JPH0581350B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] This invention relates to a multi-layer automatic arc welding method, and is particularly suitable for teaching for automatic welding robots for butt welds such as joints of columns and beams. This paper relates to a multi-layer automatic arc welding method using a playback type data computer.

[Conventional technology]

In steel frame fabrication, welding accounts for the majority of the process, and although fillet welding in the longitudinal direction of columns and beams is increasingly automated, welding at joints with many welds is difficult due to the short weld line. Much of the work is still done by hand, making it difficult to secure skilled technicians in the future. For this reason, various attempts have been made to use automatic welding robots for welding joints, but since typical welding robots are introduced for a limited number of parts, the robots have only a few welding conditions. In most cases, specific welding conditions are stored in the internal memory of the robot itself.

[Problem to be solved by the invention]

Welding of steel frame joints is performed by multi-layer welding, but multi-layer welding requires a huge number of welding conditions, including combinations of plate thickness, groove angle, root gap value, wire diameter, etc. The number of passes required for this also varies considerably depending on the conditions. Therefore, conventional welding robots have a limited memory capacity, which limits the welding conditions that can be stored. Therefore, auxiliary storage devices such as digital cassette tapes are used to compensate for the lack of storage capacity.
When loading welding conditions from tape to main memory as necessary, tape has the disadvantage that it takes a long time to load the information because it is stored and read out serially. In addition, some conventional welding robots are equipped with an arc sensor method for tracing, but in multi-layer welding, the principle is to trace the weld line detected in the first layer, so the tracing accuracy of the first layer will affect the final bead. In addition to greatly affecting the appearance and presence or absence of welding defects, it does not have an effective correction function for thermal deformation that occurs after the first layer welding, and conditions such as joint misalignment, skin gaps, and tapered root gap that tend to occur in butt welding. To deal with the problem, it was necessary to install an effective image sensor such as a television camera.

This invention solves the drawbacks of the prior art as described above, and allows the measurement results of the groove shape to be input along with conditions such as plate thickness by an operator's hand operation, for example, without using a tracing sensor, and to determine the start and end of welding. The aim is to provide a multi-layer automatic arc welding method that can realize a small, portable welding robot that automatically performs multi-layer welding simply by teaching. It is an object of the present invention to provide the above-mentioned welding method which can automatically handle the welding process simply by inputting the groove precision at the start and end ends.

[Means to solve the problem]

In the multi-layer automatic arc welding method according to the first invention of the present application, in order to achieve the above-mentioned problems, the welding speed is adjusted for each welding type such as side butt, downward butt, and fillet, and for each groove shape and angle. , multiple welding conditions according to welding current, weaving width, wire feed speed, number of layers, etc. are determined in advance by experiments using various plate thickness values and root gap values as parameters, and are compiled into a database. The positions of the start and end of the straight welding section are set by teaching, and during this teaching, the respective root gap values at the start and end of the straight welding section are memorized, and the root gap correction values in the middle of the straight welding section are set as described above. The welding conditions are obtained by linear interpolation from the memorized values, and the welding conditions corresponding to the plate thickness of the joint to be welded and the root gap value at the starting end are read out from the database, and welding is started from the starting end of the welding section using the read welding conditions. Then, each time the welding distance from the start end reaches a predetermined value during welding, the welding conditions corresponding to the root gap correction value obtained by the linear interpolation are read from the database and the welding conditions are corrected. Automatic welding is performed to the end of the straight welding section while keeping the surplus amount for each pass almost constant by changing speed and weaving width.

Furthermore, in the multi-layer automatic arc welding method of the second invention of the present application, in addition to the features of the first invention, the locus of the center of the route is calculated from the position information of the starting end and the ending end at the time of teaching and the root gap value of each end, and the trajectory of the torch is The movement locus is made to follow the center locus.

[Effect]

The welding method of the present invention is particularly suitable for multilayer welding robots for straight welding sections of several tens of centimeters such as joint sections. Therefore, it is mainly applied to horizontal welding and downward welding, but it is of course also applicable to fillet welding and narrow gap welding.

In reality, it is almost impossible to completely incorporate logarithmic welding conditions into software due to software construction problems and main memory storage capacity.

For example, even if the amount of welding in a cross-sectional calculation is obtained, it does not necessarily satisfy the appearance, penetration, defects, etc. of the weld. Even when welding downward, insufficient penetration, overlapping, undercuts, etc. may occur, and when welding sideways, problems such as bead sagging and appearance remain.

Therefore, in the present invention, we actually use a test piece,
For example, we use a method in which conditions considered to be optimal for various plate thicknesses, root gap widths, bevel angles, etc. are created by experts, and these are stored and used as a database.

That is, in the present invention, the plate thickness of the joint to be welded and the root gap values at the start and end of welding are input, and the corresponding welding conditions are searched from the database stored on a floppy disk, for example. Therefore, the torch is taught an easy-to-target location, such as the corner between the base material and the backing metal, and an optimal torch movement trajectory is internally calculated based on the input root gap value. When the root gap values at the ends are different, the welding speed and weaving width are changed in response to the change in the root gap value as welding progresses, so as to keep the excess height for each pass constant.

To create a database of welding conditions, we measure the change in the amount of welding due to changes in welding current and voltage for each wire diameter, determine the target position from the amount of welding on a computer, create a database, and create a database of the welding conditions. It is better to actually perform welding work and confirm that there are no defects using ultrasonic inspection, etc., and then use the data. In addition, it is best to create a database of welding conditions using plate thickness and root gap as search keys, and register them separately on floppy disks according to the welding position (downwards, sideways, etc.) and groove angle. This is because although the thickness and root gap vary depending on the object to be welded, the welding posture and groove angle do not change much. As a result, if the design standards for the welding object change, the floppy disk will have to be replaced during welding, making it difficult for on-site operators to make selection mistakes.

Beveling for butt welding is performed using gas cutting or a beveling machine (mechanical cutting machine), etc., and the beveling angle is kept almost in accordance with the design standards by setting the tip and using a cutting mill. Normally, the root gap differs between the starting and ending ends due to the parallelism of the rail of the gas cutting machine and the welding part cutting line (bevel line), assembly with other parts, finished dimensions, etc. This may result in a taper gap. Conventional multi-layer automatic arc welding methods cannot handle such changes in the gap, and in order to deal with this, a method is to set appropriate teaching points according to changes in the root gap and change the welding conditions there. is usually adopted. In this case, for each gap change, for example, in a weld line with a total length of 500 mm, the root gap at the beginning and end changes from 6 mm to 8 mm, and from 6 mm to 10 mm.
The number of points to be taught and the selection of welding conditions used are different, and if the welding conditions are not changed minutely, the bead width will change in steps, which may cause welding defects.

Therefore, the present invention first measures and inputs each root gap value at the start end and end end, and then teaches the origin of the groove cross section at the start end and end end, so that changes in the groove cross section along the welding line can be measured in a straight line on a personal computer. The welding conditions are simulated spatially through interpolation, and as the welding progresses, the welding conditions are corrected to correspond to the interpolated root gap value, and the torch trajectory, welding speed, and amount of change in weaving width are determined based on these welding conditions. Perform welding work. That is, even if there is a change in the groove cross section, the amount of excess build-up is kept constant by changing the welding speed and weaving width, and the necessary penetration is ensured. In this case, virtual points are created by dividing the taught starting and ending points at regular distances in the direction of welding progress, and the welding speed and weaving width are changed at each virtual point to create a step-like change. to pseudo-linearize.

In order to better understand the features and advantages of the present invention, preferred embodiments of the present invention will be described below with reference to the drawings.

〔Example〕

FIG. 1 shows the configuration of a multilayer welding robot system used for carrying out the present invention. The system according to this embodiment is of a teaching playback type, and is configured to be connected to each other by cables and hoses. In Fig. 1, 1 is a controller box, 2 is a welding power source, 3 is a water cooler, 4 is a shielding gas cylinder, 5 is a welding wire feeding device, and 6 is a welding wire feeder.
1 is a traveling rail, 7 is a robot body, 8 is a welding torch, and 9 is a teaching box.

The robot main body 7 has a four-axis configuration consisting of three orthogonal axes (X, Y, Z) and a torch rotation axis (k), and a specific example thereof is shown in detail in FIG.

In FIG. 2, the robot body 7 has guide wheels 77 that roll along the travel rail 6.
an outer case 71 having on the bottom surface, and an outer case 7
1, an inner box 73 supported by a slide guide (not shown) and moved forward and backward in the X-axis direction by a feed shaft mechanism 72, a Y-axis unit 74 attached to the tip of the inner box 73, and a Y-axis unit 74. A K-axis unit 76 that moves up and down in the Y-axis direction by a feed shaft mechanism 75 is provided. A rack screw 61 is attached to the rail 6 in the longitudinal direction, and a motor 71a fixed to the outer case 71 engages with the rack 61 via a pinion gear, and its rotation causes the robot body 7 to move in the Z direction.
It is moved along the rail 6 in the axial direction. Also, a feed shaft mechanism 7 for moving the inner box 73 forward and backward.
2 is a motor 73a supported by the outer case 71;
driven by. Furthermore, the K-axis unit 7
A feed shaft mechanism 75 that moves the robot 6 up and down is driven by a motor 74a supported by a Y-axis unit 74. A torch weaving motor 76a is attached to the K-axis unit 76 for rotating the torch 8 around the K-axis parallel to the Z-axis. In addition, 78
is the torch 8 rotated by this motor 76a.
This is a torch bracket that holds the torch, and the torch axis is eccentric from the K-axis center. As a result, even when the rail 6 is placed directly on the workpiece for horizontal T-type welding, interference between the rotating shaft motor 76a and the workpiece can be avoided, and the torch can be moved more easily than when the torch is held at the center of rotation. It is designed so that it can be laid down horizontally. Furthermore, although the trajectory of the wire tip does not change when weaving, the welding arc tends to fly in the axial direction of the wire, so the wire becomes nearly horizontal to the groove, which also prevents the bead from sagging. It's getting old. on the other hand,
In the case of downward welding, start the motor 7 from just above the bead.
Since 6a is biased, the motor can be protected from the radiant heat of welding.

Preferably, a pulse motor is used for each axis motor, and in particular, for the pinion drive of the K-axis motor 76a, it is preferable to use a harmonic drive gear or the like, which has almost no backlash, to ensure smooth torch weaving during welding.

Further, in FIG. 2, reference numeral 79 is a support for the torch cable, which is used to prevent the torch cable from getting caught on the rail or the like when the main body 1 moves on the rail 6.

The rail 6 is set to the pull welding member by a clamp or magnet (not shown), but its structure does not need to be explained here.

The configuration of the welding machine is such that a water cooling circulation device 3 is attached to a normal welding machine power supply 2, and a water cooling type torch 8 is used. The controller 1 uses, for example, a 16-bit personal computer, and is connected to a drive device 11 for a floppy disk 10. The teaching box 9 is used to perform all operations such as operating the robot during teaching and selecting and confirming welding conditions.

Now, in multi-layer build-up welding in the system of this embodiment, the floppy disk 10 is selected according to the welding posture and groove angle, the controller 1 is applied, the measurement results of plate thickness and root gap value are inputted from the teaching box 9, Welding is performed by developing the welding condition data of the floppy disk based on the teaching data of the welding start and end points. In this case, in order to perform the desired welding, it is desirable to appropriately grasp and safely control the wire feed amount.

For example, the specific gravity of welding wire is ρ (for iron wire, ρ = 7.85 g/cm ³ ), and the wire unit weight is W (wire diameter
For 1.2mm, W = 8.88g/m, for wire diameter 1.4mm, W = 12.45g/m, for wire diameter 1.6mm, W =
15.90g/m), the welding speed is W _v [mm/sec], the bead weld cross-sectional area is S _i (where i is 1, 2, 3...n: number of layers), and the wire feed rate (speed) is W When _f [mm/sec], it can be easily expressed as S _i =(α·W·W _f )/(ρ·W _v ) [mm ² ]. Here, α is a coefficient representing the rate of decrease due to spatter and slag. For multilayer stacking, the sum of S _i in the above equation is calculated for the number of layers n, and welding is performed assuming that this sum fills the groove cross section S _p as specified. That is, it can be expressed as follows.

S _p = _o 〓 ^j=1 S _iThis method, which can be called a batch prediction, enables control with sufficient accuracy by accurately managing the wire feed amount, which simplifies the welding control mechanism and The advantage is that the work speed can be increased.

For a tapered root gap, as mentioned above, by first inputting the root gap values at the start and end of the straight weld section and teaching the origin positions of the groove cross section at the start and end, the groove can be adjusted in the controller 1. The cross section is simulated spatially, and changes in the torch trajectory, welding speed, and weaving width are determined according to the welding conditions so that the amount of excess metal is kept constant and the necessary penetration is obtained. Perform welding work. In this case, linear interpolation between the two taught starting and ending points is performed in the welding direction (Z) between the two points.
axial direction) and performs pseudo-linearization that connects a plurality of divided temporary feeding points. This will be explained in detail with reference to FIG. In Figure 3,
For example, with a downward curved bevel, the bevel angle is 35 degrees, and the plate thickness is 30.
mm, the start end root gap is 6 mm, and the end root gap is 10 mm. The teaching points are only two points, the starting end and the ending end in FIG. 3, and these teaching points are the corners of the base material 31 and the backing metal 32 in this case. The above point is the origin of the XYZ orthogonal coordinates (0,
0, 0), and as a condition for the first pass, for example, the torch target position at the starting end is set to point (3, 0, 0) by giving half the shift amount of the root gap (X ₁ = 3 mm).
The weaving width W _w1 at the starting end is equal to the root gap, W _w1 = 6 mm, and the welding speed W _v = 300.
It shall be defined as mm/min. Since the starting end root gap R _g1 = 6 mm and the ending end root gap R _g2 = 10 mm, the end shift amount X ₂ is calculated from the ratio of the starting end root gap, ₆ : ₃ = 10: 5mm and the welding length is 500mm, from the starting end target position (3, 0, 0) to the end tracing position (5, 0, 500)
The straight line connecting the points up to the point becomes the torch locus of the first pass.
Similarly, the weaving width is set to 6:6=10:W _w2 W _w2 =6*10/6=10 mm, and gradually increases from the starting end W _w1 = 6 mm to the ending end W _w2 = 10 mm.

If the entire length is welded at the same welding speed, the amount of deposited metal will be constant, so the amount of surplus will not be constant. Therefore, the welding speed is adjusted according to the change in the root gap. For example, when the root gap widens toward the end as shown above, the welding speed is adjusted by W _v − W _v * (R _g2 − R _g1 ) * β. Slow down. Note that β is a parameter related to welding speed, and the speed changes by, for example, about 3 to 5% per 1 mm difference in root gap value, but it may vary due to changes in ambient temperature depending on the season, changes in base material temperature due to the presence or absence of residual heat, wires used, etc. Therefore, it is better to be able to input corrections from the teaching box 9.

In this way, the locus, weaving width, and welding speed are determined by similar calculations at intermediate feeding points as well.
Since the number of moving pulses of the virtual points is known through teaching, it is sufficient to equally divide the virtual points at intervals of several tens of millimeters when the welding line length is about 500 mm, but if the taper of the root gap Depending on the degree, the beat shape may be stepped, so it is preferable to set short intervals within a range that suits the resolution of control of the Y-axis motor 74a, K-axis motor 76a, etc.

From the second pass onwards, the root gap value at the shift position in the Y-axis direction is calculated in the same way to obtain the welding conditions. For example, if the Y-axis shift amount is 5 mm, the root gap of the second pass is 6 + tan 35 degrees at the starting end. *5=9.5mm At the end, it becomes 10+tan35°*13.5mm.

Calculations are performed in the same way in subsequent passes.

〔Effect of the invention〕

As described above, according to the present invention, without using a tracing sensor, the operator measures the groove and inputs it at hand along with conditions such as plate thickness, and automatically teaches the start and end of welding. It is possible to realize a small, portable welding robot that performs multi-layer welding on the welding section, and it can also automatically handle tapered root gaps in straight welding sections by simply inputting the groove accuracy at the start and end of the welding section. , it is possible to produce a remarkable effect in automating the multi-layer welding work of steel frame joints, which has a low automation rate.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram showing the configuration of a robot system used to carry out the present invention, Fig. 2 is a sectional view showing a specific embodiment of the robot body, and Fig. 3 shows a calculation method using linear interpolation for a tapered root gap. It is an explanatory diagram. 1: Controller box, 2: Welding power source,
3: water cooler, 4: shield gas cylinder, 5: welding wire feeder, 6: running rail, 7: robot body, 8: welding torch, 9: teaching box, 71: outer case, 72: feed shaft mechanism, 7
3: Inner box, 74: Y-axis unit, 75:
Feed axis mechanism, 76: K axis unit.

Claims (1)

[Claims] 1. Welding speed, welding current, weaving width, wire feeding speed, number of layers, etc. for each type of welding such as side butt, downward butt, and fillet, and for each type of groove shape and angle. A plurality of welding conditions are obtained in advance through experiments using various plate thickness values and root gap values as parameters and are compiled into a database, and the positions of the start and end of the straight weld section of the joint to be welded are set by teaching. At the same time, during this teaching, each root gap value at the start and end of the straight welding section is memorized, and a root gap correction value in the middle of the straight welding section is obtained by linear interpolation from the memorized values, and the plate thickness of the joint to be welded is determined. Welding conditions corresponding to and root gap value are read from the database, welding is started from the starting end under the read welding conditions, and each time the welding distance from the starting end reaches a predetermined value during welding, the straight line is By reading the welding conditions corresponding to the root gap correction value obtained by interpolation from the database and correcting the welding conditions, the straight line welding is performed while keeping the excess amount for each pass almost constant by changing the welding speed and weaving width. A multi-layer automatic arc welding method that is characterized by automatic welding to the end of the section. 2. The multi-layer automatic arc according to claim 1, wherein a locus of the center of the route is calculated from the positional information of the starting end and the end at the time of teaching and root gap values of each end, and the moving locus of the torch is made to follow the center locus. Welding method.