Welding equipment and welding method

Abstract

To easily acquire an actual rotation position of a cylindrical barrel and properly perform groove welding.SOLUTION: A welding device 1 comprises: a rotation mechanism 2 which integrally rotates two cylindrical barrels 9 in which a groove is formed in a butting part; a torch part 31 which repeatedly forms a weld bead to the whole circumference of the groove by continuously performing groove welding at a prescribed welding position in the circumferential direction with a center axis J1 as the center in parallel with rotation of the cylindrical barrels 9; a welding position movement part 41 which moves a formation position of the weld bead by the torch part 31 in the axial direction; a detection part 42 which detects at a measurement position detection object parts 96, 97 provided on the outer circumferential surface of the cylindrical barrels 9 with the position in the circumferential direction immediately before reaching the welding position as the measurement position in rotation of the cylindrical barrels 9; and a control part 10 which controls the welding position movement part 41 on the basis of detection of the detection object parts 96, 97 by the detection part 42. The welding device 1 can easily acquire the actual rotation position of the cylindrical barrel 9 and properly perform groove welding.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a welding apparatus and a welding method. [Background technology]

[0002] Conventionally, two cylindrical bodies are joined by welding. For example, a groove is formed at the butt joint of the two cylindrical bodies, and groove welding is performed while the two cylindrical bodies are rotated integrally. In this case, multi-layer welding is also performed in which weld beads are repeatedly formed around the entire circumference of the groove.

[0003] Also known is a technique for performing groove welding with high accuracy by measuring the cross-sectional shape of the groove in parallel with the rotation of a cylindrical barrel. For example, the welding apparatus of Patent Document 1 is provided with a non-contact shape measurement sensor that measures the cross-sectional shape of the groove of the welding workpiece. Welding technology data, including welding condition data, is prepared in advance, and correction data for the welding technology data is calculated based on the measurement data from the sensor. The first layer is welded based on the welding technology data, and the second and subsequent layers are welded based on the correction data for the welding technology data. The welding apparatus of Patent Document 2 is provided with a welding condition database in which welding conditions suitable for the groove shape are registered in advance, and a measurement means for measuring the actual groove shape at the weld location. Multiple discrete measurement positions are set along the weld line, and the actual groove shape at each measurement position is measured by the measurement means. Continuous welding conditions are set along the weld line using the measurement results and the welding condition database. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 9-155543 [Patent Document 2] JP 2016-10810 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, in a welding device that performs groove welding while rotating a cylindrical barrel, it may not be possible to accurately obtain the rotational position of the cylindrical barrel. For example, in a welding device that rotates the cylindrical barrel by rotating a roller that contacts the outer surface of the cylindrical barrel, if slippage occurs at the contact surface between the cylindrical barrel and the roller, the rotational position of the cylindrical barrel calculated from the rotation amount of the roller will differ from the actual rotational position of the cylindrical barrel. Therefore, a method is needed to easily obtain the actual rotational position of the cylindrical barrel and perform groove welding appropriately.

[0006] The present invention has been made in consideration of the above-mentioned problems, and has as its object to easily obtain the actual rotational position of a cylindrical shell and to perform groove welding appropriately. [Means for solving the problem]

[0007] A first aspect of the present invention is a welding device comprising two cylindrical barrels arranged in an axial direction parallel to a central axis, with a groove formed at the butt joint of the two cylindrical barrels, a rotation mechanism that rotates the two cylindrical barrels integrally, a torch unit that performs groove welding continuously at predetermined welding positions in a circumferential direction about the central axis while parallel to the rotation of the two cylindrical barrels, thereby repeatedly forming weld beads around the entire circumference of the groove, a welding position movement unit that moves the position where the weld bead is formed by the torch unit in the axial direction, a detection unit that detects a detectable part provided on the outer peripheral surface of at least one of the two cylindrical barrels at the measurement position, with the circumferential position just before the two cylindrical barrels reach the welding position as a measurement position during the rotation of the two cylindrical barrels being the circumferential position, and a control unit that controls the welding position movement unit based on the detection of the detectable part by the detection unit. When repeatedly forming the weld bead, a circumferential position of the groove at which the relative position of the weld bead in the axial direction with respect to the groove should be changed is set as a path transition position, and the control unit identifies a timing at which the path transition position passes the welding position based on detection of the detection target portion by the detection unit. do.

[0011] Aspects of the present invention 2 teeth , melt A connection device comprising: a rotation mechanism for integrally rotating the two cylindrical barrels, the rotation mechanism being configured to rotate the two cylindrical barrels in an axial direction parallel to a central axis, with a groove formed at the butt joint between the two cylindrical barrels; a torch unit for continuously performing groove welding at predetermined welding positions in a circumferential direction about the central axis while parallel to the rotation of the two cylindrical barrels, thereby repeatedly forming weld beads around the entire circumference of the groove; a welding position moving unit for moving the position at which the weld bead is formed by the torch unit in the axial direction; a detection unit for detecting a detectable portion provided on the outer peripheral surface of at least one of the two cylindrical barrels at the measurement position, the circumferential position just before the two cylindrical barrels reach the welding position during the rotation of the two cylindrical barrels being set as a measurement position; and a control unit for controlling the welding position moving unit based on the detection of the detectable portion by the detection unit. The detection unit is a non-contact shape measuring machine that measures the cross-sectional shape of the groove at the measurement position, and also detects the cross-sectional shape of the detection target portion that has reached the measurement position.

[0012] Aspects of the present invention 3 is the aspect 2 The control unit determines the axial position at which the weld bead should be formed for each portion of the groove that passes through the welding position, using multiple cross-sectional shapes measured by the detection unit for multiple portions located in the vicinity of each portion of the groove.

[0013] Aspects of the present invention 4 is the aspect 2 (Aspect 2 or 3 ) welding device, wherein the control unit determines the axial position at which the weld bead should be formed for each portion of the groove that passes through the welding position based on a cross-sectional shape measured immediately before by the detection unit for each portion, and in a specific state in which the groove is unclear in the cross-sectional shape measured immediately before, the control unit determines the axial position at which the weld bead should be formed for each portion based on the cross-sectional shape used in the previous formation of the weld bead for each portion of the groove.

[0014] Aspects of the present invention 5 is the aspect 4 In the welding device, in the specific state, the control unit acquires the axial position of the groove at the measurement position based on the cross-sectional shape of the detected portion measured by the detection unit, and determines the axial position at which the weld bead should be formed for each portion of the groove.

[0015] Aspects of the present invention 6 is the aspect 2 (Aspect 2 Or 5 ) is a welding device, wherein the control unit determines welding heat input conditions for each portion of the groove passing through the welding position using the groove depth obtained for the entire circumference of the groove so that the depth of the surface of the weld bead to be formed is constant around the entire circumference of the groove.

[0016] Aspects of the present invention 7 teeth , melt A connection device comprising: a rotation mechanism for integrally rotating the two cylindrical barrels, two cylindrical barrels being arranged around a central axis in an axial direction parallel to the central axis, with a groove formed at the butt joint of the two cylindrical barrels; a torch unit for continuously performing groove welding at predetermined welding positions in a circumferential direction around the central axis in parallel with the rotation of the two cylindrical barrels, thereby repeatedly forming weld beads around the entire circumference of the groove; a welding position moving unit for moving the position at which the weld bead is formed by the torch unit in the axial direction; a detection unit for detecting a detectable part provided on the outer peripheral surface of at least one of the two cylindrical barrels at the measurement position, with the circumferential position just before the cylindrical barrels reach the welding position as a measurement position during the rotation of the two cylindrical barrels being a measurement position; and a control unit for controlling the welding position moving unit based on the detection of the detectable part by the detection unit. a slag removal unit for removing slag generated by the groove welding; and The slag removal unit comprises a plurality of slag removal sections arranged in an arrangement direction roughly along the longitudinal direction of the groove, with their ends positioned within the groove; a support section having a member extending in the arrangement direction and supporting the plurality of slag removal sections; and a rotation mechanism that changes the inclination angle of the arrangement direction with respect to the longitudinal direction of the groove while the support section is movable in the axial direction, thereby bringing the ends of two slag removal sections located at both ends of the plurality of slag removal sections into contact with both side surfaces of the groove. An eighth aspect of the present invention is the welding device according to any one of the first to seventh aspects, wherein a plurality of detection targets are arranged in the circumferential direction on the outer circumferential surface of the at least one cylindrical barrel. A ninth aspect of the present invention is a welding apparatus according to any one of aspects 1 to 7 (or any one of aspects 1 to 8), in which the detectable part is fixed to the outer peripheral surface of at least one of the cylindrical bodies by magnetic force.

[0017] Aspects of the present invention 10 The first to third aspects are 7 Any one of the following (modes 1 to 9 The welding device may be any one of the above, wherein the torch unit has a plurality of welding torches for submerged arc welding, the plurality of welding torches are arranged adjacent to one another, and the welding device further includes a wire cutting unit that can sequentially cut the wires of the plurality of welding torches.

[0018] Aspects of the present invention 11a welding method comprising: a) a step of arranging two cylindrical barrels centered on a central axis in an axial direction parallel to the central axis, a groove formed at a butt joint between the two cylindrical barrels, and integrally rotating the two cylindrical barrels; b) a step of repeatedly forming weld beads around the entire circumference of the groove by continuously performing groove welding at predetermined welding positions in a circumferential direction centered on the central axis in parallel with the step a), c) a step of detecting a detectable portion provided on an outer peripheral surface of at least one of the two cylindrical barrels at the measurement position during the rotation of the two cylindrical barrels, with the circumferential position just before the cylindrical barrel reaches the welding position being set as a measurement position in parallel with the step a), and d) a step of moving the formation position of the weld bead in the axial direction based on the detection of the detectable portion. When repeatedly forming the weld bead, a circumferential position of the groove at which the relative position of the weld bead in the axial direction with respect to the groove should be changed is set as a path transition position, and based on the detection of the detection target portion in the step c), a timing at which the path transition position passes the welding position is identified in the step d). do. A twelfth aspect of the present invention is a welding method comprising: a) a step of arranging two cylindrical barrels centered on a central axis in an axial direction parallel to the central axis, with a groove formed at the butt joint of the two cylindrical barrels, and a step of rotating the two cylindrical barrels integrally; b) a step of repeatedly forming a weld bead around the entire circumference of the groove by continuously performing groove welding at predetermined welding positions in a circumferential direction centered on the central axis, in parallel with the step a); c) a step of detecting a detectable portion provided on the outer surface of at least one of the two cylindrical barrels by a detector disposed at the measurement position during the rotation of the two cylindrical barrels, with the circumferential position just before the cylindrical barrels reach the welding position being set as a measurement position; and d) a step of moving the formation position of the weld bead in the axial direction based on the detection of the detectable portion, wherein the detector is a non-contact shape measuring machine that measures the cross-sectional shape of the groove at the measurement position and also detects the cross-sectional shape of the detectable portion that has reached the measurement position. [Effects of the Invention]

