S232 is executed. The surface generation candidate points D23b obtained as a result of step S231 can be used as is to execute step S30, which obtains the next truss information D30. By executing thinning-out process step S232, some of the surface generation candidate points D23b are selected as surface generation determined points D23a to be used in step S30, which obtains the truss information D30. The surface generation candidate points D23b that are not selected as surface generation determined points D23a are not used in step S30, which obtains the truss information D30. As a result, the amount of data used in step S30, which obtains the truss information D30, is reduced, and the time required to execute step S30, which obtains the truss information D30, can be shortened.

[0045] A plurality of surface generation candidate points D23b are stacked (S232a). One surface generation candidate point D23b is extracted from the stack (S232b). It is determined whether a surface generation determined point D23a exists within a range of radius r centered on the extracted surface generation candidate point D23b (S232c). The "radius r" here refers to the minimum inter-point spatial distance D23d that defines the shortest distance to a second surface generation candidate point D23b different from the first surface generation candidate point D23b. If a surface generation determined point D23a exists (S232c: YES), the surface generation candidate point D23b is not selected as the surface generation determined point D23a (S232d). If a surface generation determined point D23a does not exist (S232c: NO), the surface generation candidate point D23b is selected as the surface generation determined point D23a (S232e). Then, the above-mentioned steps S232a to S232e are looped until the stack is exhausted (S232f).

[0046] As a result of executing step S20 for obtaining point cloud information D20, point cloud information D20 is obtained. Fig. 8 shows point cloud information D20 that is composed of a plurality of interior generation determined points D22a and a plurality of surface generation determined points D23a generated for an object T8 whose outer shape is a rectangle.

[0047] Truss information D30 is obtained (S30). A truss structure is a structure made up of multiple triangles, and is sometimes used as a mechanical model for structures such as bridges and buildings. A truss structure is composed of beam elements and node elements. Axial forces such as compressive stress or tensile stress act on the beam elements. The node elements connect the ends of the beam elements to each other. In the node elements, the ends of the beam elements can rotate relative to each other. In other words, no bending moment acts between beam elements connected by node elements. The ends can rotate relative to each other.

[0048] Candidate truss data D31 is generated (S31). The three-dimensional Delaunay division method is used as a process for generating the candidate truss D31a. When there are any number of points in a space, the three-dimensional Delaunay division method divides the space into tetrahedrons using those points. Only the four points that make up the tetrahedron exist inside the virtual circumscribing sphere defined for the generated tetrahedron. In other words, there are no points that do not make up the tetrahedron inside the virtual circumscribing sphere defined for the generated tetrahedron. The candidate truss D31a is generated by applying the three-dimensional Delaunay division method to the interior generation point cloud data D22 and the surface generation point cloud data D23. A specific algorithm for the three-dimensional Delaunay division method includes the following steps (1) to (8).

[0049] Step (1): Stack points that exist arbitrarily in space. Step (2): Generate a cube (superbox) that contains all the points, and divide the superbox into six tetrahedrons. Step (3): Take one point from the stack and generate a new candidate point in the superbox. Step (4): Search for a tetrahedron that contains the new candidate point. Similarly, search for tetrahedrons Y(1), Y(2), ... Y(N) that form a circumscribing sphere that contains the new candidate point. In this case, there may be multiple Ys. Step (5): From the set of tetrahedrons selected by the search, remove the triangles located between adjacent tetrahedrons to create a single polyhedron. Step (6): A new tetrahedron is generated by combining the surface triangles of the polyhedron obtained in step (5) with new candidate points located inside the surface triangles. Step (7): Repeat steps (3) to (6) until there are no more stuck points. Step (8): When there are no more stuck points, delete all tetrahedrons that have at least one vertex among the six tetrahedrons generated in step (2).

[0050] According to the space truss obtained by carrying out the above steps (1) to (8), in areas where the stress value is large, the inter-point distances between the interior generation determined points D22a and the surface generation determined points D23a that make up the point cloud information become small. In other words, in areas where the stress value is large, the density of the space truss generated with the interior generation determined points D22a and the surface generation determined points D23a as vertices increases. Note that the surface generation determined points D23a are set without reference to the stress value, so the appearance of the target-shaped object does not change.

