The invention relates to a method for assembly of thick-walled metal pipes and

5 fittings, wherein an outer end of a first fitting or pipe and an outer end of a second

fitting or pipe are in each case placed precisely against each other and wherein the outer ends are then welded together and, prior to being placed together, the outer end of the first fitting or pipe and the outer end of the second fitting or pipe are

provided with a weld preparation in the form of a chamfering plus an upright edge,

10 and the space between the chamferings is then fully welded in an automated

welding process. Thick-walled pipes have a wall thickness greater than 3

millimetres.

The method according to the preamble is known in the field and is described in

15 EP1043106. In the known automated welding process use is made of the MIG/MAG

welding method, since this method utilizes a large weld pool. In order to realize the perfect weld extending from the inner side to the outer side of the pipes or fittings, the abutting upright edges are in practice also welded through manually, wherein the TIG welding method is used. Applying this welding method produces a better- 20 quality weld than the MIG MAG welding method.

The drawbacks of manual welding-through of the abutting upright edges are

evident:

- the necessary use of highly qualified staff

25 - time loss caused by arranging the manual through-weld

- increased chance of the through-welds being rejected because of the manual

welding-through

- higher costs due to the use of different welding processes.

Owing to these drawbacks the welding-through is often omitted, and an imperfect

30 weld has to suffice.

The invention relates to a method for realizing a perfect weld which obviates the

above stated drawbacks. Particularly in the laying of long pipelines for the transport of oil or gas it is important that the produced weld be perfect, since repair of an

35 imperfect weld is time-consuming and expensive. The fact that highly qualified staff

are no longer necessary means that laying a pipeline can take place in fully continuous manner, wherein less highly qualified staff need only carry out substantially logistical tasks.

The method according to the invention has the feature that before the space between the chamferings is fully welded the abutting upright edges are welded through in the same automated welding process. The creation of a perfect weld is wholly automated by applying this method, and the use of highly qualified staff is no longer necessary. Because the upright edges are welded with the same welding process, it is no longer necessary to apply two different welding methods. It is recommended to use the TIG method for the whole welding process, since the TIG method produces a better-quality weld than the IG/MAG method. The weld preparation can be realized here in a separate automated process prior to the actual welding process. A favourable realization of the inventive method has the feature that an arc voltage of a welding arc of the welding process is measured during the welding process, wherein an amplitude and a frequency spectrum of the arc voltage are determined and can be utilized to control the welding process. A further favourable realization of the inventive method has the feature that during the welding-through the welding current is regulated on the basis of the frequency spectrum of the arc voltage, such that a weld pool caused by the welding current fills the whole seam between the upright edges. The thickness of the abutting upright edges preferably amounts to a maximum of substantially three millimetres and a minimum of substantially one millimetre, and in the welding process the TIG welding method is used for both the full welding and the welding-through. Owing to this maximum thickness the welding current can be regulated during the welding-through such that a weld pool caused by the welding current fills the seam between the upright edges on the basis of the frequency spectrum of the arc voltage until the inner side of the pipe is reached. A further favourable realization, with which a reference point can automatically be obtained on the weld, has the feature that a point where the seam is at least substantially filled is stored and that welding-through then takes place until this point is passed again, after which the space between the chamferings is fully welded. A further favourable realization has the feature that during the full welding a reciprocating movement of the welding arc is controlled on the basis of the amplitude of the arc voltage.

A further favourable realization with which a perfectly finished weld can be realized has the feature that, if it is determined that an amplitude of the reciprocating movement is greater than a predetermined amplitude, the continuous welding process is continued for one revolution, wherein a closing layer is arranged which is slightly wider than the width of the combined weld preparations. The closing layer is here arranged symmetrically around the actual weld on the basis of the known position of the seam between the upright edges.

It is known from WO 95/34400 that a welding current of the welding arc can advantageously be controlled on the basis of the spectrum of the arc voltage. This method is however only suitable for thin-walled metal pipes wherein the wall thickness is smaller than 3 millimetres.

It is known from JP 60072674 that a reciprocating movement of a welding arc can be controlled on the basis of the amplitude of the arc voltage.

The invention also relates to a device for assembly of thick-walled metal pipes and/or fittings, comprising a fixing member for placing and fixing two outer ends of pipes and/or fittings precisely against each other, which pipes and/or fittings are provided with a weld preparation in the form of a chamfering plus an upright edge, and an automatic welding machine in which a fitting or pipe can rotate or which can rotate around an outer end of a fitting or pipe, wherein the automatic welding machine is adapted to fully weld a space between the chamferings in an automated welding process. The inventive device has the feature that the automatic welding machine is also adapted to weld through the abutting upright edges in the same automated and continuous welding process.

A favourable embodiment of the device according to the invention has the feature that the automatic welding machine is provided with a sensor for measuring an arc voltage of a welding arc of the welding process during the welding process, in addition to a processor for determining an amplitude and a frequency spectrum of the arc voltage. The welding process can then be controlled on the basis of the measured amplitude and the measured frequency spectrum.

A further favourable embodiment of the device according to the invention has the feature that the automatic welding machine is provided with a regulating member adapted to regulate a welding current on the basis of the frequency spectrum of the arc voltage during the welding-through and to control an actuator, which realizes a reciprocating movement of the welding arc, on the basis of the amplitude of the arc voltage during the full welding.

In a very favourable embodiment the thickness of the abutting upright edges amounts to a maximum of substantially three millimetres and a minimum of substantially one millimetre, and the welding process is adapted to perform the TIG welding method for both the full welding and the welding-through.

A further favourable embodiment of the device according to the invention has the feature that the processor and the regulating member comprise a computer. The computer is preferably also adapted to determine a length of the welding arc and to regulate a supply of welding wire and protective gas to the welding arc so that the computer can control the overall welding process.

The invention will now be further elucidated with reference to the following figures, wherein: Fig. 1 shows schematically a pipe and a fitting to be welded thereto, and a

possible embodiment of an automatic welding machine according to the invention;

Fig. 2 shows the welding head in more detail;

Fig. 3 shows schematically a possible control of the automatic welding machine.

Fig. 1 shows a cross-section of pipe 1 and a fitting 2 to be welded thereto. The outer ends of pipe 1 and fitting 2 are provided with a weld preparation in the form of a chamfering 3 plus an upright edge 4, and pipe 1 and fitting 2 are supported by supports 5a,5b such that the outer ends for welding are placed in precise and stable manner against each other. Placed on pipe 1 is an automatic welding machine 6, a carriage 7 of which can rotate around a divisible guide 8 clamped onto pipe 1 using a drive 9. Mounted on carriage 7 is a rail 10 over which a welding head 11 can move using a drive 12. Carriage 7 and welding head 11 are controlled using a control 13, and welding head 11 receives its welding current and protective gas from a supply 14. When a weld is made, welding head 11 is positioned above the weld, after which carriage 7 is rotated slowly around pipe 1. A welding electrode 15 is then lowered and a welding current regulated such that a weld pool 16 is created which fills substantially the whole seam between upright edges 4. As soon as this is the case, the position of this point on the weld is stored, after which the whole seam is welded through until this stored point is passed once again. The space between chamferings 3 is then fully welded, wherein welding head 11 makes reciprocating movements using drive 12. The welding process is stopped when it is determined that the amplitude of the reciprocating movement is greater than the width of the weld preparation.

Fig. 2 shows welding head 11 in more detail, with rail 10 and drive 12. Welding head 11 comprises a welding electrode 15 which is mounted in a drive 17 with which it can be driven inward or outward during use in order to thus adjust the distance between welding electrode 15 and weld pool 16. Placed around welding electrode 17 is a gas cup 18, via which a protective gas is supplied from supply 14 during the welding. Welding electrode 15 is connected via a power cable 19 to a welding transformer or a pulsed power source forming part of supply 14. Welding head 11 further comprises a supply of welding wire 20, provided with a drive 21 which provides for the supply of welding wire to weld pool 16. During arranging of a weld pipe 1 or automatic welding machine 6 is rotated at a predetermined speed. Welding electrode 15 is then lowered until a welding current is measured in power cable 19. Welding electrode 15 is then pulled inward until a formed welding arc has a predetermined length, wherein the welding current is regulated in a per se known manner on the basis of the frequency spectrum of a measured arc voltage such that upright edges 4 are welded wholly through. As soon as a complete through-weld is detected somewhere, the position of this point on the weld is stored, after which the whole seam is welded through until this stored point is passed once again.

Electrode 15 then begins to move reciprocally between chamferings 3 using drive 12, wherein welding wire 20 is added at a predetermined speed. The space between chamferings 3 is in this way wholly welded through. When the space between chamferings 3 has been completely filled, this being apparent from the amplitude of the reciprocating movement, a closing layer is then also welded which has a slightly greater width than the combined weld preparations. The welding process then stops automatically. Fig. 3 shows in the form of a block diagram a possible embodiment of a control 13 which is disposed close to automatic welding machine 6. During the welding the electrode 15 is moved close to the seam between upright edges 4 such that an arc is formed which creates a weld pool 16 of molten metal. The welding current must be chosen such that weld pool 16 fills substantially the whole seam without molten metal being able to leak out. The arc voltage is measured for this purpose using a sensor 22 and fed to an analog/digital converter 23. The thus obtained signal is fed to a computer 24 which separates the signal into a direct current component and an alternating current component superimposed thereon. The condition of the weld pool is derived from the alternating current component and the welding current in power cable 19 is regulated on the basis thereof during the welding-through.

Determined from the direct current component during the full welding is whether welding electrode 15 is approaching a chamfered edge 3. If this is the case, the direction of movement of drive 12 is then reversed. The length of the welding arc is also determined from the direct current component and then adjusted if necessary using drive 17.