Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Novel Joining Techniques for Repair in the Power Industry


Novel Joining Techniques for Repair in the Power Generation Industry

D. Howse , W. Lucas  and W. Thomas 

Paper presented at EPRI Welding and Repair Technology for Power Plants Conference, Point Clear, Alabama, USA, 26 - 28 June 2002


Within the power generation industry, outage reduction is a key issue in terms of improving plant availability and increasing revenue. More specifically, within the nuclear industry, there is a need to remove personnel wherever possible from radiation areas in order to reduce dose uptake. There is, therefore, a general requirement to develop novel high productivity joining techniques that can, in some instances, also be used remotely. The paper describes recent work carried out to evaluate the replacement of conventional gas tungsten arc, gas metal arc and shielded metal arc processes for repair in power plants by alternative joining processes such as advanced arc welding processes, Nd:YAG laser welding and friction techniques. Although most procedures for repair rely upon conventional arc welding techniques, TWI has also pioneered work in other areas of arc welding. Advanced arc welding technologies such as the use of active fluxes for enhanced penetration, underwater welding techniques, moving contact arc welding and in-process monitoring and control are discussed, as is, the Nd:YAG laser process, which is highly suitable for remote operation as light can be transmitted to the work area by fibre optic cable and focused at the point of application to produce deep penetration welds. The use of friction hydro pillar processing (FHPP), and friction stitch welding as a repair technique is reviewed. In view of the general importance of nuclear waste encapsulation, this paper also addresses some of the recent advances in electron beam welding and friction stir welding for this application.

Developments in arc welding technology


The processes most widely used for repair and refurbishment in nuclear environments are gas tungsten arc (GTA), gas metal arc (GMA) and shielded metal arc (SMA) welding. [1,2,3] Although both GTA and GMA processes are currently used for remote applications they both have limitations. The GTA process can be used remotely but, GTA welding does not give deep penetration welds, is relatively slow, and can need require a number of welding passes to make a single welded joint. The process can also be prone to variable quality due to cast-to-cast variation and requires a separate wire feed system. GMA welding is a higher productivity welding process but also requires local wire feed and its use has been limited to C-Mn steel applications. Recent developments in power source technology have improved the performance and quality of GMA welding to the point where they could be considered for high quality remote applications. In considering the use of novel processes for remote applications, particularly if they are to replace highly flexible manual welding techniques, the use of sensing technology and adaptive control must also be considered. Vision sensing, in-process control and offline programming techniques are all currently employed for robotic or automatic welding and there is a requirement to evaluate their suitability for remote repair.

GTA welding with active fluxes

Gas Tungsten Arc (GTA) welding is a widely used welding process by which an arc is struck between a non-consumable electrode and the workpiece creating the heat to make the joint. The main advantage of the process is to produce high quality welds, but it also has two major limitations. The first is that the deposition rates are lower than other consumable electrode arc welding processes, and that for stainless steels, the parent material composition can affect the depth of penetration achieved by altering the flow of the molten pool during welding. Active fluxes (A-TIG fluxes) that increase the penetration of GTA welds offer a means of significantly increasing the productivity of the welding process, and are capable of welding up to 6mm thickness carbon manganese or stainless steels in a single pass, without filler material ( Fig.1).

Fig.1. The characteristic appearances of the conventional GTA arcs and GTA with active fluxes and the comparative depths of penetration in 6mm thick stainless steel: Fig.1a) without flux;

Fig.1. The characteristic appearances of the conventional GTA arcs and GTA with active fluxes and the comparative depths of penetration in 6mm thick stainless steel: Fig.1a) without flux;

Fig.1b) with flux.


The active flux process can be applied in both manual and mechanised welding operations. However, because of the need to maintain a short arc length to achieve deep penetration, it is more often applied in mechanised applications. Specific advantages claimed for the active flux process, compared with the conventional GTA process, include: [4]

  • Increased depth of penetration e.g. up to 12mm thick stainless steel can be welded in a single pass compared with typically 3mm with conventional GTA.
  • Overcomes the problem of cast to cast variation e.g. deep penetration welds can be produced in low sulphur (less than 0.002%) content stainless steels which would normally form a wide and shallow weld bead with conventional GTA.
  • Reduces weld shrinkage and distortion e.g. the deep narrow weld in a square edge closed butt joint will produce less distortion than a multi-pass weld in the same thickness material but with a V-joint

The claims for a substantial increase in productivity are derived from the reduction in the welding time either through the reduction in the number, passes or the increase in welding speed. Economic data for the process can be seen in Tableable 1.