[0051] Depending on the external shape of the target-shape object, a distorted candidate truss D31e that should not exist in reality may be generated for the original shape T9A shown in Fig. 9(a), as shown in Fig. 9(b) as a shape T9B. In other words, the candidate truss data D31 includes a candidate truss D31a for simulating the target-shape object and a distorted candidate truss D31e that should not exist in reality. This is due to the nature of the Delaunay decomposition method, and the distorted candidate truss D31e may appear even in the normal calculation results of the Delaunay decomposition method. Therefore, the deterministic truss D32a is obtained by excluding the distorted candidate truss D31e from the candidate truss data D31.

[0052] Definite truss data D32 is acquired (S32). FIG. 10 is a detailed flowchart of step S32 for acquiring the definite truss D32a. The intersection number determination method is used as the process for acquiring the definite truss D32a. When a point and an object exist in space, the intersection number determination method determines whether the point exists inside the object (inside / outside determination). As shown in FIG. 11, half lines L11a and L11b are generated connecting points B11a and B11b (truss reference points D32c) starting from the points (points P11a and P11b) for which inside / outside determination is to be performed. Then, the number of points where the half line L11a intersects with the object surface T11s and the number of points where the half line L11b intersects with the object surface T11s are counted. The number of points C11a and C11b where the half line L11a intersects with the object surface T11s is two (an even number). In this case, it is found that the candidate truss midpoint P11a exists outside the target-shape object T11. On the other hand, the number of intersection points D11a between the half-line L11b and the object surface T11s is 1 (odd number). In this case, it is found that the candidate truss midpoint P11b exists inside the target-shape object T11.

[0053] Following the explanation of the intersection number determination method, a more specific algorithm will be explained. First, in step (S32) of acquiring a definite truss D32a, candidate truss midpoints P11a and P11b are set for the candidate truss D31a. Then, using the intersection number determination method, it is determined whether or not this candidate truss midpoint D32b (points P11a and P11b) is inside the target-shape object. If the candidate truss midpoint D32b is inside the target-shape object T11, as in point P11b, the candidate truss D31a is adopted as the definite truss D32a that constitutes the target-shape object T11. If it is outside the target-shape object T11, as in point P11a, the candidate truss D31a does not constitute the target-shape object T11 and is a distorted candidate truss D31e, so it is not adopted as the definite truss D32a.

[0054] Candidate truss midpoints P11a, P11b, etc. are set (S321). A truss judgment line D32d (semi-rays L11a, L11b) connecting the candidate truss midpoints P11a, P11b and the truss reference point P11n is generated (S322). The number of truss intersection points (points C11a, C11b, D11a) where the semi-rays L11a, L11b, which are the truss judgment line D32d, intersect with the object surface T11s is counted (S323). Then, it is determined whether the truss intersection point number D32e is an odd number (S324). If the truss intersection point number D32e is not an odd number (S324: NO), it is a distortion candidate truss D31e, and is not selected as the confirmed truss D32a. If the truss intersection point number D32e is an odd number (S324: YES), it is selected as the confirmed truss D32a.

[0055] By executing the above steps S10, S20, and S30, truss information D30 including multiple definite trusses D32a is obtained. This truss information D30 is shape data for obtaining a part that can keep the generated stress value within an allowable range while maintaining a predetermined external shape.

[0056] <Shape data generation program> The shape data generating program PR causes a computer 100 having the configuration shown in FIG. 12 to execute a shape data generating method.

[0057] The processor 101 executes an operating system, application programs, etc. The memory 102 stores the shape data generation program PR and various data for generating shape data. The memory 102 passes data to the processor 101 in response to a request from the processor 101. Data is also written to the memory 102 by the processor 101. The memory 102 may be configured with a read-only memory (ROM) and a random access memory (RAM). The computer 100 may include various elements for executing the shape data generation program PR. For example, the computer 100 may include a communication control unit 103, an input unit 104, and an output unit 105. The input unit 104 may include a keyboard, a mouse, a touch panel, a microphone for voice input, etc. The output unit 106 may include a display, a printer, etc. For example, the output unit 105 may display the execution results of the shape data generation program PR. The shape data generation program PR may be provided in a form recorded on a tangible recording medium such as a CD-ROM, a DVD-ROM, or a semiconductor memory. The shape data generating program PR may be provided as a data signal via a communication network.