[0019] According to the present invention, the actual rotational position of the cylindrical shell can be easily obtained, and groove welding can be performed appropriately. [Brief explanation of the drawings]

[0020] [Figure 1] FIG. 1 is a diagram showing a configuration of a welding device. [Figure 2] FIG. [Figure 3] FIG. 2 is a diagram showing cross-sectional shapes at a plurality of positions of the groove. [Figure 4] FIG. 2 is a diagram showing a flow of a welding process by a welding device. [Figure 5] FIG. 2 is a diagram showing the arrangement of a reference magnet and a corrector magnet. [Figure 6A] FIG. 10 is a diagram showing a reference magnet. [Figure 6B] FIG. 10 is a diagram showing a reference magnet. [Figure 7] FIG. 10 is a diagram showing the vicinity of a path transition position in a groove. [Figure 8] FIG. [Figure 9A] FIG. 10 is a diagram for explaining correction of the welding direction position. [Figure 9B] FIG. 10 is a diagram for explaining correction of the welding direction position. [Figure 10] FIG. 10 is a diagram showing a groove and a correction magnet. [Figure 11A] FIG. 2 is a diagram showing the cross-sectional shapes of a groove and a correction magnet. [Figure 11B] FIG. 2 is a diagram showing the cross-sectional shapes of a groove and a correction magnet. [Figure 12] FIG. 10 is a diagram showing a flow of adaptive control processing. [Figure 13A] FIG. [Figure 13B] FIG. [Figure 13C] FIG. [Figure 14] FIG. 4 is a diagram showing the groove depth of each portion of the groove in the welding direction. [Figure 15] FIG. 2 is a diagram showing the configuration of a wire cutting unit. [Figure 16A] 10A and 10B are diagrams for explaining cutting of a wire by a wire cutting unit. [Figure 16B] 10A and 10B are diagrams for explaining cutting of a wire by a wire cutting unit. [Figure 17] FIG. 2 is a diagram showing the configuration of a slag removal unit. [Figure 18A] 4A and 4B are diagrams illustrating the structure of a first support base and a rotation unit. [Figure 18B] 4A and 4B are diagrams illustrating the structure of a first support base and a rotation unit. [Figure 19A] 10A and 10B are views showing a support body guide and a support body. [Figure 19B] 10A and 10B are views showing a support body guide and a support body. [Figure 20A] FIG. 10 is a view showing a plurality of second slag removal units. [Figure 20B] FIG. 10 is a view showing a plurality of second slag removal units. [Figure 21A] 10A and 10B are diagrams showing other examples of the support body guide portion. [Figure 21B] 10A and 10B are diagrams showing other examples of the support body guide portion. DETAILED DESCRIPTION OF THE INVENTION

[0021] FIG. 1 is a diagram showing the configuration of a welding apparatus 1 according to one embodiment of the present invention. The welding apparatus 1 of FIG. 1 is an apparatus that joins multiple cylindrical bodies 9 by submerged arc welding. The welding apparatus 1 may perform welding other than submerged arc welding. The welding apparatus 1 includes a rotation mechanism 2, a torch unit 31, a flux supply device 32, a welding position movement unit 41, a detection unit 42, a slag removal unit 5, a wire cutting unit 61, and a control unit 10. The control unit 10 includes an automatic control panel 11 and a computer 12. The control unit 10 is electrically connected to each component of the welding apparatus 1 and is responsible for overall control of the welding apparatus 1.

[0022] In the welding device 1, two cylindrical barrels 9, each centered on a central axis J1 and having approximately the same outer diameter, are arranged in a direction parallel to the central axis J1 (hereinafter referred to as the "axial direction"). Typically, the two cylindrical barrels 9 are formed from metal. The opposing end faces of the two cylindrical barrels 9 (i.e., the end faces facing the axial direction) come into contact with each other as a butt joint. A groove is formed in the butt joint, opening radially outward from the central axis J1. The groove extends in the circumferential direction from the central axis J1, and is provided around the entire circumference of the cylindrical barrels 9.

[0023] The rotation mechanism 2 has a plurality of rollers 21. In one example, two rollers 21 arranged apart in the left-right direction in FIG. 1 form a roller pair, and the plurality of roller pairs are arranged in the axial direction. The two cylindrical barrels 9 are placed on the plurality of roller pairs. In this way, the cylindrical barrels 9 are supported from below by the plurality of rollers 21 that contact the outer circumferential surface of each cylindrical barrel 9. Some of the rollers 21 are connected to a motor, for example, via a reducer, and the rotation of the rollers 21 causes the two cylindrical barrels 9 to rotate integrally about the central axis J1. Typically, the rotation speed of the cylindrical barrels 9 is approximately constant.

[0024] An encoder (roller encoder or rotary encoder) 22 is provided on one roller 21, and the amount of rotation of the roller 21 is acquired. This indirectly acquires the rotational position of the cylindrical barrel 9 in the circumferential direction. The control unit 10 can identify the position of each part of the groove in the circumferential direction based on the output value of the encoder 22. As described above, the two cylindrical barrels 9 rotate integrally, and therefore, in the following description, the two cylindrical barrels 9 will also be simply referred to as "cylindrical barrel 9." As will be described later, three or more cylindrical barrels 9 may be arranged in the axial direction.

[0025] The torch unit 31 has a plurality of welding torches 311 and a flux dispersion tube 316. The plurality of welding torches 311 and the flux dispersion tube 316 are arranged in the circumferential direction. The welding apparatus 1 of FIG. 1 is provided with two welding torches 311 for submerged arc welding. The tips of the two welding torches 311 and the tip of the flux dispersion tube 316 are positioned close to each other and face the groove near the top of the cylindrical body 9.

[0026] In the example of FIG. 1, the cylindrical barrel 9 rotates counterclockwise around the central axis J1 (see arrow C in FIG. 1). The two welding torches 311 are disposed downstream of the flux spreader pipe 316 in the rotation direction of the cylindrical barrel 9. In other words, each portion of the groove in the circumferential direction passes through the flux spreader pipe 316 and the two welding torches 311 in sequence. Each welding torch 311 and the cylindrical barrel 9 are electrically connected to a welding power source (not shown). Groove welding is performed near the tips of the two welding torches 311. In the following description, the positions near the tips of the two welding torches 311 in the circumferential direction are referred to as "welding positions." The welding positions are circumferential positions where welding is performed on the groove by the torch units 31. The number of welding torches 311 provided in the welding device 1 may be one or three or more.

[0027] Each welding torch 311 is supplied with wire, which is a filler material, from a wire supply unit 361. FIG. 1 shows only one wire supply unit 361. A wire feeder (not shown) and a wire breakage detection unit 362 are provided between the welding torch 311 and the wire supply unit 361. The wire feeder feeds the wire from the wire supply unit 361 into the welding torch 311. The wire breakage detection unit 362 detects wire breakage when there is no more wire passing through. When wire breakage detection unit 362 detects wire breakage, the welding process is stopped by the control unit 10.

[0028] The flux supply device 32 includes a flux replenishment unit 321, a hopper unit 322, a level sensor 323, and an automatic valve 324. The hopper unit 322 stores flux. The flux replenishment unit 321 automatically replenishes flux into the hopper unit 322. The level sensor 323 detects a flux shortage when the flux in the hopper unit 322 falls below a predetermined level due to a malfunction of the flux replenishment unit 321 or the like. When a flux shortage is detected, the control unit 10 issues a warning to the operator. The hopper unit 322 is connected to a flux spraying pipe 316 via an automatic valve 324. When the automatic valve 324 is opened, flux is supplied from the hopper unit 322 to the flux spraying pipe 316, and the flux is supplied from the tip of the flux spraying pipe 316 to the groove of the cylindrical body 9. When the automatic valve 324 is closed, the supply of flux to the groove is stopped. The cross-sectional shape of the groove, which will be described later, is obtained by sensing the bead at a position not covered with flux.

[0029] The detector 42 is positioned facing the groove of the cylindrical barrel 9 and measures the cross-sectional shape of the groove along the radial and axial directions. The detector 42 is, for example, a non-contact shape measuring device, preferably a laser scanning sensor or a sensor using a laser slit. The detector 42 is positioned upstream of the torch unit 31 in the rotation direction of the cylindrical barrel 9, and each portion of the groove in the circumferential direction passes through the detector 42 and the torch unit 31 in order. The circumferential position of the groove measured by the detector 42 is called the "measurement position." This measurement position is the circumferential position of the cylindrical barrel 9 just before it reaches the welding position. The distance between the welding position and the measurement position in the direction perpendicular to the axis is preferably 120 mm or more to prevent flux from covering the measurement position. The detector 42 can detect the surface position (i.e., the cross-sectional shape of the surface) of the cylindrical barrel 9 to a certain extent on both outer sides of the groove in the axial direction.

[0030] Welding position movement unit 41 has, for example, a ball screw and a motor, and can move torch unit 31 and detection unit 42 together in two mutually perpendicular directions. These two directions are the axial direction and the radial direction. The relative positions of multiple welding torches 311, flux spray tubes 316, and detection unit 42 do not change before and after movement by welding position movement unit 41. In FIG. 1, welding position movement unit 41 is indicated by a dashed-line block.

[0031] The slag removal unit 5 is provided at a position circumferentially away from the welding position and the measurement position, and is disposed facing the groove of the cylindrical body 9. The slag removal unit 5 removes slag generated in the groove by groove welding. The wire cutting unit 61 is selectively disposed at a cutting position close to the tips of the multiple welding torches 311 and at a retracted position away from the tips (see FIG. 15 described below). The wire cutting unit 61 cuts the wires protruding from the tips of the multiple welding torches 311. The configurations and operations of the slag removal unit 5 and the wire cutting unit 61 will be described later.

[0032] In the welding device 1, a plurality of cameras (e.g., CCD cameras) 661, 662 are also provided around the cylindrical barrel 9, making it possible to check the state of welding by the torch unit 31. The welding device 1 having the above configuration can be remotely controlled, and even when welding a large cylindrical barrel 9 with an outer diameter of several meters, for example, the operator can operate it from the ground without climbing to the top of the cylindrical barrel 9.

