Control information correction method, control information correction device, and program

The control information modification method adjusts the stacking start point in additive manufacturing by considering previous pass loci to prevent slag-induced arc defects, enhancing weld bead quality.

JP2024160435AActive Publication Date: 2024-11-14KOBE STEEL LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
JP2023075407
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-01
Publication Date
2024-11-14
Estimated Expiration
2043-05-01

AI Technical Summary

Technical Problem

In additive manufacturing using arc welding, slag adherence to previously formed weld beads can shift the stacking start point, leading to improper alignment and defective bead formation due to poor arc generation.

Method used

A control information modification method that adjusts the stacking start point based on the locus of previous passes, using machine learning or predefined patterns to determine an optimal shift position within a specified range, avoiding areas with slag and ensuring proper arc generation.

Benefits of technology

This method effectively aligns the stacking start point to prevent arc defects, improving the quality of weld bead formation in additive manufacturing processes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024160435000001_ABST
    Figure 2024160435000001_ABST
Patent Text Reader

Abstract

To shift a lamination start point of a lamination path to an appropriate shift position in consideration of the locus of a lamination path laminated prior to the lamination path, when lamination molding is performed by laminating weld beads.SOLUTION: A control information correction method for correcting control information of lamination molding performed by laminating weld beads includes the steps of: acquiring the loci and the orders of a plurality of lamination paths of the weld beads; specifying a vicinity range of a lamination start point of a first lamination path, on the basis of a first locus as a locus of the first lamination path among the plurality of lamination paths; extracting a second locus as a locus of a second lamination path laminated prior to the first lamination path among the plurality of lamination paths and as a locus overlapping the vicinity range, on the basis of the orders of the plurality of lamination paths; setting a shift position of the lamination start point to be within the vicinity range, on the basis of the first locus and the second locus; and correcting the control information, on the basis of the shift position.SELECTED DRAWING: Figure 3
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a control information modifying method, a control information modifying device, and a program. [Background technology]

[0002] Patent Document 1 describes an arc welding device in which a teach pendant teaches a position different from the welding start point as the arc try point, a hard disk stores various conditions including at least position data related to the arc try point, and a robot control device, when performing welding, moves the welding torch to the arc try point instead of the welding start point, generates an arc using a predetermined operating pattern, and then causes the welding torch to reach the welding start point to perform the actual welding. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2014-213375 A Summary of the Invention [Problem to be solved by the invention]

[0004] For example, when performing additive manufacturing by laminating weld beads using an arc, slag may adhere to the surface of an already formed weld bead, making it difficult to generate an arc, and therefore the lamination start point of a lamination pass may be shifted. In such a case, if the lamination start point of a certain lamination pass is shifted without taking into consideration the trajectory of the lamination pass that is laminated before that lamination pass, it will not be possible to shift it to an appropriate shift position.

[0005] The object of the present invention is to enable, when performing additive manufacturing by stacking weld beads, the starting point of a stacking pass to be shifted to an appropriate shift position taking into account the trajectory of the stacking pass that is stacked prior to that stacking pass. [Means for solving the problem]

[0006] With this objective in mind, the present invention provides a control information correction method for correcting control information for additive manufacturing performed by stacking weld beads, the control information correction method including the steps of: acquiring trajectories and order of multiple lamination passes of the weld bead; identifying a proximity range of the lamination start point of the first lamination pass based on a first trajectory, which is the trajectory of a first lamination pass among the multiple lamination passes; extracting a second trajectory, which is the trajectory of a second lamination pass that is stacked before the first lamination pass among the multiple lamination passes and which overlaps with the proximity range, based on the order of the multiple lamination passes; setting a shift position of the lamination start point within the proximity range based on the first trajectory and the second trajectory; and correcting the control information based on the shift position.

[0007] In the setting step, the shift position may be set using the correspondence between the first trajectory, the second trajectory, and the shift position. In this case, in the setting step, the shift position may be set by searching for pattern data in which the correspondence is patterned using the first trajectory and the second trajectory, or the shift position may be set by inputting the first trajectory and the second trajectory into a machine learning model that has learned the correspondence.

[0008] In the setting step, a feature amount indicating a relative positional relationship between the first trajectory and the second trajectory may be calculated, and the shift position may be set using a correspondence relationship between the feature amount and the shift position. In this case, in the setting step, the feature amount may be input to a machine learning model that has learned the correspondence relationship, thereby setting the shift position.

[0009] In the setting step, the shift position may be set in an area other than the prohibited portion within a nearby range including the prohibited portion where setting of the shift position is prohibited.

[0010] In the setting step, the shift position may be set based on the first trajectory and a second trajectory having a contour of the weld bead.

[0011] The present invention also provides a control information correction device that corrects control information for additive manufacturing performed by stacking weld beads, the control information correction device comprising: an acquisition unit that acquires trajectories and order of multiple lamination passes of the weld bead; an identification unit that identifies a proximity range of a lamination start point of a first lamination pass based on a first trajectory that is the trajectory of a first lamination pass among the multiple lamination passes; an extraction unit that extracts a second trajectory that is the trajectory of a second lamination pass that is stacked before the first lamination pass among the multiple lamination passes and that overlaps with the proximity range based on the order of the multiple lamination passes; a setting unit that sets a shift position of the lamination start point within the proximity range based on the first trajectory and the second trajectory; and a correction unit that corrects the control information based on the shift position.

[0012] Furthermore, the present invention also provides a program for enabling a control information correction device, which corrects control information for additive manufacturing performed by stacking weld beads, to realize the following functions: acquiring the trajectories and order of multiple lamination passes of the weld bead; identifying a proximity range of the lamination starting point of a first lamination pass based on a first trajectory, which is the trajectory of a first lamination pass among the multiple lamination passes; extracting a second trajectory, which is the trajectory of a second lamination pass that is stacked before the first lamination pass among the multiple lamination passes and which overlaps with the proximity range, based on the order of the multiple lamination passes; setting a shift position of the lamination starting point within the proximity range based on the first trajectory and the second trajectory; and correcting the control information based on the shift position. Effect of the Invention

[0013] According to the present invention, when additive manufacturing is performed by stacking weld beads, it is possible to shift the starting point of a stacking pass to an appropriate shift position that takes into account the trajectory of the stacking pass that is stacked prior to that stacking pass. [Brief description of the drawings]

[0014] [Figure 1] 1 is a diagram showing a schematic configuration example of a metal additive manufacturing system according to an embodiment of the present invention; [Diagram 2]FIG. 2 is a diagram illustrating an example of a hardware configuration of a stacking planning apparatus according to the present embodiment. [Diagram 3] 1 is a block diagram showing an example of a functional configuration of a stacking planning device according to an embodiment of the present invention; [Figure 4] 1A to 1C are diagrams illustrating an example of a layered object manufactured in accordance with an embodiment of the present invention. [Diagram 5] FIG. 5 is a diagram showing an example of a path of a welding torch for manufacturing the additive manufacturing object of FIG. 4. [Figure 6] FIG. 13 is a diagram illustrating an example of a method for setting the width of a neighborhood range. [Figure 7] FIG. 13 is a diagram showing an example of a neighborhood range that is generated based on a set width. [Figure 8] FIG. 13 is a diagram showing a method for extracting past passes and welding passes to be performed; [Figure 9] 1A to 1L are diagrams showing examples of shift patterns. [Figure 10] FIG. 1A is a diagram showing an example of an additive manufacturing object in which a lower layer may not exist where a shift position should be set, and FIG. 1B is a diagram showing an example of an exceptional shift pattern in which a lower layer does not exist where a shift position should be set. [Figure 11] FIG. 13 is a diagram showing an example of a shift pattern including information on a bead width. [Figure 12] 13A and 13B are diagrams showing setting of division points when a welding pass is divided at a plurality of division points. [Figure 13] 2 is a block diagram showing an example of a functional configuration of a control device according to the present embodiment. FIG. [Figure 14] 4 is a flowchart showing an example of the operation of the stacking planning device in the present embodiment. [Figure 15] 10 is a flowchart showing an example of a control program correction process performed by the stacking planning device in the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0016] [Metal Additive Manufacturing System Configuration] FIG. 1 is a diagram showing a schematic configuration example of a metal additive manufacturing system 1 according to the present embodiment. As shown in the figure, the metal additive manufacturing system 1 includes a welding robot (manipulator) 10, a CAD device 20, a lamination planning device 30, and a control device 50. The lamination planning device 30 writes a control program for controlling the welding robot 10 to a removable recording medium 70 such as a memory card, and the control device 50 is capable of reading out the control program written to the recording medium 70.

[0017] The welding robot 10 includes an arm 11 having a plurality of joints, and performs welding work by operating according to a control program read by a control device 50. The welding robot 10 also includes a welding torch 13 for forming an additive manufacturing object 100 via a wrist 12 at the tip of the arm 11. In the case of the metal additive manufacturing system 1, the welding robot 10 moves the welding torch 13 while melting a filler metal (wire) 14 made of mild steel, to manufacture the additive manufacturing object 100. Specifically, the welding torch 13 supplies the filler metal 14 and generates an arc while flowing a shielding gas to melt and solidify the filler metal 14, and stacks a plurality of layers of weld beads (hereinafter simply referred to as "beads") 101 on the base material 90 to manufacture the additive manufacturing object 100. Note that an arc is used as a heat source for melting the filler metal 14 here, but a laser or plasma may also be used. In addition, the welding robot 10 also includes a feeder for feeding the filler metal 14, but a description of this will be omitted.

[0018] The CAD device 20 has a function of designing a model using a computer, and also has a function of storing three-dimensional data obtained by the design (hereinafter referred to as "three-dimensional CAD data").

[0019] The stacking planning device 30 creates a stacking plan for the layered object 100 based on the three-dimensional CAD data held by the CAD device 20. In other words, it determines the trajectory of the welding torch 13 and also determines the welding conditions for welding by the welding robot 10. Then, it generates a control program for controlling the welding robot 10 so as to form a bead 101 along the determined trajectory under the determined welding conditions, and outputs this control program to the recording medium 70.

[0020] The control device 50 reads and holds a control program from the recording medium 70. Then, by operating this control program, the welding robot 10 is controlled so as to form a bead 101 according to the stacking plan created by the stacking plan device 30, that is, along the trajectory determined by the stacking plan device 30 and under the welding conditions determined by the stacking plan device 30.

[0021] The metal additive manufacturing system 1 may also include a positioner (not shown) that positions the additive object 100 for modeling. For example, when manufacturing a cylindrical additive object 100 by stacking multiple layers of beads around a central axis support material (not shown), the positioner may be a rotational positioner that rotates the additive object 100 about the central axis. Alternatively, the positioner may be a linear positioner that moves the additive object 100 in a straight line.

[0022] Furthermore, the metal additive manufacturing system 1 may include a movement device (not shown) for moving the welding robot 10 in parallel.

[0023] [Hardware configuration of stacking planning machine] FIG. 2 is a diagram showing an example of a hardware configuration of the stacking planning device 30. As shown in FIG. As shown in the figure, the stacking planning device 30 is realized by, for example, a general-purpose PC (Personal Computer) or the like, and includes a CPU 31 which is a calculation means, a main memory 32 which is a storage means, and a magnetic disk device (HDD: Hard Disk Drive) 33. Here, the CPU 31 executes various programs such as an OS (Operating System) and application software, and realizes each function of the stacking planning device 30. The main memory 32 is a storage area that stores various programs and data used for executing the programs, and the magnetic disk device 33 is a storage area that stores input data for the various programs and output data from the various programs. The stacking planning device 30 also includes a communication I / F 34 for communicating with the outside, a display mechanism 35 including a video memory, a display, etc., an input device 36 including a keyboard, a mouse, etc., and a driver 37 for reading and writing data from and to the recording medium 70. Note that Fig. 2 merely illustrates a hardware configuration in the case where the stacking planning device 30 is realized by a computer system, and the stacking planning device 30 is not limited to the illustrated configuration.

[0024] 2 can also be regarded as the hardware configuration of the control device 50. However, when describing the control device 50, the CPU 31, main memory 32, magnetic disk device 33, communication I / F 34, display mechanism 35, input device 36, and driver 37 in FIG. 2 will be expressed as CPU 51, main memory 52, magnetic disk device 53, communication I / F 54, display mechanism 55, input device 56, and driver 57, respectively.

[0025] [Background and Overview of the Present Invention] When manufacturing an additive manufacturing object by stacking multiple beads formed by melting and solidifying a filler metal using an arc, a process of generating an arc (arc start) is required each time a bead is formed. However, when attempting to form a next bead on top of an already formed bead, slag may adhere to the surface of the already formed bead due to melting and solidification of the filler metal. Since slag is usually made of an insulator, it becomes difficult to generate an arc in the area where the slag is attached, and as a result, poor bead formation may occur due to poor arc generation. In particular, when the target position is on top of a bead in the same layer, the slag cannot be avoided by removing it layer by layer. In such a case, removing the slag after each pass is possible, but it is inefficient.

[0026] Therefore, in this embodiment, when manufacturing an additive manufacturing object in which beads formed by melting and solidifying a filler metal using an arc are formed within a layer for multiple welding passes, the occurrence of poor bead formation due to poor arc generation is suppressed. In robot welding, a function is known that allows welding to be performed that reflects the number of welding passes, the order of welding passes, the relative positions, and the shift amount of the welding start position that have been set in advance by teaching the groove shape of the welding target and the trajectory of the first pass. In additive manufacturing, the order and arrangement of welding passes differ depending on the object to be manufactured, and cannot be set uniformly from the welding direction, so it is necessary to determine the shift direction for each individual.

[0027] [Functional configuration of stacking planning device] 3 is a block diagram showing an example of a functional configuration of the stacking planning device 30 according to the present embodiment. As shown in the figure, the stacking planning device 30 according to the present embodiment includes a CAD data acquisition unit 41, a CAD data division unit 42, a stacking planning unit 43, a control program generation unit 44, a control program correction unit 45, and a control program output unit 46.

[0028] The CAD data acquisition unit 41 acquires, from the CAD device 20, three-dimensional CAD data that represents the three-dimensional shape of the layered object 100. The CAD data dividing unit 42 generates a plurality of layer shape data each representing the shape of each layer by dividing (slicing) the three-dimensional CAD data acquired by the CAD data acquiring unit 41 into a plurality of layers. At that time, the CAD data dividing unit 42 may convert the three-dimensional CAD data into an internal format that makes it easy to divide the data into a plurality of layers.

[0029] The lamination planning unit 43 generates a lamination plan including the position of the welding torch 13 and the welding conditions when depositing beads that match the height and width of each layer of the multiple layer shape data generated by the CAD data division unit 42. In order to generate such a lamination plan, in addition to the height and width of the bead, a model that approximates the cross-sectional shape of the bead is required. These may be calculated and estimated from the actual values ​​of a measurement experiment or the cross-sectional area of ​​the amount of deposited metal. For example, bead-on-plate welding or vertical lamination of several layers is performed while changing the amount of deposition by varying the welding speed and wire feed speed under several conditions, and the results of measuring the height and width per layer under each condition are stored in a database. Then, a welding speed and deposition amount that satisfy the desired height and width to be laminated when laminating are selected, and an estimated shape of each layer is calculated from the measurement results as needed, and the position of the welding torch 13 is determined. The calculation method for the cross-section of the deposition may be changed depending on the material of the filler material 14 and the state of the shape of the already laminated portion. Lamination that includes the molded object is planned by this calculation method.

[0030] The control program generating unit 44 generates a control program for controlling the welding robot 10 so as to perform welding in accordance with the lamination plan generated by the lamination plan unit 43 .

[0031] The control program modification unit 45 modifies the control program generated by the control program generation unit 44. In this embodiment, the control program is used as an example of control information, and the control program modification unit 45 is provided as an example of a control information modification device that modifies the control information. Specifically, the control program correction unit 45 corrects the control program using a path information acquisition unit 451, a neighborhood range generation unit 452, a placement information extraction unit 453, a shift position determination unit 454, a trajectory division unit 455, and a shift position output unit 456.

[0032] The path information acquisition unit 451 acquires information about the welding path (hereinafter, referred to as “path information”) in chronological order from the trajectory data of the welding robot 10 included in the control program generated by the control program generation unit 44. Here, the pass information includes information on the trajectory of the welding pass and the layering order, and the information on the trajectory of the welding pass includes at least coordinate information of the target position at which welding torch 13 layers the beads. The pass information may additionally include information such as the welding current, welding voltage, wire feed speed, movement speed of welding torch 13, tilt angle of welding torch 13, and posture of welding robot 10 holding welding torch 13 when stacking the beads. Furthermore, the pass information may additionally include information such as the bead shape (bead height, bead width) planned for each welding pass.

[0033] 4 shows an example of the layered object 100 manufactured in this embodiment. As shown in the figure, the layered object 100A has a structure in which a filled portion 110b is filled inside a frame (wall) 110a, and filled portions 110c to 110e are further filled inside the filled portion 110b.

[0034] 5 shows an example of a trajectory of welding torch 13 to form additive manufacturing object 100A. In the figure, the thin dashed arrow indicates that welding torch 13 passes through trajectory 102a to form bead 101a, then passes through trajectory 102b to form bead 101b, and then passes through trajectories 102c and 102d to form beads 101c and 101d. In addition, the thick dashed arrow indicates that welding torch 13 will now pass through trajectory 102e from welding start position Ps toward welding direction D.

[0035] In the present embodiment, a path information acquisition unit 451 is provided as an example of an acquisition unit that acquires the trajectories and sequences of a plurality of lamination passes of a weld bead.

[0036] The vicinity range generating unit 452 generates a vicinity range that specifies the vicinity of the welding start position from the trajectory of the welding pass corresponding to the specified stacking sequence (k). Here, the shape of the vicinity area may be a planar shape such as a square, rectangle, circle, etc. Since the vicinity area is generated for each layer (for each height), the shape may be a planar shape or a curved shape with a constant height direction. For example, a planar shape may be used when beads are layered on a flat surface, and a curved shape may be used when beads are layered on the side surface of a cylinder. In addition, since the stacking height settings may differ between adjacent beads, the shape of the vicinity range may be a three-dimensional figure such as a cube, a rectangular solid, or a sphere. However, in the following description, the vicinity range will be described by taking as an example a square shape divided into a 3 × 3 mesh in a planar shape with a constant height direction. In such a case, if the welding direction and the direction in which the meshes are arranged are to be matched, the vicinity range generating unit 452 may obtain the welding start position and welding direction from a trajectory corresponding to the specified stacking order, and calculate the vicinity range from the welding start position and welding direction.

[0037] 6 shows an example of a method for setting the width of the neighborhood range generated by the neighborhood range generating unit 452. In the figure, a position away from the welding start position Ps by the amount of wire bending (Lw) and half (Lb / 2) of the width of the bead 101 (hereinafter referred to as "bead width") of the lamination conditions used in the molding is set as the target position Pr of the past pass, and the neighborhood range is set to include the target position Pr of the past pass. In other words, half the width of the neighborhood range (Ln / 2) is set to be larger than the sum of the amount of wire bending (Lw) and half the bead width (Lb / 2).

[0038] 7 shows an example of a neighborhood range generated by the neighborhood range generating unit 452 based on this set width. In the figure, the bead width (Lb) of the lamination conditions is set to a value longer than the amount of bending of the wire (Lw), and the width (Ln) of the neighborhood range is set to three times this bead width (Lb). The neighborhood range generating unit 452 then divides the neighborhood range into a 3×3 mesh for use by the arrangement information extracting unit 453, which will be described later. In the figure, the welding start position Ps is set in the central mesh, and the welding direction D is set to the downward direction of the paper.

[0039] In this embodiment, the trajectory of the welding pass corresponding to the specified lamination order (k) is used as an example of the first trajectory, which is the trajectory of the first lamination pass among the multiple lamination passes, and the welding start position is used as an example of the lamination start point of the first lamination pass. Also, in this embodiment, a vicinity range generating unit 452 is provided as an example of an identifying unit that identifies a vicinity range of the lamination start point based on the first trajectory.

[0040] The placement information extraction unit 453 extracts placement information of the trajectory of a past pass that overlaps with the vicinity range generated by the vicinity range generation unit 452 from among the trajectories of the welding passes corresponding to the specified stacking order (k), i.e., the welding passes that are stacked before the welding pass being calculated (hereinafter referred to as "past passes"). Specifically, the placement information extraction unit 453 extracts placement information indicating the presence and distribution of past pass teaching points within the neighborhood range generated by the neighborhood range generation unit 452 from among the teaching points of the past pass obtained by dividing the trajectory of the past pass of the same layer as the welding pass being calculated. The division of the trajectory of the past pass is performed by the trajectory division unit 455 described later. In addition, when a shift of the welding start position is implemented in the past pass, the placement information extraction unit 453 may extract the placement information including information on the shift.

[0041] 8 shows a method for extracting past passes and a welding pass to be welded in the future. In the figure, the thin dashed arrow indicates that welding passes 103b, 103c, and 103d of beads 101b, 101c, and 101d have been extracted as past passes. Also, the thick dashed arrow indicates that welding pass 103e directed from welding start position Ps toward welding direction D has been extracted as a welding pass to be welded in the future.

[0042] In this embodiment, the trajectory of a past pass overlapping with the vicinity range is used as an example of the second trajectory, which is the trajectory of a second lamination pass that is laminated before the first lamination pass among the multiple lamination passes and overlaps with the vicinity range. Also, in this embodiment, a placement information extraction unit 453 is provided as an example of an extraction unit that extracts the second trajectory based on the order of the multiple lamination passes.

[0043] The shift position determination unit 454 determines a shift position obtained by shifting the welding start position based on the information of the welding pass being calculated obtained by the vicinity range generation unit 452 and the arrangement information extracted by the arrangement information extraction unit 453. Specifically, the shift position determination unit 454 determines the shift position according to the welding start position and welding direction obtained by the vicinity range generation unit 452 and the presence and distribution of teaching points of past passes within the vicinity range extracted by the arrangement information extraction unit 453. Here, the vicinity range is divided into 3×3 meshes, and a combination of the welding start position, welding direction, teaching points of past passes, and shift positions within the meshes is set in advance as a shift pattern, so that the shift position determination unit 454 determines the shift position. That is, the shift position determination unit 454 extracts a corresponding shift pattern by collating the welding start position and welding direction obtained by the vicinity range generation unit 452 and the distribution of teaching points of past passes extracted by the arrangement information extraction unit 453 with pre-stored shift patterns.

[0044] 9(a) to (l) show examples of shift patterns. In each diagram, a combination of a bead of a previous pass (hereinafter, referred to as a "previous bead") B, a welding start position Ps, a welding direction D, and a shift position Pt is set. For example, in Fig. 9(a), the welding start position Ps is set to the right of the previous bead B, so the shift position Pt is set further to the right of the welding start position Ps. On the other hand, in Fig. 9(b), the welding start position Ps is set to the left of the previous bead B, so the shift position Pt is set further to the left of the welding start position Ps. Also, when FIG. 9(c) is rotated 180°, the welding start position Ps and the previous bead B will be the same as in FIG. 9(e). On the other hand, when FIG. 9(c) is rotated 180°, the shift position Pt will not be the same as in FIG. 9(e). This is because the welding direction D with respect to the previous bead B is different between FIG. 9(c) and FIG. 9(e). Thus, in FIG. 9(a) to (l), the shift position Pt is set taking into consideration the welding direction D. However, the shift position Pt may be set without taking into consideration the welding direction D. 9(k) and (l), the shift position Pt does not exist, which means that the welding start position Ps is not shifted.

[0045] In this embodiment, this function is provided in the shift position determination unit 454 as an example of a setting unit that sets the shift position of the stacking start point within the vicinity range based on the first locus and the second locus. Also, in this embodiment, this function is provided in the shift position determination unit 454 as an example of a setting unit that sets the shift position using the correspondence between the first locus, the second locus, and the shift position. Furthermore, in this embodiment, a shift pattern is used as an example of pattern data in which the correspondence is patterned, and this function is provided in the shift position determination unit 454 as an example of a setting unit that sets the shift position by searching the pattern data using the first locus and the second locus.

[0046] Also, the shift position determination unit 454 may determine the shift position using a feature value weighted by the relative position of the teaching point of the past pass to the welding start position, instead of collating the combination of the welding start position, teaching point of the past pass, and the shift position with the shift pattern. Here, a specific feature value may be Σ{(1 / Li)×(Xi, Yi)}, where the relative position of the i-th teaching point of the past pass to the welding start position is (Xi, Yi) and the distance between the welding start position and the teaching point is distance Li. That is, the shift position determination unit 454 may calculate this feature value from the welding start position and the teaching point of the past pass, and extract the shift pattern with the closest feature value. In this way, even if there is no completely matching shift pattern among the stored shift patterns, it is possible to extract a similar shift pattern. In particular, when the teaching point of the past pass in the shift pattern includes the shift position, the number of combinations becomes huge, making it difficult to comprehensively store all the shift patterns, and therefore a determination method using the feature value is preferable.

[0047] In this embodiment, this function is provided in the shift position determination unit 454 as an example of a setting unit that calculates a feature amount indicating a relative positional relationship between the first trajectory and the second trajectory and sets the shift position using the correspondence between the feature amount and the shift position.

[0048] Furthermore, when it is desired to omit the definition of the feature amount, a model may be generated by machine learning the correspondence between the distribution of teaching points of past passes and the shift position, and the shift position determination unit 454 may output the shift direction from the welding start position to the shift position to the model. As a machine learning method, a well-known method such as deep learning in which neural networks are connected in a multilayer, random forest, support vector machine, genetic algorithm, etc. may be used. In addition, as the learning data, for example, the welding start position, welding direction, distribution of teaching points of past passes, and correspondence with the shift direction of each shift pattern may be used, or the correspondence between the feature amount Σ{(1 / Li)×(Xi, Yi)} of each shift pattern and the shift direction may be used. Alternatively, as the learning data, a pattern image in which the shift position is taught may be used.

[0049] In this embodiment, this function is provided in the shift position determination unit 454 as an example of a setting unit that sets a shift position by inputting the first trajectory and the second trajectory into a machine learning model that has learned the correspondence relationship. Also, in this embodiment, this function is provided in the shift position determination unit 454 as an example of a setting unit that sets a shift position by inputting a feature amount into a machine learning model that has learned the correspondence relationship.

[0050] Although not taken into consideration in FIGS. 9(a) to (l), there may be cases where there is no lower layer at a location where a shift position should be set, such as the four corners or ends of a block.

[0051] Fig. 10(a) shows an example of an additive manufacturing object 100B in which it may happen that a lower layer does not exist at a location where the shift position should be set. Fig. 10(a) shows a case where past beads B11-B13 have already been formed, and welding from a welding start position Ps in a welding direction D will now be performed. In this case, as shown by the circled C, the shift position is set at the end of a block where no lower layer exists.

[0052] If the shift position is set in a place where there is no lower layer, watermarks will occur, leading to defective shapes. Therefore, it is necessary to consider exceptional shift patterns, such as making it impossible to set the shift position or restricting the amount of shift to the shift position.

[0053] FIG. 10(b) shows an example of such an exceptional shift pattern. In FIG. 10(b), the welding start position Ps and the previous bead B13 are at the positions shown, and the welding direction D is directed in the direction shown, so according to the example of FIG. 9(e), the shift position Pt should be set to the position shown. However, since there is no lower layer in the mesh in the bottom row shown by the diagonal hatching in FIG. 10(b), the shift position Pt is not set. This is shown by the arrow in the mesh in the bottom row from the shift position Pt to the welding start position Ps being a dashed line. Then, Ptnew, which is shifted as far as possible within the range in which there is a lower layer from the welding start position Ps in the direction of the shift position Pt, is set as the new shift position. In this way, it is possible to adjust the shift position in order to accommodate cases where lamination is performed from the end of the base material or where lamination is performed on an overhanging portion.

[0054] In this embodiment, an exceptional shift pattern is used as an example of a nearby range including a prohibited portion where setting of a shift position is prohibited, and this function of the shift position determination unit 454 is provided as an example of a setting unit that sets a shift position in an area other than the prohibited portion within the nearby range.

[0055] In addition, since the bead width varies depending on the welding conditions, the shift position determination unit 454 may determine the shift position using the contour of the past bead by adding the welding conditions and bead width of the past pass to the calculation in addition to the teaching points of the past pass.

[0056] An example of a shift pattern including bead width information is shown in Fig. 11. Fig. 11 shows a case where previous beads B21 and B22 have already been formed, and welding will now be performed from welding start position Ps in welding direction D. In this case, since the contour of bead B22 is located farther away than the contour of bead B21, shift position determination unit 454 sets shift position Pt on the side of bead B22. In this way, it is possible to adjust the shift position while taking into consideration the bead shape.

[0057] In the present embodiment, this function is provided in shift position determination unit 454 as an example of a setting unit that sets the shift position based on a first locus and a second locus having the contour of a weld bead.

[0058] When the welding pass for which the shift position determination unit 454 has determined the shift position and the next welding pass are in the same layer, the trajectory division unit 455 updates the former welding pass as a past pass. Specifically, the trajectory division unit 455 divides the trajectory of the welding path processed by the vicinity range generation unit 452 at multiple division points and adds them to the data of the teaching points of the past path. Here, the interval between the division points is set to half the bead width or less so that the division points of the trajectory that cross the boundary of the vicinity range are present within the vicinity range. In particular, when the shift position determination unit 454 performs the determination using a mesh, the interval between the division points is set to 1 / 3 or less of the mesh size.

[0059] FIG. 12(a) and (b) show how such division points are set. In FIG. 12(a), the bead of the welding pass processed by the vicinity range generating unit 452 is the past bead B31, which is divided at multiple division points as shown in the figure. However, here, the interval between the division points is longer than 1 / 3 of the mesh size. If the division of the trajectory is too coarse, it will not be recognized as being within the vicinity range. Therefore, the interval between the division points needs to be made finer. Therefore, in Fig. 12(b), the interval between division points is set to 1 / 3 or less of the mesh size, which makes it possible to correctly determine the trajectory passing through the mesh.

[0060] The shift position output unit 456 outputs the shift position determined by the shift position determination unit 454 to the control program generated by the control program generation unit 44, thereby correcting the control program.

[0061] In this embodiment, shift position output section 456 is provided as an example of a correction section that corrects control information based on the shift position.

[0062] The control program output unit 46 outputs the control program generated by the control program generation unit 44 and modified by the control program modification unit 45 to the recording medium 70.

[0063] [Control device functional configuration] 13 is a block diagram showing an example of a functional configuration of the control device 50 according to the present embodiment. As shown in the figure, the control device 50 according to the present embodiment includes a control program acquisition unit 61, a control program storage unit 62, and a control program execution unit 63.

[0064] The control program acquisition unit 61 acquires the control program recorded on the recording medium 70 . The control program storage unit 62 stores the control program acquired by the control program acquisition unit 61 . The control program executing unit 63 reads out and executes the control program stored in the control program storage unit 62. As a result, the control program executing unit 63 executes the control program generated by the control program generating unit 44 and modified by the control program modifying unit 45, thereby controlling the welding robot 10.

[0065] [Stacking Planner Operation] FIG. 14 is a flowchart showing an example of the operation of the stacking planning device 30 in this embodiment.

[0066] In the lamination planning device 30, first, the CAD data acquisition unit 41 acquires three-dimensional CAD data from the CAD device 20 (step 301). Next, the CAD data dividing unit 42 divides the three-dimensional CAD data acquired in step 301 into a plurality of layers to generate layer shape data (step 302). Next, the stacking plan unit 43 generates a stacking plan from the layer shape data generated in step 302 (step 303). Next, the control program generating unit 44 generates a control program (step 304) based on the lamination plan generated in step 303. Specifically, the control program is generated to control the welding robot 10 so as to form a bead according to the lamination plan. Thereafter, the control program correction unit 45 performs a control program correction process for correcting the control program generated in step 304 (step 305). The control program correction process will be described later in detail. Finally, the control program output unit 46 outputs the control program generated in step 304 and modified in step 305 to the recording medium 70 (step 306).

[0067] FIG. 15 is a flowchart showing an example of the control program correction process in step 305 of FIG.

[0068] As shown in the figure, when the control program correction process is started, in control program correction unit 45, first, path information acquisition unit 451 acquires path information from the trajectory data included in the control program generated in step 304 (step 321). Here, the path information includes information on the trajectory of the welding pass and the stacking order, and the information on the trajectory of the welding pass includes information on the welding start position and welding direction of the welding pass.

[0069] Next, the control program correction unit 45 sets the welding pass index k to 1 (step 322). That is, the control program correction unit 45 focuses on the first welding pass. Then, the control program correction unit 45 performs the process of steps 323 to 329 for each index k while incrementing the welding pass index k by 1 up to the number of welding passes.

[0070] That is, the vicinity range generating unit 452 generates a vicinity range of the welding start position of the kth pass (step 323). Here, the vicinity range generating unit 452 generates, for example, a square range divided into a 3×3 mesh centered on the welding start position of the welding pass acquired in step 321 as the vicinity range. In this case, the width of the vicinity range may be, for example, three times the bead width in the lamination conditions. Next, the placement information extraction unit 453 extracts placement information of teaching points of past passes included in the neighborhood range of the kth pass (step 324). Here, the placement information extraction unit 453 identifies teaching points of past passes based on, for example, information on the stacking order of the welding passes acquired in step 321. Then, the placement information extraction unit 453 extracts placement information indicating the presence and distribution of teaching points of past passes that overlap with the neighborhood range generated in step 323 among the teaching points of the past passes. Next, the shift position determination unit 454 determines the shift position of the welding start position of the kth pass (step 325). Here, the shift position determination unit 454 determines the shift position based on, for example, the welding start position and welding direction of the welding pass acquired in step 321 and the arrangement information of the teaching point of the past pass extracted in step 324.

[0071] Next, the control program correcting unit 45 determines whether or not the next welding pass, that is, the (k+1)th pass, exists (step 326). If it is determined in step 326 that a next welding pass exists, the control program correcting unit 45 determines whether or not the next welding pass exists in the same layer (step 327). If it is determined in step 327 that the next welding pass exists in the same layer, the trajectory dividing unit 455 divides the trajectory of the kth pass at a plurality of division points, and stores the plurality of division points as teaching points of the past pass (step 328). On the other hand, if it is determined in step 327 that the next welding pass does not exist in the same layer, the control program correction unit 45 resets (deletes) the teaching points of the past passes in the layer including the kth pass (step 329).

[0072] Thereafter, the control program correction unit 45 adds 1 to the welding pass index k (step 330). That is, the control program correction unit 45 focuses on the next welding pass and returns the process to step 323.

[0073] Then, it is assumed that the control program correction unit 45 repeats such processing and determines that the next welding pass does not exist in step 326. Then, the shift position output unit 456 outputs the shift position determined in step 325 to the control program generated in step 304 of Fig. 14 (step 331).

[0074] (Control device operation) In the control device 50, first, the control program acquisition unit 61 acquires the control program from the recording medium 70 and stores it in the control program storage unit 62. In this state, when the additive manufacturing object 100 is actually manufactured using the welding robot 10, the control program execution unit 63 reads out the control program stored in the control program storage unit 62 and executes it.

[0075] [Variations] In the above description, in the lamination planning device 30, the control program generated by the control program generation unit 44 is corrected by the control program correction unit 45, but this is not limited to the above. In the control device 50, the control program acquired from the lamination planning device 30 by the control program acquisition unit 61 may be corrected by the control program correction unit in the control device 50 before modeling.

[0076] Although not mentioned above, additive manufacturing may be performed based on a modified control program in conjunction with a slag removal process for a portion or the entire surface of each layer.

[0077] [Advantages of this embodiment] In this embodiment, before the manufacturing process, the positional relationship between each welding pass and the previous passes is calculated based on the trajectory data of the welding robot 10 used for the manufacturing process, and the shift position of the welding start position is determined based on the presence and distribution of previous passes within a predetermined range. This reduces poor welding start caused by the target position being on the slag on the adjacent bead.

[0078] In the present embodiment, the shift position of the welding start position is determined based on a pre-defined shift pattern, which reduces the burden on the operator and suppresses arc start errors during modeling. [Explanation of symbols]

[0079] 1...metal additive manufacturing system, 10...welding robot, 20...CAD device, 30...lamination planning device, 41...CAD data acquisition unit, 42...CAD data division unit, 43...lamination planning unit, 44...control program generation unit, 45...control program correction unit, 451...path information acquisition unit, 452...neighborhood range generation unit, 453...placement information extraction unit, 454...shift position determination unit, 455...trajectory division unit, 456...shift position output unit, 46...control program output unit, 50...control device, 61...control program acquisition unit, 62...control program storage unit, 63...control program execution unit, 70...recording medium

Claims

1. A control information correction method for correcting control information for additive manufacturing performed by stacking weld beads, comprising the steps of: obtaining a trajectory and sequence of a plurality of lamination passes of the weld bead; specifying a vicinity range of a lamination start point of a first lamination pass based on a first trajectory that is a trajectory of the first lamination pass among the plurality of lamination passes; extracting a second trajectory, which is a trajectory of a second lamination pass that is laminated before the first lamination pass among the plurality of lamination passes based on an order of the plurality of lamination passes and is a trajectory that overlaps with the vicinity range; setting a shift position of the stacking start point within the vicinity range based on the first trajectory and the second trajectory; modifying the control information based on the shift position; A method for modifying control information, comprising:

2. 2. The control information correction method according to claim 1, wherein in the setting step, the shift position is set using a correspondence relationship between the first locus, the second locus, and the shift position.

3. 3. The control information modifying method according to claim 2, wherein in the setting step, the shift position is set by searching for pattern data in which the correspondence is patterned using the first locus and the second locus.

4. 3. The control information correction method according to claim 2, wherein in the setting step, the shift position is set by inputting the first trajectory and the second trajectory into a machine learning model that has learned the correspondence relationship.

5. 2. The control information correction method according to claim 1, wherein in the setting step, a feature amount indicating a relative positional relationship between the first trajectory and the second trajectory is calculated, and the shift position is set using a correspondence relationship between the feature amount and the shift position.

6. The control information correction method according to claim 5 , wherein in the setting step, the shift position is set by inputting the feature amount into a machine learning model that has learned the correspondence relationship.

7. 2. The control information correcting method according to claim 1, wherein in the setting step, the shift position is set in an area other than a prohibited portion within the vicinity range including a prohibited portion where setting of the shift position is prohibited.

8. 2 . The control information correcting method according to claim 1 , wherein in the setting step, the shift position is set based on the first locus and the second locus having a contour of the weld bead.

9. A control information correction device that corrects control information for additive manufacturing performed by stacking weld beads, comprising: An acquisition unit that acquires a trajectory and a sequence of a plurality of lamination passes of the weld bead; an identifying unit that identifies a vicinity range of a lamination start point of a first lamination pass based on a first trajectory that is a trajectory of the first lamination pass among the plurality of lamination passes; an extraction unit that extracts a second trajectory, which is a trajectory of a second lamination pass that is laminated before the first lamination pass among the plurality of lamination passes based on an order of the plurality of lamination passes and is a trajectory that overlaps with the vicinity range; a setting unit that sets a shift position of the stacking start point within the vicinity range based on the first trajectory and the second trajectory; a correction unit that corrects the control information based on the shift position; A control information modifying device comprising:

10. A control information correction device that corrects control information for additive manufacturing performed by stacking weld beads, A function for acquiring the trajectory and sequence of multiple lamination passes of a weld bead; a function of identifying a vicinity range of a lamination start point of a first lamination pass based on a first trajectory that is a trajectory of the first lamination pass among the plurality of lamination passes; a function of extracting a second trajectory, which is a trajectory of a second lamination pass that is laminated before the first lamination pass among the plurality of lamination passes based on an order of the plurality of lamination passes and is a trajectory that overlaps with the vicinity range; a function of setting a shift position of the stacking start point within the vicinity range based on the first trajectory and the second trajectory; modifying the control information based on the shift position; A program to achieve this.

Citation Information

Patent Citations

  • Method for automatic welding of consumable electrode system welding machine

    JP1992251671A

  • Shaped-object manufacturing method, lamination control apparatus, and program

    JP2020192584A

  • Method for manufacturing laminated molding and display device

    JP2022117176A

  • Layered manufacturing method, layered manufacturing device, and layered manufacturing system

    JP6989059B1

  • Arc-welding apparatus

    JP2014213375A