[0058] 13, the processor 101 reads out the shape data generation program PR from the memory 102. The shape data generation program PR executed by the processor 101 causes the processor 101 to operate as a discrete stress data generation unit 10, a point cloud information generation unit 20, and a truss information generation unit 30.

[0059] The discrete stress data generation unit 10 executes a step (S10) of obtaining discrete stress data D11. The discrete stress data generation unit 10 has a stress analysis unit 11. Note that when discrete stress data D11 generated by another computer is accepted, the discrete stress data generation unit 10 may be omitted.

[0060] The stress analysis unit 11 executes stress analysis processing (S11). The stress analysis unit 11 reads model shape data D101, model material data D102, constraint condition data D103, and load condition data D104 from the memory 102. The stress analysis unit 11 generates discrete stress data D11 using the read data. The stress analysis unit 11 writes the discrete stress data D11 to the memory 102.

[0061] The point cloud information generation unit 20 executes a step (S20) of obtaining point cloud information D20. The point cloud information generation unit 20 has a stress function calculation unit 21 (first calculation unit), an internally generated point cloud data calculation unit 22 (second calculation unit), and a surface generated point cloud data calculation unit 23 (third calculation unit). The stress function calculation unit 21 executes a step (S21) of obtaining a stress function D21. The stress function calculation unit 21 reads out discrete stress data D11 from the memory 102. The stress function calculation unit 21 generates a stress function D21 using the discrete stress data D11. The stress function calculation unit 21 writes the stress function D21 to the memory 102.

[0062] The internally generated point cloud data calculation unit 22 executes a step (S22) of generating internally generated point cloud data D22 (see FIG. 3). The internally generated point cloud data calculation unit 22 reads out the stress function D21 from the memory 102. The internally generated point cloud data calculation unit 22 generates the internally generated point cloud data D22 using the stress function D21. The internally generated point cloud data calculation unit 22 writes the internally generated point cloud data D22 to the memory 102.

[0063] The surface generation point cloud data calculation unit 23 executes a step (S23) of generating surface generation point cloud data D23 (see FIG. 5). The surface generation point cloud data calculation unit 23 reads the model shape data D101 from the memory 102. The surface generation point cloud data calculation unit 23 generates the surface generation point cloud data D23 using the model shape data D101. The surface generation point cloud data calculation unit 23 writes the surface generation point cloud data D23 to the memory 102.

[0064] The truss information generation unit 30 (fourth calculation unit) executes a step (S30) of obtaining truss information D30. The truss information generation unit 30 includes a candidate truss calculation unit 31 and a definitive truss calculation unit 32.

[0065] The candidate truss calculation unit 31 executes a step (S31) of generating a candidate truss D31a. The candidate truss calculation unit 31 reads point cloud information D20 from the memory 102. Specifically, the candidate truss calculation unit 31 reads interior generation point cloud data D22 and surface generation point cloud data D23 from the memory 102. The candidate truss calculation unit 31 generates a candidate truss D31a using the interior generation point cloud data D22 and the surface generation point cloud data D23. The candidate truss calculation unit 31 writes the candidate truss D31a to the memory 102.

[0066] The definite truss calculation unit 32 executes a step (S32) of generating a definite truss D32a (see FIG. 10 ). The definite truss calculation unit 32 reads the candidate truss D31a from the memory 102. The definite truss calculation unit 32 generates a definite truss D32a from the candidate truss D31a. The definite truss calculation unit 32 writes the definite truss D32a to the memory 102.

[0067] <Action and effect> Incidentally, the recent spread of so-called 3D printers is creating an environment in which individuals can create things relatively easily. For example, in the field of architecture, houses are being built using architectural 3D printers. In the field of space, plans are being considered to build a lunar base using 3D printers. Modeling using a 3D printer requires modeling data such as STL files, and there is a need for research and development into CAE (Computer Aided Engineering: design support systems) to prepare modeling data.

[0068] In the medical field, a method for creating artificial joints using three-dimensional printing technology with metal materials is being considered. Artificial joints created using three-dimensional printing technology have the advantage of being able to be manufactured to fit the skeletal structure of each individual patient. Furthermore, artificial joints created using three-dimensional printing technology can also be designed to have a structure that is highly compatible with the human body, such as by imparting a porous structure to the surface. However, because artificial joints are used in the human body, they are often required to be strong enough to withstand use without breaking and without affecting the human body. In other words, an artificial joint that is too strong may have an adverse effect on the human body. Therefore, it is desirable for artificial joints to have both an external shape that fits the patient's skeletal structure and a strength comparable to that of the patient's skeletal structure.

[0069] For example, topology optimization is a technique that focuses on the shape and strength of a molded object. Topology optimization is a technique that derives a material distribution that satisfies required engineering conditions from the material distribution set as an initial condition. More specifically, topology optimization derives a shape that satisfies required strength from a shape set as an initial condition. Topology optimization replaces the structural design problem with a material distribution problem, thereby optimizing the structure while allowing for changes in shape. The shape data generation method and shape data generation program of this embodiment have the following three advantages over topology optimization.

[0070] The first advantage of the shape data generation method and shape data generation program of this embodiment is that the external shape of the target model can be specified. As the name suggests, topology optimization obtains an optimal shape by changing the topology. For example, topology optimization can be thought of as removing material from areas where stress is relatively low and therefore no material placement is deemed necessary. Depending on the stress distribution, this can result in material being removed near the surface of the object. This can result in a change in the appearance of the object.

[0071] In contrast, the shape data generation method and shape data generation program of this embodiment change the internal shape but do not change the surface shape. The shape data generation method and shape data generation program of this embodiment use the pre-optimization shape as the external shape of the object, and then densely arrange trusses in areas with relatively high stress. As a result, requirements can be met by changing the internal shape without changing the surface shape. Therefore, the shape data generation method and shape data generation program of this embodiment are advantageous for generating molding models for objects whose appearance must not be changed, such as the aforementioned artificial joint.

[0072] The second advantage is the computational cost. Because the topology optimization method solves an optimization problem, the calculations can become complicated depending on input conditions such as the object shape and required conditions such as design variables. In contrast, the shape data generation method and shape data generation program of this embodiment generate a point cloud using the PDS method according to the stress distribution, and then generate a truss. Therefore, input conditions such as the object shape and required conditions such as design variables do not affect the difficulty of the calculations.

[0073] A third advantage is that no specialized knowledge is required. There are several types of topology optimization methods, including the density method, homogenization method, and level set method. Each type has its own advantages and disadvantages. Furthermore, because topology optimization methods are highly versatile, they can be applied to engineering problems other than the problem of controlling the strength of an object. Therefore, users must determine the appropriate method for their purpose, which requires specialized knowledge. In contrast, the shape data generation method and shape data generation program of this embodiment are applicable to the problem of controlling the strength of an object. As a result, there is no need to set complex design conditions, and neither specialized knowledge nor any other knowledge is required. In other words, this CAE is well-suited to the current situation where the widespread use of 3D printers has made it relatively easy for individuals to create things, and it can be said that this CAE can easily obtain shape data that meets requirements.

[0074] In short, the shape data generation method and the shape data generation program of this embodiment can adjust the strength of an object without changing its appearance, and can also reduce material costs by removing unnecessary internal regions. Therefore, the shape data generation method and the shape data generation program of this embodiment can be suitably applied to situations such as the design of artificial joints.

[0075] To summarize the contents described so far, the shape data generation method of this embodiment includes step S21 of obtaining a stress function D21 having coordinate values ​​defined in a target-shape object as independent variables and discrete stress data D11 generated in a target-shape object for which predetermined constraint condition data D103 and load condition data D104 have been set as dependent variables; step S22 of obtaining internally generated point cloud data D22 using the stress function D21, which indicates the stress distribution generated inside the target-shape object by the density of multiple internally generated determined points D22a; step S23 of obtaining surface generation point cloud data D23 including multiple surface generation determined points D23a for defining the outer shape of the target-shape object without being based on the stress distribution generated inside the target-shape object; and step S30 of obtaining truss information D30 including determined trusses D32a connecting the internally generated determined points D22a and / or the surface generation determined points D23a using the internally generated point cloud data D22 and the surface generation point cloud data D23.

