JP2586091B2 - Swing rotary arc welding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to an oscillating rotary arc welding method employing oscillating rotary arc welding, and in particular, an appropriate penetration width can be obtained with respect to variations in groove width. The present invention relates to variable control of the welding speed in order to variably control the swing width as described above, to compensate for the amount of welding for variations in the groove width, and to keep the bead height constant.

[Prior art] According to a high-speed rotating arc welding method in which an arc is rotated by rotating a welding wire at a high speed around a rotation axis, the physical effect of the arc is dispersed in the periphery, the peripheral dispersion of penetration, flattening. Advantages such as formation of a bead (curved bead) or improvement of a wire melting rate by a rotating centrifugal force are obtained, and it is particularly effective when used for narrow groove welding of a thick plate.

Further, this kind of high-speed rotating arc welding has a property that the arc voltage E and the welding current I change when the groove width changes.

Hereinafter, the change characteristics of the voltage E and the welding current I will be described.

Figure 9 is a side view of a rotating arc torch 3 tip and groove 2 when performing high-speed rotating arc welding, the welding direction in the figure in a direction toward the surface from the plane rear surface in the plane perpendicular to, l _a is The arc length, C _f , R, L indicates the position of the wire 5 during rotation, C _f is the position of the wire 5 in front of the welding direction, R is 90 degrees clockwise to the right in the welding direction, L Indicates the position of the left wire 5 in the counterclockwise direction in the welding direction. Cw indicates the rotation direction of the wire 5.

Figure 10 is a view of the welding unit shown in FIG. 9 from the rotational axis O direction, C _r behind the wire for welding direction Z 5
Is the rotation angle of the wire 5 with respect to the welding direction Z, θ
Indicates a rotation angle when the position of the wire 5 coincides with the center line WC of the groove.

As shown in FIGS. 9 and 10, when the wire 5 rotates about the rotation axis O at a constant wire feeding speed, the wire 5 and the groove wall are changed depending on the position of the wire 5 at the time of rotation. different distance δ between the arc length l _a is changed. Arc length l _a
Changes, the load characteristics change, and the welding current I and the voltage E between the electrode 1 and the groove (hereinafter referred to as arc voltage) also change.

The changes in the welding current I and the arc voltage E indicate changes based on a sine wave corresponding to the position of the wire 5. This is because when the wire 5 rotates, the distance δ changes based on the sine wave according to the position of the wire 5.

This relationship holds not only for the consumable electrode but also for the non-consumable electrode. Also, the relationship is that the groove shape of the welding target is V
This holds true for both the character groove and the narrow gap groove.

FIGS. 11 (a) and 11 (b) show the waveforms of the arc voltage E and the welding current I which change according to the position of the rotating wire 5, that is, the arc. In the figure, (a) shows the waveform of the arc voltage E, and (b) shows the waveform of the welding current I, and each waveform has a vertically inverted shape.

The waveform of the welding current I shown in FIG. 11 (b) can be obtained only by a welding power source having a constant voltage characteristic, and the arc voltage E
Can be obtained in any of the welding power sources having the constant voltage characteristic and the constant current characteristic.

In FIGS. 11 (a) and 11 (b), the waveforms shown by solid lines are the center line WC of the groove and the rotation axis O as shown in FIGS.
If the bets are shifted △ X, if not out waveform indicated by the broken line, namely a case where the position C _f and the position connecting the C _r line of the wire 5 is coincident with the center line WC of the groove.

Figure 11 (a), as indicated by the broken line in (b), when the rotational axis O with the center line WC of the groove is not shifted, the waveform around the position C _f of the wire 5 is symmetrical but the rotation axis O is displaced from the center line WC of the groove waveform around the position C _f of the wire becomes uninteresting.

In addition, by detecting and correcting the non-target of this waveform, the shift amount Δx in the X-axis direction can be corrected. That is, the waveform is divided right and left with respect to the welding direction around the point C _f , and the divided waveforms are respectively taken out from the point C _{f for a} certain angle φ _o , and the area of the waveform formed between the angles φ _o (integral value ) S
_L and S _R are obtained and the welding torch is corrected in the X-axis direction so that the areas S _L and S _R become equal, whereby the rotation axis O can be made coincident with the center line WC of the groove. This kind of groove copying using the arc confidence as a sensor is called an arc sensor type groove copying.

[Problem to be Solved by the Invention] As a welding method for forming a wide bead, a swing welding method of performing welding while swinging (weaving) a welding torch in a groove width direction is known. This type of swing welding method has not yet been introduced into the above-described rotary arc welding method, and its introduction has been desired. (Hereinafter, in the rotary arc welding method, a welding method in which a welding torch is weaved is referred to as an oscillating rotary arc welding method.) In realizing this oscillating rotary arc welding method, in order to make automatic welding unattended, the welding process must be performed. Automatic control of the swing width according to the fluctuation of the groove width, which is constantly changing, is indispensable. This is because if the groove width changes, the required penetration width also changes.

Accordingly, a main object of the present invention is to provide an oscillating rotary arc welding method capable of obtaining appropriate penetration by controlling the oscillating width with respect to the variation of the groove width.

In addition, it is a part of the subject of the present invention to compensate the welding amount by controlling the welding speed according to the controlled swing width, and to keep the bead height constant.

[Means for Solving the Problems] In order to achieve the above object, the oscillating rotary arc welding method according to the present invention rotates a welding wire at high speed around an axis with a predetermined radius of rotation to rotate an arc. ,
The welding torch is reciprocated in the groove width direction within the groove, and detects an arc voltage or a welding current, and holds a preset arc voltage or a welding current while performing an arc sensor type groove copying. As described above, when performing arc welding while controlling the welding torch position in the height direction with respect to the base material to be welded, an average value ± g of a change in the size of the groove width that can be predicted in the groove width direction is set in advance. A single position of mechanically controllable swing is set so as to be equal to the absolute value of the average value ± g, and a predetermined groove width, which is a design value of the groove width, is determined in advance. The initial value of the swing width for obtaining the obtained penetration is given as an integral multiple of the swing unit amount, and the value corresponding to the change in the groove width by the change characteristic of the arc voltage or the welding current with respect to the change in the groove width. Is detected, and the detected value is set to 0, + g according to the magnitude. , -G, and the swing width is controlled by increasing or decreasing the integer so that the penetration is maintained at a predetermined constant value with respect to a change in the groove width based on the result of the determination. It is.

At this time, the controlled swing width is set to W _w , and the cycle is controlled by controlling the welding speed in accordance with the controlled swing width to compensate for the welding amount and maintain the bead height constant. When the previous swing width is W _wo , the wire feed speed is V _fo , the desired bead height is h, the initial value of the welding speed is V _o , and the welding speed to be controlled is V, the following formula is used. controls the welding speed based on _{V fo / {(V fo /} V o) + (W w -W wo) h}.

[Operation] According to the present invention, an average value ± g of the empirically predicted change in the groove width is set in advance, and the detection of the groove width change based on the change characteristics of the arc voltage or the welding current is performed by changing the value. 0, + g,
Perform an approximate detection that is considered to be any of the discrete values of -g.

Further, a unit amount of swing that can be mechanically controlled is set to be equal to the absolute value | g | of the average value ± g, and the initial value of the swing width is given as an integral multiple of this swing unit amount. Can be The swing width to be controlled is given by increasing or decreasing the integer according to the detected value of the groove width transition 0, + g, -g.

Further, in addition to the swing width control, a change in the welding amount accompanying a change in the swing width is compensated by controlling the welding speed in consideration of the wire feed amount.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

In the following description, a method for detecting an arc voltage using a constant current power supply as a welding power supply will be described. However, when a constant voltage power supply is used, the welding current may be detected instead of the arc voltage. , Which can be selected according to the external characteristics of the welding power source.

FIG. 1 is an explanatory view showing the basic principle of the oscillating rotary arc welding method of the present invention, showing the trajectory of the tip of a welding wire and the axis of a rotating shaft.

In the drawing, WR and WL are swing turning points as viewed from the rotation axis O, respectively, WR indicates a swing right end, and WL indicates a swing left end. W _W is the swing width, that is, the distance between the swing right end WR and the swing left end WL in the groove width direction, and the center thereof is the groove center line WC.

In the following description, when it is necessary to specify the swing position WR, WL, WC of the rotating shaft O at the time of indicating the wire rotation position, the swing position symbol is added to the rotation position symbol in parentheses. Indicated by (For example, R _(WR) indicates the rotating wire position R when the rotating shaft O is at the swinging right end WR.) Also, e ₁ ... ₅ are the trajectories of the tip of the welding wire 5 due to the rotation. Let e _{1 to} e ₅ be one swing squill. Reference numeral 1 denotes a locus drawn by the rotation axis O due to the swing.

When an arc is generated and welding is developed, the welding wire 5 rotates in the cw direction about the rotation axis O while swinging speed V _W about the rotation axis O in the width direction of the groove 2, It moves in the welding direction Z at the welding speed V while swinging back and forth at the swing width W _W. Thereby, the locus 1 of the rotation axis O is obtained.

In this case, penetration width is formed is determined by the magnitude of the swing width W _W, the swing width W _W to obtain the necessary penetration width is determined in accordance with the groove width G.

For example, the design value of the groove width G G _O (hereinafter, referred to as a reference groove width G _O) and standard fluctuation width database values W _O the swing width W _W to obtain the necessary penetration width with respect. The reference swing width data database value W _O can be empirically determined according to the size of the reference groove width G _O. Now, assuming that the groove width G changes by ΔG with respect to the reference groove width G _O to become G = G _O + ΔG, the swing width W to be controlled is W
_W is given by W _W = W _O + ΔG (1)

Thus, standard fluctuation width W _O in groove width change ΔG swing width W _W to be controlled be added simply a is obtained.

However, in actual control, it is difficult to accurately detect the size of the groove width change ΔG, and the swing width W _W is only about several mm. It is mechanically difficult to continuously adjust the swing width W _W by following it.

Therefore, in the present invention, when detecting the groove width transition ΔG, an average value ± g of the magnitude of the groove width transition ΔG that can be empirically predicted is set in advance. In this case, the actual groove width change Δ
G is detected as a value corresponding to the size of the groove width change ΔG based on the change characteristics of the arc voltage described in the above-described conventional technique, and the detected value is compared with a certain set band. One of three types of ΔG = 0, ΔG = + g, ΔG = −g is selected. That is, the groove width change ΔG is 0, +
It is assumed that it is one of b and -g.

Also, the mechanically controllable swing unit amount W _unit is set to be equal to the absolute value | g | of the predicted groove width change ± g (W _unit = | g |), and the reference swing width database is set. as the value W _O,
It takes standard fluctuation width computation for conversion value W ₁ which is expressed as an integer multiple of the swing unit amount W _Unit, the swing width W _W to be controlled the integer according to the determination value of the groove width displacement ΔG Is increased or decreased.

Here, the swing unit amount W _unit is a controllable single position in the swing width direction determined by the accuracy of the motor that gives the swing, and the swing speed V _W [mm / se
c] and the number of revolutions of the arc (the number of revolutions of the wire) N [H _Z ], the relational expression of W _unit = V _W · (1 / N) · m (2) can be obtained.

Incidentally, m is the rotational speed of Osamuru arc swing unit quantity W _Unit, determined by the value of the swing unit amount W _Unit previously given. That is, the rotation speed of the arcing between the distance W _Unit moved oscillation speed _{V W, m = W unit /} {V W · (1 / N)} (3) and expressed. However, m is appropriately processed, for example, rounded off to take an integer.

Further, the conversion value W _1/2 for the reference oscillation width calculation of a half cycle is n times the oscillation unit amount W _unit , that is, the conversion value W ₁ for the reference oscillation width calculation is 2n times the oscillation unit amount W _unit. Expressed as W ₁ = 2n · W _unit (4)

n is an empirically determined reference swing width database value
From W _O , n = (W _O / 2) / W _unit = W _O / 2 · W _unit (5) However, n is appropriately processed, for example, rounded off to take an integer. Also, W _unit
Since | = g |, the swing width W _W when the groove width change ΔG is regarded as ΔG = ± g can be expressed as W _W = 2 (n ± 1) · W _unit (6).

By substituting the above equation (2) into equations (4) and (6), W = 2V _W · (1 / N) · m · n (7) W _W = 2V _W · (1 / N) · m · (N ± 1) (8) is obtained.

(8) is controlled to provide a swing width W _W to be controlled, the initial value is expressed by equation (7). Whether ± 1 of the fourth term on the right side of the equation (8) is positive or negative is determined by the determined groove width transition ΔG. That is, +1 is taken when ΔG = + g, and −1 when ΔG = −g. Also, ΔG
When it is considered that = 0, equation (7) is held.

Accordingly, the reference groove width G _O , the reference swing width database value W _O empirically obtained, the predicted groove width change ± g, the swing unit amount
Given W _unit , swing speed V _W , and the number of rotations N of the arc, the magnitude of the groove width change ΔG is determined, and the swing width W _W can be controlled based on the determination result in accordance with equation (8). .

In the above description, two types of values, a reference swing width database value W _O empirically obtained and a reference swing width calculation conversion value W ₁ given by the equation (7), are used as the reference swing width. and which is, if a method is accumulated data according to the present invention, the value obtained empirically also (7) only may be used standard fluctuation width computation for conversion value W ₁ given by equation.

If the groove width G changes, the required welding amount also changes. In the present invention, the welding amount is compensated by controlling the welding speed, and the bead height is kept constant.

Hereinafter, the welding speed control will be described with reference to FIG. FIG. 2 is a side sectional view of the groove 2 shown in FIG. In the figure, 1a indicates the base material surface, and 2b indicates the groove bottom surface.

First, in the swing cycle, swing width W _W's swing across wire position L _(WL), the following relationship with respect to the groove width G in R _(WR).

G = W _W + 2ΔW (9) In the equation (9), ΔW is a distance between the rotation axis O and the groove wall surface at the swing end, and is basically an arc length to be set and a wire rotation diameter. If the set values of the arc voltage and the welding current are constant, they are constant even if the development width G fluctuates. This has been confirmed experimentally.

On the other hand, the wire feeding speed in one cycle of this swing is
_Assuming that V _f , the welding speed is V, and the cross-sectional area of the formed deposited metal is A, there is a relationship of V _f = A · V (10).

Further, assuming that the vertical distance between the bead surface and the wire positions at both ends of the swing is Δh, the bead height is h, and the groove angle is θ, A = ｛(W _W + 2ΔW) −2Δh · tan θ−h · tan θ ｝ H
(11) If this depth of groove 2 (or the thickness of the base material 1) ₁ and the included angle θ is constant, the deposition rate metals always provide a constant bead height even variations of a groove width G Cross section A
But shows that can be calculated based on the value of the swing width W _W obtained from the swing width control.

Therefore, if the groove angle θ, the desired bead height h, the wire feed speed _Vf, and the above constant ΔW are set in advance,
An appropriate welding speed V can be determined according to the value of the swing width W _W obtained from the swing width control.

Next, a specific calculation for obtaining the welding speed V will be described.

Generally, welding speed V, welding current I, wire feed speed V _f
The welding conditions such as are set as initial values in advance in accordance with the groove shape, and thereafter the welding conditions are changed while observing the fluctuation state of the groove width. According to this general method, the initial values of the wire feeding speed V _f and the welding speed V are respectively set.
V _fo , V _o, and when welding is started under these conditions and one cycle of rocking is performed, the cross-sectional area A of the weld metal obtained at that time is obtained.
_O is given by A _O = V _fo / V _O (12) Next, the swing width of the front of the swing cycle comprising a swing width W _W and W _WO. At this time, for example, W _W is larger than W _WO (W _W
> W _WO ).

In such a case, in order to secure a constant bead height h, the welding cross-sectional area must be increased by ΔA = A- _AO . That is, in the above-mentioned equation (11) representing A, all the parameters except _WW are constants, so that ΔA = A−A _O = (WW− _{W WO} ) h (13) The welding speed V for ensuring the bead height h is given by the following equation from the equations (11) and (12).

_{V = V fo / A = V} fo / (A O + ΔA) = V fo / {(V fo / V o) + (W W -W WO) h} (14) Using this equation (14), open If the desired bead height h is given in a groove whose tip width fluctuates, the initial welding speed V _O and the wire feed speed V _fo determined before welding, the swing width W _W and the swing width of the previous cycle are determined. the difference between W _WO, i.e., W _W -W _WO proper welding speed V in the swing width W _W from is determined by calculation.
This method can be applied irrespective of the set value if the groove angle θ ₁ is constant or does not significantly change. Therefore, the method can be applied not only to the groove shape shown in FIG. 2 but also to an asymmetric groove. .

FIG. 3 is a block diagram showing a control yarn of the swing rotary arc welding method according to the preferred embodiment of the present invention.

In the drawing, a groove 2 is formed in a member 1 to be welded, and a rotating arc torch (welding torch) 3 is formed by a rocking arm 6 while rotating a welding wire 5 in a direction of an arrow CW by a motor 4. While being weaved (oscillated) in the direction, the welding carriage 7 feeds the welding along the groove 2 in the direction perpendicular to the paper surface, and a welding bead (not shown) behind it.
Is formed.

Further, the swing by the swing arm 6 is performed via a transmission mechanism 9 that converts the rotational movement of the swing motor 8 that can be driven forward and backward into a linear movement.

The method of swinging the rotating arc torch 3 is as follows: (a) The pitch P is fixed.

(B) to a constant swing speed V _W.

Here, in order to keep the load of the swing motor 8 constant, (b) will be adopted.

The rotating arc torch 3 is supported by both moving mechanisms 10X and 10Y in a groove width direction (X-axis direction) and a height direction (Y-axis direction). The X-axis moving mechanism 10X corrects the position of the torch 3 in the X-axis direction, that is, performs groove tracing control. The Y-axis moving mechanism 10Y moves the torch 3 in the Y-axis direction so that the arc length is always constant. The position of the torch 3 is corrected, that is, constant arc length control is performed.

These moving mechanisms 10X and 10Y are driven by an X-axis motor 11X and a Y-axis motor 11Y, respectively.

The welding current / voltage is supplied to the torch 3 from the welding power source 12. The welding power source 12 is either a constant current power source or a constant voltage power source depending on the welding purpose. An arc voltage detector 13 is connected to the welding power source 12 in parallel. This detector 13E
It is provided as needed for control described later.

The wire feed motor 14 is controlled by a feed motor control device 15, the swing width of the swing arm 6 is controlled by a swing control device 16, and the welding speed of the welding carriage 7 is controlled by welding. This is performed by the speed control device 17.
Further, the X-axis motor 10X and the Y-axis motor 10Y are controlled by the X-axis controller 18X and the Y-axis controller 18Y.

Control commands to the feed motor control device 15, the swing control device 16, and the welding speed control device 17 are given from a main processing device 19 such as a microcomputer, and the main processing device 19 sets each constant and set value. An input device 20 for inputting is attached, and a groove width change ΔG from a groove width detecting device 21 is further provided.
The detection signal and the wire rotation speed detector 22 attached to the motor 4
, And receives a wire rotation number detection signal.

Next, the operation of the swing width control of the above embodiment will be described with reference to FIG.
This will be described with reference to the flowchart in FIG. The calculation in this flowchart is performed by the swing width calculation unit 192 in the main processing device 19.

First, a reference swing width database value W _O , which is empirically determined according to the reference groove width G _O , an arc rotation speed N, a swing speed V _W , and a swing unit determined according to a predicted groove width ± g. The value of the quantity W _unit is input from the input device 20 to the main processor 19, and is stored in the memory.
It is stored in 191 (step SA).

Next, the equations (3) and (5) are calculated, and the integer values of m and n are obtained. These m and n are stored in the memory 191 (step SB), and the reference value swing given by the equation (7) is obtained. width calculating conversion value W ₁ is calculated (step SC).

After the above calculations, welding is started (step SD).
Here, the arc rotation diameter position at the start of welding is shown in FIG.
e is ₁ .

When the swing of the Ns cycle is completed (step SE), the main processing unit 17 holds n in the expression (7) based on the detection signal of the groove width displacement detection device 21, or (8)
In the equation, it is determined whether the fourth term on the right side holds n + 1 or n-1 (step SF), and the swing width _WW is calculated based on the determination result (step S).
G).

Based on this, one cycle of swing welding is performed (step SH).

Here, it is determined whether or not the welding is completed (Step 1). Based on the result of the determination, the welding is completed, or the operation of (Step SF) → (Step S1) is repeated until the welding is completed. It is.

Such control of the swing width W _W is performed by the control on the variation of the groove width G in the weld, the proper penetration width can be obtained regardless of the variation of the groove width G.

Note that in (step SA) of the description, it is assumed to input the standard fluctuation width database values W _O which is obtained empirically according to the reference groove width G _O, reference rocking represented by equation (7) If data Dohaba calculation conversion value W ₁ is obtained in advance (step SA) in V _W, n, m, and enter the n (step SB) ~ (step SC) may be omitted.

The Ns cycle in step SE corresponds to a waiting time for performing control after the welding pool (molten pool) has become stable after the arc start. For example, it is considered that 15 cycles are employed. But
It can be changed as appropriate depending on the welding conditions.

Next, the operation of the welding speed control of the above embodiment will be described with reference to the flowchart of FIG. The calculation in this flowchart is performed by the welding speed calculator 194 in the main processing device 19.

First, the desired weld bead height h, the initial value V _{O of} the welding speed,
The set value _Vfo of the wire feeding speed is input from the input device 20 to the main processing device 19 and stored in the memory 193 (step S
a).

Next, the calculation of A _O = V _fo / V _O is performed, and is similarly stored in the memory 193 (step Sb).

Further, the standard fluctuation width computation for conversion value W ₁ given in step SC of the swing width control flowchart of FIG. 4 are stored in the memory 193 (step Sc). The standard fluctuation width computation for conversion value W ₁ and W _WO.

After the above, welding is started (Step Sd).

Here arc rotational diameter position at the start of welding and e ₁ in the first view.

When the swing of the Ns cycle is completed (Step Se), the flowchart SG of the swing width control in FIG.
Swing width W _W given by is stored in the memory 193 (step sf).

This swing width W _W is _defined as W _W1 .

Welding speed in more welding operation is V _O.

Next, main processor 19 calculates (W _W1 −W _WO ) h (step Sg), and further calculates welding speed V based on equation (12) relating to welding speed V (step Sg). Sh).

The welding speed V is obtained as described above, and based on this, the welding in the half swing cycle is performed (step S).
j).

Here, it is determined whether or not welding is completed (step
Sk), the welding is terminated according to the result of the determination, or the above-described operation of (Step Sf) → (Step Sj) is repeated until the welding is completed. Also, in step Se
The Ns cycle is equivalent to a waiting time for performing control after the welding pool (molten pool) is in a stable state after the arc start. For example, it is given that 15 cycles are employed. Can be changed as appropriate.

As described above, according to this embodiment, the bead height is maintained constant by a variably controlling the welding speed V with respect to the change of the swing width W _W.

The swing reversal command by the main processor 19 when performing the swing width control and the welding speed control is performed based on the wire rotation speed detected by the wire rotation speed detector 22. From the above equation (7), the number of rotations of the arc while the rotation axis O moves by the reference swing width calculation conversion value W ₁ is 2 (m · n).
Given by From the above equation (8), the number of rotations of the arc during the movement of the rotation axis O by the swing width W _W is 2 (m · n ±
Given in 1). Therefore, when the swing width to be controlled is determined, the timing at which the reversal command is given can be determined by detecting the rotation speed of the arc, that is, the wire rotation speed.

Next, an example of the detection direction of the groove width change ΔG based on the change characteristics of the arc voltage will be described.

As shown in FIG. 1, the wire 5 has left and right groove walls 2L, 2R.
_{Are the} wire rotation positions L _(WL) and R _(WR) .

Here, the distance δ _L between the wire rotation position L _(WL) and the groove wall 2L
And the distance δ _R between the wire rotation position R _(WR) and the groove wall 2R is equal (δ _L = δ _R ). Therefore, according to the arc characteristics described in the above prior art, the arc voltage at the rotation position L _(WL) is Detection value E
_L becomes the detected value of arc voltage in _(WL) and the rotational position _{R (WR) E R (WR} ) and equal _{(E L (WL) = E} R (WL)). This detection value
_{E L (WL), E R} (WR) is by the variation characteristic when the groove width G is changed, while maintaining the relationship of _{E L (WL) = E R} (WR), the magnitude of the value change I do.

Therefore, if the values of E _{L (WL)} and E _{R (WR)} are detected, it is possible to detect the positive or negative with respect to the reference groove width G _O , but here, in consideration of the voltage drop in the base material, The detection value on the swing center WC is also detected. Since the swing center WC is the center of the groove 2, the distance δ _CL between the wire rotation position L _(WL) and the swing center WC in the groove width direction and the wire rotation position R _(WR) -the swing center WC Since the distance δ _{CR in} the groove width direction is equal to (δ _CL = δ _CR ), the detected value E _{cf (wc)} of the arc voltage at the wire rotation position C _{f (WC)} on the swing center WC and The following relationship is established between the detected values E _{L (WL)} and E _{R (WR)} .

{(ER _(WR) + _{EL (WL)} ) / 2} -Ecf _(wc) = C (14) where C is a constant determined by the size of the groove width G.

 This equation (14) is described as follows for simplification.

(E _s −E _cf ) (14a) (However, E _s = (E _{R (WR)} + E _{L (WL)} ) / 2, E _cf indicates E _{cf (wc)} .) The value of the expression (14a) Indicates a value corresponding to the actual groove width G.

Further, if the value of the expression (14a) in the reference groove width G _O is E _O , this E _O indicates a value corresponding to the size of the reference groove width G _O.

The difference between the above equation (14a) and this E _O , (E _s −E _cf ) −E _O = ΔE (15) indicates a value corresponding to the size of the groove width transition ΔG. This ΔE is compared with a set dead band having an upper limit E ₁ and a lower limit E ₂ , and E ₂
≦ ΔE ≦ ΔG = 0 when the E _1, when the ΔE> E ₁ ΔG = +
g, when the ΔE <E ₂ regarded as ΔG = -g.

Next, while referring to FIG.
A more specific detection method for obtaining the following will be described. FIG. 6 shows an example of a circuit block of the groove width displacement detecting device 21 shown in FIG.

Above _{E R (WR), E L} (WL), to detect E _cf the _(WC) instantaneously, because the influence of noise or the like also have difficulties, each voltage waveform in the range of the integration angle phi _O here Integrate and detect as integrated value. Further, this integration is performed α times (for α rotations of the arc) before and after the rotation axis O is on WR, WL, and WC, respectively. In this case, while the arc rotates α times, the moving distance of the rotation axis O in the groove width direction due to the swing is (V _W / N) α
Therefore, the sections in which the above detection is performed are WR ± ( _VW · α / 2N) WL ± ( _VW · α / 2N) WC ± ( _VW · α / 2N).

In FIG. 6, an arc voltage E is detected by an arc voltage detector 13E, and a difference E− between a detected arc voltage E and a reference voltage E _O which is an average value of the arc voltage E set in a presetter 211 in advance. E _O is calculated by the operational amplifier 212. The calculated value E-E _O is divided by switch 213 and output. The timing of division by the switch 213 is performed by a command signal from the controller 214.

The controller 214 calculates the swing position WL, WC, WR of the rotation axis O set in the swing position setting device 215 and the number of processes α set in the process number setting device 216,
Detection interval WL ± (V _W・ α / 2N), WR ± (V _W・ α / 2N), WC ±
( _VW · α / 2N) is obtained, and these values are compared with the actual swing position of the wire axis detected by the wire swing position detector 217. Then, the swing position of the wire shaft core is detected in the detection section.
While in WL ± ( _VW・ α / 2N), wire rotation position detector 218
The angle of rotation φ of the wire detected in
Integral angle setter 219 with a fixed angle φ _O in the range of 0 °
And the output φ _{O of} the wire, the rotation axis of the wire is WL ± (V
_W · α / 2N), a section of φ _O around the wire rotation position L is defined as an L section, and the waveform of this section is changed by the switch 213.
Are output from the L side. Similarly, the section of phi _O around the wire rotation position C _f between the swinging position of the wire axis is in the detection zone _{WC ± (V W · α /} 2N) as C _f interval, the waveform of the section Is output from the _Cf side of the switch 213, and the section of φ _O around the wire rotation position R while the swing position of the wire axis core is in the detection section WR ± (V _W · α / 2N) is R. As a section, the waveform in this section is output from the R side of switch 213. L side switching unit 213, C _f side, the output from the R side is respectively integrated by the integrator 220, 221, 222. The integrators 220, 221 and 222 perform waveform integration on the rotation of the arc for α times (the number of processes set in the setting unit 216) output via the controller 214, and output the outputs S _L , S _R and S _cf. Output to the storage units 223, 224 and 225, respectively. The memory units 223, 224, and 225 store the signals S _L , S _R , and S _cf input from the integrators 220, 221, and 222 repeatedly α times while adding S _L , S _R to the adder 226 and S _cf to the differential amplifier. Output to one end of 227. The output S _L + S _R of the adder 226 is 1/2 by the arithmetic unit 228
Is multiplied by the average value of S _L and _{S R (S L + S R} ) / 2 = S S is output. The output S _S of the arithmetic unit 228 is input to the other end of the differential amplifier 227, the differential amplifier 227, the difference between S _S and _{S cf} (S S -S _cf) is determined. This difference, (S _S −S _cf ), is output to one end of the differential amplifier 229. Further, corresponding to the _(S S -S _cf) when the initial value of the storage unit 230 _(S S -S _cf), i.e. the standard fluctuation width to obtain a proper penetration by the reference GMA width are given the value S _O to the stored and held, the value is given to the other end of the differential amplifier 229, the differential amplifier _{229, (S S -S cf)} -S O = Δ
S is required. The value of ΔS is a value corresponding to the groove width change ΔG. Further, ΔS is given to the discriminator 231.

Meanwhile, the band having an upper limit S ₁ and a lower limit S ₂ that is set in the dead band setting unit 232 are compared with ΔS by discriminator 231, the discriminator 2
31 is 0 when S ₂ ≦ ΔS ≦ S ₁ , + g when ΔS> S ₁ , and ΔS <S ₂
In the case of, -g is output.

The reason why the range of 180 ° angle phi _O, which is set to the integrator angle setter 219 in the above description from 5 °, the angle phi _O 5
This is because when the angle is less than or equal to °, detection is difficult due to the influence of noise. The band set in the dead zone setting unit 232 is appropriately set according to the magnitude of the predicted groove width change ± g.

Next, control of the torch movement in the X-axis direction, that is, groove tracing control will be described. In the groove copying control, the sixth
The signals S _L and S _R obtained in the circuit block of the groove width displacement detection device 21 shown in the figure can be used. S _L and S _R detected at the left and right edges of the groove are equal (S _L = S _R ) if the target position of the arc is appropriate. Thus, the difference between the S _L and S _R S _L
-S _R is obtained, and the groove following control can be performed by correcting the position of the torch 3 in the X-axis direction so that the difference S _L -S _R becomes zero.

FIG. 7 shows an example of a circuit block of an X-axis control device 18X that performs such drive control in the X-axis direction.

The signals S _L and S _R output from the storage units 223 and 225 in FIG. 6 while repeatedly storing and holding α times are input to both input terminals of the differential amplifier 181X in FIG. This differential amplifier
From 181X, the difference S _L -S _R of S _L and S _R is output, the difference output is input to the X-axis motor drive controller 182x. The X-axis motor drive controller 182X moves the torch 3 via the X-axis motor 11X in the X-axis direction so that the difference output becomes zero.
In this way, the groove tracing control is performed by correcting the position of the torch 3 in the X-axis direction.

Next, constant arc length control by the Y-axis controller 18Y will be described.

FIG. 8 shows an example of a circuit block of the Y-axis control device 18Y for performing such control.

In FIG. 7, the detected arc voltage E from the arc voltage detector 13 is input to one end of a differential amplifier 181Y. on the other hand,
To the other end of the differential amplifier 181Y, a reference value previously set to an appropriate value by the setting device 182Y is input. Amplifier 18
1Y outputs the difference between the reference value and the detection value described above,
This difference output is input to the Y-axis motor drive controller 183Y.
The Y-axis motor drive controller 183Y moves the torch 3 in the Y-axis direction via the Y-axis motor 11Y so that the difference output becomes zero, that is, the arc voltage E becomes a reference value.
As a result, the arc voltage E is kept constant, and the arc length is kept constant.

Therefore, when the torch 3 moves in the X-axis direction, the torch 3
Moves in the Y-axis direction along the groove wall. According to this control method, no matter how the groove shape changes, the tip of the torch 3 always reverses and swings within the groove width while maintaining a constant distance from the base material surface 1a or the groove bottom surface 2b. Will be repeated.

[Effects of the Invention] As described above, according to the present invention, since the swing width is variably controlled based on the change in the groove width, appropriate penetration can be obtained even when the groove width changes. .

Further, since the welding speed is variably controlled based on the change in the swing width to compensate for the welding amount, there is an effect that a constant bead height can be maintained even when the groove width changes.

[Brief description of the drawings]

FIG. 1 is a top view of a welded portion for explaining the principle of the oscillating rotary arc welding method of the present invention, FIG.
3 is a block diagram showing a control system according to the oscillating rotary arc welding method of the present invention, FIG. 4 is a flowchart showing an operation of oscillating width control, FIG. 5 is a flowchart showing an operation of a welding speed control, 6 is a block diagram showing an example of a configuration of a groove width detecting device, FIG. 7 is a block diagram showing an example of a circuit block that performs groove copying, and FIG. 8 is an example of a circuit block that performs constant arc length control. FIG. 9 is a top view of a welded portion of a conventional rotary arc welding method, FIG. 10 is a side view thereof, and FIGS. 11 and 12 show an arc voltage E which changes according to the position of a rotating wire. 4 is a graph showing waveforms of welding current I and welding current I. In the figure, 1 is a member to be welded, 2 is a groove, 3 is a rotating arc torch, 19 is a main processing device, 20 is an input device, 21 is a groove width detecting device, 22 is a wire rotation speed detector, and 192 is a swinging device. The dynamic width calculator 194 is a welding speed calculator. In the drawings, the same reference numerals indicate the same or corresponding parts.

Claims (2)

(57) [Claims] An arc is rotated by rotating a welding wire around an axis at a predetermined radius of rotation at a high speed, a welding torch is reciprocally oscillated in a groove in a groove width direction, and an arc voltage or welding current is applied. And performing arc welding while controlling the position of the welding torch in the height direction with respect to the base material to be welded so as to maintain a preset arc voltage or welding current while performing the groove scanning of the arc sensor system. At this time, the average value ± g of the change in the size of the groove width in the groove width direction that can be predicted is set in advance, and mechanical control can be performed so as to be equal to the absolute value | g | of this average value ± g. Set the unit amount of the swing, with respect to the reference groove width is a design value of the groove width, giving an initial value of the rocking width to obtain a predetermined penetration as an integral multiple of the rocking unit amount, The arc voltage or the variation of the groove width Detects the value corresponding to the change in the groove width based on the change characteristics of the welding current, discriminates this detected value to 0, + g, or -g according to the magnitude, and based on the discrimination result, An oscillating rotary arc welding method, wherein the oscillating width is controlled by increasing or decreasing the integer so that the penetration is maintained at a predetermined constant value with respect to a change in the width. 2. When controlling the swing width by the swing rotary arc welding method according to claim 1, the controlled swing width is W _w , the swing width one cycle before is W _wo , and the wire is The feed rate is V _fo , the desired bead height is h,
The initial value of the welding speed V _o, when the welding speed should be controlled to by V, welding speed based on the following _{equation, V = V fo / {(} V fo / V o) + (W w -W wo) h} Oscillating rotary arc welding, characterized by controlling