GB2045473A - Control of arc welding - Google Patents

Abstract

For welding a first element 11 (Fig. 5) onto a face of a second element 10 with a butt weld, an arc from a welding head 25 is directed into one angle between the two elements 10, 11 and the weldpool in the opposite angle is optically viewed by means 27, 28, 29 to provide a control signal dependent on the radiation from the molten weldpool forming the underbead to control the power supply 36 to the welding head 25 and/or the speed of drive 26 thereto. The position of the arc on the weldpool is controlled by a magnetic deflection coil 45 in response to radiation from the surface of the second element opposite to the face to which the joint is made to prevent penetration of that element. <IMAGE>

Description

SPECIFICATION Improvements in or relating to methods of and apparatus for arc welding This invention relates to methods of and apparatus for arc welding and is concerned more particularly with the wdlding of a first element onto an elongate region of a surface of a second element where the weld has to extend over the whole of the region where the first element abuts the second element.

If a tee butt weld is considered, where the first element has an elongate surface region which abuts and is to be welded to a surface region of the second element, a good weld requires full penetration, that is to say fusion of the metal through the thickness of the first element (i.e. the narrower dimension of said elongate surface region) when the first element is adjacent the surface of the second element. Penetration of the second element however is undesirable since any such penetration may impair the integrity of the second element. This problem arises in many constructions. As one example, it may be required to produce a weld between a tube and a flange with full penetration of the flange but without fusion into the tube bore. As another example, a sleeve joint may have to be made between inner and outer tubes without impairing the integrity of the inner tube.

Tungsten inert gas (TIG) welding is widely used for this type of welding, relative movement being effected between the welding arc and the workpiece along the length of the required weld. One of the main problems with TIG welding in this way is that the shape and position of the welds commonly have significant variance. These variations in shape and position produce weld defects with possible rejection of welds. The cause of the problem is not fully understood but instabilities may arise in the electromagnetic convection within the weld pool, stemming from complex arc/root interactions. A very considerable improvement is obtained by use of feedback control for automatical control of the relative movement between the arc and workpiece and/or of the magnitude of the arc current.In particular, in a weld in which full penetration is required, it is possible to utilise optical radiation from the rear surface of the weld to measure the size of the penetration. This radiation may be used to control automatically the welding speed, that is to say the rate of relative movement between the workpiece and welding head, and/or the current in the welding arc. Preferably both the welding speed and welding current are controlled simultaneously in accordance with the measured radiation, as is described for example in the specification of Application No. 47575/77.

In applying such feedback control to the welding of a first element which abuts against part of the surface of a second element, and in which the weld has to penetrate through the first element, the welding arc may be directed onto one face of the first element and to control the welding speed and the welding current automatically in accordance with the radiation from the opposite face of the first element. The present invention is concerned with the problem of preventing any breakthrough of the welding pool through the second element. Typically the weld pool might reach to within some 2 or 3mm away from the face of the second element remote from the welding face.

According to one aspect of the present invention, in a method of making a weld between a first element and a second element, wherein the first element abuts an elongate region of a surface of a second element, with the first element extending outwardly from that surface, and the second element having a thickness, normal to said surface, which is small compared with the dimensions of said surface; wherein the weld is effected by an arc directed onto said elements with relative movement between the arc and the elements along the length of the longer dimension of said surface region, the arc producing a weld pool penetrating the first element, transversely to said longer direction, from the arc to a rear surface of said first element, and with the weld pool only partially penetrating the second element from the surface facing the arc to an opposite surface thereof; and wherein (a) the speed of relative movement between the arc and the elements and/or the power input to the arc is controlled automatically by a first feedback control responsive to radiation from the molten weld pool on said rear surface of the first element, and (b) the position of the arc on said weld pool in a direction transverse to the line of relative movement is controlled automatically by a second feedback control responsive to radiation from said opposite face of the second element to prevent penetration of the second element. In this method, penetration through the second element is prevented by the automatic control of the second feedback loop which effects a deflection of the arc. In practice this deflection need only be quite small so that the arc still remains directed at the molten weld pool extending through the first element.The radiation from the rear face of the first element need not therefore be significantly affected by such movement of the arc and thus radiation from said opposite face of the second element therefore can be used in a feedback control loop (said second feedback control) to control the position of the arc without interfering with the proper operation of the first feedback control loop which is used to control the power input and/or the welding speed.

The second control loop may be arranged to provide an analog output control of the arc position but a simple two-state control of the arc position may be employed. Such a twostate control system will displace the arc a short distance away from said other face of the second element when the radiation through said second element exceeds a predetermined value. The actual displacement is preferably less than, and may typically be only 10% of the width of the molten weld pool which is being observed for the control of the weld speed and/or power input.

The heat from the arc reaches said other face of the second element via the weld pool and the solid metal between the pool and the face from which the radiation is observed.

From the view point of the control system, the most important part of the control chain is the one which possesses the slowest response and this will normally be the solid metal zone between the molten weld pool and the reserved surface of the second element. Control theory shows that, for a system to be stable, the magnitude of the gain must be less than unity at the frequency where the phase shift is 180 . This puts a limit on the gain in the feedback loop which therefore limits the accuracy of the control system. As is known however in control theory, higher gains can be utilised by introducing a frequency sensitive network to reduce the gain to a safe level near the critical frequency or alternatively to incorporate in the control loop, an element which saturates so that the system oscillates about the critical point.It has been found that using such techniques, a stable output may be obtained with a small arc movement; in one typical example the arc movement was restricted to plus or minus 3mm.

The invention finds particular application in a welding operation wherein the first element is an element of substantially uniform thickness between two planar surfaces and when the second element is also an element of substantially uniform thickness between two planar surfaces, the thickness of each element being small compared with the dimensions of its planar surfaces and wherein the weld is a butt weld between an edge face of the first element and one of the planar surfaces of the second element.

The second element may comprise a tube.

The first element in this case may also comprise a tube extending around the second element and having an inwardly extending end portion with its end surface butt welded to the external cylindrical surface of the first element.

The invention also includes within its scope, arc welding apparatus for making a weld between a first element and a second element by the above-described method, which apparatus comprises a first control means arranged to control the speed of relative movement between the arc and said elements and/or to control the power input to the arc, by a first feedback control loop, in response to radiation from the molten weld pool on said rear surface of the first element so as to tend to keep that radiation substantially constant, and a second control means arranged to effect movement of the position of the arc on said weld pool in a direction transverse to the line of relative movement of the arc with respect to the elements, and second control means including a second feedback control loop and means responsive to radiation from said opposite face of the second element and being arranged to prevent weld pool penetration of the second element.

The invention furthermore includes within its scope apparatus for arc welding of a first element to a second element, wherein the first element has an elongate edge surface is of substantially uniform thickness along that edge surface, which edge surface abuts and is to be welded to a planar surface of a second element, which planar surface is larger than the aforesaid edge surface, the second element furthermore having a uniform thickness over the region to be welded, this thickness being small compared with the dimensions of said planar surface, said apparatus comprising a welding head, means for effecting relative movement between the welding head and the elements so that an arc from the head directed into the angle between said planar surface and a surface of the first element is moved along the line of the weld producing a molten weld pool penetrating the thickness of the first element but not penetrating the second element and wherein there is provided a first control system for controlling the rate of relative movement between the welding head and the elements along the line of the weld, which first control system incorporates radiation detecting means detecting radiation from the rear face of the first element and means responsive to the magnitude of such radiation arranged to control the speed of such relative movement and/or the welding power to maintain the radiation from said rear face substantially constant and a second control system responsive to radiation from the opposite surface of the second element to that onto which the arc is directed, which second control system effects displacement of the arc within the weld pool region but transverse to the line of weld so as to prevent penetration of the weld pool through the second element. The second control loop conveniently controls the energisation of an electromagnetic deflection coil for deflecting the arc towards or away from said rear surface.

In the following description, reference will be made to the accompanying drawings in which: Figure 1 shows, partly cut-away, part of a thermal sleeve joint such as might be welded using the method and apparatus of the present invention; Figure 2 is a diagrammatic section through part of the joint for explanatory purposes; Figures 3 and 4 are diagrammatic sections through two differently shaped thermal sleeve joints; Figure 5 is a diagram for explaining the present invention; and Figure 6 is a prtly cut-away respective view of a joint together with part of the welding apparatus for effecting the weld.

Referring to Fig. 1, there is shown diagrammatically an inner tube 10 with an outer sleeve 11, the top end of which is turned inwardly towards the tube 10. This top end of sleeve 11 is butt welded to the outer surface of tube 10 to form an annular tee joint between the sleeve and the tube. This form of joint is used for example in a heat exchanger where the separation of the two fluids must be ensured. In such a joint, the inner tube 10 remains integral provided the weld can be effected without impairment of this tube. Conventional fusion welding would be too imprecise. The weld pool might penetrate right through the inner tube 10 with the possible impairment of the integrity of the tube of the joint.

Referring to Fig. 2, part of such a tee joint is shown in section on a larger scale. If the welding arc is directed from the top right hand corner towards the junction of the elements 10, 11, a weld bead 12 will be formed along this junction. It is required that the joint should be welded through the full thickness of the element 11 and therefore the weld pool must penetrate right through the element 11, as shown at 13, through to the rear face 15.

There will necessarily be penetration of the weld pool into and partial melting of the element 10 as shown at 14. It is required to ensure that this weld pool only partially penetrates into the element 10 and does not extend as far as the inner face 16.

Figs. 3 and 4 show slightly different geometries and illustrate that, although the geometry shown in Fig. 3 is simple to machine from the point of view of manufacturing the outer sleeve 11, the underbead region is inherently compressed with consequent increased penetration through the element 10 compared with the geometry shown in Fig. 4 where the outer sleeve extends at right angles towards the face of the element 10.

Referring to Fig. 2, the penetration of the weld through element 11 can be detected by sensing the radiation from the underbead region, that is to say by looking in the direction of the arrow 18. This radiation comes from the underface of element 11 over the region of the weld pool.

Fig. 5 illustrates diagrammatically one form of control apparatus for controlling the welding of a joint such as the tee joint shown in Fig. 1. The elements 10 and 11 are shown in section with a welding head 25 directing an arc downwardly onto the junction between the two elements. Relative movement is effected between the welding head 25 and the workpiece along the line of the weld by movement of the welding head or, more conveniently, by movement of the workpiece. In the case of the tubular joint in Fig. 1, the workpiece is rotated to effect the required movement. In Fig. 5, this rotation is shown diagrammatically as being effected by means of an electric motor 26.

The radiation from the underbead region of the weld pool is sensed by means of a silicon photodiode 27, the radiation being guided from the underbead region by optical fibres 28 and passed through a filter 29 having a pass band of radiation of wavelength of 0.5 i 0.05 microns. This filter 29 selects the required signal from the back of the weld whilst eliminating the unwanted longer wavelength radiation from the heat affected zones adjacent the weld region. The output from the filter 29 and detector 27, in such a construction, is nearly proportional to the area of the smooth surface of the molten weld pool and is utilised to control the size of the weld by regulating the heat input to the weld and the welding speed.For this purpose the output of the photodiode 27 is amplified in an amplifier 31 and compared with in a comparator 32 with a size reference signal from the input 33 to give an error signal which is fed to two amplifiers 34, 35. In this particular embodiment, this error signal controls both the speed of relative movement u and the arc current I according to the relationship u = a In + b where a, b and n are constants.

Reference may be made to the specification of Application No. 47575/77 for a further description'of a system controlling both welding speed and arc current. As shown in Fig.

5, the output from the amplifier 34 forms an error signal controlling the power supply source 36 feeding the welding head 25 and thus controls the arc current. The output from the amplifier 35 is applied as a control to the aforementioned motor 26 to control the welding speed. It will be noted that the control of welding speed and welding current is a feedback control and this control loop will be referred to hereinafter as the first control loop.

Radiation from the rear surface of element 10 is detected by a silicon photodiode 40 and, after amplification by amplifier 41 is compared in a comparator 42 with a position reference signal from an input 43 to provide an error signal which is amplified by amplifier 44 fed to an electromagnetic deflection coil 45. The coil 45 is arranged to produce a magnetic field extending along the direction of the weld and thus energisation of this coil effects a deflection of the welding arc in a direction orthogonal to the field and the arc.It is convenient to arrange the control system to give a two state output, for example in which the deflection coil 45 is energised only when the detected radiation exceeds a predetermined magnitude, the coil energisation being such then as to deflect the arc to the right in Fig. 5, that is to say away from the rear face of the element 10 so as thereby to reduce the penetration of the weld pool into the element 10. The amount of deflection of the arc need only be quite small and may correspond typically to about 10% of the width of the weld pool face viewed from the underbead side of the weld. Such deflection can therefore be effected without having any significant effect on the first control loop controlling the power to the welding head 25 and the welding speed.This position control is also a feedback control and this control loop will be referred to as the second control loop.

As previously explained, to enable a high gain to be employed in the second control loop, this control loop may include a frequency sensitive network so that the overall gain remains at a safe level and critical frequency where the phase shift is 1 80'. Alternatively an element which saturates may be incorporated in the feedback loop so that the system oscillates about the critical point. This latter system can readily be arranged to give the required accuracy of control whilst still having a sufficiently high response speed to deal with any sudden perturbations.

The control system may be used for forming an annular weld as is required in the regions of Figs. 1 to 4 or a straight line weld, which is, in effect, an annular joint unrolled. Fig. 6 illustrates in further detail part of the apparatus in an arrangement for producing a straight line weld. Referring to Fig. 6, there is shown a first element 60 (corresponding to the element 10 of Fig. 1) and a second element 61 (corresponding to the element 11 of Fig. 1) with a weld being effected along the line of the junction of these two elements. A TIG welding head is shown diagrammatically at 62. In this particular arrangement, the welding head is static and the workpiece is moved along the line of the weld.Sensing of the radiation from the underbead surface of element 61 is effected by means of a plurality of optical fibres 63 each having an associated silicon'diode photocell, the photocells being electrically connected in parallel. The plurality of sensors are used because the optical fibres are, for convenience, mounted on the workpiece and move therewith. As before an optical filter may be provided between the light guides and the photodetector. The output in the paralleled photodetectors is fed into a weld pool size control circuit which controls the welding current and the weld speed as described with reference to Fig. 5. The radiation output from the rear face of element 60 is detected by means of a position detector 65 in cluding an optical fibre 66 leading to a photodiode which senses the radiation. The output of this photodiode is utilised, as described with reference to Fig. 5, to control the energisation of a coil 67 on an electromagnetic core 68 for producing a magnetic field between pole pieces 69, this field being parallel to the direction of the weld and at right angles to the arc from the welding head. Normally the coil 67 is unenergised but when the radiation detailed by detector 65 exceeds a predetermined value, the coil is energised to deflect the arc a short distance across the weld pool transverse to the line of the weld and in a direction away from the rear face of element 60. It is found in practice that a small deflection of the arc in this way does not affect the radiation detected by the optical fibres 63 and hence does not affect the control of the weld speed and weld power.

Claims (15)

1. A method of making a weld between a first element and a second element where the first element abuts an elongate region of a surface of a second element, with the first element extending outwardly from that surface, and the second element having a thickness, normal to said surface, which is small compared with the dimensions of said surface; wherein the weld is effected by an arc directed onto said elements with relative movement between the arc and the elements along the length of the longer dimension of said surface region, the arc producing a weld pool penetrating the first element, transversely to said longer direction, from the arc to a rear surface of said first element, and with the weld pool only partially penetrating the second element from the surface facing the arc to an opposite surface thereof; and wherein (a) the speed of relative movement between the arc and the elements and/or the power input to the arc is controlled automatically by a first feedback control responsive to radiation from the molten weld pool on said rear surface of the first element, and (b) the position of the arc on said weld pool in a direction transverse to the line of relative movement is controlled automatically by a second feedback control responsive to radiation from said opposite face of the second element to prevent penetration of the second element.

2. A method as claimed in claim 1 wherein the first control is responsive to the magnitude of the radiation from the molten weld pool on the rear surface of the first element.

3. A method as claimed in claim 2 wherein the first control is responsive to the magnitude of the radiation at a preselected band of frequencies.

4. A method as claimed in any of claims 1 to 3 wherein the second control is responsive to the magnitude of the radiation from said opposite face.

5. A method as claimed in any of the preceding claims wherein the position of the arc is controlled by said second control by displacement of the arc by a distance less than the width of the weld pool.

6. A method as claimed in any of the preceding claims wherein the first element is an element of substantially uniform thickness between two planar surfaces and wherein the second element is also an element of substantially uniform thickness between two planar surfaces, the thickness of each element being small compared with the dimensions of its planar surfaces and wherein the weld is a butt weld between an edge face of the first element and one of the planar surfaces of the second element.

7. A method as claimed in claim 6 wherein the second element is a tube.

8. A method as claimed in claim 7 wherein the first element is a tube extending around the second element and having an inwardly extending end portion with its end surface butt-welded to the external cylindrical surfaces of the first element.

9. A method as claimed in any of claims 1 to 5 wherein the first and second elements are each plates of substantially uniform thickness, with the weld effected between an edge surface of the plate constituting the first element and one of the planar surfaces of the second element.

10. Arc welding apparatus for making a weld between a first element and a second element by the method of any of claims 1 to 9 which apparatus comprises a first control means arranged to control the speed of relative movement between the arc and said elements and/or to control the power input to the arc, by a first feedback control loop, in response to radiation from the molten weld pool on said rear surface of the first element so as to tend to keep that radiation substantially constant, and a second control means arranged to effect movement of the position of the arc on said weld pool in a direction transverse to the line of relative movement of the arc with respect to the elements, said second control means including a second feedback control loop and means responsive to radiation from said opposite face of the second element and being arranged to prevent weld pool penetration of the second element.

11. Apparatus for arc welding of a first element to a second element, wherein the first element has an elongate edge surface and is of substantially uniform thickness along that edge surface, which edge surface abuts and is to be welded to a planar surface of a second element, which planar surface is larger than the aforesaid edge surface, the second element furthermore having a uniform thickness over the region to be welded, this thickness being small compared with the dimensions of said planar surface, said apparatus comprising a welding head, means for effecting relative movement between the welding head and the elements so that an arc from the head directed into the angle between said planar surface and a surface of the first element is moved along the line of the weld producing a molten weld pool penetrating the thickness of the first element but not penetrating the second element and wherein there is provided a first control system for controlling the rate of relative movement between the welding head and the elements along the line of the weld, which first control system incorporates radiation detecting means detecting radiation from the rear face of the first element and means responsive to the magnitude of such radiation arranged to control the speed of such relative movement and/or the welding power to maintain the radiation from said rear face substantially constant and a second control system responsive to radiation from the opposite surface of the second element to that onto which the arc is directed, which second control system effects displacement of the arc within the weld pool region but transverse to the line of weld so as to prevent penetration of the weld pool through the second element.

12. Apparatus as claimed in either claim 10 or claim 11 wherein an electromagnetic deflection system is provided for effecting the controlled deflection of the arc transverse to the line of the weld.

13. a method of making a weld substantially as hereinbefore described with reference to the accompanying drawings.

14. A welded joint made by the method of any claims 1 to 9 or claim 13.

15. Apparatus for arc welding substantially as hereinbefore described with reference to the accompanying drawings.