Hybrid laser/GMA welding of new land pipelines (August 2003)

⁽¹⁾ TWI Ltd., Granta Park, Great Abington, Cambridge CB1 6AL, UK
⁽²⁾ BP Exploration Operating Company Ltd, Upstream Technology Group, Bldg 200 - 2.52, Chertsey Road, Sunbury-on-Thames, TW16 7LN, UK

Paper presented at NOLAMP 2003 Conference, Trondheim, Norway, 4-6 August 2003

Abstract

Contemporary linepipe welding relies on automated arc welding stations carrying out root, fill and capping passes, with a number of welding stations comprising the weld spread. Reducing the number of these stations could potentially provide great economic benefits for the process as a whole. To this end, investigators have looked to incorporate methods that can provide deeper penetration, such as laser welding. This study has investigated the use of hybrid Nd:YAG laser/MAG welding to produce a deep penetration fill pass.

This paper summarises the trials that have been carried out as part of a project for BP Exploration Operating Company Ltd on the laser/arc hybrid welding of X60 and X80 C-Mn steel linepipe. Initially, a factorial style approach was adopted to determine the influence of process parameters on weld quality. ^[1] From this a welding procedure was devised which has been revised in these trials and implemented on 30 and 48inch linepipe in the 5G position. The quality and mechanical properties of these welds were also determined.

Keywords: hybrid, pipeline, welding, testing

1 Introduction

1.1 Background

Substantial work directed towards pipeline cost reduction has been undertaken in the last few years. A wide range of opportunities for cost savings have been targeted including pipeline conceptual studies and terrain evaluation, pipeline design, specification of materials, procurement of materials and services, construction and operation. ^[2] Pipeline welding is a considerable part of the latter and hence has a considerable influence on the overall cost of the pipeline.

A review project was carried out by TWI for BP, and contractor CRC-Evans, to identify potential options for reducing welding costs associated with land lay of pipelines. ^[3] This work identified improvements to existing arc welding technology, such as tandem GMAW and also highlighted the potential of high power Nd:YAG lasers in reducing costs of pipeline fabrication.

There are two factors that directly affect the cost efficiency of welding. Firstly, the speed of welding, since this largely determines the pace and timescale for construction, and secondly, the number of people employed on site as welders, fitters and associated logistics support staff since this element contributes significantly to the total cost of welding. The use of lasers has the potential to increase the effective welding speed and reduce the number of welding stations required. However these benefits have to be offset with the capital and running cost of the lasers.

1.2 Hybrid Laser/Arc Welding

Recently, work has been carried out to investigate the use of hybrid laser/arc processing for improving the performance of laser welding applications. ^[4] The advantages of this approach may be summarised as:

• Greater tolerance to joint fit up.
• Increased heat input thus giving increased weld travel speeds or penetration.
• Greater control of weld microstructure. Introduction of filler wire to the molten pool during laser/MAG hybrid processing, allows the microstructure of the weld to be modified and controlled. Potentially this provides positive benefits in terms of meeting weld metal toughness requirements and reducing solidification cracking susceptibility.

Earlier work at TWI has also shown that hybrid Nd:YAG laser/MAG welding is more suitable for pipe welding than autogenous laser welding. ^[5,6] This paper presents some recent results of work carried out at TWI highlighting the benefits of using Nd:YAG lasers, with power at the workpiece of up to 8.9kW, combined with the MAG arc welding process for girth welding of pipeline steels.

The objectives of this programme of work were to determine robust and comprehensive pipe girth welding parameters for a potentially commercial process using hybrid Nd:YAG laser-arc welding and established GMAW techniques. Incorporated within this there was a need to determine joint quality and mechanical properties and evaluate these in terms of relevant standards (e.g. API 1104 and BS 4515).

2 Experimental Programme

2.1 Materials

Table 1 lists the composition of the materials used. The parent materials were 15.9mm thickness, 762mm diameter, API 5L X60 grade, and 16mm thickness, 1219mm diameter, API 5L X80 grade linepipe steel.

Table 1: Parent steel and consumable wire grades and compositions

1. Standard designations: EN 440 - T 42 2 M M1, AWS A5.18-93: ER70L-6M
2. Standard designations: EN 440 - G 46 5 M G0, AWS A5.18-93: ER70S-6

2.2 Equipment

The laser apparatus used comprised of:

• Two 3kW Trumpf HL3006D Nd:YAG lasers.
• One 4kW Trumpf HL4006D Nd:YAG laser.
• A 3-in-1 Beam Combining Unit [BCU) manufactured by HIGHYAG Lasertechnologie.
• A 3mW diode laser used to align the processing head with the workpiece. This was fed through the output head in the same way as the high power Nd:YAG laser output.
• Output housing containing recollimating and focusing lenses.
• An optional co-axial gas shield attached below the output housing.

The output from the lasers was transmitted into the BCU using a step index fibre optic of core diameter 0.6mm and length 30m. A single fibre optic of core diameter 1mm and length 15m transmitted the combined laser power from the BCU to the laser output housing.

A Lincoln Electric Powerwave 455 MAG power source was used. A synergic pulsed welding programme appropriate for 1mm diameter wire and Ar-CO ₂ shielding gas mixtures was selected for the welding trials.

A purpose built fixture was used to hold both the laser focusing head and MAG torch ( Fig.1). The fixture permitted accurate and repeatable control of the relative position and orientation of the two heat sources. Figure 1 shows the laser output head and MAG torch attached to the hybrid fixture, with the laser co-axial gas shield and MAG shroud removed.

The hybrid fixture was mounted on a Kawasaki Heavy Industries JS-30 arc-welding robot. The robot controller was linked to both the laser and MAG equipment, allowing independent control of the start/stop for each process during the production of test welds. Welds were made by traversing the output housing across a stationary workpiece.

2.3 Previously determined procedures

Extensive preliminary procedure development was carried out prior to the trials reported here. ^[1,5] Initially a factorial style approach was adopted to determine the effect of certain parameters on weld quality. Following this a provisional procedure was devised that was then applied in position. Figure 2 shows a weld carried out using this procedure. The procedure development carried out in the trials reported here follows on directly from this preliminary work.

The procedure determined from the previous phases of this work was:

• Processing speed of 1m/min.
• GMA process leading.
• 35° to vertical GMA incidence angle to the workpiece
• Laser incident at right angles to the workpiece.
• 8.9kW laser power.
• 4.8kW GMA process power.
• GMA heat input of 0.23KJ/mm (assuming GMA efficiency of 0.8).
• Laser heat input of 0.42KJ/mm (assuming laser efficiency of 0.78).

The contractor CRC-Evans, as part of their contribution to the project, set a productivity target of 1m/min. To imitate the weld preparation that was envisaged would be used in the field, an internal root run was made using a procedure similar to that used by CRC-Evans.

2.4 Procedure development

The welding trials were carried out in three stages:

1. 45° sections of X60 linepipe were welded in the flat (1G), vertical up (3G) and overhead (4G) positions.
2. 5G vertical up welds in X60 linepipe.
3. 5G vertical up welds in X80 linepipe.

A smaller root face was chosen in this phase of trials to ensure penetration in position. Hence the procedure described in the previous section was altered in the first stage of the trials to:

• Processing speed of 1.7m/min dependent on position.
• 8.9kW Laser process power.
• 3.3kW GMA process power.
• GMA heat input of 0.093KJ/mm (assuming GMA efficiency of 0.8).
• Laser heat input of 0.245KJ/mm (assuming laser efficiency of 0.78).

Following the welding trials subsequent GMA passes were made on certain welds to allow full size Charpy specimens to be taken.

3 Results

3.1 Trials on X60 linepipe

Fully penetrating welds were made in the flat (1G), vertical up (3G) and overhead (4G) positions (Fig.3-5). Radiographic examination of these welds showed no solidification defects in the vertical up (3G) and overhead (4G) positions. In the flat (1G) position however, there were some surface breaking centreline defects.

Table 2 presents the Charpy impact energy values and the hardness testing results. The 5G vertical up girth welds made in the X60 steel were acceptable to BS4515-1:2000 Charpy energy requirements with the lowest value being 34J (5 x 10mm sub-size specimens notched at the weld metal centreline). The X60 5G vertical up welds satisfied the standard in the 6 and 12 o'clock positions regarding hardness. The weld failed to meet the BS4515-1:2000 requirement in the 9 o'clock position.

Table 2: Mechanical properties for welds carried out on X60 linepipe C-Mn steel

Notes
* Subsize specimens (5mm x 10mm)
** Full size specimens (10mm x 10mm)

Cross-weld tensile tests were also carried out on the X60 welds. BS4515-1:2000 acceptance was met with all specimens failing in the parent material. Porosity requirements were also met without exception with all values being below 0.3% compared to the maximum allowable level of 2%.

3.2 X80 5G Vertical Up Welds

Figure 6 is a macrograph of a cross-section of a weld carried out on X80 in the flat (1G) position. The root, hybrid filler and GMA filler passes can be clearly identified. Initially the welding speeds used for the X60 were used on the X80 linepipe but full penetration was not achieved from 7.30 to 10.30 so the speed was reduced.

Table 3 presents the Charpy energy and the hardness test results for the welds made in X80. The results exceed the requirements of BS4515-1:2000 with a minimum recorded Charpy energy of 90J, from full-size Charpy impact specimens notched on the fusion line. The heat affected zone hardness was within the requirements of BS4515-1:2000 with a maximum of 302HV10. The weld metal hardness however, failed to meet the requirements of BS4515-1:2000 with a maximum recorded value of 297HV10. Similar to the welds in X60, the standard was satisfied concerning porosity with all welds containing less than 0.3%. The weld metal hardness however, failed to meet the requirements of BS4515-1:2000 with a maximum recorded value of 297HV10. Similar to the welds in X60, the standard was satisfied concerning porosity with all welds containing less than 0.3%.

Table 3: Mechanical properties for welds carried out on X80 linepipe C-Mn steel (Charpy values are for full size 10 x 10mm specimens)

4 Discussion

For these trials, the weld preparation was chosen and machined on the pipe at an external contractor prior to practical welding work being carried out and there was no scope for optimising the preparation during the welding trials. Welds made at a travel speed of 1.0m/min were over-penetrating with poor visual appearance. When the weld speed was increased to produce a smaller, more controlled, weld bead at a laser power of 8.9kW, it was difficult to achieve consistent penetration especially in the 10 o'clock position. An increase in the root face thickness or a reduction in the laser power used would be required to achieve an effective processing window in all positions around large diameter pipe welds.

A compound preparation was derived from the recommendations of previous work. ^[7] The preparation was designed to alter the surface profile of the weld at a certain unit mass of filler wire deposited per unit length. Because the weld speed was increased, the amount of filler wire deposited per metre was reduced. This resulted in an underfill of the pre-machined preparation and a limited amount of surface breaking defects were seen the majority of which were in the 12 o'clock position.

Comparisons can be made between the X60 and X80 welds. Both materials had welds that were low in porosity (values were all below 0.3%). Both materials also produced welds with good toughness when tested at -10°C. Both the X60 and X80 steels produced welds with hardness above 275HV10. However, it should be noted that the welding speed was increased to cope with the small root face thickness and, that if the welds were made at the target speed of 1.0m/min, the heat input would be greater and softer microstructures would be produced in the weld metal.

Although fully penetrating girth welds were made on these large diameter pipes, the author does not consider that the procedure is fully optimised for the laser power used to make the welds. It would be recommended that a larger root face is adopted for the preparation which could be penetrated in a stable manner at a travel speed of 1.0m/min. Welded ligaments of 9.5mm were achieved in this phase of the work. Previous phases of work have shown that depths of up to 12mm can be achieved at this laser power ( Fig.2). Also, the internal GMA root weld adds thickness to the amount of material that needs to be penetrated to make the weld. As a fully penetrating weld procedure has now been adopted to avoid solidification cracking, the advantages of using an internal GMA root pass needs to be considered in relation to the disadvantages in increasing the thickness of the material to be penetrated to make the weld.

5 Conclusions

The following conclusions can be drawn:

• Deep penetration (~9.5mm) welds were made in large diameter pipes of both API 5L X60 and X80 grades of steel using a vertical up 5G procedure with the hybrid Nd:YAG/GMA process and 8.9kW of laser power delivered to the workpiece.
• The Charpy impact energy results produced at -10°C were acceptable to BS4515:2000. The minimum Charpy energy was 34J (sub-size 10 x 5mm specimens) which is equivalent to 68J for a full size specimen.
• The cross joint tensile performance of the welds made in API 5L X60 grade steel was also acceptable with all specimens failing in the parent material.
• Welds made in both the API 5L X60 and X80 grades had measured hardness in the weld metal greater than 275 but below 300HV10. Although these values exceed those permitted by BS4515:2000, it should be noted that satisfactory weld properties may still be achieved at this hardness level in hybrid laser/GMA welds. Heat affected zone hardness (HAZ) values were acceptable to BS4515:2000, all being below 350HV10.

6 Areas for further development

Following the trials reported here, further welds were produced using the original procedures developed within the previous phase of work ^[7] for 1m/min welding speed, ^[8] on the X60 grade steel ( Table 1). Instead of fixing the plates together using a root weld the plate was tacked together using GMAW. This resulted in a reduction in the required penetration and hence the root face chosen was increased accordingly. A macrograph of a weld produced using no root pass and with a greater throat is shown as Fig.7. From this macrograph it can be determined that the penetration through the thickness of the parent material was 9.5mm compared to <8mm when using a root pass. It is obvious when comparing the previous welds with that shown in Fig.7 that the weld has a better shape resulting from the improved balance between laser power and penetration.

The mechanical performance of the weld shown in Fig.7 was similar to those carried out with a root pass with the Charpy energy and hardness results satisfying and not satisfying BS4515-1:2000 respectively. The minimum Charpy energy was 179J (sub size 10 x 7.5mm specimens) equivalent to 239J at full size. The weld metal failed with an average hardness of 288HV5 and a maximum of 295HV5. The HAZ satisfied the standard at an average of 318HV5 and a maximum of 328HV5. These welds did not contain any solidification defects.

Further work would define the final procedure for deep penetration fill pass welding using hybrid laser/GMA welding at 1m/min. Start/stop procedures would also be determined in this phase.

7 References

1. Howse, D.S.; Scudamore, R.J.; Woloszyn, A.C.; Booth, G.S.; Howard, R.: Development of the Laser/MAG Hybrid Welding Process for Land Pipeline Construction, International Conference on the Application and Evaluation of High-Grade Linepipes in Hostile Environments, Japan, 7-8 November 2002.
2. Espiner, R.: BP's Pipeline Cost Reduction Project, Proc. Conf New Developments in Land Pipeline, Pipeline Industries Guild, December 2001.
3. Booth, G.S.; Hammond, J.: Low Cost Welding Techniques for Pipelines. Proc.Conf. Onshore Pipeline Cost Reduction, 'Pipes & Pipelines International, UK and Clarion Technical Conferences, US', Amsterdam, Paper 11, April 2000.
4. Dilthey, U.; Wieschemann, A.: Prospects by Combining and Coupling Laser Beam and Arc Welding Processes, IIW Doc XII-1565-99, 1999, pp. 29-43.
5. Booth, G.S. et al: Hybrid Nd:YAG Laser/Gas Metal Arc Welding for New Land Pipelines, Proc. Conf. 'International Conference on Pipeline Technology', Wollongong, March 2002.
6. YAGPIPE - Proposal for a Joint Industry Project on Laser Pipeline Welding, TWI Proposal PR4258/EWI proposal 43531CPQ-5, June 2000.
7. Scudamore, R.J. et al: Phase VI Hybrid Welding Results for September to November 2002. TWI Report No. 12940/40/03, February 2003.
8. Moore, P.M.: Investigation into the Microstructures and Properties of Laser and Laser/Arc Hybrid Welds in Pipeline Steels, University of Cambridge, June 2003.

8 Acknowledgements

The authors would like to acknowledge the contributions of Ashley Spencer and Frank Nolan for carrying out the bulk of the welding trials and Philippa Moore.