[0033] Here, we will explain the tracking control in the welding device 1. Fig. 2 is a diagram showing a cross section of the groove 91 along the radial and axial directions, and shows the state of the groove 91 during the welding process. In the welding device 1, the torch unit 31 continuously performs groove welding at the welding position while the cylindrical body 9 rotates, thereby repeatedly forming weld beads 92 around the entire circumference of the groove 91. In practice, multi-layer welding is performed while changing the axial position of the weld bead 92 relative to the groove 91 every time the cylindrical body 9 makes one revolution.

[0034] Depending on the circumferential roundness and axial horizontality of the cylindrical barrel 9, the position of the cylindrical barrel 9 may shift axially due to rotation of the cylindrical barrel 9. In this case, the position of the groove 91 in the axial direction will not be constant. The same applies when the groove 91 itself is formed in a meandering manner. Furthermore, as the formation of the weld bead 92 is repeated, the position of the surface of the weld bead 92 located at the outermost position in the radial direction (hereinafter referred to as the "bottom of the groove 91") changes, which changes the distance between the bottom of the groove 91 and the welding torch 311. Therefore, in the welding device 1, a tracing control is performed in which the control unit 10 automatically controls the welding position moving unit 41 based on the cross-sectional shape of the groove 91 measured by the detection unit 42. Through the tracing control, the welding torch 311 is positioned at a constant relative position to the groove 91 in the axial direction while the cylindrical barrel 9 makes one revolution. Furthermore, the distance between the bottom of the groove 91 and the welding torch 311 is maintained constant. The tracing control will be described in detail below.

[0035] FIG. 3 is a diagram illustrating the cross-sectional shape of the groove 91 at multiple locations in the circumferential direction. The detector 42 measures the cross-sectional shape of the groove 91 at a constant measurement period R (sensing period). The upper row of FIG. 3 illustrates times Ta, Tb, Tc, Td, and Te for each measurement period R at which the cross-sectional shape of the groove 91 is measured. The middle row illustrates the groove 91 viewed along the radial direction, with the symbols Ya, Yb, Yc, Yd, and Ye indicating the circumferential positions of the groove 91 that pass through the measurement positions at each time Ta to Te (hereinafter referred to as "measurement points"). The middle row also illustrates the axial direction with an arrow, and the relative direction of travel of the torch 31 and the detector 42 relative to the groove 91 of the rotating cylindrical body 9 is also illustrated by an arrow as the "welding direction" (similar to other drawings illustrating the groove 91 viewed along the radial direction). The welding direction is substantially the same as the circumferential direction and is also referred to as the weld line direction. In the welding device 1, the coordinates of the welding direction are set for each portion of the groove 91 in the circumferential direction. The lower row shows the cross-sectional shape of the groove 91 measured by the detector 42 at each of times Ta to Te. In the cross-sectional shape in the lower row, the up-down direction corresponds to the axial direction, and the left-right direction corresponds to the radial direction. As described above, the detector 42 measures the surface position of the cylindrical body 9, and therefore, in the cross-sectional shape of the groove 91 with the weld bead 92 formed therein, the bottom of the groove 91 indicates the surface of the weld bead 92 located on the outermost side. Note that the measurement period in the detector 42 does not necessarily have to be constant.

[0036] For example, for measurement point Yb of groove 91, whose cross-sectional shape is measured at time Tb, the average cross-sectional shape is calculated by averaging the cross-sectional shapes of measurement point Yb and measurement points Ya and Yc adjacent to measurement point Yb (i.e., the cross-sectional shape obtained by averaging the radial positions at each axial position). Then, when measurement point Yb of groove 91 passes the welding position, the axial and radial positions (also called the tracing position) at which welding torch 311 should be placed are determined from the average cross-sectional shape, and welding position moving unit 41 places welding torch 311 at the axial and radial positions. In the example of FIG. 2, when forming the central weld bead 92, welding torch 311 is placed at an axial position that is the center of groove 91 in the average cross-sectional shape. When forming the left weld bead 92, welding torch 311 is placed at an axial position shifted leftward from the center of groove 91. When forming right weld bead 92, welding torch 311 is placed at an axial position shifted to the right from the center of groove 91. The radial position at which welding torch 311 should be placed is a position that is a predetermined distance from the bottom surface of groove 91 (surface of weld bead 92) that faces welding torch 311.

[0037] As described above, the measurement position is located upstream of the welding position in the rotation direction, i.e., ahead of the welding direction, so the cross-sectional shapes of measurement points Ya to Yc are obtained before (just before) measurement point Yb passes the welding position. Similarly, when measurement point Yc of groove 91, whose cross-sectional shape is measured at time Tc, passes the welding position, an average cross-sectional shape is obtained by averaging the cross-sectional shapes of measurement points Yb to Yd, and the axial and radial positions at which welding torch 311 should be placed are determined from this average cross-sectional shape.

[0038] In this way, the axial and radial positions at which the welding torch 311 should be positioned relative to each measurement point on the groove 91 in the welding direction are determined by averaging (moving averages) the cross-sectional shapes of the measurement point and measurement points located in the vicinity of the measurement point. Note that the average cross-sectional shape does not necessarily have to be determined. For example, for each cross-sectional shape, coordinate values ​​of two corners P11 and P14 on the opening side of the groove 91 and two corners P12 and P13 on the bottom side are acquired, and the average of the coordinate values ​​of each corner P11 to P14 (hereinafter also referred to as "feature points P11 to P14") is determined for each cross-sectional shape. Then, instead of the average cross-sectional shape, the axial and radial positions at which the welding torch 311 should be positioned are determined from a shape obtained by connecting the positions of the average coordinate values ​​of each feature point P11 to P14. The feature points P11 to P14 may be identified by a well-known method, such as by determining points at which the inclination changes by a predetermined value or more.

[0039] Furthermore, for a position between each measurement point and an adjacent measurement point, the cross-sectional shape for that position is determined by an interpolation calculation (e.g., linear interpolation) using two average cross-sectional shapes for these two measurement points. For example, for each position between measurement point Yb and measurement point Yc, the cross-sectional shape for that position is determined by an interpolation calculation using the average cross-sectional shape of measurement point Yb and the average cross-sectional shape of measurement point Yc. Then, when that position passes the welding position, the axial and radial positions at which welding torch 311 should be placed are determined from the cross-sectional shape for that position, and welding torch 311 is placed at those axial and radial positions.

[0040] Incidentally, the detection unit 42 may measure an incorrect cross-sectional shape due to the surface condition of the groove 91, etc. In the example of Fig. 3, an incorrect cross-sectional shape is measured for measurement point Ye of the groove 91, whose cross-sectional shape is measured at time Te. In this case, for example, when calculating the average cross-sectional shape for measurement point Yd, if the cross-sectional shapes of measurement points Yc to Ye are used, the average cross-sectional shape will be abnormal.

[0041] Therefore, the cross-sectional shape of measurement point Ye, which has an invalid cross-sectional shape, is replaced with, for example, the average cross-sectional shape for measurement point Yc obtained immediately before, and the average cross-sectional shape for measurement point Yd is obtained using the cross-sectional shapes of measurement points Yc and Yd and the average cross-sectional shape of measurement point Yc. The same applies when the cross-sectional shape cannot be measured at one measurement point (when the cross-sectional shape is missing) due to a malfunction or the like. To determine whether each measurement point has an invalid cross-sectional shape, for example, the amount of deviation of the positions of characteristic points P11 to P14 between the cross-sectional shape of the measurement point and the cross-sectional shape of the measurement point adjacent to the measurement point on the rear side of the measurement point in the welding direction is calculated. If the deviation is equal to or greater than a threshold value (e.g., 5 mm), the cross-sectional shape of the measurement point is determined to be invalid. Note that if measurement points with invalid cross-sectional shapes continue for a predetermined distance (e.g., 300 mm) or more, the control unit 10 stops the welding process.

[0042] Next, the welding process for the cylindrical barrel 9 using the welding apparatus 1 will be described with reference to FIG. 4. In the welding process, first, one reference magnet 96 and multiple compensation magnets 97 are attached to the outer circumferential surface of the cylindrical barrel 9 shown in FIG. 1 (step S11). The one reference magnet 96 is placed near the groove 91 at an arbitrary position in the welding direction. The multiple compensation magnets 97 are arranged in the welding direction near the groove 91. The multiple compensation magnets 97 are preferably arranged at approximately regular intervals (e.g., 2 to 3 m), but the intervals may vary. Both the reference magnet 96 and the compensation magnets 97 are cylindrical and fixed to the outer circumferential surface of the cylindrical barrel 9 by magnetic force. In a preferred processing example, as shown in FIG. 5, the reference magnet 96 is placed near one edge of the groove 91, and the multiple compensation magnets 97 are placed near the other edge of the groove 91. In other words, the reference magnet 96 and the compensation magnets 97 are placed on opposite sides of the groove 91. As will be described later, the reference magnet 96 and the corrector magnet 97 are parts to be detected by the detection unit 42. The reference magnet 96 and the corrector magnet 97 preferably used in this processing example are heat-resistant magnets that do not lose much magnetic force even at high temperatures (e.g., 300°C).

[0043] Next, rotation of the cylindrical barrel 9 by the rotation mechanism 2 is started (step S12). In the torch unit 31, groove welding is continuously performed at the welding position in parallel with the rotation of the cylindrical barrel 9, thereby repeatedly forming weld beads 92 around the entire circumference of the groove 91 (step S13). At this time, the above-mentioned tracking control is performed, and the weld beads 92 are formed at a constant relative position to the groove 91 in the axial direction while the cylindrical barrel 9 makes approximately one rotation. In addition, when the reference magnet 96 and the corrector magnet 97 pass the measurement position, the detection unit 42 detects the reference magnet 96 and the corrector magnet 97 (step S14). Below, the processing based on the detection of the reference magnet 96 will be described first, and then the processing based on the detection of the corrector magnet 97 will be described. Note that while FIG. 4 shows steps S13 and S14 arranged in series, in reality, the detection of the reference magnet 96 and the corrector magnet 97 in step S14 is performed in parallel with the repeated formation of the weld beads in step S13. The same applies to the relationship between step S13 and step S15 described later.

[0044] FIG. 6A is a diagram illustrating the detection of the reference magnet 96 by the detection unit 42. In FIG. 6A, the positions in the welding direction of the groove 91 where the cross-sectional shape is measured (i.e., the positions of the measurement points) are indicated by dashed lines. As described above, the detection unit 42 can detect the surface position (i.e., the cross-sectional shape of the surface) of the cylindrical barrel 9 up to a certain range on both sides of the groove 91 in the axial direction. Therefore, at each measurement point that overlaps with the reference magnet 96 in the axial direction, the cross-sectional shape of the reference magnet 96 is detected along with the cross-sectional shape of the groove 91. In FIG. 6A, point E11 indicates the edge of the upper end surface of the reference magnet 96 (the edge located on the groove 91 side), which is identified from the cross-sectional shapes measured at each measurement point.

[0045] The reference magnet 96 has a cylindrical shape, and the control unit 10 determines an approximate circle that passes through multiple points E11. This makes it possible to identify the center of the reference magnet 96 with greater accuracy than when a prismatic reference magnet 96 is used, as shown in FIG. 6B. The diameter of the reference magnet 96 may be input in advance, and the control unit 10 may determine the center of the reference magnet 96 using only the coordinates of two points E11. The detection unit 42 does not necessarily need to detect the entire cross-sectional shape of the reference magnet 96, and may only detect a portion of the cross-sectional shape of the reference magnet 96 (the cross-sectional shape including the above-mentioned edges).

[0046] FIG. 7 is a diagram illustrating the vicinity of the pass transition position in the groove 91. The pass transition position is a position in the welding direction (circumferential direction) of the groove 91 where the axial relative position of the weld bead 92 with respect to the groove 91 should be changed. The pass transition position is set near the reference magnet 96. The control unit 10 acquires the distance in the welding direction from the time point when the center of the reference magnet 96 passes the measurement position until the pass transition position reaches the welding position (hereinafter referred to as the "delay distance"). After the center of the reference magnet 96 passes the measurement position, when the encoder 22 detects a movement of the delay distance, it is determined that the pass transition position has reached the welding position. As a result, even if slippage occurs between the contact surface between the roller 21 of the rotation mechanism 2 and the outer circumferential surface of the cylindrical barrel 9 or between the roller 21 and the encoder 22 (hereinafter collectively referred to as "slippage of the roller 21, etc."), the timing at which the pass transition position passes the welding position can be accurately determined based on the detection of the reference magnet 96 by the detection unit 42.

[0047] 7 shows groove 91 immediately before the formation of one weld bead 92 (hereinafter referred to as "current weld bead 92"), and the path along which welding torch 311 moves relative to the weld bead 92 when the current weld bead 92 is formed is indicated by a dashed arrow. Also, weld bead 92 formed previously (hereinafter referred to as "previous weld bead 92") is indicated by parallel diagonal lines, and the path transition position of previous weld bead 92 is indicated by reference symbol P31. The axial position of previous weld bead 92 changes relative to groove 91 from path transition position P31 toward the welding direction, and the change in the axial position is completed at the position indicated by reference symbol P32.

[0048] For current weld bead 92, position P33, which is a predetermined distance L1 (e.g., several mm) away from position P32 in the welding direction, is set as the pass transition position. Then, when the pass transition position passes the welding position, welding position movement unit 41 moves welding torch 311, and the formation position of weld bead 92 is moved in the axial direction (step S15). As described above, in welding apparatus 1, pass transition is performed to change the relative axial position of weld bead 92 based on the detection of reference magnet 96 by detection unit 42. As shown in FIG. 7 , by overlapping previous weld bead 92 and current weld bead 92 by distance L1, it is possible to eliminate welding defects at the pass transition portion. Note that distance L1 may be changed appropriately depending on the axial position of weld bead 92, etc., and may be zero.

[0049] In the welding device 1, the above-described tracking control and path transition are automatically performed, and weld beads 92 are repeatedly formed around the entire circumference of the groove 91. As a result, as shown in FIG. 8, most of the groove 91 is filled with the weld beads 92. In FIG. 8, a group of weld beads 92 filling the groove 91 is indicated by parallel diagonal lines. When most of the groove 91 is filled, the cross-sectional shape measured by the detection unit 42 becomes unclear about the groove 91 (hereinafter referred to as a "specific state"). In other words, the characteristic points in the cross-sectional shape become unclear, and the position of the groove 91 cannot be accurately identified. As a result, normal tracking control cannot be performed. Tracking control in a specific state will be described below. Note that the determination of whether or not the specific state exists may be performed using any method. For example, if the maximum depth in the cross-sectional shape of the groove 91 is equal to or less than a threshold value (e.g., 15 mm), the specific state is determined.

[0050] Here, when forming weld bead 92 at each portion of groove 91 in the welding direction, control unit 10 stores the cross-sectional shape used for that portion when forming weld bead 92 the previous time (hereinafter simply referred to as the "previous cross-sectional shape"). In FIG. 8, characteristic points P11 to P14 of the previous cross-sectional shape at the cross-sectional portion are indicated by black dots. Also, even in a specific state, the cross-sectional shape of groove 91 is measured by detection unit 42, and in FIG. 8, the cross-sectional shape measured immediately before for that portion (hereinafter simply referred to as the "current cross-sectional shape") is indicated by a thick dashed line.

[0051] When the specific state is reached, the control unit 10 determines the depth D1 of the groove 91 from the current cross-sectional shape. The depth D1 is the radial distance between the deepest position in the current cross-sectional shape and the outer peripheral surface of the cylindrical barrel 9. In a shape connecting the characteristic points P11 to P14 of the previous cross-sectional shape, points P15 and P16 are determined, where the points P15 and P16 intersect with a line indicating the position of the depth D1 (a line parallel to the line indicating the outer peripheral surface of the cylindrical barrel 9, and shown by a dashed line in FIG. 8 ). A new cross-sectional shape is obtained with the points P15 and P16 as the two characteristic points on the bottom side. In other words, the shape connecting the characteristic points P11, P15, P16, and P14 is the cross-sectional shape to be referenced when forming the weld bead 92 at that portion of the groove 91. Then, the axial and radial positions at which the welding torch 311 should be placed are determined from the cross-sectional shape, and the welding position moving unit 41 places the welding torch 311 at the axial and radial positions. As described above, in the tracking control in a specific state, the axial and radial positions at which the weld bead 92 should be formed for each portion of the groove 91 are determined based on the cross-sectional shape used to form the previous weld bead 92 for that portion.

[0052] However, if slippage of the rollers 21 or the like occurs, the rotational position of the cylindrical barrel 9 acquired by the encoder 22 differs from the actual rotational position of the cylindrical barrel 9. As a result, in the tracking control in the specific state described above, when forming a weld bead 92 at each portion of the groove 91 in the welding direction, it becomes impossible to accurately identify the cross-sectional shape used when forming the previous weld bead 92 at that portion (i.e., the previous cross-sectional shape). Therefore, in the welding device 1, processing is performed based on the detection of the correction magnet 97. In the processing based on the detection of the correction magnet 97, correction of the welding direction position and correction of the axial position are performed. Below, correction of the welding direction position will be described first, and then correction of the axial position will be described.

[0053] 9A and 9B are diagrams for explaining correction of the welding direction position. The upper parts of Fig. 9A and Fig. 9B show multiple correction magnets 97 and the groove 91, and the lower parts show the cross-sectional shape of the groove 91. Fig. 9A corresponds to a non-specific state, i.e., a normal state, and Fig. 9B corresponds to a specific state.

[0054] In a normal state, each time the cross-sectional shape of the groove 91 is measured by the detector 42, the coordinates of the measurement point in the welding direction are determined based on the output value of the encoder 22 and stored together with the cross-sectional shape. Furthermore, as described above, normal tracing control and path transition are performed. Furthermore, each time each corrector magnet 97 passes the measurement position, the coordinates of the center of the corrector magnet 97 in the welding direction are determined based on the output value of the encoder 22, in a manner similar to the detection of the reference magnet 96 described with reference to FIG. 6A. The coordinates of the corrector magnet 97 are stored together with the cross-sectional shape relative to the center position of the corrector magnet 97 (e.g., the cross-sectional shape obtained by the interpolation calculation described above). In FIG. 9A, the cross-sectional shapes relative to the center positions of the corrector magnets 97 are shown in the upper row and in the lower row. In a normal state, the acquisition of the coordinates and cross-sectional shape of each measurement point and the coordinates and cross-sectional shape of each corrector magnet 97 (center) are repeated. In practice, the coordinates and cross-sectional shape of each measurement point, as well as the coordinates and cross-sectional shape of each corrector magnet 97, are updated each time the cylindrical barrel 9 rotates once.

[0055] 8, when forming weld bead 92 for each portion of groove 91 in the welding direction, the cross-sectional shape used when forming weld bead 92 the previous time for that portion (i.e., the previous cross-sectional shape) is used. In the specific state, when each corrector magnet 97 passes the measurement position, the welding direction coordinate of the portion of groove 91 passing the measurement position, i.e., the coordinate of that portion specified by the output value of encoder 22, is corrected to the welding direction coordinate acquired for that corrector magnet 97 one revolution before.

[0056] As a result, even if slippage of rollers 21 or the like occurs, the welding direction coordinates identified by encoder 22 are corrected for each interval between corrector magnets 97 in the welding direction. As a result, during tracking control in a specific state, when forming weld bead 92 at each portion of groove 91, it is possible to identify with some degree of accuracy the cross-sectional shape used when forming weld bead 92 at that portion the previous time. In FIG. 9B corresponding to the specific state, the previous cross-sectional shape for the position of each corrector magnet 97 shown in the upper row is shown in the lower row. As described above, by controlling welding position moving unit 41 while correcting the welding direction position based on the detection of corrector magnets 97 by detection unit 42 (step S15), it is possible to perform tracking control in the specific state with high accuracy.

[0057] Next, correction of the axial position in a specific state will be described. Fig. 10 is a diagram showing the groove 91 and two corrective magnets 97a, 97b. Fig. 11A is a diagram showing the cross-sectional shapes of the groove 91 and corrective magnet 97a measured by the detection unit 42, and Fig. 11B is a diagram showing the cross-sectional shapes of the groove 91 and corrective magnet 97b measured by the detection unit 42. In Figs. 11A and 11B, the upper part shows the previous cross-sectional shape, and the lower part shows the current cross-sectional shape.

[0058] 11A, at the position of corrector magnet 97a in the welding direction, the axial positions of groove 91 and corrector magnet 97a in the current cross-sectional shape are unchanged from the previous cross-sectional shape. In this case, the axial position at which weld bead 92 should be formed relative to that position can be appropriately determined based on the previous cross-sectional shape by the process described with reference to FIG. 8. In FIG. 10, welding torch 311 during welding relative to the position of corrector magnet 97a in the welding direction is indicated by a solid circle, and the path along which welding torch 311 moves relative to it is indicated by a dashed arrow.

[0059] 11B, the axial position of corrector magnet 97b in the current cross-sectional shape has changed (shifted) from the previous cross-sectional shape in the welding direction. In fact, the axial position of groove 91 has also changed. In this case, if the axial position at which weld bead 92 should be formed relative to that position in the welding direction is determined based only on the previous cross-sectional shape, weld bead 92 will be formed within the range indicated by arrow A1 in FIG. 11B. As a result, weld bead 92 may be formed at a position shifted from groove 91.

[0060] Therefore, the control unit 10 corrects the axial position based on the detection of the corrective magnet 97. Specifically, in the previous cross-sectional shape shown in the upper part of FIG. 11B, the axial position E21 of the edge (the edge on the groove 91 side) of the upper end surface of the corrective magnet 97b is identified. In addition, in the current cross-sectional shape shown in the lower part of FIG. 11B, the axial position E22 of the edge of the upper end surface of the corrective magnet 97b is identified. Thereafter, the difference V1 between the axial position E22 of the edge in the current cross-sectional shape and the axial position E21 of the edge in the previous cross-sectional shape is calculated as the positional deviation amount (drift amount).

[0061] 11B, a shape obtained by moving (shifting) the previous cross-sectional shape in the axial direction by difference V1 is obtained, and the axial position at which weld bead 92 should be formed relative to the position of corrector magnet 97b is determined based on this shape. In FIG. 10, corrector magnet 97b, welding torch 311, and groove 91 corresponding to the current cross-sectional shape at the position of corrector magnet 97b in the welding direction are shown by dashed two-dot lines. In this way, by controlling welding position moving unit 41 while correcting the axial position based on the detection of corrector magnet 97 (step S15), it is possible to appropriately form weld bead 92 relative to groove 91.

[0062] In welding device 1, while performing the above-described copying control and path transition, the formation of weld bead 92 is repeated around the entire circumference of groove 91. When groove 91 is entirely filled with weld bead 92, the rotation of cylindrical barrel 9 by rotation mechanism 2 is stopped (step S16). This completes the welding process by welding device 1.

[0063] In the above process example, welding is performed while maintaining constant the current, voltage, and welding speed (hereinafter referred to as "welding heat input conditions") applied to welding torch 311. However, automatic adjustment of the welding heat input conditions (hereinafter referred to as "adaptive control") may be performed based on the cross-sectional shape of groove 91 measured by detector 42. FIG. 12 is a diagram showing the flow of adaptive control processing. FIGS. 13A to 13C are diagrams for explaining adaptive control, showing a cross-section of groove 91. As described above, by forming weld beads 92 around the entire circumference of groove 91 while performing tracing control and path transition, multiple weld beads 92 arranged in the axial direction within groove 91 are formed as weld bead layer 93, as shown in FIG. 13A. Weld bead layer 93 may be formed from a single weld bead 92. In adaptive control, the welding heat input conditions are adjusted when forming a weld bead layer 93 following one weld bead layer 93 (hereinafter referred to as the "previous weld bead layer 93").

[0064] Specifically, when forming the first weld bead 92 in the previous weld bead layer 93, the cross-sectional shape of each portion of groove 91 in the welding direction is acquired. Furthermore, when forming each weld bead 92 included in the previous weld bead layer 93, the wire feed amount during welding of each portion of groove 91 in the welding direction is acquired by the wire feed unit. Control unit 10 adds a thickness estimated from the wire feed amount for that portion during the formation of all weld beads 92 included in the previous weld bead layer 93 to the cross-sectional shape of each portion of groove 91 acquired when forming the first weld bead 92, thereby approximating the cross-sectional shape of that portion to a trapezoid. The relationship between the wire feed amount and the thickness of weld bead 92 is obtained in advance. In this manner, when forming the next weld bead layer 93, the groove width W2 and groove depth D2 (see FIG. 13A ) of groove 91 in which the previous weld bead layer 93 was formed are determined over the entire circumference of groove 91 (step S21). The groove width W2 is the length of the radially inner side of the trapezoid that represents the cross-sectional shape, and the groove depth D2 is the height of the trapezoid in the radial direction.

[0065] FIG. 14 is a diagram showing groove depth D2 at each portion of groove 91 in the welding direction. In the top row of FIG. 14, the change in groove depth D2 in the welding direction is indicated by a solid line with the same symbol D2. The second to fourth rows of FIG. 14 will be described later. Once groove width W2 and groove depth D2 are determined for the previous weld bead layer 93, the average value of groove width W2 around the entire circumference of groove 91 is determined. In this processing example, a lookup table showing the range of welding heat input conditions that are less likely to cause welding defects for multiple groove widths is obtained in advance through experiments or the like and stored in control unit 10. By referring to the lookup table using the average value of groove width W2, the range of welding heat input conditions for forming the next weld bead layer 93 is determined (step S22). In addition, when a welding heat input condition range is determined based on standards or the like for a product to be manufactured by groove welding of the cylindrical body 9, for example, the range where the welding heat input condition range overlaps with the welding heat input condition range identified using a reference table is determined as the welding heat input condition range for forming the next weld bead layer 93.

[0066] Next, average (approximately central) welding heat input conditions are identified within the range of welding heat input conditions for the next weld bead layer 93. Assuming that the next weld bead layer 93 is formed under the average welding heat input conditions, a groove depth (hereinafter referred to as the "estimated groove depth") that would be obtained for the next weld bead layer 93 is estimated (step S23). In this processing example, the relationship between a plurality of welding heat input conditions and the thickness of the weld bead 92 formed under the welding heat input conditions is determined in advance through experiments or the like and stored in control unit 10. The estimated groove depth is obtained using this relationship. In FIG. 13B, the weld bead 92 of the weld bead layer 93 (hereinafter also referred to as the "virtual weld bead layer 93") virtually formed under the above average welding heat input conditions is indicated by a two-dot chain line. The estimated groove depth D3 and groove width W3 of the virtual weld bead layer 93 are indicated by arrows. In the second row from the top of FIG. 14, the change in estimated groove depth D3 in the welding direction is indicated by a two-dot chain line with the same reference symbol D3.

[0067] Once estimated groove depth D3 is obtained, an average value M1 of estimated groove depth D3 around the entire circumference of groove 91 is determined. In Fig. 13C and the second diagram from the top in Fig. 14, the position of average value M1 of estimated groove depth D3 is indicated by a dashed line. Thereafter, for each portion of groove 91 in the welding direction, the difference between average value M1 of estimated groove depth D3 and estimated groove depth D3 is determined as the excess or deficiency (thickness) of virtual weld bead layer 93 (step S24). In the third diagram from the top in Fig. 14, the area indicating the excess or deficiency of virtual weld bead layer 93 is shaded.

[0068] Then, for each portion of groove 91 in the welding direction, welding heat input conditions modified from the average welding heat input conditions are determined so that the thickness of weld bead layer 93 increases or decreases depending on the excess or deficiency of virtual weld bead layer 93 (step S25). In other words, for each portion of groove 91, the welding heat input conditions are determined based on the difference between groove depth D2 of the previous weld bead layer 93 and average value M1 of estimated groove depth D3. At this time, the welding heat input conditions for that portion in the welding direction are limited to within the welding heat input condition range for the next weld bead layer 93 determined in step S22. Therefore, if the welding heat input conditions required to form a weld bead layer 93 with a thickness corresponding to the difference between groove depth D2 and average value M1 of estimated groove depth D3 are equal to or greater than the upper limit of the welding heat input condition range, the welding heat input conditions are changed to the upper limit of the welding heat input condition range. Similarly, if the welding heat input conditions required to form a weld bead layer 93 of a thickness corresponding to the above difference are below the lower limit of the welding heat input condition range, the welding heat input conditions are changed to the lower limit of the welding heat input condition range.

[0069] In forming next weld bead layer 93, when each portion of groove 91 in the welding direction passes the welding position, the current and voltage applied to welding torch 311 are changed according to the welding heat input conditions for that portion. As a result, next weld bead layer 93 is formed around the entire circumference of groove 91, with the groove depth approximating average value M1 of estimated groove depth D3 (step S26). In the fourth row from the top of FIG. 14, a thick solid line D4 indicates the change in the groove depth measured by detector 42 in the welding direction for next weld bead layer 93. In the portion where solid line D4 is less than average value M1 of estimated groove depth D3, the welding heat input conditions are changed to the upper limit of the range of welding heat input conditions. In the portion where solid line D4 is greater than average value M1 of estimated groove depth D3, the welding heat input conditions are changed to the lower limit of the range of welding heat input conditions.

[0070] As described above, adaptive control makes it possible to maintain a substantially constant depth of the surface of weld bead layer 93. In practice, similar to the tracking control in a specific state, adaptive control can be performed with high accuracy by controlling welding position moving unit 41 while correcting the welding direction position based on the detection of corrective magnet 97 by detection unit 42. When a new weld bead layer 93 is formed, steps S21 to S26 described above are repeated.

[0071] As described above, in the welding apparatus 1, the torch unit 31 continuously performs groove welding at predetermined welding positions in the circumferential direction while rotating the cylindrical barrel 9, thereby repeatedly forming weld beads 92 around the entire circumference of the groove 91. Furthermore, during the rotation of the cylindrical barrel 9, the circumferential position immediately before reaching the welding position is set as the measurement position, and the detection unit 42 detects a detection target (in the above example, the reference magnet 96 or the correction magnet 97) provided on the outer peripheral surface of the cylindrical barrel 9 at the measurement position. This makes it easy to obtain the actual rotational position of the cylindrical barrel 9 even when slippage occurs on the rollers 21, etc. Furthermore, a welding position moving unit 41 is provided that moves the position where the torch unit 31 forms the weld bead 92 in the axial direction, and the control unit 10 controls the welding position moving unit 41 based on the detection of the detection target by the detection unit 42. As a result, path transition, tracking control in specific states, adaptive control, and the like can be performed accurately, i.e., groove welding can be performed appropriately.

[0072] Preferably, a plurality of detection targets (in the above example, correction magnets 97) are arranged in the circumferential direction on the outer circumferential surface of the cylindrical barrel 9. By providing a plurality of detection targets in this manner, the rotational position of the cylindrical barrel 9 can be acquired at fairly short intervals, allowing groove welding to be performed more appropriately.

[0073] Preferably, the detectable part is fixed to the outer peripheral surface of the cylindrical barrel 9 by magnetic force. This allows for easy detachable fixation of the detectable part. Furthermore, it is possible to prevent the detectable part from coming off the cylindrical barrel 9 due to the influence of heat when the cylindrical barrel 9 is heated during welding or heat generated by welding. Note that, depending on the temperature of the cylindrical barrel 9 during welding, the detectable part may be fixed with adhesive tape or the like. Furthermore, depending on the application of the cylindrical barrel 9, the detectable part may be a portion formed on the outer peripheral surface of the cylindrical barrel 9 or a portion fixed by welding or the like.

[0074] Preferably, when repeatedly forming weld bead 92, a circumferential position of groove 91 at which the axial relative position of weld bead 92 with respect to groove 91 should be changed is set as the path transition position. Then, control unit 10 identifies the timing at which the path transition position passes the welding position based on detection of a detected part (reference magnet 96 in the above example) by detection unit 42. This allows path transition to be performed accurately at the path transition position.

[0075] Preferably, detection unit 42 measures the cross-sectional shape of groove 91 at the measurement position and also detects the cross-sectional shape of the detection target portion that has reached the measurement position. In this way, by detecting the detection target portion using a shape measuring device used to measure the cross-sectional shape of groove 91, the number of parts in welding device 1 can be reduced.

[0076] Preferably, control unit 10 determines the axial position where weld bead 92 should be formed for each portion of groove 91 that passes through the welding position, using multiple cross-sectional shapes measured by detection unit 42 for multiple portions located near that portion of groove 91. This makes it possible to more or less appropriately determine the axial position where weld bead 92 should be formed, even if there is an abnormality or the like in the cross-sectional shape measured for that portion.

[0077] Preferably, in a normal state, control unit 10 determines the axial position at which weld bead 92 should be formed for each portion of groove 91 that passes through the welding position, based on the cross-sectional shape of that portion measured immediately before by detection unit 42. Also, in a specific state in which groove 91 is unclear in the cross-sectional shape measured immediately before, control unit 10 determines the axial position at which weld bead 92 should be formed for that portion based on the cross-sectional shape used to previously form weld bead 92 for each portion of groove 91. This makes it possible to appropriately determine the axial position at which weld bead 92 should be formed in the specific state.

[0078] Preferably, in the specific state, control unit 10 acquires the axial position of groove 91 at the measurement position based on the cross-sectional shape of the detected portion measured by detection unit 42, and determines the axial position at which weld bead 92 should be formed for each portion of groove 91. This allows weld bead 92 to be formed at an appropriate position within groove 91 even if the axial position of groove 91 is shifted.

[0079] Preferably, the control unit 10 determines the welding heat input conditions for each portion of the groove 91 that passes through the welding position using the groove depth obtained for the entire circumference of the groove 91 so that the depth of the surface of the formed weld bead 92 is constant around the entire circumference of the groove 91. In practice, the amount of deposited metal can be adjusted so that the difference in groove depth between the welding start and end is small. This reduces (ideally, eliminates) the additional welding work required by the worker when the depth of the groove 91 varies.

[0080] Next, the wire cutting unit 61 will be described. The wire cutting unit 61 cuts the wire when slag adheres to the tip of the wire of the welding torch 311, when the tip of the wire is to be sharpened at the start of groove welding (including when groove welding is resumed after a temporary stop), and other such cases. FIG. 15 is a diagram showing the configuration of the wire cutting unit 61, showing the wire cutting unit 61 and the welding torch 311 as viewed along the axial direction. As described above, the welding apparatus 1 is provided with a plurality of welding torches 311 for submerged arc welding, and the plurality of welding torches 311 are arranged adjacent to one another. In the example of FIG. 15, the two welding torches 311 are tilted so that their tips are close to one another. Specifically, the two welding torches 311 are tilted so that the central axes J2 of the two welding torches 311 approximately intersect below their tips (on the cylindrical body 9 side).

[0081] The wire cutting unit 61 includes a cutting unit main body 62 and a main body moving mechanism 63. The main body moving mechanism 63 has an air cylinder (not shown) and a support arm 631 to the end of which the cutting unit main body 62 is fixed. The main body moving mechanism 63 swings the support arm 631 about a swing axis K1 parallel to the axial direction by driving the air cylinder. This allows the cutting unit main body 62 to be selectively positioned at a cutting position close to the tips of the two welding torches 311 (the position of the cutting unit main body 62 shown by the solid line in FIG. 15) and a retracted position away from the tips (the position of the cutting unit main body 62 shown by the two-dot chain line in FIG. 15).

[0082] The cutting unit main body 62 includes an air nipper 621 and a nipper swinging unit 626. The air nipper 621 has two blade portions 622, and is selectively placed in a state where the two blade portions 622 are in contact with each other or in a state where the two blade portions 622 are spaced apart by driving an air cylinder (not shown). The nipper swinging unit 626 is fixed to an end of the support arm 631, and the air nipper 621 is attached to the nipper swinging unit 626 so as to protrude in the axial direction. As shown by solid lines in FIG. 15 , when the cutting unit main body 62 is placed in the cutting position, the two blade portions 622 of the air nipper 621 are positioned below the two welding torches 311. The nipper swinging unit 626 swings the air nipper 621 around a swing axis K2 that is parallel to the axial direction by driving an air cylinder (not shown). When viewed along the axial direction, the oscillation axis K2 roughly coincides with the position where the central axes J2 of the two welding torches 311 intersect. Note that the drive sources for the main body moving mechanism 63, the air nipper 621, and the nipper oscillation part 626 may be motors or the like other than the air cylinder.

[0083] When the wire 312 of the welding torch 311 is cut by the wire cutting unit 61, the following operations are automatically performed, for example, by an operator instructing the control unit 10 via the input unit. First, the cutting unit main body 62 is moved from the retracted position to the cutting position by the main body moving mechanism 63. In the air nippers 621 placed at the cutting position, the two blade portions 622 are spaced apart, as shown in FIG. 16A , and the central axis J2 is positioned below one of the welding torches 311 and between the two blade portions 622. The wire feeder feeds the wire 312 of the welding torch 311 by a predetermined length, and the wire 312 is positioned between the two blade portions 622, as shown by the two-dot chain line in FIG. 16A . Then, the two blade portions 622 are brought into contact with each other, thereby cutting the wire 312. After the two blade portions 622 are spaced apart, the wire feeder slightly returns the wire 312 to the inside of the welding torch 311.

[0084] Next, nipper swinging unit 626 swings air nipper 621, so that its central axis J2 is positioned below the other welding torch 311 and between the two blade portions 622, as shown in FIG. 16B. Wire 312 of welding torch 311 is fed by a predetermined length by the wire feeder, and wire 312 is positioned between the two blade portions 622, as shown by the two-dot chain line in FIG. 16B. Then, wire 312 is cut by bringing the two blade portions 622 into contact with each other. After the two blade portions 622 are separated from each other, wire 312 is slightly returned into welding torch 311 by the wire feeder. After air nipper 621 is returned to the orientation shown in FIG. 16A, cutting unit main body 62 is returned to the retracted position. This completes cutting of wire 312 by wire cutting unit 61.

[0085] As described above, welding device 1 is provided with wire cutting unit 61 that can sequentially cut wire 312 of multiple welding torches 311. This allows the number of parts in welding device 1 to be reduced compared to when multiple wire cutting units are provided to match the number of welding torches 311. Furthermore, wire 312 can be easily cut by remote control without the operator having to climb to the top of cylindrical barrel 9.

[0086] Next, the slag removal unit 5 will be described. The slag removal unit 5 removes slag generated in the groove 91 by groove welding in parallel with the rotation of the cylindrical drum 9. FIG. 17 is a diagram showing the configuration of the slag removal unit 5. In the example of FIG. 1, the slag removal unit 5 is disposed in a position facing the lower part of the cylindrical drum 9, but in FIG. 17, the slag removal unit 5 is illustrated so that the longitudinal direction of the groove 91 (i.e., the tangent direction to the outer peripheral surface of the cylindrical drum 9) is approximately aligned with the horizontal direction of the page.

[0087] The slag removal unit 5 includes a first removal unit 51, a second removal unit 52, a first support unit 53, a second support unit 54, a rotating unit 55, and an ejector 59. The first removal unit 51, the second removal unit 52, and the ejector 59 are arranged along the longitudinal direction of the groove 91. Each portion of the groove 91 where a weld bead 92 is formed at the welding position passes through the first removal unit 51, the second removal unit 52, and the ejector 59 in that order. In FIG. 17, the direction of movement of each portion of the groove 91, i.e., the rotation direction of the cylindrical barrel 9, is indicated by an arrow. In the first removal unit 51 and the second removal unit 52, slag adhering to the groove 91 is peeled off from the cylindrical barrel 9, and the slag is sucked and removed by the ejector 59.

[0088] The first removal unit 51 has a plurality of first slag removal units 511. The plurality of first slag removal units 511 are arranged at regular intervals in an arrangement direction that roughly follows the longitudinal direction of the groove 91. The first slag removal units 511 are, for example, chippers, and vibrate rod-shaped chisels 512 that protrude toward the groove 91 using compressed air. The chisels 512 are approximately perpendicular to an axial direction that is perpendicular to the paper surface of FIG. 17. The chisels 512 are inclined with respect to the longitudinal direction of the groove 91 so that they are positioned forward (toward the second removal unit 52) ​​in the moving direction of the groove 91 as they move away from the groove 91. The end (tip) of the chisel 512 is positioned within the groove 91. In the example of FIG. 17, three first slag removal units 511 are provided, but the number of first slag removal units 511 may be two, four, or more. The same applies to the second slag removal unit 521 described below.

[0089] The first support unit 53 has a first support base 531, a plurality of guide units 532, and a plurality of biasing units 536. The first support base 531 supports a plurality of guide units 532 (spline shafts 533, which will be described later). The structure of the first support base 531 will be described in detail later. Each guide unit 532 is, for example, a ball spline, and has a spline shaft 533 and an outer cylinder 534, which is a movable body. In FIG. 17, only the spline shaft and outer cylinder of one guide unit 532 are labeled. The spline shaft 533 extends approximately parallel to the longitudinal direction of the chisel 512. The first slag removal unit 511 is fixed to the outer cylinder 534. The guide unit 532 supports the first slag removal unit 511 so that it can move along the spline shaft 533. The biasing unit 536 is attached to the spline shaft 533. The biasing portion 536 has an elastic member such as a spring, and presses (biases) the first slag removal portion 511 toward the groove 91 together with the outer cylinder 534 .

[0090] The second removal unit 52 has a plurality of second slag removal units 521. The plurality of second slag removal units 521 are arranged at regular intervals in an arrangement direction that is approximately along the longitudinal direction of the groove 91. The second slag removal unit 521 is, for example, a needle scaler, and uses compressed air to reciprocate a plurality of needles 522 that protrude toward the groove 91. The plurality of needles 522 are approximately perpendicular to the outer peripheral surface of the cylindrical barrel 9, and the ends of the needles 522 are positioned within the groove 91.

[0091] The second support portion 54 has a second support base 541, a plurality of guide portions 542, and a plurality of biasing portions 546. The second support base 541 has a member extending in the arrangement direction of the second slag remover 521. 542The guide portion 542 (a spline shaft 543 described later) is fixed to and supported by the corresponding member of the second support base 541. Each guide portion 542 is, for example, a ball spline, and has a spline shaft 543 and an outer cylinder 544, which is a movable body. In FIG. 17, only the spline shaft and outer cylinder of one guide portion 542 are labeled. The spline shaft 543 extends approximately parallel to the longitudinal direction of the multiple needles 522. The second slag removal portion 521 is fixed to the outer cylinder 544. The guide portion 542 supports the second slag removal portion 521 so that it can move along the spline shaft 543. The biasing portion 546 is attached to the spline shaft 543. The biasing portion 546 has an elastic member such as a spring, and, together with the outer cylinder 544, presses (biases) the second slag removal portion 521 toward the groove 91.

[0092] 18A and 18B are diagrams for explaining the structure of first support base 531 and turning unit 55, and show first support base 531 and turning unit 55 as viewed radially from the outside in the radial direction. In Figures 18A and 18B, support body guide portion 58, which will be described later, is not shown.

[0093] 18A and 18B, the first support base 531 is a parallel link mechanism and has two long links 537 and three short links 538. The two long links 537 are parallel to each other, and the three short links 538 are also parallel to each other. The three short links 538 are arranged along the long links 537. Ends of two short links 538 are connected to both ends of each long link 537 by joints, and an end of the remaining short link 538 (i.e., the central short link 538) is connected to the center of the long link 537 by a joint.

[0094] A first slag removal section 511 is fixed to the center of each short link 538 via a guide section 532 (see FIG. 17) or the like. In FIGS. 18A and 18B, the position of the first slag removal section 511 is indicated by a hatched rectangle, and the direction from the base of the chisel 512 to the end is indicated by an arrow bearing the same reference numeral 512. The multiple first slag removal sections 511 are supported by a first support base 531 in an arrangement direction that is approximately along the longitudinal direction (rotation direction) of the groove 91. The two long links 537 of the first support base 531 are members that extend in the arrangement direction. In FIGS. 18A and 18B, the position of the second slag removal section 521 is also indicated by a hatched circle.

[0095] The rotation unit 55 includes a first support base rotation mechanism 56, a second support base rotation mechanism 57, a support body 550, and a support body guide 58 (see FIG. 17). The support body 550 is a member extending in a direction along the longitudinal direction of the groove 91. A central short link 538 of the first support base 531 is fixed to the support body 550. When the rotation unit 55 is viewed in the radial direction as shown in FIGS. 18A and 18B, the central short link 538 is approximately perpendicular to the longitudinal direction of the support body 550. The support body 550 also supports the second support base 541 rotatably about a rotation axis K4 that is approximately parallel to the radial direction. In the rotation unit 55 shown in FIGS. 18A and 18B, when viewed in the radial direction, the rotation axis K4 approximately overlaps with the central second slag removal section 521.

[0096] The first support base rotation mechanism 56 has an air cylinder 561. The body of the air cylinder 561 is supported by the support body 550. The tip of the piston rod of the air cylinder 561 is connected to one long link 537 of the first support base 531. When the air cylinder 561 retracts the piston rod, the arrangement direction of the multiple first slag removal units 511 is approximately parallel to the longitudinal direction of the support body 550, as shown in FIG. 18A. Furthermore, the three short links 538 are approximately perpendicular to the longitudinal direction, and the direction of the chisel 512 (see arrow 512) is approximately parallel to the longitudinal direction. When the air cylinder 561 pushes out the piston rod, the arrangement direction of the multiple first slag removal units 511 is inclined with respect to the longitudinal direction of the support body 550, as shown in FIG. 18B. At this time, the three short links 538 are maintained in a state where they are approximately perpendicular to the longitudinal direction. That is, the direction of the chisel 512 is maintained substantially parallel to the longitudinal direction.

[0097] The second support base rotating mechanism 57 has an air cylinder 571. The body of the air cylinder 571 is supported by the support body 550. The tip of the piston rod of the air cylinder 571 is connected to the second support base 541. When the air cylinder 571 retracts the piston rod, the arrangement direction of the multiple second slag removal units 521 is approximately parallel to the longitudinal direction of the support body 550, as shown in FIG. 18A. When the air cylinder 571 pushes out the piston rod, the arrangement direction of the multiple second slag removal units 521 becomes inclined with respect to the longitudinal direction of the support body 550, as shown in FIG. 18B.

[0098] 19A and 19B are diagrams for explaining the structure of the support body guide portion 58, and show the support body guide portion 58 and the support body 550 as viewed from the radial outside along the radial direction. In FIGS. 19A and 19B, the first removal portion 51 and the second removal portion 52 are schematically shown by dashed rectangles, and the first support portion 53 and the second support portion 54 are not shown. In addition, the reference position where the groove 91 should be positioned in the axial direction is shown by a dashed line labeled N1. The same applies to FIGS. 21A and 21B described below.

[0099] The support body guide unit 58 has a linear guide 581 and a support pin 586. The linear guide 581 has a rail 582 and a block 583. The rail 582 extends in the axial direction and is fixed to a support base (not shown). The block 583 is movable along the rail 582. The block 583 is connected to the support body 550 via the support pin 586, and the support body 550 is rotatably supported relative to the block 583. In this manner, the support body 550 is movable in the axial direction and rotatable relative to the block 583. As shown in FIG. 19B , even when the groove 91 has a meandering portion when viewed radially or when the position of the groove 91 in the axial direction is deviated from the reference position N1, the support body guide unit 58 having the above structure can align the longitudinal direction of the support body 550 with the groove 91.

[0100] 20A and 20B are views showing the plurality of second slag removal parts 521 as viewed along the longitudinal direction of the groove 91. As described above, the second slag removal parts 521 are pressed toward the groove 91 by the biasing parts 546 (see FIG. 17). Furthermore, the air cylinder 571 of the second support base rotation mechanism 57 pushes out the piston rod, causing the second support base 541, i.e., the arrangement direction of the plurality of second slag removal parts 521, to be inclined with respect to the longitudinal direction of the support main body 550 (see FIG. 18B).

[0101] At this time, the ends of the two second slag removal parts 521 located at both ends (i.e., the ends of the needles 522) among the multiple second slag removal parts 521 contact both side surfaces of the groove 91, thereby restricting the tilt in the arrangement direction. In detail, when the air cylinder 571 pushes out the piston rod and one of the two second slag removal parts 521 contacts one side surface of the groove 91, the support body 550 moves axially and the second support base 541 further tilts, and the other second slag removal part 521 contacts the other side surface of the groove 91. As a result, as shown in FIG. 20A , the ends of the two second slag removal parts 521 are positioned near two corners on the bottom side of the groove 91 of any width. Furthermore, the end of the central second slag removal part 521 is positioned approximately in the center between both side surfaces of the groove 91. In FIG. 19A, the arrow F indicates that the end of the second slag removal portion 521 abuts on the side surface of the groove 91 substantially perpendicularly (similarly in FIG. 19B).

[0102] In actual groove welding, as shown in FIG. 20B , weld beads 92 are present within groove 91, and the width of groove 91 at the surface of the outermost weld bead 92 (i.e., the bottom surface of groove 91) changes as the weld bead 92 is repeatedly formed. Even in this case, in slag removal unit 5, ends of the two second slag removal parts 521 located at both ends are automatically positioned near the positions where the bottom surface of groove 91 intersects with both side surfaces. In addition, the end of the central second slag removal part 521 is positioned approximately in the center between both side surfaces of groove 91. Even in cases such as when groove 91 has a meandering portion as viewed radially as shown in FIG. 19B , the longitudinal direction of support body 550 is aligned with groove 91, so that the end of second slag removal part 521 abuts the side surface of groove 91 approximately perpendicularly (see arrow F).

[0103] As described above, in the second support base rotating mechanism 57, the extrusion action of the piston rod by the air cylinder 571 can bring the ends of the two second slag removal parts 521 located at both ends of the multiple second slag removal parts 521 into contact with both side surfaces of the groove 91. This allows the second slag removal parts 521 to properly remove slag even near the corners on the bottom side of the groove 91, where slag is difficult to remove. Note that the second support base rotating mechanism 57 can also be considered a pressing mechanism that presses the ends of the second slag removal parts 521 against the side surfaces of the groove 91 (the same applies to the first support base rotating mechanism 56).

[0104] As with the second support base rotating mechanism 57, the first support base rotating mechanism 56 also allows the ends of two of the multiple first slag removal units 511 located at both ends to contact both side surfaces of the groove 91 approximately perpendicularly by pushing out the piston rod with the air cylinder 561. This allows the first slag removal units 511 to appropriately remove slag from the vicinity of the corners on the bottom side of the groove 91. As described above, in the first support base 531, which is a parallel link mechanism, even if the arrangement direction of the multiple first slag removal units 511 is inclined with respect to the longitudinal direction of the support main body 550, the orientation of the chisels 512 remains approximately parallel to the longitudinal direction. This allows slag to be efficiently removed by vibration of the chisels 512.

[0105] As described above, the slag removal unit 5 includes a plurality of slag removal sections 511, 521 arranged in an arrangement direction roughly along the longitudinal direction of the groove 91 with their ends positioned within the groove 91, support sections 53, 54 having members extending in the arrangement direction and supporting the plurality of slag removal sections 511, 521, and rotation mechanisms 56, 57 that change the inclination angle of the arrangement direction with respect to the longitudinal direction of the groove 91 while the support sections 53, 54 are movable in the axial direction, thereby bringing the ends of the two slag removal sections located at both ends of the plurality of slag removal sections 511, 521 into contact with both side surfaces of the groove 91. This makes it possible to properly remove slag up to the vicinity of the corners within the groove 91.

[0106] 21A and 21B are diagrams showing another example of a support body guide portion. Support body guide portion 58a shown in FIGS. 21A and 21B has a support arm 585 and a support pin 586. One end of support arm 585 is rotatably (swingably) supported by a support base 589. The other end of support arm 585 is connected to support body 550 via support pin 586. Support body 550 is rotatable relative to the other end of support arm 585. With support body guide portion 58a having the above structure, as shown in FIG. 21B, even when groove 91 has a meandering portion when viewed radially, the longitudinal direction of support body 550 can be aligned with groove 91. As a result, in the slag removal unit 5 having the support body guide portion 58a, similar to the slag removal unit 5 having the support body guide portion 58 of Figures 19A and 19B, the ends of the slag removal portions 511, 521 can be brought into contact with both side surfaces of the groove 91 in the first removal portion 51 and the second removal portion 52, making it possible to properly remove slag up to near the corners within the groove 91.

[0107] The above welding device 1 and welding method can be modified in various ways.

[0108] In the above embodiment, by arranging the reference magnet 96 and the correction magnet 97 on opposite sides of the groove 91 (i.e., on the outer peripheral surfaces of different cylindrical barrels 9), it is possible to easily distinguish between the reference magnet 96 and the correction magnet 97 in the cross-sectional shape measured by the detection unit 42, but the reference magnet 96 and the correction magnet 97 can also be arranged on the same cylindrical barrel 9. In this case, it is preferable to make the outer diameter and height of the reference magnet 96 different from those of the correction magnet 97. As described above, it is sufficient that the detected part is provided on the outer peripheral surface of at least one of the two cylindrical barrels 9.

[0109] Depending on the content of the control by the control unit 10, only one of the reference magnet 96 and the corrective magnet 97 may be used.

[0110] From the viewpoint of accurately identifying the centers of the reference magnet 96 and the corrector magnet 97, the reference magnet 96 and the corrector magnet 97 may have a cylindrical shape or an elliptical cylindrical shape, etc. Depending on the accuracy required for the processing using the reference magnet 96 and the corrector magnet 97, the reference magnet 96 and the corrector magnet 97 may have a rectangular prism shape or the like, as shown in FIG. 6B .

[0111] In the detection unit 42, the detection of the detection target portion and the measurement of the cross-sectional shape of the groove 91 may be performed by separate measuring machines. In this case, a measuring machine other than a non-contact shape measuring machine may be used to detect the detection target portion. For example, the detection target portion can also be detected by a camera that captures an image of the outer peripheral surface of the cylindrical barrel 9.

[0112] In the welding device 1, three or more cylindrical barrels 9 may be arranged in the axial direction. In this case, a torch unit 31, a detection unit 42, a welding position movement unit 41, etc. may be individually provided for each groove 91, and groove welding for the three or more cylindrical barrels 9 may be performed in parallel. Also, a reference magnet 96 and a correction magnet 97 may be provided near only one groove 91, and the detection results of the reference magnet 96 and the correction magnet 97 may be used in processing the other grooves 91.

[0113] In the slag removal unit 5, only one of the first removal section 51 and the second removal section 52 may be used.

[0114] The above-described copying control and adaptive control may be used in longitudinal welding, for example, to weld together the ends of rolled plate-like members when manufacturing the cylindrical shell 9. The same applies to the wire cutting section 61 and the slag removing unit 5.

[0115] The configurations in the above-described embodiment and each modification may be combined as appropriate as long as they are not mutually contradictory. [Explanation of symbols]

[0116] 1. Welding equipment 2 Rotation mechanism 5. Slag removal unit 9 Cylindrical body 10 Control Unit 31 Torch section 41 Welding position moving part 42 Detector 53,54 Support part 56,57 Support base rotation mechanism 61 Wire cutting section 91 Bevel 92 Weld Bead 96 Reference Magnet 97, 97a, 97b Correction magnets 311 Welding Torch 511,521 Slag removal section J1 (Cylindrical body) central axis P31 Path transition position S11~S16, S21~S26 steps

Claims

1. 1. A welding device comprising: two cylindrical barrels arranged in an axial direction parallel to a central axis, with grooves formed at the butt joints of the two cylindrical barrels, and a rotation mechanism that rotates the two cylindrical barrels integrally; a torch unit that performs groove welding continuously at predetermined welding positions in a circumferential direction around the central axis while rotating the two cylindrical barrels, thereby repeatedly forming weld beads around the entire circumference of the groove; a welding position moving unit that moves a position where the weld bead is formed by the torch unit in the axial direction; a detection unit that detects a detection target provided on an outer peripheral surface of at least one of the two cylindrical barrels at a measurement position in the circumferential direction immediately before the rotation of the two cylindrical barrels reaches the welding position; and a control unit that controls the welding position moving unit based on the detection of the detection target portion by the detection unit; Equipped with When repeatedly forming the weld bead, the circumferential position of the groove at which the relative position of the weld bead in the axial direction with respect to the groove should be changed is set as a pass transition position, The welding device is characterized in that the control unit identifies the timing at which the path transition position passes the welding position based on the detection of the detection target portion by the detection unit.

2. A welding device, two cylindrical barrels arranged in an axial direction parallel to a central axis, with grooves formed at the butt joints of the two cylindrical barrels, and a rotation mechanism that rotates the two cylindrical barrels integrally; a torch unit that performs groove welding continuously at predetermined welding positions in a circumferential direction around the central axis while rotating the two cylindrical barrels, thereby repeatedly forming weld beads around the entire circumference of the groove; a welding position moving unit that moves a position where the weld bead is formed by the torch unit in the axial direction; a detection unit that detects a detection target provided on an outer peripheral surface of at least one of the two cylindrical barrels at a measurement position in the circumferential direction immediately before the rotation of the two cylindrical barrels reaches the welding position; and a control unit that controls the welding position moving unit based on the detection of the detection target portion by the detection unit; Equipped with The detection unit is a non-contact shape measuring machine that measures the cross-sectional shape of the groove at the measurement position, and also detects the cross-sectional shape of the detection portion that has reached the measurement position.

3. 3. The welding device according to claim 2, The control unit determines the axial position where the weld bead should be formed for each portion of the groove that passes through the welding position, using a plurality of cross-sectional shapes measured by the detection unit for a plurality of portions located in the vicinity of each portion of the groove.

4. 3. The welding device according to claim 2, The control unit determines the axial position at which the weld bead should be formed for each portion of the groove passing through the welding position based on the cross-sectional shape measured immediately before by the detection unit for each portion, In a specific state where the groove is unclear in the cross-sectional shape measured immediately before, the control unit determines the axial position at which the weld bead should be formed for each portion of the groove based on the cross-sectional shape used in the previous formation of the weld bead for each portion of the groove.

5. 5. The welding device according to claim 4, A welding device characterized in that, in the specific state, the control unit acquires the axial position of the groove at the measurement position based on the cross-sectional shape of the detected portion measured by the detection unit, and determines the axial position at which the weld bead should be formed for each portion of the groove.

6. 3. The welding device according to claim 2, The control unit determines welding heat input conditions for each portion of the groove that passes through the welding position using the groove depth obtained for the entire circumference of the groove so that the surface depth of the weld bead to be formed is constant around the entire circumference of the groove.

7. A welding device, two cylindrical barrels arranged in an axial direction parallel to a central axis, with grooves formed at the butt joints of the two cylindrical barrels, and a rotation mechanism that rotates the two cylindrical barrels integrally; a torch unit that performs groove welding continuously at predetermined welding positions in a circumferential direction around the central axis while rotating the two cylindrical barrels, thereby repeatedly forming weld beads around the entire circumference of the groove; a welding position moving unit that moves a position where the weld bead is formed by the torch unit in the axial direction; a detection unit that detects a detection target provided on an outer peripheral surface of at least one of the two cylindrical barrels at a measurement position in the circumferential direction immediately before the rotation of the two cylindrical barrels reaches the welding position; and a control unit that controls the welding position moving unit based on the detection of the detection target portion by the detection unit; a slag removal unit for removing slag generated by the groove welding; Equipped with The slag removal unit A plurality of slag removal units arranged in an arrangement direction approximately along the longitudinal direction of the groove, with ends disposed within the groove; a support portion having a member extending in the arrangement direction and supporting the plurality of slag removal portions; a rotation mechanism that changes the inclination angle of the arrangement direction with respect to the longitudinal direction of the groove while the support portion is movable in the axial direction, thereby bringing the ends of two slag removal portions located at both ends of the plurality of slag removal portions into contact with both side surfaces of the groove; A welding device comprising:

8. 8. A welding apparatus according to claim 1, A welding device characterized in that a plurality of detection parts are arranged in the circumferential direction on the outer peripheral surface of at least one of the cylindrical bodies.

9. 8. A welding apparatus according to claim 1, The welding device is characterized in that the detected part is fixed to the outer peripheral surface of the at least one cylindrical body by magnetic force.

10. 8. A welding apparatus according to claim 1, the torch unit has a plurality of welding torches for submerged arc welding, the plurality of welding torches are disposed adjacent to one another; The welding device is The welding device further comprises a wire cutting unit capable of sequentially cutting the wires of the plurality of welding torches.

11. A welding method comprising: a) a step of arranging two cylindrical barrels centered on a central axis in an axial direction parallel to the central axis, forming grooves at the butt joints of the two cylindrical barrels, and integrally rotating the two cylindrical barrels; b) In parallel with the step a), groove welding is continuously performed at predetermined welding positions in the circumferential direction around the central axis, thereby repeatedly forming weld beads around the entire circumference of the groove; c) a step of detecting a detection target provided on an outer peripheral surface of at least one of the two cylindrical barrels at a measurement position in the circumferential direction immediately before the rotation of the two cylindrical barrels reaches the welding position, in parallel with the step a); d) moving a formation position of the weld bead in the axial direction based on the detection of the detection target portion; Equipped with When repeatedly forming the weld bead, the circumferential position of the groove at which the relative position of the weld bead in the axial direction with respect to the groove should be changed is set as a pass transition position, A welding method characterized in that, in the step d), the timing at which the path transition position passes the welding position is identified based on the detection of the detection target portion in the step c).

12. A welding method comprising: a) a step of arranging two cylindrical barrels centered on a central axis in an axial direction parallel to the central axis, forming grooves at the butt joints of the two cylindrical barrels, and integrally rotating the two cylindrical barrels; b) In parallel with the step a), groove welding is continuously performed at predetermined welding positions in the circumferential direction around the central axis, thereby repeatedly forming weld beads around the entire circumference of the groove; c) a step of detecting a detection target provided on an outer peripheral surface of at least one of the two cylindrical barrels by a detection unit arranged at a measurement position in the circumferential direction immediately before the two cylindrical barrels reach the welding position during rotation of the two cylindrical barrels, in parallel with the step a); d) moving a formation position of the weld bead in the axial direction based on the detection of the detection target portion; Equipped with A welding method characterized in that the detection unit is a non-contact shape measuring machine that measures the cross-sectional shape of the groove at the measurement position, and also detects the cross-sectional shape of the detection portion that has reached the measurement position.

Images