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

Subscribe >
Skip to content

Advanced Welding Processes for Fusion Reactor Fabrication

By A Sanderson, C S Punshon and J D Russell

Paper presented at the 5th International Symposium of Fusion Nuclear Technologies, ISFNT-5, Rome, September 1999 and published in Fusion Engineering and Design, 49-50, 2000, pp. 77 - 87.


The huge scale and complexity of proposed fusion reactor structures has placed new demands upon fabrication technology and material performance. The need to weld substantial material thicknesses with severe service performance requirements, often in a multi-positional mode, over extended distances, certainly presents a formidable series of challenges. Of course, conventional welding processes can offer some solutions but welding times will be very long and component distortion potentially large.

Fortunately, these challenges can be addressed, at least in part, by recent advances in electron beam and laser beam welding technology. Also recent developments in friction welding processes could offer benefits for the manufacture of some components.

Specifically in the case of electron beam welding, TWI has developed a means of producing deep, narrow welds in a relatively poor vacuum. This new process variant called Reduced Pressure Electron Beam Welding is being studied for fusion reactor vessel fabrication. Process advantages will be described. The work undertaken by TWI on behalf of the UKAEA for the EU Task No. T301/1 programme will be covered by a poster presentation at this conference. Recent advances in high power Nd:YAG laser welding and Friction Stir welding developments, which have taken place at TWI, will also be described.

1. Introduction

The construction of large Tokamak reactors, especially the thick section vacuum vessel and toroidal field coil casings, presents many challenges. The components are extremely large and require both special means of manufacture and further development of joining methods. Traditional welding processes such as narrow gap tungsten arc and submerged arc welding are capable of welding the thicknesses involved but only in a multi-pass mode. Unfortunately, although the welding speed during each individual run is relatively fast, the need to apply numerous passes to fill the joint gap, slows the overall joining rate and can lead to excessive distortion particularly in austenitic stainless steels where the expansion coefficient is large.

Fortunately, welding technology has not stood still; advances in potential methods of electrical power generation have, to a modest degree, been paralleled by significant breakthroughs in the means of joining thick section materials. Notably over the past 40 years, electron beam welding has been taken from a very thin section welding process to the point where joints in even 300mm thick material can be attempted in a single pass. Similarly laser welding has advanced in leaps and bounds; numerous types of laser have been developed and although currently lacking the ultimate thick section welding capability of the electron beam welding process the penetration performance is advancing at a substantial pace.

Perhaps more surprising is the emergence of a fundamentally new method of joining thick section material, which 10 years ago had not even been thought possible. Friction stir welding, in which a spinning mechanical tool is used to plasticise the abutting faces of metal to produce a solid phase joint (i.e. with no melting) is a reality. Again the thicknesses and material which can be joined by this means advance yearly and although perhaps not directly applicable to reactor vessel or toroidal coil casing fabrications at the present time, within the current perceived fusion reactor programme, it might well advance to the point where some components could be fabricated by this means.

This paper will describe the welding process advances which have occurred. Emphasis will be placed on special electron beam techniques since this is where the majority of the work is currently taking place at TWI in support of the fusion programme, but advances in Nd:YAG laser welding will be described and finally the friction stir welding process will be explained with examples of its current capability.

Regarding the electron beam welding process, this paper will concentrate on the equipment and process development which have lead to the possible use for vessel and coil casing fabrication. For details of the recent and ongoing application of EBW on behalf of the UKAEA in support of the EU Task No. T301/1 programme, conference participants are directed to the poster paper presented by Lawrence Jones of the EFDA team.

2. Electron beam welding

The so called power beam welding sources, namely electron beam and laser beam welding uniquely differ from arc welding sources. The power density in the focus spot is typically in excess of 10 4 watts/mm 2 whereas electric arcs seldom achieve much better than 10 2 watts/mm 2 . As a consequence these beam sources are capable of not only surface melting but vaporisation of metal allowing the energy to be delivered in depth. Within a matter of seconds or less the beam establishes a vapour filled, liquid lined, keyhole. Translation of the keyhole along the plane abutting joint surfaces allows the materials to be fused creating a deep narrow, near parallel, fusion zone of minimal width. Even in 60mm stainless steel, for instance, with a well focused electron beam it is possible to achieve a parallel fusion zone of 4mm width or less. This characteristic 'keyhole' welding method minimises the heat input, the degree of shrinkage and hence distortion and for conventional components, allows a high joining rate by virtue of the single pass capability.

High power electron beams cannot readily be generated in anything other than a good vacuum (e.g. <5x10 -5 mbar) but are currently applied either in 'high vacuum' (~5x10 -4 mbar) 'partial vacuum' (~5x10 -2 mbar) or even 'non-vacuum' i.e. at atmospheric pressure. TWI has developed all manner of electron beam generation and delivery devices ranging from high vacuum to non-vacuum systems at beam power levels up to100kW. Instinctively, it might be thought that for thick section fusion reactor welding, particularly bearing in mind the sheer size of the components, a non-vacuum device would be the best option. However, when a beam of electrons is projected into a gas at atmospheric pressure, multiple electron scattering occurs and the beam intensity dissipated fairly rapidly. Some benefit can be achieved by displacing air by helium; this low atomic cross section gas reduces scattering but nevertheless the working range over which sufficient power density can be maintained is typically less than 30mm compared with >1000mm at 5x10 -3 mbar. In addition even at short working distance, gas and metal plasma disrupts the electron beam; consequently the welds tend to be quite wide and penetration depth currently limited to some 50mm in steel. Although research and development is still continuing on non-vacuum electron beam welding, in the near term the process is not suitable for many of the welding operations presently envisaged in the fusion reactor fabrication programme.

On the other hand, the prospect of totally encompassing huge reactor components in a vacuum canopy in order to effect the welding is not practicable. Moreover, even the use of locally sealed canopies at vacuum levels of 5x10 -3 mbar or better would be expected to be fraught with sealing difficulties. Fortunately, significant advances have been made in methods of generating and launching beams into high pressure gas in recent years, largely as a result of intensive research and development carried out by TWI for the Swedish Nuclear Fuel and Waste Management Company and for a leading offshore pipeline laying company, Saipem of Italy.

Both of these applications of EBW, which will be described in more detail later, require the welding of very large components, and of course in the case of pipeline girth welding the components are effectively of infinite length. Particularly in the latter case, welding speed is of the essence to meet pipe laying rate requirements and adequate seal integrity would be difficult to maintain if attempts were made at high vacuum or even partial vacuum levels.

The nuclear and offshore projects, therefore, gave impetus to the development of a new EBW process variant - Reduced Pressure EBW - where only a minimal reduction in chamber operating pressure is necessary to permit a highly focussed beam, at power levels of at least up to 100kW to be projected substantial distances. Typically, the process operates at 0.5mbar but has produced welds in even 150mm thick steel at over 1mbar background pressure which have the same deep narrow profile as high vacuum welds. Beam projection distance and protection of the material being welded are enhanced, as in the case of non-vacuum EBW, by the use of a helium gas shield, which in fact maintains a local welding environment in the vicinity of the welds of in excess of 1mbar. Under these conditions the process is not sensitive to small vacuum canopy seal leaks and the time to reach operating pressure, even for sizeable pumped volumes is a small fraction of that required to achieve a high vacuum level.

2.1 Reduced pressure electron beam welding equipment

In order to project an electron beam from the gun electrode region typically held at a pressure of 5x10 -6 mbar into the welding region at a pressure in excess of 5x10 -1 mbar, it is necessary to transmit the beam through a series of fine bore nozzles and to continuously pump the cavities between the nozzles. In essence this is relatively simple to achieve, but to give optimum separation of the various pumped stages, the nozzle bores need to be of the order of a few millimetres in diameter and have sufficient length to restrict the flow of gas. Conversely the electron beam must pass cleanly through the nozzles, since even a small power loss would cause substantial heating and damage. For a given beam power level, this can be readily achieved, but over the full current range of the equipment (say 0-500mA for a 200kV system) the beam shape and focal position need to be extremely stable. Up until recently this has only been achieved for a modest power rating (e.g. <25kW). At high powers the design of the electron gun becomes more critical.

The new technology which has permitted this to be achieved at beam powers up to 100kW is a Radio Frequency (RF) excited diode gun. This enabling technology consists of an RF transformer, the primary of which is a single turn aerial winding mounted in the interior of the gun column enclosure and the secondary, a single turn winding embedded in the gun cartridge, which is maintained at the accelerating voltage of the system. Both primary and secondary windings are held in resonance usually at a frequency of 84MHz. The power transferred to the gun cartridge is used to heat a tungsten ribbon filament, which is in series with the resonant circuit, to electron emission temperature. The high voltage produced across the inductive and capacitive element of the secondary circuit is then used to accelerate electrons (every half cycle of the RF) onto the back face of the main electron gun cathode. This cathode is mounted in a rigid refractory holder ensuring precise axial and lateral accuracy. Also, since the cathode button is made from a relatively thick disc of lanthanum hexaboride, the cathode is not subjected to thermal distortion and generally will last for many tens of hours of welding even at high power level.

Unlike conventional guns, the beam power is controlled by changing the cathode temperature rather than by adjusting the voltage on a grid electrode. This avoids large deviations in the axial position of focus which are characteristic of gridded guns, but more importantly avoids the risk of grid voltage breakdown which, in a conventional triode system, can cause huge excursions in beam power and focus and in the case of Reduced Pressure welding equipment, severe nozzle damage. Moreover this unique design of electron gun requires only one high voltage feed cable plus one RF connection.

For mobile welding heads, whether they be orbiting a static pipeline or moving relative to a large vessel, the use of a single core high voltage cable greatly reduces the diameter of the cable and hence the flexibility of the umbilical connection. The use of an RF excited diode gun also simplifies the electron beam power control since there is only one control parameter, viz cathode temperature.

2.1.1 Current applications

As mentioned above, the most prominent current applications of Reduced Pressure EBW are associated with two quite different industrial fields. Although making use of the special advantages of the process, they place different emphasis on certain aspects as described below:- Encapsulation of high level nuclear waste

Growing global concern over the accumulation of high level nuclear waste has encouraged most countries operating nuclear power plants to be actively involved in investigating the means of final safe disposal. In Sweden, a concept of burying the unprocessed spent fuel elements in large thick section steel honeycomb capsules encapsulated in a thick wall copper canister has been pioneered. The current design thickness of the canister is 50mm and it is anticipated that when these are buried some 500m deep in specially prepared clay lined granite repositories, leakage of radioactivity into the ground water will not occur for in excess of 100,000 years. To ensure precise sealing of the canister, special welding procedures using Reduced Pressure EBW have been developed. Apart form the process advantages previously mentioned, the electron beam optics and helium gas flow employed are used to control the beam energy density distribution in such a way as to create a near parallel sided weld with a well radiused tip. No other system has yet achieved this effect in copper; fusion zone shapes in copper are notoriously pointed and contain numerous cavities. The presence of such cavities, in view of the extremely long corrosion life of the canisters, is not permitted.

To date, following several decades of process and equipment development, numerous trial canisters have been fabricated and sealed. Fig.1 shows the set-up used for linear welding of 5m long canister half shells at a pressure of 5 x 10 -3 mbar and Fig.2 the lowering of a 5m canister into the Reduced Pressure welding facility at TWI for experimental lid sealing. A typical copper lid weld and a section of the fusion zone is shown in Fig.3; note the rounded weld tip which avoids root flaws. Welding power required for such a weld is usually some 75-85kW.


In April 1998 work was commenced on the installation of a 220kV, 100kW Reduced Pressure facility at Oskarshamn in Sweden. This unit is mounted on a large vacuum chamber into which the canisters are inserted from below. Fig.4 shows the electron gun assembly and the arrangement of the canister, lid and lid handling equipment. Several experimental lids have already been welded in the canister laboratory which has been constructed to verify the EBW process before construction of the full Encapsulation Station.

Fig.4. Offshore pipe welding


For oil and gas pipeline welding offshore, Reduced Pressure EBW offers four major advantages - very short cycle time, reduced criticality of vacuum seals compared with conventional EBW, high weld integrity and good weld reproducibility.

Although the beam power requirements in the case of offshore pipe welding project is only some 35kW, with steel pipe thickness usually approximately 40mm, welding speed and minimal cycle time are of the essence. The high operating pressure will enable the vacuum chamber to be evacuated in under 1 minute and welding speed is largely controlled by steel quality; the lower the impurity content, the higher the welding speed. The use of EBW allows the girth joints to be joined in a single pass at a speed of approximately 500mm/min with the pipe axis essentially vertical. This welding mode is more difficult for arc welding processes which are conventionally applied with the pipe lying horizontally. In this case, multi-pass welding can be undertaken simultaneously at several welding stations spread along the length of the laying barge. This maximises the joining rate. For deep water pipe laying this is much more difficult because of the laying stresses imposed on the pipe. Welding in the 'J' mode, Fig.5, where the pipe is lowered essentially vertically downwards is therefore preferred. However, it is then not feasible to stack the welding stations vertically since the pipe strake is many tens of metres in length. Hence Reduced Pressure EBW is much favoured.

A laboratory prototype equipment as depicted in Fig.6 has already been built and used to produce several hundred test welds. This equipment subsequently underwent a complete refurbishment in the summer of 1999 in order to test all the components of an advanced twin headed gun column unit destined for land trials in Italy prior to installation and operation on a barge.


2.1.2 Reduced pressure electron beam welding of fusion reactor components

The above examples, it is hoped, have served to illustrate the advanced nature of current Reduced Pressure Electron Beam Welding applications. In the case of fusion reactor fabrication, the problems are in many respects more challenging, the scale of the components much larger and of course the requirement to weld on site involves the need for multi-positional welding. However, during the work carried out for the EU Task GB8-T301 confidence has continued to grow in the feasibility of applying the process.

Two particular applications of RPEB welding relating to the fusion reactor construction are under investigation at TWI. The first relates to the vacuum vessel fabrication, the second the manufacture and assembly of the toroidal field coil casings. Vacuum vessel

The objectives of the preliminary study were to establish a facility and to begin to examine the feasibility of using the Reduced Pressure Electron Beam (RPEB) process for producing single sided butt welds in 60mm thick 316L type austenitic stainless steel in a range of welding positions from flat position to overhead. The aim was to identify the maximum weld penetration depth or thicknesses of material which could be welded satisfactorily in each welding position and to consider concepts for the practical application of the process for joining containment vessel sectors.

The current design of the large Tokamak Vacuum Vessel (VV) requires that a number of sectors are joined together on-site, Fig.7. This is to permit the vessel to be 'threaded' through the toroidal field coils during initial construction and because of the size and weight constraints. In addition, it is anticipated that during the service life of the vessel, remote maintenance operations will take place which in turn will necessitate removal of sectors of the vessel and their subsequent replacement by means of splice plates. The reference design is for a double skinned vessel with two wall thicknesses of 60mm separated by a gap of up to some 600mm. Because access is limited to the plasma internal space, to reach the thermal shield which insulates the super-conducting coils, 'splice plates' of 120-180mmwidth are required between the adjacent VV sector walls. It is proposed that the sectors are joined at the mid point between each toroidal field coil in a vertical plane using a welding head mounted on a versatile robotic manipulator known as the intersector weld/cut robot (IWR). It is further proposed that the IWR will be employed for both the original construction of the vessel and also remote maintenance operations.


As part of the ITER R&Dmp;D programme, full scale prototype vacuum vessel sectors were manufactured (by the Japan home team), and welded together using the reference welding process, narrow gap TIG welding (NGTIG). This exercise indicated that whilst high quality welds can be achieved with this process together with satisfactory dimensional accuracy, the joint completion rate is low and could substantially prolong the time scale for completion of the whole vessel. For this reason EFDA are considering alternative welding methods with the potential for increasing the joint completion rate. Those processes currently under consideration are Nd:YAG laser welding and Reduced Pressure electron beam welding. As it is impractical to manipulate the containment vessel, satisfactory performance in all welding positions, (in a vertical plane) from flat position to overhead is required. A preliminary feasibility study was carried out at TWI to examine the practicality of employing Reduced Pressure EB Welding for producing the Tokamak vessel intersector field joints.

The details of the IWR and vessel fabrication concept are described in more detail in the poster paper presented by Lawrence Jones of EFDA.

The experimental facility is shown in Fig.8, and comprises a small Reduced Pressure welding chamber operating at a pressure down to approximately 0.1mbar fitted with a 100kW RPEB electron gun. The entire chamber is designed to rotate and thereby permit welding to be carried out over a range of welding positions from flat position to overhead. The results of the preliminary welding tests carried out illustrated that for flat position welding, 60mm penetration could be achieved reliably, Fig.9. It was established that 40mm penetration could be achievable in the vertical and near vertical positions. Welding overhead was found, as expected, to be more problematical, although a weld penetration depth of 30mm was obtained without the weld metal dripping out.



In conclusion, it was considered that the results were sufficiently encouraging for the work to proceed to the second stage of the programme which is designed to optimise welding procedures for application to the vacuum vessel fabrication. Toroidal field coil case

The reference design of the ITER fusion reactor employed twenty toroidal field coils, Fig.7. Similarly, the latest proposed reactor design involves 'D' shaped toroidal field coils which will consist of super-conducting windings mounted in a supporting structure to minimise movement of the windings and a coil case to provide rigidity and a cryogenic operating environment. It has been proposed that the coil cases and radial plate supporting the windings are produced by fabrication from 316LN stainless steel. The size of the components and accuracy to which they are required to be manufactured impose great demands on traditional welding fabrication practice and thus the use of EB welding has been proposed as a potential method to be employed, together with narrow gap arc welding, for these heavy wall thickness components.

The ability of the EB welding process to produce single pass welds in heavy thicknesses is well documented; 300mm thick C-Mn steel and 450mm thick aluminium alloys have been welded successfully. However, the weldability of the material selected (316LN stainless steel) for the coil case and radial plate structures, as well as for many other components of the ITER reactor had not been previously investigated at the thicknesses required for these applications particularly in the vertical-up welding position which has been proposed for the closing welds. Therefore a study was carried out to examine the quality of welds that could be produced using the EB process and to determine the maximum thickness at which appropriate quality could be achieved reliably. The programme of work was carried out on behalf of Belleli Energy to meet these aims.

The work demonstrated that a reliable weld penetration depth of 50mm could be achieved with consistent quality and that the cryogenic properties of the welds were satisfactory. In consequence, this led to the development of RPEB welding procedures for producing a root pass of 50mm penetration which could provide a low distortion, high accuracy preliminary weld pass over which the fill passes could be made with high productivity welding methods e.g. narrow gap submerged arc welding.

3. Nd:YAG Laser welding

Nd:YAG lasers have been commercially available for over 30 years. Until recently, these lasers were only available with an average power of up to 2kW and were developed for either continuous wave operation or with some peak power enhancement (up to 5kW at 2kW average power). At this power level, the industrial applications were limited to thin sheet steel welding, particularly in the automotive industry. However, recently introduced higher power lasers operating up to 4kW are expected to be beneficial in raising productivity and are also attractive for welding aluminium alloys and thicker steel sections, particularly for complex geometries.

Encouraging results and improved penetration performance with Nd:YAG lasers with an output power of up to 5kW has now been achieved and there is interest in extending the power capabilities of these lasers further. Progress in this important Nd:YAG laser technology obviously increases the possibility of use for reactor component fabrication.

Conventional Nd:YAG laser-welding applications have been concerned mostly with high speed or precision welding of thin section materials. However, the introduction of Nd:YAG lasers with 4-5kW of power opens up many new welding application opportunities for thicker section applications in industries which up to now have not considered this process viable.


Such applications can be found in shipbuilding, off-highway vehicles, power generation and petro-chemical industries. For these industries, distortion due to arc welding is being increasingly recognised as a major cost in fabrication. This realisation led to at least three shipyards introducing high power CO 2 laser welding in attempt to substantially reduce distortion and improve overall fabrication accuracy.

High power Nd:YAG laser welding with the benefit of fibre optic beam delivery offers even more potential benefits for thick section welding application and a programme has been carried out at TWI to assess the potential for welding of structural steel. The main joint types of interest are butt welds and T-joints in various linear and circular forms depending on the application.

Due to the limitation at present of typically 4kW at the workpiece, the welding speeds for single pass welding of butt joints in 10mm thick steel are not very high (typically 0.3m/min) as can be seen Fig.10.

However, TWI has recently developed a system in which 3,4kW Nd:YAG lasers are combined together into a single fibre with a delivered power of up to 10kW. A multi-industry process development and applications project was launched in order to exploit this unique facility for welding, cutting and surface treatment with beam powers of 4-10kW.

The majority of the work carried out to-date has been concerned with applications and materials from the automotive and aerospace industries but there is also a significant involvement from the structural steel and nuclear industries where materials up to 20mm thick are being investigated. In these industries work on initial fabrication of structures and components is being carried out as well as techniques for reactor dismantling and repair.

Most of the work being carried out has not yet been released by the industrial companies involved in the project, but it can be reported that high quality welds can be made at least up to 15mm in a single pass and work is proceeding to develop the gas shielding and process technologies for thicker materials.

4. Friction stir welding

In late 1991 a very novel and potentially revolutionary welding method was conceived. The process was duly named Friction Stir Welding (FSW), and TWI filed for worldwide patent protection in December of that year. Consistent with the more conventional methods of friction welding which have been practised since the early 1950s, the weld is made in the solid phase; i.e. there is no melting.

In Friction Stir Welding (FSW), Fig.11, a cylindrical, shouldered tool with a profiled probe is rotated and slowly plunged into the joint line between two pieces of sheet or plate material, which are butted together. The parts have to be clamped onto a backing bar in a manner that prevents the abutting joint faces from being forced apart. Friction heat is generated between the wear resistant welding tool and the material of the workpieces. This heat causes the material to soften without reaching the melting point and allows traversing of the tool along the weld line. The plasticised material is transferred from the leading edge of the tool to the trailing edge of the tool profile and is forged by the intimate contact of the tool shoulder and the pin profile. This produces a solid phase bond between the two pieces.


4.1 Process advantages

The process advantages result from the fact that the FSW process (as all friction welding of metals) takes place in the solid phase below the melting point of the materials to be joined. The benefits therefore include the ability to join materials, which are difficult to fusion weld, for example 2000 and 7000 aluminium alloys. Other advantages are as follows:-

  • Low distortion, even in long welds
  • Excellent mechanical properties as proven by fatigue, tensile and bend tests
  • No fume, no weld porosity and no spatter
  • Can operate in all positions and is energy efficient
  • No consumables or shielding gas required
  • Makes use of existing machine tool technology

4.2 Materials and thicknesses


Friction stir welding can be used for joining many types of materials and material combinations, if tool materials and designs can be found which operate at the forging temperature of the workpieces. Up to the present day, TWI has concentrated most of its efforts to optimising the process for the joining of aluminium and its alloys. A major Group Sponsored Project undertaken for TWI's Industrial Members demonstrated that 2000, 5000, 6000 and 7000 and 8000 series alloys could be successfully welded to yield reproducible, high integrity welds within defined parametric tolerances. This work primarily investigated welding of wrought and extruded alloys. However, subsequent studies have shown that cast to cast, and cast to extruded alloy combinations, in similar and dissimilar aluminium alloys are equally possible.

Single pass butt joints with aluminium alloys have been made in thicknesses ranging from 1.2mm to 50mm without the need for a weld preparation. Parameters for butt welding of most aluminium alloys have been optimised in a thickness range from 1.6mm to 10mm. Special lap joining tools have also been developed for aluminium with thicknesses of 1.2mm to 6.4mm. Thicknesses of up to 100mm can be welded using two passes, one from each side, with 6082 aluminium alloy, see Fig.12.

Continuing development of the FSW tool, its design and materials have allowed preliminary welds to be successfully produced in copper and its alloys, lead, titanium and its alloys, magnesium to aluminium, zinc, plastics and mild steel. Work is continuing on the development of the process for higher melting point materials. Already it has proved possible to weld in a single pass 25mm of alloy steel using this process.

5. Concluding remarks

Fusion reactor vacuum vessel components are both large and heavy. This dictates that an on-site method of assembly of the toroidal field coils and vacuum vessel sectors is required. In any case, the planned need for sector removal and replacement dictates that the vacuum vessel fabrication method should be capable of on-site operation, and multi-positional welding.

Trial sector welding has already been carried out in Japan using multi-pass narrow gap tungsten inert gas welding but some doubts still remain about the productivity of this approach and the cumulative distortion of the vessel as sectors are added.

Power beam processes potentially offer distinct advantages of higher welding speed and much lower distortion and in recent years advances in beam power and penetration capability even for multi-positional welding place both Reduced Pressure EBW and Nd:YAG laser welding in a favourable position to meet the requirement of vessel fabrication. Of course, Nd:YAG laser welding has the great advantages of fibre optic delivery and atmospheric pressure operation, combined with a relatively lightweight welding head but presently power levels are limited for the material thicknesses involved. However, the rapid increase in power capability of these lasers, plus the feasibility of using beam combiners to deliver even higher powers down a single fibre, offers considerable promise for the future.

Reduced Pressure electron beam welding, on the other hand, can easily achieve the penetration required in all welding positions and requires only a relatively rough vacuum level compared with conventional electron beam welding. Moreover, for at least flat-position and vertical-up welding, the vessel wall reference thickness does not present any problems. However, for overhead and near-overhead welding positions, any deep welding process will obviously be influenced by gravity and may require the use of more than one pass to achieve full fusion. Nevertheless, if the majority of the vessel sectors can be welded in a single pass, this would still represent a substantial improvement in joint completion time combined with lower distortion, compared with multi-pass arc welding processes.

For the toroidal field coil casing, it is probable that much of the welding can be achieved off-site employing more ideal welding attitudes for power beam processes. Also, Reduced Pressure EBW in particular could penetrate in a single pass very thick sections, but the high nitrogen content of the 316LN stainless steel chosen for the coil casing components, tends to cause weld porosity problems somewhat limiting the single-pass weld depth at the present time.

Finally, Friction Stir welding does not at the present time appear to be directly applicable to either vessel welding or toroidal field coil casing fabrication but this is a rapidly developing process which may offer more immediate possibilities for some of the other fusion reactor components.

6. Acknowledgements

UKAEA, ITER, Saipem, SKB and Belleli Energy are acknowledged for their support in funding the work described. The authors also wish to acknowledge the contribution on Friction Stir Welding made by Wayne Thomas, David Nicholas and colleagues of the TWI Friction Welding Group.

Presented with permission from Elsevier

For more information please email: