Defect monitoring device, defect monitoring method, welding support system, and welding system

Abstract

To provide a defect monitoring device, a defect monitoring method, a welding support system and a welding system which specifies a portion having the risk of defect occurrence during welding and prevents the defect from occurring on the portion during welding.SOLUTION: A defect monitoring device 170 includes a shape profile acquisition part 33 which acquires a shape profile of an existing deposited bead, a feature quantity extraction part 35 which extracts a feature quantity of concave shape formed by a plurality of existing deposited beads contained in the shape profile, a defect position specifying part 37 specifying a defect candidacy portion on which the occurrence of the welding defect is estimated in accordance with the extracted feature quantity and a control part 11 which, when a welding device forms a deposited bead anew, causes the shape profile acquisition part 33 to update the shape profile and repeatedly causes the feature quantity extraction part 35 to execute extraction of feature quantity and causes the defect position specification part 37 to execute specification of the defect candidacy portion.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a defect monitoring device, a defect monitoring method, a welding support system, and a welding system.

Background Art

[0002] In arc welding, a technique is known for detecting defects generated in a welded structure and determining whether proper welding work has been performed. For example, Patent Document 1 discloses a technique for acquiring a plurality of pieces of information regarding the behavior of a welder, the shape of a molten pool, and the wire protruding length during semi-automatic welding, and determining the quality of welding work (Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] For example, when forming a multi-layered shaped object by laminating weld beads formed by melting and solidifying a filler material, depending on the shape and arrangement of the weld beads, defects such as cavities may occur inside the shaped object after shaping. For example, when filling the inner region of a wall portion formed by laminating weld beads in a frame shape with weld beads, the target position of the weld bead may be narrowed by adjacent existing weld beads, or the inclination of the side surface of the existing weld bead may be large, resulting in minute depressions between the base and the weld bead. Then, the molten metal of the weld bead may not be able to flow into the entire planned filling region of the target position, and local cavities (unwelded portions) may occur. This can occur not only in laminated shaping but also in ordinary welding.

[0005] To identify defects such as unwelded areas, non-destructive testing can be performed along the entire length of the weld bead after welding, or the location of the defect can be identified by cutting the welded structure. Another possible method is to compare the welded structure with a production log recording the welding process. However, all of these methods are time-consuming and cumbersome, making them impractical. Furthermore, if a structure that is to be manufactured is cut, restoring it to its original state would be extremely time-consuming and laborious.

[0006] Furthermore, it is desirable to be able to evaluate the degree to which defects in unwelded areas affect the quality of the structure. This makes it easier to decide whether to repair defects or continue welding, ultimately leading to a better balance between productivity and quality. One example of a specific evaluation indicator is defect size, but for structures with complex shapes, it is difficult to adequately evaluate defect size using non-destructive testing methods such as ultrasonic testing.

[0007] Therefore, the present invention aims to provide a defect monitoring device, a defect monitoring method, a welding support system, and a welding system that identify areas where defects are likely to occur during welding and prevent the occurrence of defects. [Means for solving the problem]

[0008] The present invention consists of the following configuration. (1) A defect monitoring device that predicts the occurrence of welding defects from historical information of the welding bead formation by the welding device when a structure is formed by stacking multiple welding beads formed by melting and solidifying filler material by a welding device, A shape profile acquisition unit that acquires the shape profile of an existing welding bead, A feature extraction unit that extracts feature quantities of recess shapes formed by a plurality of existing welding beads included in the shape profile, A defect location identification unit identifies candidate defect locations where welding defects are expected to occur, based on the extracted feature quantities. When the welding apparatus forms a new weld bead, the control unit causes the shape profile acquisition unit to update the shape profile, and repeatedly performs the extraction of feature quantities by the feature quantity extraction unit and the identification of candidate defect locations by the defect location identification unit. A defect monitoring device equipped with the following features. (2) A defect monitoring method for predicting the occurrence of welding defects from historical information of the welding bead formation by the welding apparatus when a structure is formed by stacking multiple welding beads, which are formed by melting and solidifying filler material with a welding apparatus, The process of obtaining the shape profile of the existing weld bead, A step of extracting characteristic quantities of recess shapes formed by a plurality of existing weld beads included in the shape profile, A step of identifying candidate defect locations where welding defects are expected to occur according to the extracted feature quantities, The welding apparatus updates the shape profile when it newly forms the welding bead, and repeatedly performs the extraction of feature quantities and the identification of candidate defect locations. A defect monitoring method comprising the following features. (3) The defect monitoring device described in (1), An instruction information generating device that generates instruction information to improve the welding defects at the detected candidate defect locations, A welding support system equipped with (4) The welding support system described in (3), The welding apparatus that forms the welding bead, A bead processing device for processing potential defect locations in the welded bead of the formed structure, A welding system equipped with [the following features]. [Effects of the Invention]

[0009] According to the present invention, it is possible to prevent defects from occurring by identifying areas where defects are likely to occur during additive manufacturing. [Brief explanation of the drawing]

[0010] [Figure 1]FIG. 1 is an overall configuration diagram of a welding system. [Figure 2] FIG. 2 is a schematic functional block diagram of a control unit. [Figure 3] FIG. 3 is a flowchart showing a procedure for monitoring the occurrence of defects. [Figure 4] FIG. 4 is a schematic diagram showing a welding torch and a shape detection unit, and a bead formation track of a weld bead. [Figure 5A] FIG. 5A is a diagram schematically showing a cross-sectional shape of existing weld beads formed adjacent to each other, and is a cross-sectional view when the existing weld beads B are properly spaced apart. [Figure 5B] FIG. 5B is a diagram schematically showing a cross-sectional shape of existing weld beads formed adjacent to each other, and is a cross-sectional view when the existing weld beads are arranged close to each other. [Figure 6] FIG. 6 is an explanatory diagram showing an example of information on defect candidate locations displayed on a display unit. [Figure 7] FIG. 7 is a graph showing an example of determination results of defect candidate locations along a bead formation track. [Figure 8] FIG. 8 is an explanatory diagram showing an example of other feature amounts in a cross-section orthogonal to the bead formation direction of an existing weld bead. [Figure 9] FIG. 9 is an explanatory diagram showing a state in which the interval between the existing weld beads shown in FIG. 8 is changed. [Figure 10] FIG. 10 is an explanatory diagram showing a state in which the interval between the existing weld beads shown in FIG. 8 is changed. [Figure 11] FIG. 11 is an explanatory diagram showing an example of other feature amounts in a cross-section orthogonal to the bead formation direction of an existing weld bead. [Figure 12] FIG. 12 is an explanatory diagram showing an example of feature amounts when an existing weld bead is approximated as a trapezoid in a cross-section orthogonal to the bead formation direction of the existing weld bead. [Figure 13] FIG. 13 is an explanatory diagram showing an example of feature amounts of an existing weld bead in a cross-section orthogonal to the bead formation direction of the existing weld bead. [Figure 14]Figure 14 is an explanatory diagram showing an example of characteristic quantities when a welding bead is formed in the filling portion inside the wall of an existing welding bead in a cross-section perpendicular to the bead formation direction. [Figure 15] Figure 15 is an explanatory diagram showing the pitch of each welding bead when multiple welding beads are stacked in a cross-section perpendicular to the bead formation direction of an existing welding bead. [Figure 16] Figure 16 is an explanatory diagram showing how the shape profile is approximated by a curve. [Figure 17] Figure 17 is an explanatory diagram showing the results of predicting the shape of a fabricated object with existing weld beads layered on top of it. [Figure 18] Figure 18 is an explanatory diagram showing the factors influencing defects. [Figure 19] Figure 19 is a plan view showing the existing weld bead and the planned surface for bead formation of the new weld bead. [Figure 20] Figure 20 is a cross-sectional view along the line XX-XX shown in Figure 19. [Figure 21] Figure 21 is a cross-sectional view along the line XXI-XXI shown in Figure 19. [Figure 22] Figure 22 is a plan view illustrating an example of post-processing. [Figure 23] Figure 23 is a cross-sectional view along the line XXIII-XXIII shown in Figure 22. [Figure 24] Figure 24 is a plan view showing the process of remelting the protruding portion. [Figure 25] Figure 25 is a cross-sectional view along the line XXV-XXV in Figure 24. [Modes for carrying out the invention]

[0011] The configuration examples of the present invention will be described in detail below with reference to the drawings. Here, additive manufacturing, which creates a laminated object by stacking welding beads, will be used as an example, but the present invention can also be applied to general welding methods such as fillet welding and butt welding.

[0012] <Configuration of the welding system> Figure 1 is an overall diagram of the welding system. The welding system 100 comprises a welding apparatus 110, a bead processing apparatus 130, and a welding support system 150. Here, the welding support system 150 is exemplified as being included in the control unit 11 of the welding apparatus 110, but it may be configured separately from the welding apparatus 110.

[0013] (Welding equipment) First, let's explain the configuration of the welding apparatus 110. The welding apparatus 110 comprises a control unit 11, welding robots 13 each connected to the control unit 11, a robot drive unit 15, a filler material supply unit 17, a welding power supply unit 19, a shape detection unit 21, a display unit 23, and an output unit 25.

[0014] The welding robot 13 is an articulated robot, and a welding torch 27 is mounted on its tip axis. The robot drive unit 15 outputs commands to drive the welding robot 13 and arbitrarily sets the position and orientation of the welding torch 27 in three dimensions within the range of freedom of the robot arm. In addition, a continuously supplied filler material (welding wire) M is supported at the tip of the welding torch 27.

[0015] The welding torch 27 is a gas metal arc welding torch having a shield nozzle (not shown) from which shielding gas is supplied. The arc welding method can be either a consumable electrode type such as shielded metal arc welding or carbon dioxide arc welding, or a non-consumable electrode type such as TIG welding or plasma arc welding, and is appropriately selected according to the object (structure) to be fabricated. For example, in the case of a consumable electrode type, a contact tip is placed inside the shield nozzle, and the filler material M to which the melting current is supplied is held by the contact tip. The welding torch 27 holds the filler material M and generates an arc from the tip of the filler material M in a shielding gas atmosphere.

[0016] The filler material supply unit 17 includes a reel 17a around which the filler material M is wound. The filler material M is sent from the filler material supply unit 17 to a feeding mechanism (not shown) attached to a robot arm or the like, and is fed to the welding torch 27 by the feeding mechanism, which feeds the material in forward and reverse directions as needed.

[0017] Any commercially available welding wire can be used as the filler material M. For example, welding wires specified in JIS Z 3312 (Mag welding and MIG welding solid wires for mild steel, high-tensile steel, and low-temperature steel) and JIS Z 3313 (arc welding flux-cored wires for mild steel, high-tensile steel, and low-temperature steel) can be used. Furthermore, depending on the required properties, filler materials M such as aluminum, aluminum alloys, nickel, and nickel-based alloys can also be used.

[0018] The welding power supply unit 19 supplies the welding current and welding voltage to the welding torch 27 in order to generate an arc from the tip of the torch.

[0019] The shape detection unit 21 is provided on or near the tip axis of the welding robot 13, and its measurement area is near the tip of the welding torch 27. The shape detection unit 21 may be other detection means provided at a different location from the welding torch 27.

[0020] In this configuration, the shape detection unit 21 is moved together with the welding torch 27 by the driving of the welding robot 13, and measures the shape of the weld bead B and the base material that forms the weld bead B. For example, a laser sensor that acquires height data from the reflected light of the irradiated laser beam can be used as the shape detection unit 21. Alternatively, other detection means such as a camera for 3D shape measurement may be used as the shape detection unit 21.

[0021] The display unit 23 is a display such as a liquid crystal panel or an organic EL panel, and may be a touch panel with a UI (user interface) input function that allows various processes to be performed by touching it with a finger or pen. The display unit 23 displays information necessary for various operations of the welding system 100. The display unit 23 may also display a 3D image formed by assembling line segments at each cross-section of the laser reflected light measured by the shape detection unit 21, display information on potential defect locations (described later), display welding conditions, etc.

[0022] Furthermore, the control unit 11 may be electrically connected to the output unit 25. Various types of information, such as information to be output to the display unit 23 and the molding program, are output to the output unit 25.

[0023] According to the welding apparatus 110 with the above configuration, a molding program corresponding to the object to be fabricated is transmitted from the control unit 11 to the robot drive unit 15. The molding program consists of numerous command codes and is created based on an appropriate algorithm according to various conditions such as the shape data of the object to be fabricated (CAD data, etc.), material, and heat input.

[0024] The robot drive unit 15 executes the received molding program and drives the welding robot 13, filler material supply unit 17, and welding power supply unit 19, etc., to form a weld bead B according to the molding program. In other words, the robot drive unit 15 drives the welding robot 13 to move the welding torch 27 along the trajectory of the welding torch 27 (bead formation trajectory) set in the molding program. At the same time, it drives the filler material supply unit 17 and the welding power supply unit 19 according to the set welding conditions to melt and solidify the filler material M at the tip of the welding torch 27 by arc. As a result, a weld bead B is formed on the base plate P along the trajectory of the welding torch 27. The weld beads B are formed adjacent to each other, forming a weld bead layer consisting of multiple weld beads B. A subsequent layer of weld beads is stacked on top of this weld bead layer, and so on, to create a three-dimensional object WK of the desired shape.

[0025] (Bead processing equipment) The bead processing device 130 processes the welding bead B by driving a robot arm 53, on which a processing tool 51 is attached to the tip shaft, with a processing drive unit 55. The processing drive unit 55 receives a drive command from the control unit 11 and moves the robot arm 53 to position the processing tool 51 in the desired position and orientation. Then, the processing tool 51 removes a portion of the welding bead B, which will be described later.

[0026] (Control Unit) Figure 2 is a schematic functional block diagram of the control unit 11. In addition to having the function of comprehensively controlling the robot drive unit 15, filler material supply unit 17, welding power supply unit 19, shape detection unit 21, and display unit 23 shown in Figure 1, the control unit 11 may also include a molding program generation unit 31 and a welding support system 150.

[0027] The molding program generation unit 31 determines a bead formation trajectory representing the formation sequence of the welding beads B for forming the object, according to the input object to be molded, molding conditions, etc., and generates the aforementioned molding program.

[0028] The welding support system 150 predicts the occurrence of welding defects from the history information of the welding bead B formed by the welding device 110 when creating a fabricated object by layering multiple welding beads B formed by melting and solidifying filler material M by the welding device 110, and generates instruction information to improve the predicted welding defects. The welding support system 150 includes a defect monitoring device 170 and an instruction information generation device 190. Details of the defect monitoring device 170 and the instruction information generation device 190 will be described later.

[0029] The control unit 11, having the functions described above, is composed of a computer device including a processor such as a CPU, memory such as ROM and RAM, and a storage unit such as an HD (hard disk drive) or SSD (solid state drive), although these are not shown in the figures. Each component of the control unit 11 operates according to the commands of the CPU and performs its respective function. The control unit 11 may also be located separately from the welding apparatus 110 and connected to the welding apparatus 110 remotely via a communication means such as a network. The molding program may be created by another device and input to the control unit 11 via communication or a storage medium, in addition to being created by the control unit 11.

[0030] (Defect monitoring device) The defect monitoring device 170 predicts the occurrence of welding defects when a multiple welding beads, formed by melting and solidifying filler material M using the welding device 110 shown in Figure 1, are stacked in layers to create a fabricated object. The device predicts the occurrence of welding defects based on the history information of the welding beads formed by the welding device 110. In other words, when the welding device 110 forms welding beads B according to a predetermined fabrication program, the device predicts the locations where welding defects are likely to occur using feature quantities that represent the shape of the welded area. This defect monitoring device 170 comprises a shape profile acquisition unit 33, a feature quantity extraction unit 35, a defect location identification unit 37, a formation area identification unit 39, and a defect size prediction unit 41. The functions of each of the above-described parts will be outlined below using the fabrication example shown.

[0031] <Defect Monitoring Methods> Figure 3 is a flowchart showing the procedure for monitoring the occurrence of defects. (Shape profile acquisition process) First, the welding apparatus 110 shown in Figure 1 moves the welding torch 27 along the bead formation trajectory, which is the movement trajectory of the welding torch 27 set in the drive program, to form a welding bead. Along with this bead formation, the shape detection unit 21 measures the shape of the welding bead and the surface on which the bead is to be formed (S1).

[0032] Figure 4 is a schematic diagram showing the welding torch 27, the shape detection unit 21, and the bead formation trajectory of the weld bead B. The welding bead B is formed sequentially as the welding torch 27 moves on the base plate P along a pre-created bead formation trajectory. Simultaneously with the movement of the welding torch 27, the shape detection unit 21 measures the surface shape of the existing welding bead B and the surface G where the bead is to be formed. The shape detection unit 21 then outputs the surface shapes of the welding bead B and the surface G where the bead is to be formed (collectively called the shape profile) to the control unit 11.

[0033] It is preferable to measure the shape profile at the same time as the formation of the weld bead B. For example, when the welding bead is layered while the workpiece is moved with the welding torch 27 fixed, the shape detection unit 21 may be placed at the fixed position, or when the welding bead is layered while the welding torch 27 is moved, the shape detection unit may be positioned so as to be fixed around the welding torch 27. This allows for efficient measurement of the shape of the formed welding bead B along the movement path while the welding bead B is formed by the movement of the welding torch 27, thereby shortening the cycle time. This measurement of the shape profile may be performed at a different time than the formation of the welding bead B, and may be performed at a desired time depending on various conditions. Hereinafter, a welding bead that is to be formed but has not yet been formed will be referred to as a "newly formed welding bead," and a welding bead that has already been formed will be referred to as an "existing welding bead."

[0034] The molded object WK shown here has a frame-shaped wall Aw formed by welding beads B, and a filling area Af that fills the area enclosed by the wall Aw with welding beads B. This filling area Af is formed after the wall Aw is formed. That is, after the wall Aw is formed, the welding beads B that will become the filling area Af inside the wall Aw are formed along the bead formation trajectories F1 to F3 shown by the dotted lines. Then, welding beads B are formed along the bead formation trajectory F4. Note that the order in which the welding beads B are formed within the filling area Af is arbitrary.

[0035] When forming the wall portion Aw, priority is given to creating the weld bead with high shape accuracy, while when forming the filling portion Af surrounded by the wall portion Aw, priority is given to filling it quickly and in large quantities with weld beads rather than accuracy. When forming weld beads in the filling portion Af, defects such as voids are particularly likely to occur, so it is important to prevent such defects from occurring.

[0036] For example, when beads are formed on bead formation trajectories F2 and F3, the shape profile acquisition unit 33 shown in Figure 2 drives the shape detection unit 21 to measure the shape profile of the existing welded bead B and acquires that shape profile. In other words, it determines the shape profile of the pair of existing welded beads B located on both sides of the bead formation trajectory F4 and the surface G on which the bead is to be formed.

[0037] (Feature extraction process) Then, the feature extraction unit 35 shown in Figure 2 extracts predetermined features from the acquired shape profile. The feature extraction unit 35 identifies the bead formation trajectory of the new weld bead to be formed and the existing weld bead adjacent to it. Then, using the information of the identified existing weld bead and the shape profile of the surface where the bead is to be formed, feature quantities are obtained (S2).

[0038] Figure 5A is a schematic diagram showing the cross-sectional shape of adjacent existing weld beads B, and is a cross-sectional view when the existing weld beads B are properly spaced apart from each other. Figure 5B is a schematic diagram showing the cross-sectional shape of adjacent existing weld beads B, and is a cross-sectional view when the existing weld beads B are placed close together.

[0039] As shown in FIG. 5A, when a pair of adjacent existing weld beads B are properly spaced apart, it is assumed that a new weld bead Ba indicated by a dotted line is formed on a bead formation planned surface G. Regarding the orthogonal direction of the bead formation direction at this time (the left - right direction in FIG. 5A), the distance between the existing weld beads at the bottom of the valley (bead formation region width) of a pair of adjacent existing weld beads B that are arranged adjacent to each other to form a valley portion is defined as a bottom interval U, and the distance between bead tops Pt1 and Pt2 is defined as a bead interval W. The bottom interval U may be the distance between the bead end points (start end and end end of bead formation) of a pair of existing weld beads.

[0040] Also, in a cross - section orthogonal to the bead formation direction of the existing weld bead B, a tangent line at a position P1 where the existing weld bead B having a bead top Pt2 contacts the bottom surface FL of the existing weld bead B is defined as L1. Also, among the intersection angles between the tangent line L1 and the bottom surface FL, the angle on the side of the existing weld bead B that contacts the tangent line L1 is defined as a root angle θ. Similarly, the root angle θ can be obtained for the existing weld bead B having a bead top Pt1. Here, although the root angles θ of the existing weld beads B of the bead tops Pt1 and Pt2 are regarded as equal to each other, their average values, maximum values, etc. may be determined as representative values of the root angle θ.

[0041] As shown in FIG. 5B, when the existing weld beads B are formed close to each other, the bottom interval Un becomes narrower than the case shown in FIG. 5A (Un < U). Also, the root angle θn of the side surface of the existing weld bead B becomes larger (θn > θ). In that case, a narrow portion K is formed between the existing weld beads B. When trying to newly form a weld bead in this narrow portion K, the molten metal of the new weld bead may not completely flow into the narrow space between the existing weld bead B and the bead formation planned surface G, and a minute space may remain to cause a cavity. This cavity becomes a defect (unwelded portion), deteriorates the welding quality, and ultimately reduces the strength of the shaped object.

[0042] Therefore, the defect monitoring device 170 monitors the characteristic quantities of the bottom interval U (Un) and the root angle θ (θn), which are the widths for forming the new weld bead described above, and identifies defect candidate locations that are likely to become defects. By using both the bottom spacing U(Un) and the root angle θ(θn) as features, an improvement in the accuracy of defect candidate detection can be expected compared to using only one feature. Furthermore, the bead spacing W may also be used as a feature.

[0043] Figures 5A and 5B show examples of forming a new weld bead between a pair of adjacent existing weld beads B. However, there are cases where existing weld beads exist only on one side. In such cases, the distance between the existing weld bead on one side and the bead formation trajectory (planned line) of the new weld bead may be treated as a feature quantity, similar to the bottom spacing U or bead spacing W described above. In other words, the feature quantity can be appropriately selected according to the arrangement of the existing weld beads B.

[0044] For extracting these features, a shape database DB1 may be used, which pre-associates shape profiles with features. The shape database DB1 registers information on the shape profile measured by the shape detection unit 21, associating it with the position (coordinates, path) of the bead formation trajectory. The shape profile information includes, for example, the surface shape of the welded bead B, the coordinate values ​​of the measured position, the bead width and bead height of the existing welded bead B, etc. For example, if the shape detection unit 21 is a laser sensor, the information on the reflection profile of the laser light can be recorded in the shape database DB1. Alternatively, the information on the shape of the existing welded bead described above may be calculated and recorded from the reflection profile information.

[0045] When using the shape database DB1, the feature extraction unit 35 shown in Figure 2 refers to the shape database DB1 to extract information on specific shape profiles that can approximate the measured shape profile, and then determines feature quantities corresponding to the extracted shape profile. These feature quantities may be obtained from the shape profile information acquired from the shape database DB1, or they may be set as feature quantities based on information about the shape of existing weld beads that has been pre-recorded in the shape database DB1.

[0046] In this way, the feature extraction unit 35 outputs the obtained feature information to the defect location identification unit 37.

[0047] (Defect location identification process) Next, the defect location identification unit 37 identifies candidate defect locations where welding defects are expected to occur, according to the extracted feature quantities (S3). In other words, according to the obtained feature quantities of the welded bead B, it identifies candidate defect locations in the newly formed welded bead where defects are expected to occur.

[0048] For example, if the bottom spacing U is outside the acceptable range greater than a predetermined threshold, or if the root angle θ is outside the acceptable range greater than a predetermined threshold, that location is set as a potential defect location. Alternatively, if both the bottom spacing U and the root angle θ exceed the acceptable range set by the thresholds described above, that location may also be set as a potential defect location. On the other hand, if the bottom spacing U is smaller than a predetermined threshold, that location may be set as a potential defect location, considering the possibility that molten metal may not flow sufficiently into the bottom.

[0049] The threshold values ​​used to determine potential defect locations can be obtained, for example, by examining the relationship between defect occurrence and feature quantities such as the bottom spacing U and root angle θ from existing element tests or simulations. Furthermore, the threshold values ​​may be set individually for each feature quantity, or they may be set by combining multiple feature quantities. For example, the threshold value may be determined by a predetermined calculation using the values ​​of the bottom spacing U and root angle θ for the cross-sectional area of ​​the weldable space on the bead formation surface G.

[0050] The information on potential defect locations extracted through the above process may be displayed on, for example, the display unit 23 shown in Figure 1, so that the operator can confirm it. Alternatively, the information on potential defect locations may be output to the output unit 25 and read by an appropriate external device.

[0051] Figure 6 is an explanatory diagram showing an example of information on potential defect locations displayed on the display unit 23. The extracted defect candidate locations Pf may be displayed on the display unit 23 along with the current weld bead formation position Pk. This makes it easier for the operator to identify defect candidate locations. The display unit 23 may also display the pass (bead formation trajectory Fn) for forming a new weld bead after the current pass in which the weld bead was formed. In this case, the operator can identify in advance the locations where defects are likely to occur in the next pass. Although not shown in the diagram, the display unit 23 may also display current welding conditions such as welding speed and filler material supply rate, as well as recommended work details such as welding conditions for defect candidate locations.

[0052] Furthermore, the 3D model displayed on the screen of the display unit 23 may be displayed in a way that makes it easier to visually understand the status of bead formation by color-coding the welding bead B that is in the process of forming according to temperature, and displaying potential defect locations with dots. In addition, if the number of potential defect locations exceeds a certain number, information such as recommended post-processing work or changes to welding conditions may be displayed. This will allow the operator to easily make visual decisions such as repairing existing welding beads, changing welding conditions for the next pass, or stopping the fabrication process.

[0053] The identification of potential defect locations described above may be performed for all passes that form the welding bead, but to reduce processing time, it may be limited to only some of the passes. For example, as shown in Figure 4 above, if the molded object WK has a frame-shaped wall portion Aw formed by welding beads B and a filling portion Af that fills the area surrounded by the wall portion Aw with welding beads B, the formation portion identification unit 39 shown in Figure 2 identifies whether the existing welding beads B are in the wall-like continuous wall portion Aw or the filling portion Af. This identification may be determined, for example, from the width of each existing welding bead obtained from the measured shape profile, the height distribution of the existing welding beads, etc., or it may be determined based on information such as the welding volume and bead width of each welding bead set in the molding plan of the molded object WK.

[0054] The formation area identification unit 39 identifies the filling area Af from the existing welding bead B, and if it determines that the next pass will be formed on the filling area Af, it identifies potential defect locations. On the other hand, if it identifies the wall area Aw from the existing welding bead B and determines that the next pass will be formed on the wall area Aw, it omits identifying potential defect locations in the subsequent pass. In this way, by identifying potential defect locations only in the case of passes that form the filling area Af, which is particularly prone to defects, the process can be simplified and the cycle time can be shortened.

[0055] Although Figure 4 shows a configuration in which the four sides of the filling section Af are surrounded by wall sections Aw, the configuration in which the filling section Af is surrounded is not limited to this. For example, the filling section Af may be placed between a pair of adjacent wall sections, resulting in a configuration surrounded by two sides, or it may be placed in a configuration surrounded by three sides, resulting in a configuration surrounded by a pair of adjacent wall sections and another wall section that connects the ends of that pair of wall sections. Furthermore, the shape of the filling section Af is not limited to a quadrilateral; it may be a polygon with five or more sides, a circle, an ellipse, or any other shape.

[0056] Furthermore, the defect size prediction unit 41 shown in Figure 2 predicts the defect size from the positional information of the extracted defect candidate locations. In other words, when defect candidate locations occur consecutively along the bead formation trajectory, it determines that these defect candidate locations constitute a single block-like defect and considers the area of ​​the consecutive defect candidate locations as having the defect size of a single defect.

[0057] Figure 7 is a graph showing an example of the defect candidate location determination results along the bead formation trajectory. The defect size prediction unit 41 determines whether the feature quantities exceed a predetermined tolerance range based on the results obtained by the defect location identification unit 37 for each position along the bead formation trajectory. Figure 7 shows an example where the value is distinguished as "1" if the feature quantities exceed the tolerance range and as "0" if they do not.

[0058] The defect size prediction unit 41 predicts the size of candidate defects from the judgment results of feature quantities continuously output from the defect location identification unit 37 along the bead formation trajectory. In other words, it extracts regions Lc1 and Lc2 where the feature quantities exceed the acceptable range and the judgment result is "1". The defect length is predicted from the length of these regions where the judgment result is "1" (length along the bead regulation trajectory). The defect size prediction unit 41 displays information on the candidate defects and the predicted defect length on the display unit 23. Alternatively, each piece of information may be output to the output unit 25.

[0059] According to this method, not only the location of potential defects but also their length can be predicted, allowing for the precise setting of specific measures to prevent the occurrence of defects, which will be described in detail later.

[0060] In the welding system 100 described above, when a new weld bead is formed by the welding device 110, the control unit 11 causes the shape profile acquisition unit 33 to update the shape profile, and repeatedly performs feature extraction by the feature extraction unit 35 and identification of potential defect locations by the defect location identification unit 37. By performing post-processing such as machining or re-welding of the existing weld bead at the identified potential defect locations using the bead processing device 130, or by changing the welding conditions of the newly formed weld bead, the occurrence of defects at the potential defect locations can be prevented.

[0061] (Other examples of features) In addition to the parameters exemplified above, various other parameters can be used as feature quantities. Figure 8 is an explanatory diagram showing examples of other feature quantities in a cross-section perpendicular to the bead formation direction of an existing weld bead. When a pair of adjacent existing welded beads B1 and B2 are formed on the base plate or the surface of an existing welded bead in the lower layer (represented by the bottom spacing U and bead spacing W), the average height H of the height from the base surface FL to the bead top Pt1 and the height from the base surface FL to the bead top Pt2 may also be used as a feature quantity. The average height H corresponds to the depth of the valley to the bottom of the valley formed by the pair of existing welded beads B1 and B2 in the lamination direction.

[0062] Figures 9 and 10 are explanatory diagrams showing how the spacing between the existing weld beads shown in Figure 8 changes. As shown in Figure 9, when the existing welded beads B1 and B2 are close enough to touch each other, the bottom spacing U becomes 0, and the average height H is the valley depth between the beads, represented by the dotted triangle Pt1-Pt2-P1(P2). Furthermore, as shown in Figure 10, when existing welded beads B1 and B2 overlap each other, the bottom spacing U is 0, and the average height H is shallower than in the cases shown in Figures 8 and 9.

[0063] Thus, by including a combination of bottom spacing U, bead spacing W, and average height H as features, the shape of the valley can be identified in a manner substantially equivalent to that of the combination of features including the root angle θ mentioned above. Alternatively, the cross-sectional area of ​​the recessed shape may be calculated using at least one of the following features: root angle θ, bead formation region width, bead spacing W, average height H (valley depth), bottom spacing U, etc. When using the cross-sectional area as a feature, evaluation can be performed while mitigating the effects of variations in the accuracy of measurement points from one measurement location to another.

[0064] Figure 11 is an explanatory diagram showing examples of other feature quantities in a cross-section perpendicular to the bead formation direction of an existing weld bead. In Figure 11, let r1 be the radial distance from the bead center Pc1 of one existing weld bead B1, and let r2 be the radial distance from the bead center Pc2 of the other existing weld bead B2. Furthermore, an arbitrary width Wa is set, and Ha is the height (height of the straight line L2) of the surface positions Pr1 and Pr2 such that the distance between existing weld beads B1 and B2 is width Wa. Alternatively, the height Ha may be set arbitrarily, with Pr1 being the surface position of existing weld bead B1 and Pr2 being the surface position of existing weld bead B2 at height Ha, and Wa being the distance between surface position Pr1 and surface position Pr2.

[0065] The above-mentioned radial distances r1, r2, and at least one of the surface positions Pr1, Pr2 relative to the height Ha can also be used as features.

[0066] Alternatively, the existing weld beads B1 and B2 may be approximated as trapezoids to set the feature quantities. Figure 12 is an explanatory diagram showing an example of feature quantities when an existing weld bead is approximated as a trapezoid in a cross-section perpendicular to the bead formation direction of the existing weld bead.

[0067] Let Ta1 be the bottom edge (bottom base) of trapezoid Db1, Tb1 be the top edge (top base), and Hb1 be the height of trapezoid Db1, which is a trapezoidal approximation of the existing weld bead B1. Let Ta2 be the bottom edge (bottom base), Tb2 be the top edge, and Hb2 be the height of trapezoid Db2, which is a trapezoidal approximation of the existing weld bead B2. The distance between the ends of trapezoids Db1 and Db2 on the underlying surface FL is defined as the bottom spacing U. The distance between the ends of the top edges of trapezoids Db1 and Db2 is defined as Wb. These parameters may also be set as feature quantities.

[0068] Figure 13 is an explanatory diagram showing an example of characteristic quantities of an existing weld bead in a cross-section perpendicular to the bead formation direction of the existing weld bead. If the cross-sectional shape of the existing weld beads B1 and B2 is a circle with an arc length longer than the semicircle shown in Figure 11, then Wc is the distance from the substrate surface FL to the surface positions Pn1 and Pn2 where the existing weld beads B1 and B2 are closest to each other, Hc is the height (straight line L3) from the substrate surface FL to these closest surface positions Pn1 and Pn2, K1 is the depth of the weld leg from the position P1 where the existing weld bead B1 touches the substrate surface FL to the position along the substrate surface FL corresponding to surface position Pn1, and K2 is the depth of the weld leg from the position P2 where the existing weld bead B2 touches the substrate surface FL to the position along the substrate surface FL corresponding to surface position Pn2. Also, φ1 and φ2 are the angles of the weld legs (angles obtained by subtracting the root angle from 180°), and U, as mentioned above, is the bottom spacing.

[0069] As shown in Figure 13, it is also possible to set at least one of Wc, Hc, K1, K2, φ1, φ2, and U as a feature.

[0070] Furthermore, the width and height (valley depth) mentioned above are lengths expressed in a coordinate system of two orthogonal axes: the bead stacking direction (e.g., the vertical direction) and the in-plane direction of the base surface FL. However, they may also be lengths expressed in a coordinate system that is inclined to these axes. Figure 14 is an explanatory diagram showing an example of characteristic quantities when a welding bead is formed in the filling portion Af inside the wall portion Aw in a cross-section perpendicular to the bead formation direction of an existing welding bead.

[0071] When an existing weld bead B2, which will form a filler portion Af, is positioned between existing weld beads B1 and B3 on one side, which will form a frame-shaped wall portion Aw, and existing weld bead B4 on the other side, a narrow portion Pnp is formed between existing weld beads B2 and existing weld beads B3. At the location including this narrow portion Pnp, cavities (defects) are more likely to occur when forming new weld beads.

[0072] However, the existing weld bead B3 was formed after the formation of the existing weld bead B2, and the orientation of the narrow section Pnp is inclined from the bead stacking direction Dh. Therefore, the common tangent L4 between the existing weld beads B2 and B3 is determined, and the point of contact between the common tangent L4 and the existing weld bead B2 is taken as Pm1, and the point of contact between the common tangent L4 and the existing weld bead B3 is taken as Pm2. Then, the inclination direction Dn that passes through the narrow section Pnp and is perpendicular to the common tangent L4 is determined. The angle between this inclination direction Dn and the bead stacking direction Dh is taken as the inclination angle ψ. Also, the distance from point of contact Pm1 to point of contact Pm2 along the common tangent L4 is taken as Wd, and the distance from the narrow section Pnp to the common tangent L4 along the inclination direction Dn is taken as the height Hd.

[0073] As shown in Figure 14, at least one of ψ, Wd, and Hd can also be set as a feature.

[0074] Furthermore, when multiple welding beads are layered periodically, the size and spacing of each existing welding bead are often constant. However, if the shape of a welding bead is suddenly disrupted, it may contain shapes that induce defects, such as narrower constrictions within the disrupted bead compared to its surroundings. Therefore, the spacing of existing welding beads may be set as a feature variable.

[0075] Figure 15 is an explanatory diagram showing the pitch of each welding bead when multiple welding beads are stacked in a cross-section perpendicular to the bead formation direction of an existing welding bead. When multiple existing weld beads B are arranged, the arrangement pitch Pa along the bead stacking direction Dh and the arrangement pitch Pb along the bead alignment direction Dr for each existing weld bead B may be determined, and each arrangement pitch Pa and Pb may be set as a feature. Alternatively, the deviation of the determined arrangement pitches Pa and Pb from the reference value may be set as a feature.

[0076] Furthermore, in addition to setting the various feature quantities described above by directly obtaining values ​​from the shape profile, they may also be obtained from the results of approximating the shape profile with curves or a specific model shape.

[0077] Figure 16 is an explanatory diagram showing how the shape profile is approximated by a curve. First, the baseline BL, which is the base surface excluding the welded bead area, is determined by curve approximation from the measured shape profile. Then, from the determined approximation curve AC, feature quantities such as the bead spacing W, bottom spacing U, and average height H of the aforementioned pair of existing welded beads are calculated.

[0078] Here, the average height H may be the average height of the height from the lowest point Pd where the approximation curve AC is even lower than the baseline BL, to the top of one bead Pt1, and to the top of the other bead Pt2.

[0079] Figure 17 is an explanatory diagram showing the results of predicting the shape of a fabricated object with existing weld beads layered on top of it. The shape of the fabricated object WK obtained by layering welding beads can be predicted by setting a shape model BM that mimics the shape of the welding beads, for example, using a polygonal shape such as a trapezoid, and then simulating layering this shape model BM. When predicting the shape of the fabricated object WK by simulation in this way, feature quantities may be set according to the difference between the predicted shape from the simulation result and the shape profile actually measured. Examples of feature quantities in this case include the cross-sectional area in a specific region between the predicted shape and the shape profile, such as the cross-sectional area of ​​a narrow part or the cross-sectional area of ​​a weldable region.

[0080] <Instructions for correcting potential defect areas> (Instruction information generation device) The welding support system 150 extracts various feature quantities from the shape profile measured as described above, and identifies potential defect locations from the obtained feature quantities. Then, based on the information of the identified potential defect locations, the welding support system 150 generates instruction information for post-processing of the existing weld bead after bead formation, and for changing the welding conditions of the newly formed weld bead (S4).

[0081] The instruction information generation device 190 shown in Figure 2 includes a post-processing condition setting unit 61 and a welding condition setting unit 63. The post-processing condition setting unit 61 sets the post-processing conditions for repairing the detected defect candidate areas by machining or remelting. The welding condition setting unit 63 sets the welding conditions for forming a weld bead at the detected defect candidate areas.

[0082] Figure 18 is an explanatory diagram showing the factors influencing defects. The welding condition setting unit 63 searches for process conditions to be adjusted based on the welding conditions set in the trajectory plan, using information on the defect candidate locations as a reference. For example, if a narrow section Pnp is formed, the welding current, welding voltage, filler material feeding rate, welding speed, or the inclination angle α (receding angle) of the welding torch 27 are set to increase the heat input during welding so that the bead penetrates to the bottom of the narrow section Pnp. The welding speed may also be increased so that the molten metal does not precede the arc generated from the welding torch 27. Furthermore, the horizontal distance δ between the narrow section Pnp and the welding torch 27 may be set. The specific adjustment amounts for each of the above process conditions should be determined appropriately based on conditions tested in bead-on-plate (BOP) element tests, etc., within a range that does not change the amount of welded material.

[0083] Furthermore, the post-processing condition setting unit 61 modifies the shape of the narrow portion Pnp of the weld bead of the previous layer by cutting and repairing it with a grinder or the like. This ensures that defects can be reliably prevented even in cases where correction by welding conditions alone is insufficient.

[0084] The post-processing condition setting unit 61 and the welding condition setting unit 63 may both set their respective conditions by referring to a pre-prepared defect prevention condition database DB2. The defect prevention condition database DB2 stores conditions for preventing defects for each expected feature quantity, associated with elemental tests or simulations. By referring to this defect prevention condition database DB2, it is possible to identify conditions that will not cause defects in a candidate defect location, according to the feature quantity of that candidate defect location.

[0085] (Examples of improvements to potential defect areas) Next, we will explain examples of measures to prevent defects from occurring based on the information of the extracted potential defect locations. The instruction information generation device 190 determines measures to prevent the occurrence of defects in the candidate defect locations, based on the extracted information about the candidate defect locations. Figure 19 is a plan view showing the existing weld bead and the planned surface for bead formation of the new weld bead. Figure 20 is a cross-sectional view along the line XX-XX shown in Figure 19, and Figure 21 is a cross-sectional view along the line XXI-XXI shown in Figure 19.

[0086] Here, as shown in Figure 19, suppose a protrusion Bp occurs in a part of the existing welded bead B adjacent to the bead formation surface G due to melt dripping. In that case, the width dn between the existing welded beads B shown in Figure 21 becomes narrower compared to the width d between the existing welded beads B in other areas shown in Figure 20, and a narrow section K is created between the beads. Also, the root angle θn of the existing welded bead B on which the protrusion Bp is formed becomes larger. This is thought to be because when high-temperature molten metal flows out more than necessary, its tip is cooled from the bottom, and then more molten metal accumulates on top of it and cools, resulting in the tip of the protrusion Bp becoming sharply shaped.

[0087] In the narrow section K, when forming a new weld bead, the molten metal may not flow to the corners of the narrow section K, making it prone to the formation of voids (defects). Therefore, to prevent the occurrence of defects, the post-processing condition setting unit 61 generates instruction information for post-processing, such as machining or remelting for repair.

[0088] Figure 22 is a plan view illustrating an example of post-processing, and Figure 23 is a cross-sectional view along the line XXIII-XXIII shown in Figure 22. As shown in Figures 22 and 23, the protruding portion Bp is removed by a processing tool 51 such as a grinding wheel or cutter shown in Figure 1, based on the instruction information generated by the post-processing condition setting unit 61. This eliminates the narrow portion K and restores the width d to approximately the same as shown in Figure 20. Therefore, even if a welding bead is formed after the protruding portion Bp is removed, the occurrence of defects is suppressed because the narrow portion K does not exist.

[0089] Figure 24 is a plan view showing the process of remelting the resulting protrusion Bp. Figure 25 is a cross-sectional view along the line XXV-XXV in Figure 24. As shown in Figures 24 and 25, based on the instruction information generated by the post-processing condition setting unit 61, the protruding portion Bp is heated and remelted by an arc from the welding torch 27 ("touching" in TIG welding). As a result, the molten metal of the protruding portion Bp flows into the narrow portion K, making the slope of the narrow portion K gentler and forming a smooth welded portion Mt. The welded portion Mt is formed from the molten protruding portion Bp itself and does not contain defects such as voids. Furthermore, when a welding bead is formed on the welded portion Mt, the smooth surface prevents the formation of voids, thus suppressing the occurrence of defects.

[0090] By repairing the protruding portion Bp through this post-processing, a cavity can be formed on the bead formation surface G without creating a void, thereby preventing welding defects.

[0091] Furthermore, when forming a welding bead at the location where a protrusion Bp has occurred, the instruction information generation device 190 may generate instruction information by the welding condition setting unit 63 to change the welding conditions for forming the welding bead and prevent the occurrence of defects.

[0092] In this case, the control unit 11 changes the welding conditions in the bead formation trajectory at the position corresponding to the protrusion Bp, according to the welding condition instruction information output by the instruction information generation device 190. By doing so, even if a welding bead is formed while the protrusion Bp remains, the occurrence of voids can be suppressed.

[0093] Specific changes to the welding conditions include the aforementioned welding current, welding voltage, filler material M supply rate, welding speed, tilt angle of the welding torch 27 (forward angle, backward angle), and changing from forward welding to backward welding. It is preferable to adjust (increase) the heat input. Of the above, for example, when changing to backward welding, the molten metal does not flow forward of the arc during bead formation, resulting in deeper penetration and reducing the effect of the protruding portion Bp. Furthermore, even in arc-leading welding, a similar effect can be obtained by increasing the welding current, welding voltage, and welding speed.

[0094] Furthermore, if defect suppression is difficult with either post-processing or modification of welding conditions, both may be implemented. In that case, the range of acceptable protrusions Bp can be expanded.

[0095] The instruction information generation device 190 may determine whether to perform the above-mentioned post-processing or change the welding conditions, and the details of the post-processing or welding conditions, by referring to the defect prevention condition database DB2 according to the characteristic quantities in the bead formation trajectory corresponding to the weld bead to be formed. In addition, the instruction information generation device 190 may adjust the content of the post-processing or change in welding conditions according to the defect length estimated by the defect size prediction unit 41.

[0096] The instruction information generation device 190 can also perform each of the above-described countermeasures, determined according to the feature quantities, in stages. For example, if the feature quantity at the defect candidate location is at Level 1, which is close to the feature quantity in the normal case (for example, the normal feature quantity when the protrusion Bp does not exist), the countermeasure is taken by changing the welding conditions and applying it to the next weld bead formed. If the deviation from the normal feature quantity is greater than Level 1, which is Level 2, the countermeasure is taken by re-welding the existing weld bead. If it is even greater than Level 2, which is Level 3, machining is performed. This minimizes the burden of repair work according to the expected degree of defect. Therefore, work efficiency is improved and the cycle time can be shortened.

[0097] Furthermore, as a control sequence after a potential defect location is extracted during additive manufacturing, for example, the following patterns (A) to (D), or a combination thereof, may be implemented. (A) If a potential defect is identified during the formation of the welding bead, the formation of the welding bead will be temporarily stopped, and the existing welding bead at the potential defect will be repaired. (B) If a potential defect location is detected during the formation of the welding bead, bead formation continues along the planned bead formation trajectory, and notification information indicating that a potential defect location has been detected is output to the display unit 23 or output unit 25 shown in Figure 1. (C) Furthermore, after (B), the formation of the welding bead is stopped at a timing specified by the operator, and the potential defect area is repaired. After the repair, the notification signal output to the display unit 23 or output unit 25 is canceled. (D) If a potential defect is identified during the formation of the weld bead, the bead formation will continue along the planned bead formation trajectory, and when forming a new weld bead at the potential defect location, the welding conditions will be changed only within the area of ​​the potential defect location.

[0098] According to the defect monitoring method described above, it is possible to identify paths where unwelded defects are likely to occur during the printing process, providing time to address them midway through. Therefore, defect occurrence can be prevented in real time, saving the effort of repairing the completed object. Furthermore, by saving the details of the actions taken during printing as log data, when inspecting defects in the completed object, the actual defect location can be compared with the information in the log data to identify the cause of the defect. In this way, a configuration with excellent traceability can be achieved.

[0099] Thus, the present invention is not limited to the embodiments described above. It is also intended and within the scope of protection to be provided for those skilled in the art to modify and apply the various configurations of the embodiments in combination with each other, based on the description in the specification and well-known art.

[0100] As described above, the following matters are disclosed in this specification: (1) A defect monitoring device that predicts the occurrence of welding defects from historical information of the welding bead formation by the welding device when a structure is formed by stacking multiple welding beads formed by melting and solidifying filler material by a welding device, A shape profile acquisition unit that acquires the shape profile of an existing welding bead, A feature extraction unit that extracts feature quantities of recess shapes formed by a plurality of existing welding beads included in the shape profile, A defect location identification unit identifies candidate defect locations where welding defects are expected to occur, based on the extracted feature quantities. When the welding apparatus forms a new weld bead, the control unit causes the shape profile acquisition unit to update the shape profile, and repeatedly performs the extraction of feature quantities by the feature quantity extraction unit and the identification of candidate defect locations by the defect location identification unit. A defect monitoring device equipped with the following features. This defect monitoring device obtains feature quantities from the shape profile of the welding bead and extracts potential defect locations based on these feature quantities, thereby identifying potential defect locations that may occur in the fabricated object each time a welding bead is formed.

[0101] (2) The defect monitoring device according to (1), wherein the characteristic quantity includes at least one of the following: the root angle, which is the angle on the welding bead side, of the angle of intersection between the tangent line at the position where the existing welding bead is in contact with the underlying surface of the welding bead in a cross section perpendicular to the bead formation direction of the welding bead, and the width of the bead formation region in the area where the new welding bead is to be formed. According to this defect monitoring device, by setting the root angle and the width where bead formation is planned as features, it is possible to identify areas with a high probability of defect occurrence.

[0102] (3) The defect monitoring device according to (1) or (2), wherein the characteristic quantity includes at least one of the following: the distance between the bead crests of a pair of existing welding beads that are arranged adjacent to each other and form a valley in a cross section perpendicular to the bead formation direction of the welding bead; the depth of the valley from the bead crest to the valley bottom; and the distance between the bottoms of the welding beads at the valley bottom. According to this defect monitoring device, by including the distance of each valley formed between adjacent pairs of existing weld beads as a feature, it is possible to reliably identify areas prone to defects as potential defect locations.

[0103] (4) The defect monitoring device according to (3), wherein the feature quantities include the cross-sectional area of ​​the recess shape calculated using at least one of the feature quantities. According to this defect monitoring device, when using cross-sectional area as a feature, evaluation can be performed while mitigating the effects of variations in the accuracy of measurement points from one measurement location to another.

[0104] (5) The defect monitoring device according to any one of (1) to (4), further comprising a formation site identification unit that identifies whether the existing welding bead is a wall-like series of wall portions or a filling portion within a region surrounded by the wall portions. This defect monitoring device can identify whether an existing weld bead is a wall or a filling section, allowing it to switch between identifying potential defect locations and not. Therefore, defect identification can be concentrated only on paths forming filling sections, which have a particularly high probability of defect occurrence, thereby simplifying the process and shortening the cycle time.

[0105] (6) A defect monitoring device according to any one of (1) to (5), further comprising a defect size prediction unit that predicts the defect size from the location information of the candidate defect location or the magnitude of the feature quantity. This defect monitoring device can predict the defect length at each potential defect location, allowing it to determine whether the defect is of acceptable size or not. This enables the adjustment of the treatment based on the defect length. Furthermore, it can accurately identify areas requiring treatment and perform treatment only on those specific areas. As a result, the process can be simplified and the cycle time can be shortened.

[0106] (7) A defect monitoring method for predicting the occurrence of welding defects from historical information of the welding bead formation by the welding apparatus when a structure is formed by stacking multiple welding beads formed by melting and solidifying filler material with a welding apparatus, The process of obtaining the shape profile of the existing weld bead, A step of extracting characteristic quantities of recess shapes formed by a plurality of existing weld beads included in the shape profile, A step of identifying candidate defect locations where welding defects are expected to occur according to the extracted feature quantities, The welding apparatus updates the shape profile when it newly forms the welding bead, and repeatedly performs the extraction of feature quantities and the identification of candidate defect locations. A defect monitoring method comprising the following features. This defect monitoring method allows for the identification of potential defect locations in a fabricated object each time a welding bead is formed. This is achieved by obtaining feature quantities from the shape profile of the welding bead and extracting potential defect locations based on these obtained feature quantities.

[0107] (8) The defect monitoring method according to (7), which involves forming a new welding bead and simultaneously acquiring the shape profile. This defect monitoring method allows for efficient acquisition of a shape profile by measuring the shape of the welding bead during the welding bead formation path. This, in turn, reduces the cycle time.

[0108] (9) The defect monitoring method according to (7) or (8), which distinguishes whether the existing welding bead is a wall-like continuous wall portion or a filled portion within a region surrounded by the wall portion, and determines the characteristic quantity only when the filled portion is formed. This defect monitoring method allows for the identification of whether an existing weld bead is a wall or a filler portion, enabling the process to switch between identifying potential defect locations and not identifying them. Therefore, defect identification can be concentrated only on paths forming filler portions, which have a particularly high probability of defect occurrence, thereby simplifying the process and shortening the cycle time.

[0109] (10) A defect monitoring method according to any one of (7) to (9), further comprising the step of predicting the defect size from the location information of the candidate defect location or the magnitude of the feature quantity. This defect monitoring method allows for the prediction of defect length at each potential defect location, enabling a determination of whether the defect is of acceptable size or not. This allows for switching the treatment method according to the defect length. Furthermore, it enables the accurate identification of areas requiring treatment, allowing treatment to be performed only on those areas. Thus, the process can be simplified and the cycle time can be shortened.

[0110] (11) A defect monitoring device described in any one of (1) to (6), An instruction information generating device that generates instruction information to improve the welding defect at the identified candidate defect location, A welding support system equipped with the following features. This welding support system helps to prevent predicted welding defects at identified potential defect locations.

[0111] (12) The welding support system according to (11), wherein the instruction information generating device includes a post-processing condition setting unit that sets conditions for repairing the candidate defect location by machining or remelting. This welding support system allows for post-processing of potential defect areas, thereby promoting the prevention of welding defects that are predicted to occur.

[0112] (13) The welding support system according to (11) or (12), wherein the instruction information generating device includes a welding condition setting unit that sets at least one of the welding conditions, welding current, welding voltage, filler material feeding rate, welding rate, and torch holding angle, for forming the welding bead at the defect candidate location. This welding support system allows for the prevention of welding defects by adjusting various welding conditions.

[0113] (14) A welding support system described in any one of (11) to (13), The welding apparatus that forms the welding bead, A bead processing device for processing potential defect locations in the welded bead of the formed structure, A welding system equipped with [the following features]. This welding system allows for the identification of areas where welding defects are likely to occur, and by processing these identified defect-prone areas, the occurrence of welding defects can be prevented. [Explanation of Symbols]

[0114] 11 Control Unit 13 Welding robots 15 Robot drive unit 17 Filler metal supply section 17a Reel 19 Welding power supply unit 21 Shape detection unit 23 Display section 25 Output section 27 Welding Torch 31 Modeling program generation unit 33 Shape Profile Acquisition Unit 35 Feature Extraction Unit 37 Defect location identification unit 39 Formation site identification part 41 Defect size prediction unit 51 Processing tools 53 Robot Arm 55 Machining drive unit 100 welding systems 110 Welding equipment 130 Bead processing equipment 150 Welding Support Systems 170 Defect Monitoring Device 190 Instruction information generation device AC trendline Af filling section Aw wall section B, B1, B2, B3, B4 Existing welded beads (welding beads) Ba New welding bead BL Baseline BM Shape Model Bp protrusion d,dn width DB1 Shape Database DB2 Defect Prevention Conditions Database Db1, Db2 trapezoid Dh bead stacking direction Dn slope direction Dr. Bead alignment direction F1,F2,F3,F4,Fn Bead formation trajectory FL Base surface G Bead formation planned surface K Narrow area L1 tangent L2,L3 straight line L4 common tangent Lc1,Lc2 area M filler metal Mt weld part P Base Plate P1,P2 position Pa,Pb arrangement pitch Pc1, Pc2 bead center Pd lowest point Pf defect candidate locations Pk welding bead formation position Pm1, Pm2 contacts Pn1,Pn2,Pr1,Pr2 Surface position Pnp narrow area Pt1, Pt2 bead top r1,r2 radial distance U,Un Bottom spacing W Bead spacing Wa width WK Modeled object (structure) θ, θn: Root angle φ1, φ2 welding leg angle ψ Inclination angle

Claims

1. A defect monitoring device that predicts the occurrence of welding defects from the history information of the welding bead formation by the welding device when a structure is formed by stacking multiple welding beads, which are formed by melting and solidifying filler material by a welding device, A shape profile acquisition unit that acquires the shape profile of an existing welding bead, A feature extraction unit that extracts feature quantities of recess shapes formed by a plurality of existing welding beads included in the shape profile, A defect location identification unit identifies candidate defect locations where welding defects are expected to occur, based on the extracted feature quantities. The welding apparatus comprises a control unit which, when it forms a new weld bead, causes the shape profile acquisition unit to update the shape profile, and repeatedly performs the extraction of feature quantities by the feature quantity extraction unit and the identification of candidate defect locations by the defect location identification unit, A defect monitoring device in which the characteristic quantity includes at least one of the following: the root angle, which is the angle on the welding bead side, of the intersection angle between the tangent line at the position where the existing welding bead is in contact with the underlying surface of the welding bead in a cross section perpendicular to the bead formation direction of the welding bead, and the width of the bead formation region in the area where the new welding bead is planned to be formed.

2. A defect monitoring device that predicts the occurrence of welding defects from the history information of the welding bead formation by the welding device when a structure is formed by stacking multiple welding beads, which are formed by melting and solidifying filler material by a welding device, A shape profile acquisition unit that acquires the shape profile of an existing welding bead, A feature extraction unit that extracts feature quantities of recess shapes formed by a plurality of existing welding beads included in the shape profile, A defect location identification unit identifies candidate defect locations where welding defects are expected to occur, based on the extracted feature quantities. The welding apparatus comprises a control unit which, when it forms a new weld bead, causes the shape profile acquisition unit to update the shape profile, and repeatedly performs the extraction of feature quantities by the feature quantity extraction unit and the identification of candidate defect locations by the defect location identification unit, A defect monitoring device in which the characteristic quantity includes at least one of the following: the distance between the bead crests of a pair of existing welding beads that are arranged adjacent to each other and form a valley in a cross section perpendicular to the bead formation direction of the welding bead, the valley depth from the bead crest to the valley bottom, and the distance between the bottoms of the welding beads at the valley bottom.

3. The defect monitoring device according to claim 2, wherein the feature quantities include the cross-sectional area of ​​the recess shape calculated using at least one of the feature quantities.

4. The defect monitoring device according to any one of claims 1 to 3, further comprising a formation site identification unit that identifies whether the existing welding bead is a wall-like series of wall portions or a filling portion within a region surrounded by the wall portions.

5. The defect monitoring device according to any one of claims 1 to 4, further comprising a defect size prediction unit that predicts the defect size from the location information of the candidate defect location or the magnitude of the feature quantity.

6. A defect monitoring method for predicting the occurrence of welding defects when forming a structure by layering multiple weld beads formed by melting and solidifying filler material with a welding device, based on the history information of the welding bead formation by the welding device, The process of obtaining the shape profile of the existing weld bead, A step of extracting characteristic quantities of recess shapes formed by a plurality of existing weld beads included in the shape profile, A step of identifying candidate defect locations where welding defects are expected to occur according to the extracted feature quantities, The welding apparatus updates the shape profile when it newly forms the weld bead, and repeatedly performs the extraction of feature quantities and the identification of candidate defect locations. A defect monitoring method that distinguishes whether the existing welding bead is a wall-like continuous wall portion or a filled portion within a region surrounded by the wall portion, and determines the characteristic quantity only when the filled portion is formed.

7. The defect monitoring method according to claim 6, further comprising the step of predicting the defect size from the location information of the candidate defect location or the magnitude of the feature quantity.

8. A defect monitoring device according to any one of claims 1 to 5, An instruction information generating device that generates instruction information to improve the welding defect at the identified candidate defect location, A welding support system equipped with the following features.

9. The welding support system according to claim 8, further comprising a post-processing condition setting unit for setting conditions for repairing the candidate defect location by machining or remelting.

10. The instruction information generation device includes a welding condition setting unit that sets at least one of the welding conditions, such as welding current, welding voltage, filler material feeding rate, welding speed, and torch holding angle, when forming the welding bead at the candidate defect location. The welding support system according to claim 8 or 9.

11. A welding support system according to any one of claims 8 to 10, The welding apparatus that forms the welding bead, A bead processing device for processing potential defect locations in the welded bead of the formed structure, A welding system equipped with [the following features].

Images