[0076] The shape data generation program PR causes the computer 100 to function as: a stress function calculation unit 21 that obtains a stress function D21 having coordinate values ​​defined in a target-shape object as independent variables and discrete stress data D11 generated in a target-shape object for which predetermined constraint condition data D103 and load condition data D104 have been set as dependent variables; an internal generation point cloud data calculation unit 22 that uses the stress function D21 to obtain internal generation point cloud data D22 that indicates the stress distribution generated inside the target-shape object by the density of a plurality of internal generation determined points D22a; a surface generation point cloud data calculation unit 23 that obtains surface generation point cloud data D23 that sets a plurality of surface generation determined points D23a for defining the outer shape of the target-shape object without being based on the stress distribution generated inside the target-shape object; and a truss information generation unit 30 that uses the internal generation point cloud data D22 and the surface generation point cloud data D23 to obtain truss information D30 including determined trusses D32a that connect the internal generation determined points D22a and / or the surface generation determined points D23a to each other.

[0077] The shape data generation method and shape data generation program PR generate internally generated point cloud data D22, which indicates the stress distribution occurring inside a target-shape object by the density of multiple internally generated fixed points D22a. A truss structure obtained using this internally generated point cloud data D22 can impart desired strength to the target-shape object simulated by the model shape data D101. Furthermore, this shape data generation method generates surface generation point cloud data D23, which includes multiple surface generation fixed points D23a for defining the outer shape of the target-shape object, without being based on the stress distribution occurring inside the target-shape object. The truss structure obtained using this surface generation point cloud data D23 does not change the appearance of the target-shape object simulated by the model shape data D101, because it is not based on the stress distribution occurring inside the target-shape object. Therefore, it is possible to obtain model shape data D101 for a part that can keep the generated stress value within an allowable range while maintaining a predetermined outer shape.

[0078] Step S22 for obtaining internally generated point cloud data includes step S221 for obtaining, using a stress function, multiple candidate-point stress values ​​D22c occurring at each of multiple internal generation candidate points D22b included in the target-shape object, steps S223 and S224 for obtaining a lowest-point-to-point density distance D22e for treating a first candidate-point stress value D22c occurring at a first internal generation candidate point D22b selected from the multiple internal generation candidate points D22b as virtual distance information from the first internal generation candidate point D22b, and step S225 for selecting the first internal generation candidate point D22b as an internal generation determined point D22a if a second internal generation candidate point D22b different from the first internal generation candidate point D22b does not exist within the range indicated by the lowest-point-to-point density distance D22e set for the first internal generation candidate point D22b. This step (S22) allows for obtaining internally generated point cloud data D22 corresponding to the stress distribution.

[0079] Step S23 for obtaining surface generation point cloud data D23 includes the steps of: (S231) setting, as multiple surface generation candidate points D23b, nodal elements defining both ends of connected elements and additional points overlapping the connected elements; and (S232) using a minimum inter-point spatial distance D23c that defines the shortest distance from a first surface generation candidate point D23b included in the multiple surface generation candidate points D23b to a second surface generation candidate point D23b, selecting the second surface generation candidate point D23b as a surface generation confirmed point if the distance from the first surface generation candidate point D23b to the second surface generation candidate point D23b is greater than the minimum inter-point spatial distance D23c. This step S23 allows for the acquisition of surface generation point cloud data D23 with a light computational load.

[0080] Step S30 of obtaining truss information includes step S31 of obtaining candidate truss data D31 using the internally generated point cloud data D22 and the surface generated point cloud data D23, and step S32 of obtaining definite truss data including definite trusses that constitute the target-shape object by removing, from the candidate trusses D31a included in the candidate truss data D31, distorted candidate trusses D31e that are not included in the target-shape object. This step S30 can prevent changes in the appearance of the target-shape object simulated by the model shape data D101.

[0081] <Modification> The present invention can be implemented in various forms, including the above-described embodiment, with various modifications and improvements based on the knowledge of those skilled in the art. Furthermore, it is also possible to configure modified examples by utilizing the technical matters described in the above-described embodiment. The configurations of the embodiments and the like may be used in appropriate combination.

[0082] <Calculation example> It has been confirmed by calculation that the shape data generation method of the embodiment can reduce stress generated inside a target-shaped object without changing the outer shape of the target-shaped object.

[0083] Figures 14(a) and 14(b) show the truss information model T14 before the shape data generation method is applied. Figure 14(b) shows an enlarged view of area T14s in Figure 14(a). The truss information model T14 shown in Figure 14(a) was obtained by applying the two-dimensional Delaunay division method to model shape data D101 including 2881 interior generation points and 240 surface generation points.

[0084] For the truss information model T14 shown in Figures 14(a) and 14(b), a fixed condition was imposed on the nodes that make up the bottom side T14a as a constraint condition. Furthermore, a load of 50 MPa acting in the negative Y direction on the nodes that make up the top side T14b was set as a load condition. Under these conditions, the stress generated inside was obtained by calculation. The calculation results are shown in Figure 14(c). The maximum tensile stress was 40.9 MPa. The maximum compressive stress was 197 MPa. In other words, the maximum stress was 197 MPa.

[0085] Next, the shape model generation method of the embodiment was performed using the model shape data D101 and the results of the above stress analysis, resulting in point cloud information D20 consisting of interior generation point cloud data D22 including 2760 interior generation determined points D22a and surface generation point cloud data D23 including 460 surface generation determined points D23a.

[0086] As a result of applying the shape model generation method of the embodiment, it was found that the number of internal generation fixed points was reduced from 2881 to 2760. As the number of internal generation fixed points is reduced, the number of trusses also decreases. The trusses are parts that are printed when the shape model is printed by a three-dimensional printing device. In other words, a reduction in the number of trusses leads to a reduction in the parts that are printed. As a result, less material is required for printing, which is also found to be advantageous in terms of manufacturing costs.

[0087] Next, step S30 was performed to obtain truss information D30 using this point cloud information D20, resulting in the truss information model T20 shown in Figures 15(a) and 15(b). Figure 15(b) shows an enlarged view of area T20s in Figure 15(a). Then, the same constraint conditions and load conditions as described above were imposed on the truss information model T15 shown in Figures 15(a) and 15(b), and a stress analysis was performed. The results of the calculation are shown in Figure 15(c). The maximum tensile stress value was 62.5 MPa. The maximum compressive stress value was 181.5 MPa. In other words, the maximum stress value was 181.5 MPa. Table 1 shows each stress value. [Table 1]

[0088] It was found that the outer shape of the truss information model T15 after the application of the shape model generation method (see Figure 15(a)) was able to maintain the outer shape of the truss information model T14 before the application of the shape model generation method (see Figure 14(a)). It was also found that the maximum stress generated in the truss information model T15 after the application of the shape model generation method was lower than the maximum stress generated in the truss information model T14 before the application of the shape model generation method. In other words, the shape model generation method made it possible to obtain a shape model that can reduce the stress generated inside the target-shape object without changing the outer shape of the target-shape object.

[0089] <Consideration> As can be seen from Table 1, when the shape model generation method of the embodiment was applied, the maximum compressive stress value was reduced. In other words, when the shape model generation method of the embodiment was applied, the maximum stress value was reduced. On the other hand, the maximum tensile stress value was increased. However, the relationship that the maximum compressive stress value was greater than the maximum tensile stress value remained the same regardless of whether the shape model generation method of the embodiment was applied or not.

[0090] 16(a) shows a truss information model T16 before the application of the shape model generation method of the embodiment. The triangular meshes constituting the truss information model T16 are all the same in size and shape. It is assumed that, as a result of imposing predetermined constraint conditions and load conditions on this truss information model T16 and performing stress analysis, a relatively large stress value occurs in region T16a and relatively small stress values ​​occur in regions T16b and T16c.

[0091] Assume that the truss information model T17 shown in FIG. 16(b) is obtained by applying the geometric model generation method of the embodiment to the truss information model T16 shown in FIG. 16(a) and the results of stress analysis. When the geometric model generation method of the embodiment is applied, a dense truss structure is generated in the region T17a where a relatively large stress value occurs, and a sparse truss structure is generated in the regions T17b and T17c where a relatively small stress value occurs. Assume that similar constraint conditions and load conditions are imposed on the truss information model T17, including the sparse and dense structures. In this case, force lines T17d and T17e are expected to be generated from the region T17a where a relatively large stress value occurs toward the regions T17b and T17c where a relatively small stress value occurs. This stress redistribution is expected to decrease the stress value occurring in the region T17a and increase the stress value occurring in the regions T17b and T17c. As a result, when the shape model generation method of the embodiment is applied, force lines T17d and T17e are generated according to the density of the truss, and stress is redistributed along these force lines T17d and T17e, resulting in a decrease in the maximum stress value.

[0092] [Note] The present disclosure includes the following contents.

[0093] The present disclosure provides: [1] "a step of obtaining a stress function in which coordinate values ​​defined on a target-shaped object are used as independent variables and stress values ​​generated on a target-shaped object to which predetermined constraint conditions and load conditions are set are used as dependent variables; a step of obtaining internally generated point cloud data that indicates a stress distribution occurring inside the target-shaped object by a density of a plurality of internally generated points using the stress function; obtaining surface generation point cloud data including a plurality of surface generation points for defining the outer shape of the target-shaped object without being based on a stress distribution occurring inside the target-shaped object; and using the internally generated point cloud data and the surface generated point cloud data to obtain truss information including truss elements connecting the internally generated points and / or the surface generated points to one another."

[0094] The present disclosure states, [2] "The step of obtaining the internally generated point cloud data includes: using the stress function to obtain a plurality of candidate point stress values ​​occurring at a plurality of internally generated candidate points included in the target shape object; a step of obtaining an inter-point evaluation distance for treating a first candidate point stress value occurring at a first internally generated candidate point selected from the plurality of internally generated candidate points as virtual distance information from the first internally generated candidate point; and when a second internally generated candidate point different from the first internally generated candidate point does not exist within a range indicated by the inter-point evaluation distance, which is virtual distance information set for the first internally generated candidate point, selecting the first internally generated candidate point as an internally generated confirmed point, which is the internally generated point.

[0095] The present disclosure states, [3] "The shape of the target-shaped object is defined by a plurality of meshes defined by a plurality of node elements and connecting elements connecting the plurality of node elements; The step of obtaining the surface generated point cloud data includes: setting the nodal elements defining both ends of the connected elements and additional points overlapping the connected elements as a plurality of surface generation candidate points; and using a lowest-point spatial distance that defines the shortest distance from a first surface generation candidate point included in the plurality of surface generation candidate points to a second surface generation candidate point different from the first surface generation candidate point, when the distance from the first surface generation candidate point to the second surface generation candidate point is greater than the lowest-point spatial distance, determining the second surface generation candidate point as the surface generation point.

[0096] The disclosure states, [4] "The step of obtaining truss information includes: obtaining candidate truss data using the internally generated point cloud data and the surface generated point cloud data; and obtaining definite truss data including a definite truss that constitutes the target-shape object by removing distorted truss elements that are not included in the target-shape object from among the candidate truss elements included in the candidate truss data.

[0097] This disclosure is [5] "a computer, a first calculation unit for obtaining a stress function having coordinate values ​​defined on a target-shape object as independent variables and stress values ​​generated on the target-shape object for which predetermined constraint conditions and load conditions are set as dependent variables; a second calculation unit that uses the stress function to obtain internally generated point cloud data that indicates a stress distribution occurring inside the target-shaped object by the density of a plurality of internally generated points; a third calculation unit that obtains surface generation point cloud data including a plurality of surface generation points for defining the outer shape of the target-shaped object without being based on a stress distribution occurring inside the target-shaped object; and a shape data generation program that functions as a fourth calculation unit that uses the internally generated point cloud data and the surface generated point cloud data to obtain truss information including truss elements that connect the internally generated points and / or the surface generated points to each other. [Explanation of symbols]

[0098] 100...computer, 101...processor, 102...memory, 10...discrete stress data generation unit, 11...stress analysis unit, 20...point cloud information generation unit, 21...stress function calculation unit (first calculation unit), 22...internal generated point cloud data calculation unit (second calculation unit), 23...surface generated point cloud data calculation unit (third calculation unit), 30...truss information generation unit (fourth calculation unit), 31...candidate truss calculation unit, 32...determined truss calculation unit, D21...stress function, D22...internal Generated point cloud data, D22a...internal generated confirmed points, D22b...internal generated candidate points, D22c...candidate point stress values, D23...surface generated point cloud data, D23a...surface generated confirmed points, D23b...surface generated candidate points, D30...truss information, D31...candidate truss data, D31a...candidate truss, D32...confirmed truss data, D32a...confirmed truss, D23c...lowest inter-point spatial distance, PR...shape data generation program, T11...target shape object.

Claims

1. a step of obtaining a stress function in which coordinate values ​​defined on the target shape object are used as independent variables and stress values ​​generated on the target shape object to which predetermined constraint conditions and load conditions are set are used as dependent variables; a step of obtaining internally generated point cloud data that indicates a stress distribution occurring inside the target-shaped object by a density of a plurality of internally generated points using the stress function; obtaining surface generation point cloud data including a plurality of surface generation points for defining the outer shape of the target-shaped object without being based on a stress distribution occurring inside the target-shaped object; and obtaining truss information including truss elements connecting the internally generated points and / or the surface generated points to one another using the internally generated point cloud data and the surface generated point cloud data.

2. The step of obtaining the internally generated point cloud data includes: using the stress function to obtain a plurality of candidate point stress values ​​occurring at a plurality of internally generated candidate points included in the target shape object; a step of obtaining an inter-point evaluation distance for treating a first candidate point stress value occurring at a first internally generated candidate point selected from the plurality of internally generated candidate points as virtual distance information from the first internally generated candidate point; 2. The shape data generation method according to claim 1, further comprising a step of selecting the first internally generated candidate point as an internally generated confirmed point, which is the internally generated point, when a second internally generated candidate point different from the first internally generated candidate point does not exist within a range indicated by the point-to-point evaluation distance, which is virtual distance information set for the first internally generated candidate point.

3. the shape of the target shape object is defined by a plurality of meshes defined by a plurality of node elements and connecting elements connecting the plurality of node elements; The step of obtaining the surface generated point cloud data includes: setting the nodal elements defining both ends of the connected elements and additional points overlapping the connected elements as a plurality of surface generation candidate points; 2. The shape data generation method according to claim 1, further comprising: a step of using a lowest-point-to-lowest-point spatial distance that defines the shortest distance from a first surface generation candidate point included in the plurality of surface generation candidate points to a second surface generation candidate point different from the first surface generation candidate point, and determining the second surface generation candidate point as the surface generation point if the distance from the first surface generation candidate point to the second surface generation candidate point is greater than the lowest-point-to-lowest-point spatial distance.

4. The step of obtaining truss information includes: obtaining candidate truss data using the internally generated point cloud data and the surface generated point cloud data; The shape data generation method according to claim 1, further comprising the step of: obtaining definite truss data including a definite truss that constitutes the target-shape object by removing distorted truss elements that are not included in the target-shape object from among the candidate truss elements included in the candidate truss data.

5. Computer, a first calculation unit for obtaining a stress function having coordinate values ​​defined on a target-shape object as independent variables and stress values ​​generated on the target-shape object for which predetermined constraint conditions and load conditions are set as dependent variables; a second calculation unit that uses the stress function to obtain internally generated point cloud data that indicates a stress distribution occurring inside the target-shaped object by the density of a plurality of internally generated points; a third calculation unit for obtaining surface generation point cloud data including a plurality of surface generation points for defining the outer shape of the target-shaped object without being based on a stress distribution occurring inside the target-shaped object; and a shape data generation program that functions as a fourth calculation unit that uses the internally generated point cloud data and the surface generated point cloud data to obtain truss information including truss elements that connect the internally generated points and / or the surface generated points to each other.

Images