Table 1: Costing analysis of conventional mechanised GTA welding compared with GTA welding with active fluxes for 6.0mm thickness stainless steel.

ItemCost $/m weld (2002 figures)
Active flux
Argon shielding gas 1.23 0.48
Welding wire 3.08 -
Labour ($30.00 per hour) 38.46 7.70
Plant 0.48 0.09
Power 0.06 0.03
Active flux - 12.00
Total 43.31 20.30

Disadvantages of using a flux include the rougher surface appearance of the weld bead and the need to clean the weld after welding. In mechanised welding operations, the as-welded surface is significantly less smooth than is normally produced with the conventional GTA process but in manual welding operations, the surface roughness is similar. On welding, there is a light slag residue on the surface of the weld which often requires rigorous wire brushing to remove.

Fluxes are now commercially available from various suppliers for C-Mn steel, stainless steel, titanium and some nickel alloys.

Underwater wet welding

Within the nuclear industry there is also some occasional requirement for underwater repair, work has been carried out to develop wet welding techniques to replace SMA processes for wet welding and cutting using flux cored arc (FCA) welding processes. [5] The advantage of the FCA process compared to SMA it that it is much more suited to being automated and, therefore, has potential for remote application.

The E O Paton Institute has recently developed an innovative wet welding technique, based on the self-shielded flux-cored arc (FCA) process, which can also be used for cutting. The FCA wires have been developed specifically for operating in direct contact with water, and the novel wire feed system can be completely immersed. When used in either welding or cutting operations, the FCA process offers potential for significant productivity benefits through use of a continually fed wire, compared with SMA where the rod electrodes must be changed at frequent intervals. Furthermore, it is claimed that the combination of flux formulation and wire composition produces the desired slag-gas forming reactions which will not only improve the weld bead profile but also reduce the pick up of hydrogen and oxygen in the weld metal.

As the FCA process appears to offer substantial benefits for cutting and welding operations, a series of welding trials was carried out at TWI to evaluate the FCA process (consumables and equipment). This was to substantiate claims for wet underwater welding and cutting with regard to the benefits in weld bead characteristics and productivity. Results of the trials were collated over a period of six months using TWI welders and welder-divers from the UK. Several applications carried out in the former Soviet Union countries have been used to illustrate the benefits of the process for underwater welding.

Examples of the welds are shown in Fig.2. As the wire is essentially a rutile type, molten metal transfers from wire to weld pool by the short circuiting mode of metal transfer and is protected during cooling by a vapour gas bubble which surrounds the arc. This bubble is more stable in the flat position than vertical but vertical welding can also be carried out.

Fig.2. FCA wet weld in 8mm thick C-Mn steel plate welded in the vertical down position:

Fig.2a) General appearance of root pass;


Fig.2b) Cross section of weld.


The FCA process can also be readily used for cutting operations. A 2.4 mm diameter wire is normally used which generates a more forceful arc and gas 'jet'. The process appears to work equally well in the vertical-down and horizontal-vertical positions. However, when cutting in the vertical-up position, it was significantly more difficult to maintain the cut opening.

It was concluded that, based on TWI's evaluation tests at Cambridge and the practical experience in the former Soviet Union, there is no doubt that the FCA system offers a substantial advantage over conventional SMA for wet welding and cutting operations, especially in those situations where a large amount of welding/cutting must be carried out. Potential savings from use of FCA welding operations compared with SMA welding should be approximately 50%. The savings will be realised from reduction in the ancillary operations e.g. electrode changing, and the slightly higher deposition rates. Although the process was designed for manual welding, it also opens up the possibility of remote operation using an ROV. However, successful application of automatic techniques will depend upon the ability of the ROV to mimic the manipulative skill of a human welder. Irrespective of the type of operation (manual or automatic), reliability of the system will be crucial in order to ensure that the economic benefits derived from continuous operation, can indeed be realised.

Moving contact arc welding

Moving contact arc welding (MCAW) is another technique developed at TWI which could offer benefits for repair or surfacing applications and offers greater flexibility in application for remote processing than manual SMA welding. [6]

The process works as follows: the current supply is made with a sliding or rolling tool to the consumable via a narrow ridge which is part of the consumable core as shown in Fig.3. Initially an arc is struck, using a fuse or fine wire wool at the end of the electrode to ionise the consumable/substrate arc gap. The flux underneath the consumable core ensures electrical insulation between base material and the Ridgeback TM consumable and maintains a controlled arc length throughout the welding operation. The arc length can be changed by altering the thickness of the flux covering or by changing the shape of the metal core. The arc burns along the consumable electrode leaving a weld deposited onto the workpiece, as illustrated in Fig.4.

Fig.3. Basic principle of MCAW using a Ridgeback TM consumable


Fig.4. Example of surface appearance of MCA weld made with Ridgeback TM consumable, BS970 grade 316L deposit onto BS970 grade 304L substrate


The MCAW is simple to operate and eliminates the need to use skilled welders as required for conventional manual processes. For this reason, reproducible and defect free weld deposits should be achievable, whilst at least matching the metallurgical properties of other arc processes. At the same time, the technique will offer a more convenient weld overlay process. Large areas can be coated in one pass, simplifying the operation and reducing the number of interfaces in the coating. The consumable can be shaped to suit the substrate geometry, resulting in accurate and optimal placement of the coating material.

Currently, applications such as welding under water or in radioactive environments require sophisticated welding equipment. These are often expensive and time consuming to operate. MCAW is seen as a way of reducing down times and set-up times when welding in such restricted situations. The practicability of restricted access conditions has already been demonstrated using a local habitat for welding underwater under dry conditions.

In-process monitoring and control

Another area which has become increasingly important in applying any welding technology for remote applications is sensing and control technology. [7,8] The advances in this area of welding have largely come about through reduced cost of camera and image processing systems. These systems use data collected in real time to derive information regarding the state of the process and control the parameters to maintain weld quality for various environmental changes.

This is of particular interest where the process has to be robust enough to cope with variable joint fit-up and manual welding cannot be applied. Camera based systems can be used to either view the joint to be welded illuminated with a laser, 'laser stripe' systems. These types of systems are particularly suited to tracking joint lines and maintaining height position.

Other camera based systems use data derived directly from the image of the molten pool and use this information for seam tracking and molten pool measurement. Using these systems the relative position of the molten pool and joint line can be measured as can the gap in the seam. Using a closed loop system the torch can be manipulated back into position to provide real time adaptive control of the process. Similarly weld parameters such as welding current or wire feed rate can be adjusted to adjust weld penetration or the cap profile.

Other systems available which provide real time tracking and weld quality data rely upon the monitoring of the arc welding parameters themselves, such as the welding current and voltage. These systems use the readings not only to adjust for relative position, i.e. tracking, in a closed loop system, but also have the capability to measure the stability of the welding process to detect defects such as porosity and missed edges.

Developments in non-arc welding techniques

Nd:YAG laser welding

Lasers offer the advantage of providing a highly concentrated heat source that has the ability to vapourise metal and produce deep penetration keyhole welds in a manner that is not possible by any arc welding process. The advantages of the process are the ability to make high quality welds in close fitting edge preparations at either high speeds or in a single pass at the same time giving very low distortion. Within the last few years, high power Nd:YAG lasers up to 6kW laser power have become commercially available and these lasers unlike CO 2 lasers have the ability to be transmitted to the workpiece via long, optical fibres. This greatly enhances the flexibility of the process and makes the use of Nd:YAG lasers possible for deep penetration re-melt repair.

Work carried out at TWI demonstrated the feasibility of using such an approach to repair burst can detector (BCD) hangers with cracking resulting from oxide jacking ( Fig.5). [9] The laser used for these trials was a 4.0kW Nd:YAG system transmitting the beam to the workpiece via a 0.6mm optical fibre and focused to produce a 0.6m diameter spot size at the workpiece. Current welding technology is for the use of GMA welding to reinforce the fillet throat and leave the existing defect in place.

Trials were carried out using linear, single and double pass autogenous re-melting of the cracked region, use of linear, single and double pass re-melting with wire feed addition and use of a weave pass technique with wire feed autogenous laser welding to completely remelt the crack defect. A procedure was also developed to spall off the oxide layer prior to welding.

Fig.5. Cracked burst can detector hanger welds.
Arrow shows location of crack.


Results showed that the cracks could easily be repaired in a single operation by using a deep penetration re-melt procedure. It was also found that the oxide layer could be left on the surface prior to welding with no detrimental effect on the weld quality and the oxide layer actually gave better results due to increased absorption of the laser at the workpiece surface. The hardness of the resulting fused zone was measured and showed a fused zone no harder than that using the original GMA procedure. An example of a cross section of an acceptable weld is shown in Fig.6

Fig.6. Cross-section of a weave pass Nd:YAG laser weld on a sample with oxide. Wire feed rate 1.6m/min, laser power 3.9 kW, travel speed 0.2m/min, laser focus 5mm above workpiece surface


Using a 4kW laser with the focal position at or around the workpiece surface, a maximum penetration of around 10mm can be achieved at relatively low speeds. Alternatively, penetration of around 3.0mm can be achieved at welding travel speeds of around 3.0m/min. Using these high levels of productivity, the Nd:YAG laser process may also have application for high productivity sealing welds in low and intermediate level waste containment.

