Method for developing additive manufacturing technology and 3D additive manufacturing system

Abstract

The present invention greatly reduces the time and cost involved in development for establishing manufacturing conditions for additive manufacturing, detecting defects during manufacture, and optimizing defect repair. The method includes: a defect-predicting step for predicting defects occurring in the manufactured product based on a combination of a plurality of design data and a plurality of manufacturing conditions; a defect detection data collection step for monitoring the manufactured product during manufacture according to a combination of a plurality of design data and a plurality of manufacturing conditions to collect defect detection data for defect detection; and a process-map-generating step for generating a process map plotting a plurality of manufacturing conditions using the predicted defects and the collected defect detection data. Furthermore, the method includes: a defect repair data collection step for monitoring the manufactured product during manufacture, repairing defects, and collecting defect repair data; and a defect repair data storage step for storing the associated defect and defect repair data using the defect repair data and repair results.

Description

[Technical field]

[0001] The present invention relates to a method for developing an additive manufacturing technology and a three-dimensional additive manufacturing system. [Background technology]

[0002] In the above technical field, Non-Patent Document 1 discloses a technology in which a reflected electron image of a surface is monitored in-process, and the user of the device classifies the image into porous, even, or uneven, and plots it. However, the user manually creates a process map and derives optimal conditions. In addition, Non-Patent Document 2 discloses a technology that predicts deformation by combining multi-scale simulations from powder scale to part scale. However, process optimization requires multiple simulations to be performed many times, and actual modeling is also required to verify the simulations. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. WO2020 / 039581 [Non-patent literature]

[0004] [Non-Patent Document 1] CR Pobel, C. Arnold, F. Osmanlic, Z. Fu, C. Korner, “Immediate development of processing windows for selective electron beam melting using layerwise monitoring via backscattered electron detection”, Materials Letters, 249 (2019) 70-72. [Non-Patent Document 2] W. King, AT Anderson, RM Ferencz, NE Hodge, C. Kamath, SA Khairallah “Overview of modeling and simulation of metal powder bed fusion process at Lawrence Livermore National Laboratory”, Materials Science and Technology, 31 (2015) 957-968. Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the technology described in the above document, since defect inspection is repeatedly performed after the object is manufactured, it takes time and cost to establish the manufacturing conditions for additive manufacturing and to optimize the detection and repair of defects during manufacturing.

[0006] An object of the present invention is to provide a technique for solving the above-mentioned problems. [Means for solving the problem]

[0007] In order to achieve the above object, the present invention Additive Manufacturing Systems teeth, Using electron beam molding (EBM) or laser beam molding (LBM), powder is melted at a specific location in a powder bed, A three-dimensional additive manufacturing system for manufacturing a three-dimensional additive object, comprising: a defect prediction unit that predicts defects that will occur in the object based on a combination of a plurality of design data and a plurality of modeling conditions used when additively manufacturing the object; and a defect detection data collection unit that monitors before and after melting of each layer of the object being modeled and collects defect detection data for detecting defects; Predicted The above Defects and Collected The above a process map generating unit that generates a process map in which the plurality of modeling conditions are plotted using the defect detection data; an additive manufacturing unit that additively manufactures a model while adjusting the plurality of modeling conditions in accordance with the process map; While adjusting the plurality of modeling conditions according to the process map A defect repair data collection unit that monitors before and after melting of each layer of the object being modeled, repairs defects detected in the object, and collects defect repair data for defect repair. and, It is an additive manufacturing system equipped with the above. Effect of the Invention

[0009] According to the present invention, it is possible to significantly reduce time and costs in establishing modeling conditions for additive manufacturing and in developing methods for optimizing defect detection and defect repair during manufacturing. [Brief description of the drawings]

