Molding history monitoring device, molding object manufacturing system, and molding history monitoring method

Abstract

The present invention includes a shape sensor that acquires the shape profile of a pre-existing weld bead along its extension direction, and a control unit comprising: a welding information acquisition mechanism that acquires welding information during the formation of an adjacent weld bead when an adjacent weld bead is formed at a position adjacent to the pre-existing weld bead; and a defect candidate extraction mechanism that infers angular feature parts with root angles above a threshold in the pre-existing weld bead based on the shape profile, and infers welding feature parts of the welding information based on the welding information, and establishes a correlation between the welding feature parts and the angular feature parts corresponding to the angular feature parts to extract them as defect candidates.

Description

Technical Field

[0001] This invention relates to a device for monitoring the design history, a manufacturing system for design objects, and a method for monitoring the design history. Background Technology

[0002] Patent document 1 discloses the following technology: In welding construction, multiple vision sensors are used to obtain multiple pieces of information related to the protruding length of the welding wire, the shape of the molten pool, and the behavior of the welding personnel, and the quality of the welding construction is determined based on this information.

[0003] In addition, Patent Document 2 discloses the following technology: the quality of butt welding is determined based on the shape of the weld bead produced when butt welding a metal plate material based on high energy density welding, based on the cross-sectional shape obtained by a cross-sectional reading sensor that obtains the two-dimensional cross-sectional shape of the weld bead formed in the groove.

[0004] Existing technical documents

[0005] Patent documents

[0006] Patent Document 1: Japanese Patent Application Publication No. 2008-110388

[0007] Patent Document 2: Japanese Patent Application Publication No. 2008-212944 Summary of the Invention

[0008] The problem that the invention aims to solve

[0009] However, in stacked molding, where heat sources such as lasers or electric arcs are used to melt metal powder or wire and form weld beads to shape the object, heat accumulates in the object due to the repeated stacking of weld beads. Furthermore, in stacked molding, instead of a flat base with walls supported from both sides like a bevel, the next weld bead is formed along an existing weld bead as the base. Therefore, it is difficult to apply the techniques of Patent Documents 1 and 2, which determine the quality of welding to a flat base with walls supported from both sides like a bevel, to stacked molding.

[0010] Therefore, the object of the present invention is to provide a molding history monitoring device, a molding manufacturing system, and a molding history monitoring method that can reliably estimate defects in moldings formed by repeated fusion welds.

[0011] Solution for solving the problem

[0012] The present invention is composed of the following structure.

[0013] (1) A modeling history monitoring device, which estimates defects based on the history information of the modeling process of a modeling object formed by melting and solidifying filler material using a welding torch, wherein,

[0014] The appearance history monitoring device has the following features:

[0015] A shape profile acquisition mechanism acquires the shape profile of a pre-set weld bead along its extension direction.

[0016] A welding information acquisition mechanism acquires welding information during the formation of an adjacent weld bead at a position adjacent to the existing weld bead; and

[0017] The defect candidate extraction mechanism infers angular features with root angles above a threshold in the established weld bead based on the shape profile, and infers welding features of the welding information based on the welding information, and establishes an association between the welding features and the angular features to extract them as defect candidates.

[0018] (2) A system for manufacturing a shaped object, wherein, while moving a welding torch, a weld bead is formed by melting and solidifying a filler material using the welding torch to shape the object, wherein...

[0019] The manufacturing system for the object is equipped with the object history monitoring device described in (1) above.

[0020] (3) A method for monitoring the molding history of an object, which infers defects based on the molding history information of the object by forming multiple weld beads formed by melting and solidifying filler material using a welding torch, wherein,

[0021] The method for monitoring the appearance resume includes:

[0022] Shape profile acquisition process: Obtain the shape profile of the pre-set weld bead along the extension direction;

[0023] Welding information acquisition and processing: when an adjacent weld bead is formed at a position adjacent to the existing weld bead, welding information during the formation of the adjacent weld bead is acquired; and

[0024] The defect candidate extraction process involves inferring angular features with root angles above a threshold in the established weld bead based on the shape contour, and inferring welding features based on the welding information. The welding features corresponding to the angular features are then associated with each other and extracted as defect candidates.

[0025] Invention Effects

[0026] This invention can reliably predict defects in shapes formed by repeated fusion welds. Attached Figure Description

[0027] Figure 1 This is a schematic outline structural diagram of the manufacturing system according to an embodiment of the present invention.

[0028] Figure 2 This is a diagram showing the overlapping weld beads formed. Figure 2 (A) is a schematic cross-sectional view showing the case where the root angle of the pre-deposited weld bead is small. Figure 2 (B) is a schematic cross-sectional view showing the case where the root angle of the pre-deposited weld bead is large.

[0029] Figure 3 This is a perspective view showing the formation of adjacent weld beads along an existing weld bead.

[0030] Figure 4 This is an explanatory diagram that schematically shows the root angle of a pre-set weld bead and the curve of the welding voltage of the adjacent weld bead.

[0031] Figure 5 It is a perspective view showing the formation of a weld bead along an existing weld bead.

[0032] Figure 6 It is a three-dimensional diagram illustrating the molten pool formed in the weld bead.

[0033] Figure 7 It is a diagram illustrating the changes in the shape of the molten pool. Figure 7 (A) is a three-dimensional diagram showing the formation of a bulge. Figure 7 (B) is a three-dimensional diagram showing the state with a recessed portion. Detailed Implementation

[0034] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0035] Figure 1 This is a schematic outline diagram of a manufacturing system 100 for a shaped object equipped with the styling history monitoring device of the present invention.

[0036] The manufacturing system 100 for this structure includes a welding robot 11, a robot controller 13, a filler material supply unit 15, a welding power source 19, and a control unit 21.

[0037] The welding robot 11 is a jointed robot, with a welding torch 23 supported on its front axis. The position and orientation of the welding torch 23 can be arbitrarily set in three dimensions within the range of the robot arm's degrees of freedom. The welding torch 23 holds the filler material (welding wire) M continuously supplied from the filler material supply unit 15 in a state where it protrudes from the front end of the torch. A shape sensor 25 is provided on the front axis of the welding robot 11 together with the welding torch 23.

[0038] The welding torch 23 has a protective nozzle (not shown) and supplies shielding gas to the welding area from the protective nozzle. As an arc welding method, it can be any of the consumable electrode types such as covered arc welding or carbon dioxide gas arc welding, or the non-consumable electrode types such as TIG welding or plasma arc welding, and the appropriate type is selected according to the shape being produced.

[0039] For example, in the case of a consumable electrode type, a conductive tip is disposed inside the protective nozzle, and the filler material M supplied with molten current is held in the conductive tip. While holding the filler material M, the welding torch 23 generates an arc from the tip of the filler material M under a protective gas atmosphere. The filler material M is fed into the welding torch 23 via a feed mechanism (not shown) mounted on a robotic arm or the like. Furthermore, as the continuously fed filler material M melts and solidifies while the welding torch 23 is moved, a weld bead 29, which is the molten solidified filler material M, is formed on the base plate 27.

[0040] The base plate 27 is made of metal plates such as steel plates, and is generally larger than the bottom surface (lowest surface) of the object W. The base plate 27 is not limited to a plate shape, and can also be a base of other shapes such as a block or a rod.

[0041] The heat source for melting the filler material M is not limited to the electric arc described above. For example, other heat sources based on methods such as heating with both electric arc and laser, heating with plasma, or heating with electron beam or laser can also be used. When heating with electron beam or laser, the amount of heat can be controlled more precisely, and the state of the weld bead can be maintained more appropriately, which helps to further improve the quality of the molded object.

[0042] The filler material M can use all commercially available welding wires. For example, it can use welding wires specified such as solid welding wires for MAG welding and MIG welding of mild steel, high-tensile steel and low-temperature steel (JIS Z 3312), and flux-cored welding wires for arc welding of mild steel, high-tensile steel and low-temperature steel (JIS Z 3313).

[0043] As filler material M, active metals such as titanium can also be used. In this case, to avoid oxidation and nitriding caused by reaction with the atmosphere during welding, the weld area needs to be in a protective gas atmosphere.

[0044] A shape sensor 25 is mounted on the welding torch 23 and moves with it. This shape sensor 25 measures the shape of the portion of the substrate that forms the weld bead 29. For example, a laser sensor that uses the reflected light of an emitted laser beam as height data can be used as the shape sensor 25. It should be noted that a three-dimensional shape measurement camera can also be used as the shape sensor 25.

[0045] The robot controller 13 receives instructions from the control unit 21, drives the various parts of the welding robot 11, and controls the output of the welding power supply 19 as needed.

[0046] The control unit 21 is a computer device equipped with a CPU, memory, storage, etc., and executes pre-prepared driver programs or driver programs created under desired conditions to drive various parts such as the welding robot 11. Thus, according to the driver programs, the welding torch 23 is moved, and multiple layers of weld beads 29 are deposited on the base plate 27 based on the created layering plan, thereby forming a multi-layered structure W. Furthermore, a database 17 is connected to the control unit 21. This database 17 stores and accumulates history information including defect candidates extracted by the control unit 21 (described later).

[0047] However, the flowability of the weld bead 29 formed during the shaping of the object W is greatly affected by factors such as the material and the melting conditions, resulting in variations in one or both of its width and height. Thus, during the shaping of the object W, when weld beads 29 are formed adjacent to existing weld beads 29, defects may occur at the overlapping portions of adjacent weld beads 29.

[0048] Figure 2 This is a diagram showing the overlapping weld beads formed. Figure 2 (A) is a schematic cross-sectional view showing the case where the root angle of the pre-deposited weld bead 29A is small. Figure 2 (B) is a schematic cross-sectional view showing the case where the root angle of the pre-deposited weld bead 29A is large.

[0049] like Figure 2 As shown in (A), when the root angle θ of the existing weld bead 29A is small, the defect generation rate between the weld beads 29A and 29B is lower when they overlap to form adjacent weld beads 29B. In contrast, as... Figure 2 As shown in (B), when the root angle θ of the existing weld bead 29A is large, when an adjacent weld bead 29B is formed by overlapping with the existing weld bead 29A, the molten metal does not flow sufficiently to the root of the existing weld bead 29A, resulting in a higher rate of defects between these weld beads 29A and 29B. For example, when the root angle θ of the existing weld bead 29A is 40° or more, the rate of defects caused by poor fusion between weld beads 29A and 29B is high.

[0050] Furthermore, the defect generation rate in the overlapping portion of adjacent weld beads 29 varies depending on the root angle θ of the established weld bead 29A, and also on the welding conditions such as welding voltage, welding current, feed rate of filler material M, feed resistance of filler material M, shielding gas flow rate, and flow condition of the molten pool when forming the weld bead 29.

[0051] Therefore, the manufacturing system 100 of this embodiment has a pattern history monitoring device for presuming defects in adjacent weld beads 29A, 29B as described above.

[0052] The shape history monitoring device extracts defect candidates based on the measurement results from the shape sensor (profile acquisition mechanism) 25, which is also installed on the welding torch 23, and information on the welding status, such as welding voltage, welding current, feed rate of filler material M, and flow condition of the molten pool, when the weld bead 29 is formed. The control unit 21 includes a welding information acquisition mechanism and a defect candidate extraction mechanism. The welding information acquisition mechanism acquires welding information, and the defect candidate extraction mechanism extracts defect candidates.

[0053] Furthermore, the resume information of candidates with this defect will be saved to database 17.

[0054] Next, the extraction of defect candidates by the design history monitoring device will be explained. Here, the case where the welding voltage when forming a weld bead 29B adjacent to the existing weld bead 29A is used as welding information will be explained.

[0055] (Shape and outline acquisition processing)

[0056] Figure 3 This is a perspective view showing the formation of an adjacent weld bead 29B along the existing weld bead 29A.

[0057] like Figure 3 As shown, weld beads 29B are formed while the welding torch 23 is moved at a position adjacent to the existing weld bead 29A. At this time, the shape of the existing weld bead 29A is measured by the shape sensor 25 in front of the welding torch 23, and the shape profile of the weld bead 29A along the extension direction is obtained.

[0058] (Welding Information Acquisition and Processing)

[0059] Using the welding information acquisition mechanism of the control unit 21, when forming a weld bead 29B adjacent to an existing weld bead 29A, the welding voltage during the formation of the weld bead 29B is acquired as welding information. It should be noted that the welding information acquisition mechanism of the control unit 21 acquires the output of the welding power source 19, for example, by monitoring.

[0060] (Defect candidate extraction and processing)

[0061] The defect candidate extraction mechanism of the control unit 21 is used to extract defect candidates.

[0062] Specifically, firstly, based on the shape profile of the established weld bead 29A obtained by the shape sensor 25, the root angle θ of the open side of the established weld bead 29A is inferred, and the portion of the root angle θ that is above a predetermined threshold is inferred as the angle feature portion Rc. The threshold for the root angle θ is, for example, 40°, which is prone to defects.

[0063] Next, based on welding information derived from the welding voltage obtained during the formation of weld beads 29B formed at adjacent locations, the welding characteristic Wc of the welding information is inferred. This welding characteristic Wc represents the portion where the welding voltage changes significantly. For example, if the waveform of the welding voltage becomes abnormally turbulent, arc disturbances or interruptions may occur, potentially having an undesirable impact on the flowability of the molten metal. Therefore, by setting thresholds for instantaneous fluctuations, the slope of these fluctuations, etc., based on past abnormal waveforms, states with a high probability of defect occurrence are extracted as the welding characteristic Wc.

[0064] Furthermore, the inferred angular feature Rc is compared with the welding feature Wc, and the welding feature Wc corresponding to the angular feature Rc is associated with the angular feature Rc and extracted as a defect candidate F.

[0065] As for the historical information obtained in order to extract defect candidate F, for example, information recorded over time (according to position) by establishing the correspondence between coordinates X, Y, Z, root angle θ and welding voltage V when forming weld bead 29B.

[0066] Figure 4 This is an explanatory diagram that schematically shows the root angle θ of the pre-established weld bead 29A and the curve of the welding voltage of the adjacent weld bead 29B.

[0067] like Figure 4 As shown, the defect candidate extraction mechanism of the control unit 21 designates the root angle θ of the pre-set weld bead 29A as an angle feature Rc when it is above a threshold value, and designates the weld feature Wc when the welding voltage of the adjacent weld bead 29B changes from a stable value. Furthermore, the portion of the angle feature Rc that generates the weld feature Wc is extracted as a defect candidate F, and the history information containing the defect candidate F (information on each position where coordinates X, Y, Z, root angle θ, and welding voltage V are respectively established) is saved to the database 17.

[0068] As explained above, according to the styling history monitoring device and styling history monitoring method of this embodiment, by taking the case where there is a feature in the welding information of the adjacent weld bead 29B at the position corresponding to the root angle θ of the established weld bead 29A that is above the threshold angle feature Rc as a defect candidate F, a defect candidate F with high reliability can be extracted.

[0069] Furthermore, according to the manufacturing system 100 for shaped objects equipped with a shaped history monitoring device, when shaping the object W, it is easy to identify the possible defects that may occur in the object W. Therefore, after the object W is shaped, the defects in the object W can be quickly repaired.

[0070] Furthermore, the defect candidate extraction mechanism of the control unit 21 preferably performs a frequency calculation process to determine the occurrence frequency of defect candidate F during the defect candidate extraction process. In this way, by calculating the occurrence frequency of defect candidate F, the parts of the molded object W that require precise inspection can be estimated based on the occurrence frequency of defect candidate F. As for calculating the occurrence frequency of defect candidate F, indicators such as the number of times it occurs during the formation of a weld bead 29 or the duration of the defect candidate's occurrence can be set. Additionally, by referring to information from the track plan, and grouping cases into molded wall sections, filled wall sections, and overhanging sections, the occurrence frequency of defect candidate F can also be calculated in each group.

[0071] Furthermore, the defect candidate extraction mechanism of the control unit 21 preferably estimates the defect size in the molded object W based on the occurrence frequency of the defect candidate F, the occurrence time of the defect candidate F, and the moving speed of the welding torch 23 during the defect candidate extraction process. In this way, by estimating the defect size in the molded object W, defects generated in the molded object W can be easily identified without performing complex inspections such as destructive testing or ultrasonic testing. Additionally, if the occurrence frequency of the defect candidate F persists for a certain duration, it can be estimated that a long, thin defect has occurred; if it is extremely short, it can be estimated that a small defect has occurred or that a defect candidate has appeared due to noise. Therefore, if the duration of the occurrence frequency of the defect candidate F is above a preset threshold, the history information containing the defect candidate F can be saved as defect information.

[0072] It should be noted that in the above embodiments, the example illustrates the case where welding voltage is used as welding information to infer the welding feature Wc, but the welding information is not limited to welding voltage. As welding information, at least one of welding voltage, welding current, feed rate of filler material M, feed resistance of filler material M, shielding gas flow rate, and flow condition of the molten pool can be used. Alternatively, these welding voltage, welding current, feed rate of filler material M, and flow condition of the molten pool can be combined.

[0073] Here, we will explain the flow conditions of the molten pool used as welding information.

[0074] Figure 5 This is a perspective view showing the formation of weld bead 29B along the established weld bead 29A.

[0075] like Figure 5 As shown, when the flow condition of the molten pool is used as welding information, a camera 26 that captures a portion of the molten pool P is mounted on the welding torch 23 together with a shape sensor 25.

[0076] Furthermore, when forming a weld bead 29B adjacent to the existing weld bead 29A, the shape of the existing weld bead 29A is measured using a shape sensor 25, and the molten pool P of the weld bead 29B is captured by a camera 26. This captured data is obtained as welding information. Based on the welding information derived from this captured data, for example, the portion where the shape of the molten pool P changes significantly is inferred to be the welding feature Wc. It should be noted that data quantified as the area of ​​the molten pool P can also be used as the captured data.

[0077] Figure 6 This is a three-dimensional diagram illustrating the molten pool P formed in weld bead 29B. Figure 7 This is a diagram illustrating the changes in the shape of the molten pool P. Figure 7 (A) is a three-dimensional diagram of the state in which the bulge Pb is formed. Figure 7 (B) is a three-dimensional diagram showing the state in which the recessed portion Pk is formed.

[0078] Figure 6 The diagram shows a molten pool P of typical shape. The molten pool P exists relative to... Figure 6 This illustrates a typical case of significant shape variation. For example, the molten pool P... Figure 7 As shown in (A), there is a case where a portion expands and a bulge Pb is formed. Additionally, as... Figure 7 As shown in (B), there is a partial depression that results in a depression Pk. Furthermore, the welding information acquisition mechanism of the control unit 21 infers the portion of the weld pool P that has bulged out and depressed out in the weld bead 29B as the welding feature portion Wc.

[0079] Furthermore, the defect candidate extraction mechanism of the control unit 21 compares the angle feature Rc inferred from the shape profile of the pre-set weld bead 29A obtained by the shape sensor 25 with the welding feature Wc inferred from the welding information composed of the shooting data of the camera 26, and establishes a correlation between the welding feature Wc corresponding to the angle feature Rc and the angle feature Rc to extract the defect candidate F.

[0080] In this way, when the flow condition of the molten pool P is used as welding information, by establishing a correlation between the welding feature Wc and the angle feature Rc in the welding information of the obtained flow condition of the molten pool P as a defect candidate F, it is also possible to extract a defect candidate with high reliability.

[0081] It should be noted that the flow condition of the molten pool P can also be the temperature, brightness, etc. In this case, the temperature sensor for detecting the temperature of the molten pool P and the brightness sensor for detecting the brightness of the molten pool P are also installed on the welding torch 23.

[0082] Thus, the present invention is not limited to the above-described embodiments. Combining the various structures of the embodiments with each other, as well as making changes and applications based on the description and well-known techniques by those skilled in the art, are also intended by the present invention and are included within the scope of the claimed protection.

[0083] As stated above, the following matters are disclosed in this specification.

[0084] (1) A modeling history monitoring device, which estimates defects based on the history information of the modeling process of a modeling object formed by melting and solidifying filler material using a welding torch, wherein,

[0085] The appearance history monitoring device has the following features:

[0086] A shape profile acquisition mechanism acquires the shape profile of a pre-set weld bead along its extension direction.

[0087] A welding information acquisition mechanism acquires welding information during the formation of an adjacent weld bead at a position adjacent to the existing weld bead; and

[0088] The defect candidate extraction mechanism infers angular features with root angles above a threshold in the established weld bead based on the shape profile, and infers welding features based on the welding information, and extracts the welding features corresponding to the angular features and the angular features as defect candidates by establishing an association between them.

[0089] According to this design history monitoring device, welding features corresponding to angle features with root angles of 100° or higher than a threshold value in an existing weld bead are linked and extracted as defect candidates. When an adjacent weld bead is formed next to an existing weld bead, the possibility of defects due to insufficient flow increases when the root angle of the existing weld bead is large. Therefore, by using features present in the welding information of adjacent weld beads at positions corresponding to angle features with root angles of 100° or higher than a threshold value in an existing weld bead as defect candidates, highly reliable defect candidates can be extracted.

[0090] (2) According to the modeling history monitoring device described in (1), wherein,

[0091] The welding information acquisition mechanism acquires at least one of the following as welding information: welding voltage, welding current, feed rate of filler material, feed resistance of filler material, shielding gas flow rate, and flow condition of the molten pool.

[0092] According to this design history monitoring device, by establishing a correlation between the welding feature and the angular feature in at least one of the welding information obtained from welding voltage, welding current, feed rate of filler material, and flow condition of molten pool, and using them as defect candidates, it is possible to extract highly reliable defect candidates.

[0093] (3) The modeling history monitoring device according to (1) or (2), wherein,

[0094] The defect candidate extraction mechanism performs a frequency calculation process to determine the occurrence frequency of the defect candidates.

[0095] Based on the design history monitoring device, by calculating the frequency of occurrence of defect candidates, the parts of the design that should be carefully inspected can be estimated based on the frequency of occurrence.

[0096] (4) The modeling history monitoring device according to (3), wherein,

[0097] The defect candidate extraction mechanism estimates the defect size in the object based on the frequency of occurrence of the defect candidates.

[0098] According to this design history monitoring device, defects in the design can be easily identified without performing complex inspections such as destructive testing or ultrasonic testing, based on the defect size estimated from the frequency of defect candidates. Furthermore, if the frequency of defect candidates persists for a certain duration, it can be inferred that a long, thin defect has occurred; if it is extremely short, it can be inferred that a tiny defect or a defect candidate has appeared due to noise.

[0099] (5) A system for manufacturing a shaped object, wherein, while moving a welding torch, a weld bead is formed by melting and solidifying a filler material using the welding torch to shape the object, wherein...

[0100] The manufacturing system for the object has any one of the object history monitoring devices (1) to (4).

[0101] According to the manufacturing system of this object, it is easy to identify potential defects in the object during the shaping process. Therefore, defects can be quickly repaired after the object is shaped.

[0102] (6) A method for monitoring the molding history of an object, which infers defects based on the molding history information of the object by forming multiple weld beads formed by melting and solidifying filler material using a welding torch, wherein,

[0103] The method for monitoring the appearance resume includes:

[0104] Shape profile acquisition process: Obtain the shape profile of the pre-set weld bead along the extension direction;

[0105] Welding information acquisition and processing: when an adjacent weld bead is formed at a position adjacent to the existing weld bead, welding information during the formation of the adjacent weld bead is acquired; and

[0106] The defect candidate extraction process involves inferring angular features with root angles above a threshold in the established weld bead based on the shape contour, and inferring welding features based on the welding information. The welding features corresponding to the angular features are then associated with each other and extracted as defect candidates.

[0107] According to this pattern history monitoring method, welding features corresponding to angle features with root angles exceeding a threshold of an existing weld bead are linked and extracted as defect candidates. When adjacent weld beads are formed adjacent to existing weld beads, the possibility of defects due to insufficient flow increases when the root angle of the existing weld bead is large. Therefore, by using features present in the welding information of adjacent weld beads at positions corresponding to angle features with root angles exceeding a threshold of an existing weld bead as defect candidates, highly reliable defect candidates can be extracted.

[0108] (7) According to the modeling history monitoring method described in (6), wherein,

[0109] In the welding information acquisition and processing, at least one of the following is obtained as the welding information: welding voltage, welding current, feed rate of filler material, feed resistance of filler material, shielding gas flow rate, and flow condition of molten pool.

[0110] According to this pattern history monitoring method, by establishing a correlation between the welding feature and the angular feature in at least one of the welding information obtained from welding voltage, welding current, feed rate of filler material, feed resistance of filler material, shielding gas flow rate, and flow condition of molten pool, and using them as defect candidates, it is possible to extract highly reliable defect candidates.

[0111] (8) The styling history monitoring method according to (6) or (7), wherein,

[0112] In the defect candidate extraction process, the occurrence frequency of the defect candidate is calculated.

[0113] According to this design history monitoring method, by calculating the frequency of occurrence of defect candidates, it is possible to infer the parts of the design that should be carefully inspected based on the frequency of occurrence.

[0114] (9) According to the modeling history monitoring method described in (8), wherein,

[0115] In the defect candidate extraction process, the defect size in the object is estimated based on the occurrence frequency of the defect candidates.

[0116] According to this design history monitoring method, defects in the design can be easily identified without performing complex inspections such as destructive testing or ultrasonic testing, based on the defect size estimated from the frequency of defect candidates. Furthermore, if the frequency of defect candidates persists for a certain duration, it can be inferred that a slender defect has occurred; if it is extremely short, it can be inferred that a tiny defect or a defect candidate has appeared due to noise.

[0117] It should be noted that this application is based on Japanese patent application (Japanese Patent Application No. 2021-13390) filed on January 29, 2021, the contents of which are referenced in this application.

[0118] Explanation of reference numerals in the attached figures

[0119] 21. Control Department (Welding Information Acquisition Agency, Defect Candidate Extraction Agency)

[0120] 23 welding torches

[0121] 25. Shape sensor (shape profile acquisition mechanism)

[0122] 29, 29A, 29B weld beads

[0123] 100 Manufacturing System (Styling History Monitoring Device)

[0124] M filling material

[0125] F Defect Candidate

[0126] P molten pool

[0127] W-shaped object

[0128] Rc Angle Feature Section

[0129] Wc welding feature area

[0130] θ is the root angle.

Claims

1. A modeling history monitoring device, which estimates defects based on the modeling history information of a model formed by melting and solidifying filler material using a welding torch, wherein, The appearance history monitoring device has the following features: A shape profile acquisition mechanism acquires the shape profile of a pre-set weld bead along its extension direction. The welding information acquisition mechanism acquires welding information during the formation of the adjacent weld bead when it forms an adjacent weld bead at a position adjacent to the existing weld bead. as well as The defect candidate extraction mechanism infers angular features with root angles above a threshold in the existing weld bead based on the shape profile, and infers welding features in the welding information that indicate a high probability of defect occurrence. It then establishes a correlation between the welding features corresponding to the angular features and extracts them as defect candidates. The welding information acquisition mechanism acquires at least one of the following as welding information: welding voltage, welding current, feed rate of filler material, feed resistance of filler material, shielding gas flow rate, and flow condition of the molten pool.

2. The modeling history monitoring device according to claim 1, wherein, The defect candidate extraction mechanism performs a frequency calculation process to determine the occurrence frequency of the defect candidates.

3. The modeling history monitoring device according to claim 2, wherein, The defect candidate extraction mechanism estimates the defect size in the object based on the frequency of occurrence of the defect candidates.

4. A system for manufacturing a shaped object, wherein, while moving a welding torch, a weld bead is formed by melting and solidifying a filler material using the welding torch to shape the object, wherein... The manufacturing system for the object is equipped with a model history monitoring device as described in any one of claims 1 to 3.

5. A method for monitoring the molding history of an object, which infers defects based on the molding history information of the object formed by melting and solidifying multiple weld beads using a welding torch, wherein, The method for monitoring the appearance resume includes: Shape profile acquisition process: Obtain the shape profile of the pre-set weld bead along the extension direction; Welding information acquisition and processing: when an adjacent weld bead is formed at a position adjacent to the existing weld bead, welding information during the formation of the adjacent weld bead is acquired; and The defect candidate extraction process involves inferring angular features with root angles exceeding a threshold in the existing weld bead based on the shape contour, and inferring welding features in the welding information that indicate a high probability of defect occurrence. The welding features corresponding to the angular features are then associated with each other and extracted as defect candidates. In the welding information acquisition and processing, at least one of the following is obtained as the welding information: welding voltage, welding current, feed rate of filler material, feed resistance of filler material, shielding gas flow rate, and flow condition of molten pool.

6. The method for monitoring the design history according to claim 5, wherein, In the defect candidate extraction process, the occurrence frequency of the defect candidate is calculated.

7. The method for monitoring the design history according to claim 6, wherein, In the defect candidate extraction process, the defect size in the object is estimated based on the occurrence frequency of the defect candidates.

Images