The Nd:YAG laser process has also been used in a defocused condition to scabble concrete. [10] Work carried out at TWI in partnership with BNFL has demonstrated that Nd:YAG lasers can be used to remove contaminated layers of the surface by traversing with a defocused laser beam to assist in the decommissioning of nuclear facilities ( Fig.7).

Fig.7. Use of defocused Nd:YAG laser beam for scabbling of concrete block. Block is 100mm high and 300mm long


Friction processes The recorded use of frictional heat for solid-phase joining techniques dates back over a hundred years. The friction welding process, however, to a large extent has been restricted to round, square, or rectangular bars. In addition to the applicability of these techniques to form attachment to structures, TWI has been working on techniques which now allow solid-phase friction welding to be applied to sheet and plate material as a viable option for plate fabrication in a range of materials. [11] Of particular interest are two techniques that have potential for the repair of defects, friction taper stitch and friction hydropillar processing.

Friction taper stitch welding Friction taper stitch welding is particularly suited to repair of cracks see Fig.8. [12] This is a solid phase welding process and involves drilling a tapered hole through the full thickness of a plate at the location of the defect. A tapered plug with a similar included angle is then friction welded into the hole. By using a series of interlinking holes long defects can be repaired. The process is portable and will run from power supplied by mobile generators. The hole plugging weld cycle time in 8mm thickness stainless steel is ~0.5 seconds. 

Fig.8. Friction taper stitch welding



Friction Hydropillar Processing Friction hydro pillar processing (FHPP) is a comparatively recent solid-phase welding technique. Invented at TWI, this technique is the focus of considerable R&D interest because of its potential in fabrication and manufacturing where it offers a number of novel production routes. The FHPP technique is still under development, but already shows promise for joining and repairing thick plate in ferrous and non-ferrous materials. Conventional fusion welding of thick section fabrications involves lengthy processing sequences and with some process large volumes of consumable material. In contrast, use of the FHPP welding technique should provide a reduction in joint preparation and weld filler metal, which will lead to significant cost savings.

The FHPP technique involves rotating a consumable rod co-axially in a circular hole, under an applied load to continuously generate a plasticised layer. The layer consists of an almost infinite series of adiabatic shear surfaces. The main features of the process are illustrated in Fig.9. During FHPP the consumable is fully plasticised at the frictional interface across the bore of the hole. This interface travels through the thickness of the workpiece. The plasticised material develops at a rate faster than the feed rate of the consumable rod. This means that the frictional rubbing surface rises along the consumable to form the dynamically recrystallised deposit material. The plasticised material at the rotational interface is maintained in a sufficiently viscous condition for hydrostatic forces to be transmitted, both axially and radially, to the bore of a parallel sided hole enabling a metallurgical bond to be achieved. Since this material is being forced hydrostatically into the surrounding bore, the diameter of the deposit material is nominally greater than the feedstock material.

Fig.9. Friction Hydropillar Processing


Waste encapsulation techniques for thick section joining

As previously mentioned, there are a number of techniques that can be used for high quality sealing welds for storage of low or intermediate level waste such as the GTA or laser processes. However, TWI has been involved with the development of two technologies for encapsulating high level waste and storage in deep level repositories. [13] The two processes under investigation are reduced pressure electron beam and friction stir welding.

The design of the canister consists of a cast nodular iron inner container with cavities to support the spent fuel assemblies. This is loaded into a copper canister that provides the corrosion barrier. The copper canister is approximately 50mm thick, 5m high and 1m diameter. Sealing of the lid is across a minimum joint plane thickness of 50mm, this thickness being the least necessary to maintain the integrity of the canister wall.

Reduced pressure electron beam welding development carried out over the last ten years on this project has produced deep penetration welds with a rounded root bottom profile to ensure that root porosity defects are maintained to an acceptable level not compromising the structural integrity of the joint ( Fig.10).

