WO1984002099A1 - Welding method and apparatus - Google Patents

Abstract

In a method and apparatus for mechanised welding of a joint in a work piece a welding electrode (7) is melted at a rate V, the joint is prepared at an angle from the horizontal, the work piece is moved past the welding electrode (7) at a speed T and the welding electrode (7) is weaved across the joint at a frequency F and width W. With V and constant the variables F and T are synchronously changed with W such that: T = (W + w)/v and F = xT + y where w, v, x and y are predetermined constants. If is allowed to vary while V is constant then w, v, x and y are all found to be directly proportional to . Similarly if is held constant while V is allowed to vary the welding equations can be modified by replacing w, v, x and y which are all found to be proportional to V.

Description

WELDING METHOD AND APPARATUS

The invention relates to a mechanised method of welding a work piece and apparatus for use in such a method.

One problem which is commonly encountered in welding joints is that joint geometries vary considerably due to joint fit up variations and pass-to-pass differences in multi-pass welding. In heavy industry, for example, the root gap, that is the minimum separation between two work pieces, can vary between 0 and 10-12 millimetres or more. A further complication is that the thermal and mechanical stresses involved in welding can move the plates during welding and result in a joint whose geometry is not only varying due to the problems of joint preparation but also continuously and unpredictably with time. If the joint is being welded manually then some compensation can be made for this by the welder. However, an increasing requirement is for a mechanised welding system and with, such a system compensation does not occur. In multi-pass welding, the joint is subjected to not only the aforementioned problems, but also to problems of tuning and retuning the welding system when changing from one pass to the next. In the past, in order to provide some compensation for these problems, the weld pool has been weaved across the width of the joint as welding proceeds. Under ideal circumstances, the thickness of metal deposited with each pass of the welding gun may be, for example 5 millimetres. In order to assure the fusion of the weld bead in a joint of varying geometry, it is necessary to adjust the width of the weave to suit the joint. If the joint opens up or another pass is started, the weave width must be increased and consequently the thickness of the bead may fall to, for example 3 millimetres. Furthermore, the basic shape of the weld bead may also change to an unacceptable form. In mechanised welding, coping with a joint at different fill rates at different positions is very difficult because the position of the gun in relation to the surface of the work piece must be continuously varied.

Thus, when conventional welding systems encounter joints of varying geometry, the problem results that the weave width can only be tuned (or varied) within a very limited range whilst maintaining the fused integrity of the joint, the bead shape integrity, and the degree of fill. Similarly, when one pass is completed, the complete welding system may need to be retuned, prior to commencing a subsequent pass in order to maintain the integrity of the joint.

In accordance with the present invention, a mechanised method of welding a work piece comprises moving the work piece and welding electrode relative to one another along a weaving path, the speed of movement (T), the weave width (W), and the basic weave frequency (F), all as hereinafter defined, being controlled in accordance with the following relationships: T = (W + w)/v and F = xT + y where w, v, x and y are constants or coefficients which may depend on θ, the angle of the weld joint, and V, the rate of melting of the welding electrode (the electrode feed rate). The speed of movement (T) is the speed with which the weaving path is traversed in the general direction of the joint. The weave width (W) is the distance nominally at right angles to the general direction of movement between successive extreme points of the weaving path. The basic weave frequency (F) is the frequency of the weaving path without end dwell periods.

In the simplest form of the welding relationships where θ and V are constant, then w, v, x and y are all constants.

In one form where the welding position θ, the angle of the joint off the horizontal, is included as a variable then w, v, x and y are coefficients dependent on θ.

The relationships between T, W and F then became: T = (W + aθ + b)/(cθ + d) and F = θ(e + gT) + hT + k where a, b, c, d, e, g, h and k are all constants. In an alternative form where the welding electrode is melted into a weld pool at a variable rate V with θ constant, then the relationships linking T, W and F now became:

T = (W + IV +m)/(nV + p) and F = V(qT + r) + sT + u where 1, m, n, p, q, r, s and u are all constants.

We have discovered that if the relationships specified above are followed then flat, consistant, well formed weld beads will be produced independent of joint fit up or pass-to-pass considerations. Thus, if any of the parameters are significantly different from those satisfying the relationships then either an unacceptably hollow weld bead will be produced (thus having a lower fill rate than the optimum), or unacceptable peaky welds will be produced (due to an excessive fill rate).

Preferably, the weave width (W) is controlled in relation to the welding position (θ) or the rate (v) so that the basic weave frequency (F) and the speed of movement (T) are automatically determined by the relationships specified. In this way, a wide range of weave widths is possible whilst maintaining the fused integrity, the bead shape integrity, and the optimum fill rate of the joint. Thus, the problem of varying joint geometries when using mechanised systems is overcome. If weave width is to be used as the controlling parameter then some method must be provided to determine what weave width is required and this can be of a conventional type, for example electrical or optical.

The appropriate constants in the group a to y must be determined for the particular welding process involved. Preferably, where two or more of a root pass, fill pass, and capping pass are to be carried out then at least the appropriate constants in the group a to y are adjusted for each type of pass.

The method in accordance with the invention is applicable not only to MIG welding but also to most other welding methods such as TIG, plasma, or flux core wire welding, and also more sophisticated processes including electron beam and laser welding.

In accordance with another aspect of the present invention, apparatus for mechanised welding of a work piece comprises a welding electrode; means for moving the work piece and the welding electrode relative to one another along a weaving path and means for controlling the speed of movement (T), the weave width (W), and the basic weave frequency (F), all as hereinbefore defined, in accordance with the following relationships: T = (W + w)/v F = xT + y where w, v, x, y are constants or coefficients which may depend on θ, the angle of the weld joint, and V, the rate of melting of the welding electrode.

Examples of a method and apparatus in accordance with the present invention will now be described with reference to the accompanying drawings, in which:-

Figure 1 is a schematic diagram of welding apparatus; Figure 2A is a diagrammatic, perspective view of part of the welding apparatus;

Figure 2B is an end view of the apparatus shown in Figure 2A; Figure 2C is a partial plan of the apparatus shown in Figure 2A;

Figure 2D is a schematic diagram of the welding plane; Figures 3A and 3B illustrate two weave patterns; Figure 4 illustrates the variation of weave width with travel speed and basic weave frequency for a given electrode feed rate; Figure 5 illustrates various bead shapes for given electrode feed rate;

Figures 6(i) to 6(iii) illustrate examples in cross section of differing multi-pass welded joints; and Figures 7(i) and 7(ii) are schematic diagrams of two arrangements of the Figure 1

The welding apparatus illustrated in Figure 1 is a metal-inert gas (MIG) apparatus with the inert gas supply omitted for clarity. The apparatus comprises a weave controller 1; a preprogrammer 2; a power source 3; a travel speed motor 4; a wire feed unit 5; and a weaving head 6. An electrode wire 7 is fed from the feed unit 5 towards a work piece 8 as is well known.

In this apparatus, the work piece 8 is moved under the electrode wire 7 by means of the motor 4 and the wire 7 is caused to oscillate transversely to the direction of movement of the work piece 8 so that a weaving path 9 (Figures 2A - 2C) is described. This process can be characterised by three parameters. Firstly, the travel speed (T) which is the effective speed along the direction of the joint as indicated by an arrow 10 in Figure 2A. Secondly, the weave width (W) which is indicated by the reference numeral 12 in Figure 3A and is equal to the distance transverse to the joint line 11 between extreme points of the weaving path 9. Thirdly, the basic weave frequency (F) which is proportional to the period between successive cycles of the weaving path 9. If end dwell periods 13 (Figure 3B) are included in the weaving path 9 then there will be a longer effective weave period between cycles and this should be distinguished from the basic weave period previously defined.

These three variables have been shown to be linked by the following pairs of equations:

W = a'T + b' and W = c'F² + d'F + e' with a' to e' constants.

It has been shown that these equations can be simplified to: F = xT + y ) ₍₁₎ and T = (W + w)/v ) with w, v, x, y constant. This pair of equations represents the basic form of synchronous welding in accordance with the present invention. In the simplest form w, v, x, y are constants which are a function of the type of pass ie they may be different for a root run than for a fill pass or for a capping pass. By preprogramming and selection of the appropriate programme in relation to the stage of joint completion, the joint can be completed without having to tune or retune any parameter except weave width W. The weave width must be continuously tuned in relation to the joint fit-up and the pass type eg the weave width will be less for the first fill pass than for the second etc provided the joint preparation does not have parallel sides.

When used in this way the variables are clearly W, T and F and are controlled in relation to W the weave width. The constants would include wire feed speed (V), welding position (θ) and end dwell periods that may or may not be included into the weave cycle. It became apparent that the constants w, v, x and y are further dependent on the welding position θ and the wire feed speed (V). In one form, keeping the wire feed speed V constant, it has been shown that w = aθ + b v = cθ + d x = gθ + h and y = eθ + k where a, b, c, d, e, g, h, k are constants. Substituting into equations (1) we get: T = (W + aθ + b)/(cθ + d) ) (2) and F = θ(e + gT) + hT + k )

In an alternative arrangement where the welding elecctrode is melted into a weld pool at a variable rate V and the angle θ of the weld joint is a constant, then it has been shown that the constants w, v, x and y can be represented by: w = IV + m v = nV + p x = qV + s and y = rV + u where l, m, n, p, q, r, s, u are constants. Again substituting into equations (1): T = (W + IV + m)/(nV + p) ) (3) and F = V(qT + r) + sT + u )

The relationships between weave width and travel speed and basic weave frequency respectively are, for a given electrode feed rate and welding position θ, illustrated in Figure 4. Thus, for example, if a satisfactory weld is achieved with a weave width W1, a travel speed T1, and a basic weave frequency F1, these values can be suitably adjusted when it becomes necessary to increase the weave width to a value W2 by controlling the travel speed T and basic weave frequency F to take the values T2 and F2 respectively. A continuous relationship is provided between each of the three parameters thus enabling a wide range of weave widths to be accommodated. Alternatively, of course, either travel speed or basic weave frequency could be controlled instead of weave width.

Figure 5 illustrates how bead profile varies when the optimum relationships are not followed. The optimum relationship is illustrated in terms of the travel speed only in this particular case. If the optimum relationship is not achieved, then a hollow bead 14 or a peaky bead 15 will result. The desirable flat bead is illustrated by reference numeral 16.

In order to determine the appropriate constants from the group a to y required by the relationships It is necessary to carry out a number of preliminary passes the results of which are fed to the synchronous welding controller 1. The constants and coefficients are set up in the preprogrammer unit 2 and thereafter the weave controller is able to control travel speed and weave width in accordance with the appropriate relationships. We have found that where different types of pass are required, as in multi-pass weldings the variation of travel speed with weave width may differ from one pass type to another. Figures 6(i) to 6(iii) show different multi-pass welded joints, each having root pass R, and at least one capping pass C. Fig 6(i) shows several intermediate fill passes while Figure 6(ii) shows a joint capped above and below and requiring two fill passes on the upper surface and one on the lower surface. The variations of travel speed (T) with weave width (W) for root passes, fill passes and cap passes are illustrated in the upper section of Figure 4. The variation of weave widths W with basic weave frequency F however also varies with the type of pass. Two possible variations of the welding control shown in Figure 1 may be used depending on the equipment configuration. Figure 7(i) shows the simplest control with θ, V and pass type being preprogrammed and the weave width (W) 71 monitored by a suitable sensor while the controller 1 gives an appropriate control signal 72 (W) to the weaving head to control the width of the weave and in synchronism gives the signals 73 (F) to the weave head to control the frequency of the weave and 74 (T) to the motor 4 to control the speed of the weaving head along the joint. Figure 7(ii) probably represents the most complex system electrically requiring three sensors. A weave width monitor provides width information 75 (W) to the controller 1 as before. 76 provides angular information 77 (θ) and the wire feed unit 5 provides information 78 on the wire melting rate V to the controller 1. In this case only the pass type is preprogrammed (79) and the controller controls the weave width synchronously controlling weave frequency and weave head speed as in the Figure 7(i) arrangement.

In one example adopting relationship (2) given above, where a joint is filled at a wire feed of 3.25 m/min, a welding plane (θ) of 90° and including a 0.8 second end dwell period 13 in the weave cycle, the values of the various constants are as follows: a = + 0.10 e = + 0.0018 b = - 64.9 g = - 0.000047 c = + 0.0024 h = + 0.17 d = - 0.99 k = - 0.44

In use, information such as end dwell period, electrode wire size, joint geometry, welding direction, weave pattern type, and the constants a to y are programmed into the synchronous preprogrammer 2. The action of the travel speed motor 4 and the weaving head 6 are then controlled by the synchronous weave controller 1, 2 in accordance with the relationships previously specified. The pass type, wire feed speed and welding position can be either preprogrammed or used in combination, as pairs or individually in conjunction with the synchronous welding controller. The invention is particularly useful in this respect since once preprogramming has been completed all variation in welding motion is controlled solely by weave width. Clearly, means must also be provided to determine the correct weave width and this can be any conventional method and may include such methods as the use of light beams, laser beams, or mechanical probes. Although the weaving path 9 has been shown having a zig-zag pattern it could comprise simple harmonic motion or a more complex pattern. The invention may also be applied to joints other than the butt joint described; for example to fillet welds. The synchronous welding method may, by application to suitable apparatus, be used on circumferential and saddle joints in addition to longitudinal joints.

Claims

Claims

1. A method for mechanised welding of a joint in a work piece comprising a welding electrode and a means for moving the work piece and the welding electrode relative to one another along a weaving path characterised in that there is provided a means for controlling the speed of movement T of the welding electrode parallel to the length of the joint; and means for controlling the weave width W measured generally perpendicularly to the length of the joint ; and a means to control the basic weave frequency F; T, W and F being controlled such that:

T = (W + w)/v and F = xT + y where w, v, x and y are functions of the angle of the joint θ and the melting rate V of the welding electrode.

2. A mechanised method of welding according to claim 1 characterised in that the rate of melting of the welding electrode V is constant and the angle of the joint θ is constant or predetermined such that w, v, x and y are all constant.

3. A mechanised method of welding according to claim 1 characterised in that the rate of melting of the welding electrode V is maintained constant and w, v, x and y are each directly proportional to the angle of the joint θ such that the welding is controlled in accordance with the following relationships: T = (W + aθ + b)/(cθ + d) and F = θ (e + gT) + hT + k where a, b, c, d, e, g, h and k are constants.

4. A mechanised method of welding according to claim 3 characterised in that the angle of the joint θ is measured and the rate of melting of the welding electrode V is pre-adjusted.

5. A mechanised method of welding according to claim 1 characterised in that the angle of the joint θ is constant or predetermined and w, v, x and y are each directly proportional to V such that the welding is controlled in accordance with the following relationships:

T = (W + IV + m)/(nV + p) and F = V(qT + r) + sT + u where l, m, n, p, q, r, s and u are constants.

6. A mechanised method of welding according to any one of claims 2 to 5 wherein the constants are predetermined in relation to the type of welding process used.

7. A mechanised method of welding according to claim 6 characterised in that the welding electrode makes more than one pass relative to the work piece and the constants are predetermined for each type of pass as hereinbefore defined.

8. A mechanised method of welding according to any one preceding claim characterised in that the weave width W is measured in relation to the joint being welded.

9. Apparatus for mechanised welding of a joint in a work piece comprising a welding electrode, means for moving the work piece and the welding electrode relative to one another along a weaving path characterised in that there are included means for controlling the speed of movement T, the weave width W and the basic weave frequency F, all as hereinbefore defined, in accordance with the following relationships:

T = (W + w)/v and F = xT + y where w, v, x and y functions of the angle of the joint θ and the melting rate V of the welding electrode, as hereinbefore defined.

10. Apparatus for mechanised welding according to claim 9 characterised in that there is included means to measure the angle θ (76).

11. Apparatus for mechanised welding according to claim 9 or 10 characterised in that there is included means to measure the width of weave W (71, 75) in relation to the joint.

12. Apparatus for mechanised welding according to any one of claims 9 to 11 characterised in that the rate of melting of the welding electrode V is maintained constant and the angle of the joint θ is constant or predetermined such that w, v, x and y are all constant.

13. Apparatus for mechanised welding according to any one of claims 9 to 11 characterised in that the rate of melting of the welding electrode V is maintained constant and w, v, x and y are each directly proportional to the angle of the joint θ and the welding is controlled in accordance with the relationships:

T = (W + aθ + b)/(cθ + d) and F = θ(e + gT) + hT + k where a, b, c, d, e, g, h and k are constants.

14. Apparatus for mechanical welding according to any one of claims 9 to 11 characterised in that the angle of the joint θ is constant or predetermined and w, v, x and y are each directly proportional to the rate of melting of the welding electrode V and the welding is controlled in accordance with the relationships:

T = (W + IV + m)/(nV + p) and F = V(qT + r) + sT + u where l, m, n, p, q, r, s and u are constants.