[0010] [Figure 1] 1 is a flowchart showing the steps of a development method for additive manufacturing technology according to the first embodiment. [Figure 2A] 11A to 11C are diagrams illustrating a development method for additive manufacturing technology according to a second embodiment. [Figure 2B] FIG. 1 is a diagram illustrating a development method for an existing additive manufacturing technology. [Diagram 3] FIG. 2B is a diagram for explaining pattern division in FIG. 2A. [Figure 4A] 2B is a flowchart showing a procedure for generating a process map in FIG. 2A. [Figure 4B] FIG. 2B is a diagram for explaining an example of process map generation in FIG. 2A. [Figure 4C] FIG. 2B is a diagram for explaining another example of the generation of the process map in FIG. 2A. [Diagram 5] FIG. 2B is a diagram for explaining the integration of part printing conditions in FIG. 2A. [Figure 6A] 2B is a flowchart showing a procedure for repairing defects during the fabrication in FIG. 2A. [Figure 6B] FIG. 2B is a diagram showing the configuration of data used for defect repair during modeling in FIG. 2A. [Figure 7A] FIG. 11 is a block diagram showing the configuration of a three-dimensional additive manufacturing system according to a third embodiment. [Figure 7B] 13 is a diagram showing the configuration of an additive manufacturing unit according to a third embodiment. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. However, the components described in the following embodiment are merely examples, and are not intended to limit the technical scope of the present invention.

[0012] [First embodiment] A method 100 for developing an additive manufacturing technique according to a first embodiment of the present invention will be described with reference to Fig. 1. The method 100 for developing an additive manufacturing technique is a development method for establishing modeling conditions, detecting defects during modeling, and optimizing defect repair.

[0013] As shown in Fig. 1, the method 100 for developing an additive manufacturing technique includes a defect prediction step S101, a defect detection data collection step S103, and a process map generation step S105. The defect prediction step S101 predicts defects that will occur in a molded object based on a combination of multiple design data and multiple molding conditions used for additive manufacturing of the molded object. The defect detection data collection step S103 monitors the molded object during molding according to the combination of multiple design data and multiple molding conditions, and collects defect detection data for defect detection. The process map generation step S105 generates a process map in which multiple molding conditions are plotted, using the predicted defects and the collected defect detection data.

[0014] Furthermore, the additive manufacturing technology development method 100 includes a defect repair data collection step S107 and a defect repair data storage step S109. The defect repair data collection step S107 monitors the object being modeled while adjusting a plurality of modeling conditions according to the process map, repairs defects detected in the object, and collects defect repair data for the defect repair. The defect repair data storage step S109 uses the collected defect repair data and repair results of the defect repair to associate defects detected in the object being modeled with the defect repair data and stores them.

[0015] According to this embodiment, it is possible to significantly reduce time and costs in establishing modeling conditions for additive manufacturing and in developing methods for optimizing defect detection and defect repair during modeling.

[0016] [Second embodiment] Next, a method for developing an additive manufacturing technique according to a second embodiment of the present invention will be described. The method for developing an additive manufacturing technique according to this embodiment shows more specifically the processing of each step of the method for developing an additive manufacturing technique according to the first embodiment.

[0017] <Issues with prerequisite technology> Hereinafter, a development method of an existing additive manufacturing technique will be described with reference to FIG. 2B.

[0018] For example, the conventional procedure for manufacturing defect-free metal parts using metal additive manufacturing is as follows. Modeling is performed under modeling conditions A (S250) set by the user (S260), the defect rate of the obtained model is measured, and it is evaluated whether the model has been obtained as required (S270). Based on the evaluation results, the user adjusts the modeling conditions, performs modeling under the adjusted condition B, and adjusts the modeling conditions again based on the evaluation results of the model. This procedure (S250 → S260 → S270) is repeated until the required defect rate is obtained. This requires a very large number of repetitive steps, which is time-consuming and costly. For example, attempts have been made to reduce the cost of experiments by using simulations instead of actual modeling tests, but this requires specialized knowledge of metal additive manufacturing and may require a supercomputer, and is often not possible for general users. In addition, simulations always require experimental verification, so the cost of experiments is not necessarily small.

[0019] <Development method of additive manufacturing technology according to this embodiment> The method for developing additive manufacturing technology in this embodiment aims to enable even users without specialized knowledge to develop defect-free parts efficiently and in a short time by standardizing the process from optimizing the manufacturing conditions for metal parts of the material and shape required by the user to part manufacturing.

[0020] Fig. 2A is a diagram for explaining a method for developing an additive manufacturing technique according to the present embodiment. The method for developing an additive manufacturing technique in Fig. 2A includes step S210 of dividing the part shape into patterns to be formed, and step S220 of searching the process map generated for each divided pattern. The method for developing an additive manufacturing technique in Fig. 2A further includes step S230 of generating part forming conditions integrated for the entire part, and step S240 of repairing defects while monitoring the surface of the object being formed.

[0021] In step S210, patterns are classified according to the part shape. For example, patterns are classified according to the scan path length and the distance from the underskin. Patterns are also classified according to the melting area and the inclination angle. In step S220, a process map is searched for for each pattern. In this case, if a process map stored in the process map database 220 exists, it is used to determine the optimal conditions. On the other hand, if a process map does not exist or is insufficient, a process map is generated and then the optimal conditions are determined.

[0022] Here, the process map may be generated using, for example, the method described in Patent Document 1. That is, the generation of the process map includes the following functions.

[0023] <Functions of defect prediction system> (Image discrimination function) Input: Deep learning based on surface property data, Training data: Surface image, Output: Determining whether something is flat or not (Defect discrimination function 1) Input: Machine learning (SVM) based on the output of image discrimination function, Training data: None, Output: Process window (Defect discrimination function 2) Input: Deep learning based on surface property data, unevenness information, temperature, etc., Teacher data: Surface images, temperature data, melting and solidification simulation and melt pool monitoring, Output: Normal or above (internal defects, spatter generation, etc.)

[0024] <Specific examples of each step> Hereinafter, a specific processing example will be described with reference to FIGS. 2A to 6B for each step in FIG. 2A.

[0025] (Pattern classification: S210) FIG. 3 is a diagram for explaining the pattern division in FIG. 2A.

[0026] 3, pattern classification is performed according to the shape of a part to be molded using pattern classification data 310 or 320. The pattern classification data 310 stores a scan path length 312 and a distance from the underskin 313, which are pattern classification conditions, in association with a pattern 311. The pattern classification data 320 stores a melting area 322 and an inclination angle 323, which are pattern classification conditions, in association with a pattern 321.

[0027] The pattern classification S210 includes the following steps: In step S301, the shape of the part to be molded is acquired. In step S303, the part shape pattern is classified using pattern classification data 310 or 320. In step 305, classification is performed into pattern 1. In step 306, classification is performed into pattern 2. In step 307, classification is performed into pattern N.

[0028] (Generating a process map: S220) FIG. 4A is a diagram for explaining an example of process map generation in FIG. 2A.

[0029] In FIG. 4A, the process window generation method in the additive manufacturing system includes a sample manufacturing process (S401), a boundary determination process (S403), and a process window generation process (S405).

[0030] The sample modeling process (S401) corresponds to "1. Data point setting" and "2. Modeling of multiple samples" in Fig. 2A, and performs additive modeling of a sample using at least two parameter sets that control additive modeling and are scattered within a process window. The sample modeling process (S401) includes step S411 for setting parameter sets that are scattered within the process window, and step S413 for additive modeling of a sample using the parameter sets.

[0031] 2A, and determines the boundaries of the evaluation results by machine learning in a process map generated by mapping the evaluation results of the quality of the additively manufactured sample. The boundary determination process (S403) includes step S431 of evaluating the quality of the additively manufactured sample from an image of the sample, and step S433 of determining, by machine learning, the boundaries in the process map onto which the quality evaluation of the sample is mapped.

[0032] The process window generation process (S405) corresponds to "5. Resetting of missing data points" and "6 to 9. Sample modeling → monitoring → process window construction" in FIG. 2A. The process window generation process (S405) repeats the sample modeling step and the boundary determination step with the boundary area including the determined boundary as a new process window, and generates the process window separated by the boundary finally determined as a process window that guarantees the quality of additive manufacturing. The process window generation process (S405) includes step S451 of determining whether the boundary accuracy is sufficient or not, and step S453 of setting the boundary area including the boundary as the new process window when the boundary accuracy is not sufficient. The process window generation process (S405) also includes step S455 of setting the process window separated by the boundary finally determined as a process window that guarantees the quality of additive manufacturing under a termination condition such as when the boundary accuracy is sufficient.

[0033] Fig. 4B is a diagram for explaining an example of a process map 410 generated by the process map generation in Fig. 2A. Fig. 4B shows the result of assigning image positions as ◯(+1) / ×(-1) and determining a decision function representing a boundary using a support vector machine. In Fig. 4B, the boundary is shown by a temporary dashed line 411 to clarify the boundary portion.

[0034] FIG. 4C is a diagram for explaining another example of the generation of the process map in FIG. 2A.

[0035] In step S461, modeling conditions (output, scanning speed) are set. In step S463, 3D point cloud data is acquired for each modeling chamber. In this embodiment, the 3D point cloud data is obtained by capturing images of the surface of the model, the modeling surface of each layer, or the surface of the powder bed of each layer with a camera when modeling a layered object, and expressing each point on the surface as (x, y) coordinates and z coordinate in the height direction.

[0036] Next, in step S465, three-dimensional surface texture parameters are calculated from the acquired three-dimensional point cloud data. Here, the three-dimensional surface texture parameters are parameters for expressing surface roughness defined in the international standard (ISO25178). In this embodiment, the minimum autocorrelation lengths Sal, Sda, Sha, Sdv, and Smr2 of the surface are selected as three-dimensional surface texture parameters suitable for predicting the relative density of the molded object from the surface texture of the additive manufacturing.

[0037] In step S467, the relative density of the object is predicted based on the calculated three-dimensional surface texture parameters. Note that the relative relationship between the calculated three-dimensional surface texture parameters and the relative density of the object is stored based on the relative density measured from the object produced by additive manufacturing and the values ​​of the three-dimensional surface texture parameters during or after the object is manufactured.

[0038] In step S469, the modeling quality is evaluated based on the modeling density predicted based on the three-dimensional surface texture parameters. The predicted value of the modeling density is used to evaluate whether the modeling density of the modeled object after modeling is formed at a desired density. If the predicted value of the modeling density is predicted to be lower than the relative density required for the model, it may be used to adjust the modeling conditions, particularly the energy density, based on the laser intensity and scanning speed. In this case, it is desirable to prepare a process map of the relative density based on the combination of the laser intensity and scanning speed in advance and select an adjustment method.

[0039] In Step S471, the modeling evaluation is repeated by changing the modeling conditions in Step S461. When predetermined modeling and evaluation data are obtained, in Step S473, a process map is generated.

[0040] The process map 420 generated in this embodiment is shown in the right diagram of FIG. 4C. In this way, by preparing the generated process map 420, point cloud data is generated from a surface image captured after at least one layer is modeled. Then, appropriate three-dimensional surface property parameters are calculated from the generated point cloud data. Based on this three-dimensional surface property parameter, the modeling density can be predicted and controlled to be the modeling density required for the modeled object. For example, if the modeling density is stable at or above the required modeling density, modeling is continued as is, and if a decrease in the modeling density is predicted, remelting is performed under the same conditions, or the laser output and scanning speed are adjusted to increase the modeling density. If the modeling density is predicted to be unstable even if it is above the required modeling density, the modeling density is transitioned to a level at which the modeling density is stable.

[0041] (Integration of part building conditions: S230) FIG. 5 is a diagram for explaining the integration of part building conditions in FIG. 2A.

[0042] In step S501, a process map for pattern n (N≧n≧1) is obtained. In step S503, part modeling conditions are searched for from each process map. In step S505, the part modeling conditions obtained as a result of the search are stored. In step S507, it is determined whether part modeling conditions have been searched for and stored for all patterns or for necessary patterns. If the stored part modeling conditions are sufficient, in step S509, integrated part modeling conditions are set.

[0043] 5 shows a table 510 used for integrating part printing conditions. The table 510 stores a generated process map 512, part printing conditions 513 searched for by the process map 512, and integrated part printing conditions 514, in association with a part shape pattern 511.

[0044] (Repairing defects during printing: S240) FIG. 6A is a diagram illustrating defect repair during fabrication in FIG. 2A.

[0045] In step S603, additive manufacturing of one layer (n layers) is instructed. In the case of n layers, defect determination and repair are performed for each n layer. In step S605, reception of the modeling surface image is awaited. When the modeling surface image is received, in step S607, surface image data is acquired. Then, in step S609, the boundary between the molten area and the non-molten area is extracted based on the surface image data. In step S611, the molten area is divided to generate small areas.

[0046] In step S613, it is determined whether the small region is on the boundary between the molten region and the non-molten region. If the small region is on the boundary between the molten region and the non-molten region, in step S615, the data of the non-molten region is complemented (replaced) with the data of the molten region. Then, in step S617, it is determined whether there is a defect in the small region of the molten region and the complemented small region of the non-molten region. If there is no defect and it is determined in step S619 that modeling has not ended, the process returns to step S603 and additive modeling of the next layer (n layers) is performed.

[0047] If there is a defect, in step S621, the defective small area is remelted as a repair process. If the modeling quality is relatively low, the entire melted area or a predetermined area including the defective small area may be remelted instead of repairing the small area. In step S623, the number of remelting times is compared with a number threshold, and if the number of remelting times is less than the number threshold, the process returns to step S605 and remelting is repeated. On the other hand, if the number of remelting times is equal to or greater than the number threshold, modeling is stopped in step S625. Note that FIG. 6A does not show the case where a defect that cannot be repaired occurs in step S617, but in that case, modeling is stopped in step S625.

[0048] Fig. 6B is a diagram showing the configuration of data used for defect repair during modeling in Fig. 2A. The upper diagram of Fig. 6B is a diagram showing the configuration of a defect determination table 650 used to detect defects.

[0049] The defect determination table 650 stores the captured surface data 651 passed from the surface image acquisition unit 505, and the modeling boundary 652 and molten region data 653 extracted from the captured surface data 651. The defect determination table 650 also stores small region data 654 obtained by dividing the molten region data 653 into small regions. The small region data 654 includes position data of the small region on the modeling surface and image data of the small region. The defect determination table 650 also stores data 655 indicating whether the small region is a boundary small region or another non-boundary small region, image data that is evaluation target data 656 of the non-boundary small region, and complementary data that is evaluation target data 656 of the boundary small region. The defect determination table 650 also stores roughness data 657 derived from the image data of the non-boundary small region and roughness data 657 derived from the complementary data of the boundary region. Finally, the defect determination table 650 stores evaluation results 659 based on a defect determination threshold value 658. Small areas with roughness data 657 equal to or greater than the threshold value 658 are determined to have a defect, and small areas with roughness data 657 less than the threshold value 658 are determined to have no defect.

[0050] Although a simple method has been described for the defect determination, machine learning (deep learning, CNN, support vector machine, random forest, naive Bayes, etc.) may also be used (see Patent Document 2).

[0051] FIG. 6B is a diagram showing the configuration of a defect repair table 660 used to repair defects.

[0052] The defect repair table 660 stores a repair count 661 counted by the repair count counter, a count threshold 662, a defect repair process in 663 when the repair count is < the count threshold, and stores a modeling stop process in 664 when the repair count is ≥ the count threshold. Then, a defect repair instruction command 631 and a set melting energy (output, scanning speed, beam diameter) 632 to be sent to the additive manufacturing device are stored in association with the defect repair process.

[0053] According to the present embodiment, it is possible to significantly reduce the time and cost required for establishing the molding conditions for additive manufacturing and for developing the optimum defect detection and repair during manufacturing. In addition, it is possible to significantly reduce the time and cost required for developing defect-free metal parts using metal additive manufacturing.

[0054] [Third embodiment] Next, a 3D additive manufacturing system according to a third embodiment of the present invention will be described. The 3D additive manufacturing system according to this embodiment is configured by incorporating the additive manufacturing technology development method of the second embodiment into the 3D additive manufacturing system. Detailed explanations of the operations that are the same as those of the second embodiment will be omitted.

[0055] <Configuration of 3D additive manufacturing system> FIG. 7A is a block diagram showing the configuration of a three-dimensional additive manufacturing system 700 according to this embodiment.

[0056] The three-dimensional additive manufacturing system 700 includes an information processing device 710 of this embodiment, an additive manufacturing device 720, and a surface image capturing unit 730. The surface image capturing unit 730 may be built into the additive manufacturing device 720 or may be installed separately. The information processing device 710 includes a defect prediction unit 711, a defect detection data collection unit 712, a process map generation unit 713, a defect repair data collection unit 714, and a database 715.

[0057] The defect prediction unit 711 predicts defects that will occur in the object based on a combination of multiple design data and multiple modeling conditions. The defect detection data collection unit 712 monitors the object being modeled according to a combination of multiple design data and multiple modeling conditions, and collects defect detection data for defect detection. The process map generation unit 713 generates a process map in which multiple modeling conditions are plotted, using the predicted defects and the collected defect detection data. The defect repair data collection unit 714 generates a process map in which multiple modeling conditions are plotted, using the predicted defects and the collected defect detection data. The database 715 includes the process map database 220 in FIG. 2A, and further serves as a modeling data storage unit for the target layered object and a defect repair data storage unit.

[0058] The additive manufacturing apparatus 720 has a manufacturing control unit 721 that controls the additive manufacturing unit 722 to manufacture an additive manufacturing object in accordance with instructions from the information processing device 710, and the additive manufacturing unit 722 that manufactures an additive manufacturing object in accordance with the control of the manufacturing control unit 721. The surface image capturing unit 730 captures images of the surface of the additive manufacturing object before and after melting of each layer. The surface image capturing unit 730 may be an X-ray camera or the like that can capture images of the internal structure of the additive manufacturing object without processing such as cutting the additive manufacturing object.

[0059] (Additive Manufacturing Department) FIG. 7B is a diagram showing the configuration of the additive manufacturing unit 722 according to this embodiment.

[0060] The additive manufacturing unit 771 in FIG. 7B is an additive manufacturing unit of the powder bed method using the laser melting method, and is called selective laser melting (SLM). On the other hand, the additive manufacturing unit 772 in FIG. 7B is an additive manufacturing unit of the powder bed method using the electron beam melting method, and is called electron beam melting (EBM). In this embodiment, electron beam additive manufacturing (EBM) is described as an example, but the present invention can be applied to selective laser melting (SLM) and has the same effect. Note that the configurations of the additive manufacturing units 771 and 772 shown in FIG. 7B are only examples and are not limited thereto. For example, the powder supply of the additive manufacturing unit 772 may be configured like the additive manufacturing unit 771.

[0061] According to this embodiment, by automating the process from optimizing the molding conditions for metal parts of the material and shape required by the user to part molding, even users without specialized knowledge can develop defect-free parts efficiently and in a short time.

[0062] [Other embodiments] Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the technical scope of the present invention. In addition, systems or devices that combine the separate features included in each embodiment in any way are also included in the technical scope of the present invention.

[0063] The present invention may be applied to a system consisting of multiple devices, or to a single device. Furthermore, the present invention is also applicable to a case where an information processing program for implementing the functions of the embodiments is supplied to a system or device and executed by a built-in processor. In order to implement the functions of the present invention with a computer, a program installed on a computer, a medium storing the program, a server for downloading the program, and a processor for executing the program are also included in the technical scope of the present invention. In particular, at least a non-transitory computer readable medium storing a program for causing a computer to execute the processing steps included in the above-mentioned embodiments is included in the technical scope of the present invention.

Claims

1. A three-dimensional additive manufacturing system that uses electron beam additive manufacturing (EBM) or laser additive manufacturing (LBM) to melt powder at a predetermined position in a powder bed to manufacture a three-dimensional additive object, a defect prediction unit that predicts defects that will occur in the object based on a combination of a plurality of design data and a plurality of modeling conditions used when additively manufacturing the object; and a defect detection data collection unit that monitors before and after melting of each layer of the object being modeled and collects defect detection data for detecting defects; a process map generating unit that generates a process map in which the plurality of modeling conditions are plotted by using the predicted defects and the collected defect detection data; and an additive manufacturing unit that additively manufactures a model while adjusting the plurality of modeling conditions in accordance with the process map; a defect repair data collection unit that monitors before and after melting of each layer of the object being modeled while adjusting the plurality of modeling conditions in accordance with the process map, repairs defects detected in the object, and collects defect repair data for defect repair; An additive manufacturing system equipped with

2. An additive manufacturing system as described in claim 1, further comprising a defect repair data storage unit that uses the collected defect repair data and the repair results of the defect repair to store defects detected in the object being manufactured in correspondence with the defect repair data.

3. The additive manufacturing system of claim 1 , wherein the defect detection data and the defect repair data include surface texture data.

4. The additive manufacturing system according to claim 1 , wherein the plurality of design data are classified according to the shape of the object.

5. The method further comprises a defect repair data storage unit that uses the collected defect repair data and a repair result of the defect repair to store a defect detected in the object being modeled in association with the defect repair data, The additive manufacturing system according to claim 1 , wherein the additive manufacturing unit repairs defects detected in subsequent models in accordance with the defect repair data stored in the defect repair data storage unit.

Images