Fig.10. Typical electron beam weld section profile


Friction stir welding (FSW) is a continuous hot-shear process involving a non-consumable, rotating probe of harder material than the substrate itself. [14] The basic principle of the process is shown in Fig.11. Essentially, a portion of a specially shaped rotating tool is entered between the abutting faces of the workpiece (i.e. the joint). The tool's rotary motion generates frictional heat which creates a plasticised region (a local active zone) around the immersed portion of the tool, the contacting surface of the shouldered region on the tool and the workpiece top surface. The shouldered region provides additional friction treatment to the workpiece as well as preventing plasticised material from being expelled. The tool is then steadily moved along the joint line, with the plasticised zone cooling behind the tool to form a solid-phase joint as the tool moves forward.

Over the past 3 years a full size canister welding machine has been designed and built. For the past 12 months this equipment has been employed for development of FSW tool technology and optimisation of weld parameters. Welds in 50mm thick copper have been produced in 120° segments of approximately 1m length, see Fig.12. Much of the recent work has concentrated on the development of tool forms and tool materials. The results being achieved using FSW have been very encouraging and development of the technique will continue at SKB and TWI.

Fig.11. Principle of friction stir welding

Fig.12. Friction Stir Welded segment in 50mm thick copper

Concluding remarks

As the requirement to carry out repair, fabrication and encapsulation of waste within the nuclear industry grows, clearly there is a need for welding techniques to be applied that suit individual applications. A range of welding techniques have been highlighted that are currently available to the engineer to provide solutions to these applications in addition to the GTA, SMA and GMA processes currently employed. As every welding project is different, clearly there will be no single process that can be universally applied. However, there is a wide range of potential welding solutions which can offer economic, reliable and safe options for the industry.


  1. Dunn J.W. et al, 'Investigation and repair of a leak at a high temperature stainless [steel] butt weld', Int. Conf. Integrity of high temperature welds, Nottingham, 3-4 November 1998. [magazinearticle]
  2. Allen D.J., 'Developments in materials and welding technology for power generation', Welding and Metal Fabrication, 64, No. 2, February 1996. pp.70, 72, 74-75. [magazine article]
  3. Dyke D., 'Warrior (manipulator) welds where no manipulator has welded before', Nuclear Engineering International, 32, No. 393, April 1997, pp. 46. [magazine article]
  4. Lucas W., 'Activating flux- Improving the performance of the TIG process', Welding and Metal Fabrication, 2000, Vol. 68, No. 2, February, pp 7-10 [magazine article]
  5. Woloszyn A. C. et al, 'Underwater Wet-FCA Welding', ICAWT 2000, Gas Metal Arc Welding for the 21 st Century, AWS, Orlando, December, 2000 [conference paper]
  6. Verhaeghe G. et al, 'Moving contact arc welding' Exploiting advances in arc welding technology, Cambridge, UK, 30-31 March 1998. [conference paper]
  7. Smith J.S. and Lucas W., 'Vision Based System for Monitoring and Control of Welding', IIW Commission XII meeting, Ljubljana, July, 2001. [conference paper]
  8. Smith J.S. and Lucas W., 'Advances in Sensing and Control Systems for Arc Welding', Recent Developments and Future trends in Welding Technology, Cranfield University, September, 2001. [conferencepaper]
  9. Sekhar N. C. et al, 'Remote Crack Repair Welding in Steels Using High Power Nd: YAG Lasers', ICALEO2000, Dearborn Mi USA October 2-5, 2000, C-72 [conference paper]
  10. Johnston E. et al, 'Evaluation of high-powered lasers for a commercial laser concrete scabbling (large-scale ablation) system', ICALEO2000, Dearborn Mi USA October 2-5, 2000, pp.210-218. [conferencepaper]
  11. Thomas W.M. and Nicholas E.D., 'Emerging friction joining technology for stainless steel and aluminium applications' 'Productivity beyond 2000': IIW Asian Pacific Welding Congress, Auckland, New Zealand,February 1996. [conference paper]
  12. Hartley P., 'Friction plug weld repair of space shuttle external tank' Welding and Metal Fabrication, September 2000, pp 6 - 8 [magazine article]
  13. Ribton C.N. and Andrews R.E., 'Canister sealing for high level waste encapsulation', 2001 International High - Level Radioactive Waste Management Conference, Las Vegas, Nevada, USA, 29 April - 3 May 2001Publisher: American Nuclear Society. [conference paper]
  14. Thomas W, 'Friction technology for the aluminium industries', Seventh International Aluminium Extrusion Technology Seminar May 16-19, 2000, Chicago, Illinois. [conference paper]

For more information please email: