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

Subscribe >
Skip to content

Friction stir welding developments (April 2002)

W M Thomas and R E Dolby


Paper presented at 6 th International Conference on Trends in Welding Research, 15 - 19 April 2002, Callaway Gardens Resort, Pine Mountain, Georgia, USA


Friction stir welding technology is described together with recent developments in tool technology for lap welding. The application of the process to tube welding is outlined.


It has long been recognised that frictional heating can be used to join, process and treat materials.

Fig.1. Friction processing technologies

A number of friction based technologies have been developed that enable the friction processes to be applied to a range of component shapes and products [1-7] (see Fig.1). For example non-rotary, orbital, linear and angular reciprocating motions have enabled the friction process to be applied to a variety of non-round components. [8,9] Practically all of these friction welding techniques are regarded as 'one shot' joining processes, whereby the joining process essentially takes place at more or less the same time across the entire joint. Thus, as the cross-sectional area of the joint becomes larger, so the forces involved in the friction welding operation increase proportionally and eventually, in practical terms, become prohibitive.

Friction stir welding (FSW) is a solid-phase continuous hot shear welding process with certain similarities to laser, electron beam and plasma arc welding in that FSW is a moving point welding source with keyhole type features. [10-15] The design and materials required for the welding tool are critical to the successful material flow and consolidation around the keyhole, and consequently weld integrity. The basic principle of the FSW process is shown in Fig.2.


Fig.2. Schematic showing the basic principle of the FSW process

Recent tool developments for friction stir lap welding

Early in the development of FSW it was realised that the form of the welding tool was critical in achieving sound welds with good mechanical properties. In general terms, the tools shown in Fig.3a and b comprise a shoulder and a probe.


Fig.3 Triflute TM family of probe variants for FSW
a) MX Triflute TM for butt welding
b) Flared-Triflute TM with tip profile for lap welding

The shoulder compresses the surface of the workpiece and contains the plasticised weld region. Heat is generated on the surface by friction between the rotating shoulder and the workpiece surface and, when welding thin sheets, this is the main source of heat.

As the workpiece thickness increases, more heat must be supplied by friction between the rotating probe and the workpiece.

Also, the main function of the probe is to ensure sufficient working of the material at the weld line and to control the flow of the material around the probe and underneath the shoulder to form a satisfactory weld. The probe generally has a flat or re-entrant fluted shape profile.

Preferably, the probe has an odd number of equally spaced flats or flutes to maintain maximum cross-section opposite to the ridge lands. [16,17] It should also be noted that all changes in section are well radiused in order to reduce stress concentration and thereby minimize the tendency of tool fracture during welding. In essence, the core of the probe is tapered to maintain an approximately uniform stress distribution arising from torsion and the forward thrust.

For butt welding, the probe is frustum shaped (see Fig.3a). Typically, the probe features for butt welding incorporate three helical flats or flutes and a coarse helical ridge (usually a coarse thread) around the lands of the probe to facilitate a downward augering effect. These flat or re-entrant features reduce the probe volume and provide a suitable swept volume to static volume ratio. The greater the volume ratio the greater the path for material flow and the more efficient is the probe. In addition, these re-entrant features, especially the helical coarse ridges around the lands, help break up and disperse the surface oxides within the joint region.

For lap welding, a probe has been developed to provide a wider region and also to help avoid problems associated with upper plate thinning at the lap weld interface. The weld interface of a butt weld is essentially perpendicular to the workpiece plate surface and penetration through the thickness to the bottom of the plate is usually the main requisite. For lap welds, the width of the weld interface and the angle at which the notch meets the edge of the weld is of fundamental importance, especially for those applications that are subjected to fatigue. [14,18,19,20,21] Figure 3b shows a Triflute TM flared probe, with straight flutes designed for lap welds. The core remains as a taper frustum but the flute lands are flared out so as to increase the tip diameter. Moreover, a tip profile has been included in the shape of a three pronged whisk. [7,12] (see Fig.4).


Fig.4. Flared-Triflute TM probe with lap welding tip profile positioned across the weld interface

These features collectively increase the difference between the swept volume and the static volume of the probe, thereby further improving the flow path around and underneath the probe. In addition this tip profile for lap welding provides an improved mixing action for oxide fragmentation and dispersal at the weld interface.

A lap joint made with a Flared-Triflute TM probe is shown in Fig.5. In this example the width of the weld region is 190% of that of the plate thickness and little upper plate thinning is apparent at the weld interface.


Fig.5. Macrosection showing a wide weld region produced using a Flared-Triflute TM probe with lap welding tip profile. Lap weld produced in 6mm thick 5083-0 condition aluminium alloy, at a weld travel speed of 4mm/sec (240mm/min)

The outer regions of the weld at the overlapping plate/weld interface show a slight upturn (see Fig.6a and b). This upturn is much less than that caused by a conventional pin type probe. A greater presence of oxide interface remnant is revealed at the retreating side (see Fig.6a) compared with the advancing side (see Fig.6b).






a) Retreating Side










b) Advancing Side






Fig.6. Detail at the extremes of the weld region. For Flared-Triflute TM type weld shown in Fig.5

Skew-stir TM & associated tool developments

The skew-stir TM variant of FSW differs from the conventional method in that the axis of the tool is given a slight inclination (skew) to that of the machine spindle, [22] as shown in Fig.7a, b and c.

The skew-stir TM technique enables the ratio between the 'dynamic' (swept) volume and the static volume to be increased by the skew motion of the tool. This can be additional to that provided by the use of re-entrant features machined into the probe. It is this ratio that is a significant factor in enabling a reduction or elimination of void formation and improving process efficiency.

The arrangement shown in Fig.7a results in the shoulder face being oblique to the axis of the skew tool and square to the axis of the machine spindle. This shoulder face remains in a fixed relationship with respect to the plate top surface. Tilting the plate or the machine spindle will produce a plate to tool tilt that can be varied to suit conditions.

The focal point of a skewed tool affects the amplitude of the orbit of the tool shoulder and probe. With the focal point at the shoulder position, i.e. at the top of the workpiece, the shoulder essentially has a rotary motion with no off-axis orbit. When the focal point is positioned slightly above the top surface of the work piece, or at any position through the thickness of the workpiece, the shoulder contact face has a off-axis orbital movement. In addition, the off-axis orbital motion of the shoulder is dependent on the angle of skew and the distance that the intersection (focal point) is away from the top of the plate. The greater the skew angle and the greater the distance that the focal point is away from the workpiece surface, the greater is the amplitude of the shoulder off-axis movement.

The skew action results in only the outer surface of the probe making contact with the extremities of the weld region. The FSW tool does not rotate on its own axis, and therefore only a specific part of the face of the probe surface is directly involved in working the substrate component material. Consequently, the inner part of the tool can be cut away to improve the flow path of material during welding, (see Fig.7a). This probe type is termed A-Skew TM.


a) Side view






b) Front view, showing tip profile





c) Swept region encompassed by skew action

Fig.7. Details of prototype A-Skew TM probe


The skew-stir TM technique provides an easier flow path than conventional FSW and a weld nugget region of width greater than the diameter of the probe. In addition the skew action provides an orbital forging action at the root of the weld, which improves weld quality in this region.

A weld made with the skew-stir TM technique is shown in Fig.8.


Fig.8. Macrosection showing a wide weld region produced using the Skew-stir TM technique with the A-skew TM probe, lap weld produced in 6 mm thick, 5083-0 condition aluminium alloy, at a weld travel speed of 4mm/sec (240mm/min). (Other process conditions were the same as for the weld shown in Fig.5) Retreating side (left) Advancing side (right)

In this example, the width of the weld region is 195% of the plate thickness and no upper plate thinning at the weld interface is apparent. Unlike the welds produced with the Flared-Triflute TM probe, ( Fig.6a and b) the welds produced with an A-Skew TM probe revealed a nominal downturn at the outer regions of the overlapping plate/weld interface ( Fig.9a and b). As with the Flared-Triflute TM welds shown in Fig.6a and b, these A-Skew welds revealed a greater presence of oxide interface remnant on the retreating side, Fig.9a, compared with the advancing side, Fig.9b, as would be expected with a conventional pin type probe.






a) Retreating side










b) Advancing side






Fig.9. Detail at the extremes of the weld region for the A-skew TM weld shown in Fig.9

Mechanical testing of lap welds was undertaken using a 'hammer 'S' bend test'. Bend testing was carried out with the weld region unrestrained. This lap 'hammer 'S' bend test' proved a discerning method for establishing basic weld integrity and freedom from weakness caused by plate thinning. Figure 10a and b show typical results achieved from welds produced with Flared-Triflute TM and A-Skew TM probes.





a) Lap weld produced using a Flared-Triflute TM probe











b) Lap weld produced using a A-skew TM probe






Fig.10. Hammer 'S' bend tested lap welds

Repeatability trials with both probes achieved consistent and good results and when compared with a conventional threaded pin type probe the following benefits were established.

  • Over 100% improvement in travel speed.
  • About 20% reduction in axial force.
  • Upper plate thinning reduced by a factor of >4.

Both Flared-Triflute TM and A-skew TM probes and their methods of use are well suited for lap and 'T' and similar welds where the interface is 90° to the machine axis, i.e. parallel with the work piece surface. Welds made with these tools, using the same process conditions, showed improvement over conventional pin type probes. Moreover, both types of tool and technique used provide an effective method of increasing the width of the weld region; which is particularly advantageous for material processing.

Tubular applications of FSW

For most applications, the FSW technique is used as a one sided welding process. The solid-phase characteristic of FSW provides all-positional welding capability. [10] Figures 11a-e show a number of FSW applications for positional welding, which give difficulty when fusion welding.

When welding flat plate it is immaterial whether the FSW process is applied to the joint internally or externally (see Fig.11a and b).






a) Welding of hollow component









b) Welding of hollow component









c) Welding stationary horizontal pipe








d) Welding stationary 45° angled pipe







e) Welding a corrugated component




Fig.11. Examples of positional Friction Stir Welding

(The G designations refer to the American Welding Society convention for positional welding [23] )

Furthermore, FSW can be achieved by traversing the tool or the workpiece. However, when welding comparatively small radii tubular components the design of the shoulder depends upon whether the tool is applied to the outside or inside of the tube.

Moreover, when welding tubular components, either the tool can be traversed around the tube (in which case all-positional welding is employed) or the tool is stationary and the tube is rotated (in which case welding is generally performed in one position).

With regard to the future, investigations at TWI are continuing to explore a number of developments. Applications include cylindrical tanks and similar tubular components. [10,24,25] Figure 12 illustrates the essential design concept being considered to enable a single or multi-headed friction stir welding facility to be used to internally butt weld line pipe and carry out remote down hole-repairs.










a) Side view














b) End view






Fig.12. FSW design concept facility for butt welding of line pipe. Comprising a hinged flexible drive head together with a longitudinal positioned 'run off' wedge plate, designed to eliminate any end of run hole filing requirements.

A minimum of three centralising roller supports is shown, although the actual number would depend on the flexibility of the tubular component being welded. Additional roller supports would act as an expansion device to force the tube against the external anvil. The necessary pressure of the tool on the inside tubular workpiece would be kept constant during the welding operation. At the end of the weld the FSW tool will then be swung in a longitudinal direction on to a removable curved wedge plate to allow the probe to clear from the workpiece. Such a technique will eliminate problems associated with the exit hole. Apart from butt welding, the approach could also be applied to the manufacture of helical welded pipe and to the repair of defects encountered in pipelines.

Welding mechanisms

The significant and increasing interest shown in FSW by aerospace, marine, rail, automotive and manufacturing industries has re-awakened interest in related friction based technologies. A greater understanding of these related technologies will provide a greater understanding of FSW. For example, friction extrusion and the extrudability of certain materials have a direct correlation with material friction stir weldability. [7] The hydrodynamic condition established in friction hydro pillar processing (FHPP), the 'third-body' effect observed with most friction variants, and certain types of wear phenomena provide a better appreciation of process parameter optimisation. [6,26] The asymmetry and cycloidal deposition characteristics of the friction surfacing and friction seam welding processes provides invaluable evidence of the underlying mechanisms that enable the FSW process to be further developed. Greater understanding will help to accommodate more difficult materials, thinner and thicker components and the achievement of increased joint integrity and process efficiency. [3] Friction surfacing, friction seam and FSW processes show some lack of symmetry at extreme process conditions. The use of optimised conditions, however, virtually ensures that differences between the advancing side and retreating side of the weld do not cause adverse effects. However, with less suitable conditions, the asymmetric nature of the process can lead to defects. In friction surfacing, lack of symmetry can lead to excess expulsion of material at the retreating edge of the deposit.

In FSW, defects, such as buried voids or a surface breaking groove that runs along the advancing side, can be found when using non-optimised conditions or poorly designed tools. The inherent lack of process symmetry causes a pressure differential around the probe such that the rotating tool tries to veer away from the retreating side of the weld towards the advancing side. Secure fixing and robust machine tool equipment will prevent any noticeable sideways deflection.


This paper has described two FSW lap-welding developments, a Flared-Triflute TM probe and an A-Skew TM probe. Both give lap welds of 190% of the plate thickness, an improvement in weld integrity and a reduction in upper plate thinning over current practice. Travel speed was doubled and a 20% reduction in axial force was apparent, when compared with conventional threaded pin type probes. The application of FSW to pipe welding is also described.


The authors wish to thank K I Johnson, D G Staines, I M Norris, E D Nicholas, I J Smith, E R Watts, A Leonard, P L Threadgill and P D Evans.


  1. Bevington J: 'Spinning tubes mode of welding the ends of wire, rods, etc, and mode of making tubes'. US patent 463134, 1891.
  2. Klopstock H and Neelands R A: ' An improved method of joining or welding metals'. British Patent Specification 572789, (1941).
  3. Nicholas E D and Thomas W M: ' Metal deposition by friction welding'. Welding Journal, (1986), p.17-27.
  4. Bedford G M: 'Friction surfacing for wear applications'. Metals and Materials, November 1990, pp.702-705.
  5. Andrews R E: 'Friction surfacing tomorrow-industrial applications of cladding'. TWI Bulletin, November/December 1996.
  6. Thomas W M, Nicholas E D, Dolby R E, Jones C J and Lilley R H: 'Friction plug extrusion'. International Patent. PCT/GB92/01540.21.8.92.
  7. Thomas W M and Nicholas E D: 'Friction stir welding and friction extrusion of aluminium and its alloys'. 3 rd World Congress on Aluminium, 15-19 April 1997, Limassol, Cyprus.
  8. Nicholas E D: 'Linear friction welding' DVS Berichte Conference on Flash Butt and Friction Welding with Allied processes, Stuttgart, December 1991, 18-24.
  9. Nicholas E D: 'Friction processing technologies'. Advances in Welding & Related Technologies, TWI, 8-9 September 1997.
  10. Thomas W M, Nicholas E D, Needham J C, Murch M G, Temple-Smith P and Dawes C J (TWI): Improvements relating to friction welding. European Patent Specification EP 0 615 480 B1.
  11. Nicholas E D: 'Developments in the friction stir welding of metals'. The 6 th International Conference on Aluminium Alloys, 1CAA-6, July 5-10, 1998, Toyohashi, Japan.
  12. Boon T, Thomas W M and Temple-Smith P: ' Friction welding apparatus and method'. Patent No. US 6,325,273 B1 (priority date December 6, 1996)
  13. Mishra R S, Mahoney M W, McFadden S X, Mara N A and Mukherjee A K: 'High strain rate superplasticity in friction stir welding processed 7075 Al alloy'. Scripta Materialia, 42 (2000), pp.163-168, 11October 1999.
  14. Cederqvist L and Reynolds A P: 'Properties of Friction Stir Welding Aluminum Lap Joints'. TWI 2 nd International Symposium on Friction Stir Welding, 26-28 June 2000, Gothenburg, Sweden.
  15. Dawes C J: 'Faster and faster - welding speed increases with tool development - one of a series of steps'. Bulletin Vol. 41 No. 4, July/August 2000, pp.51-55.
  16. Thomas W M, Nicholas E D and Smith S D: 'Friction stir welding - Tool developments'. 2001 TMS Annual Meeting and Exhibition, Aluminium Joining - Emphasising Laser and Friction Stir Welding, New Orleans,February 11-15 2001.
  17. Kallee S W, Nicholas E D and Thomas W M: ' Industrialisation of friction stir welding for aerospace structures'. Structures and technologies - challenges for future launchers. CNES 3 rd European Conference, 11-14 December 2001, Strasbourg, France.
  18. Heideman R J, Scott J W T, Smith C B, Thessin T M, Ranganathan B N and Bishofberger D F: 'Friction Stir Welding Tool'. International Patent Publication Number WO 99/58288.
  19. Brooker M J, van Deudekom A J M, Kallee S W and Sketchley P D: 'Applying Friction Stir Welding to the Araine 5 Main Motor Thrust Frame'. Fokker Paper, European Conference on Spacecraft Structures,Materials and Mechanical Testing, 29 November - 1 December 2000, Estec, Noordwijk, The Netherlands.
  20. Cederquist L and Reynolds A.P: ' Factors affecting the properties of friction stir welded aluminium lap joints'. Welding Research Supplement, December 2001, pp.281-287.
  21. Matsumoto K and Sasabe S: 'Lap Joints of Aluminium Alloys by Friction Stir Welding'. TWI 3 rd International Symposium on Friction Stir Welding, Kobe, Japan, 27-28 September 2001.
  22. Thomas W M, Braithwaite A B M and Dolby R E: 'Skew-Stir TM technology'. TWI 3 rd International Symposium on Friction Stir Welding, Kobe, Japan, 27-28 September 2001.
  23. Kearns W H: 'Engineering, Costs, Quality and Safety'. Welding Handbook, American Welding Society, 550 N.W. Lejeune Road, P.O. Box 351040, Miami, FL 33135.
  24. Andersson C-G, Andrews R E, Dance B G I, Russell M J, Olden E J and Sanderson R M: 'A comparison of copper canister fabrication by the electron beam and friction stir processes'. Swedish Nuclear Fuel andWaste Management Co (SKB) and TWI. TWI 1 st International Symposium on Friction Stir Welding, 14-16 June 1999, Thousand Oaks, California, USA.
  25. Engelhard G, Hillers T and Pellkofer D: 'Orbital Friction Stir Welding of Aluminium Pipes'. TWI 3 rd International Symposium on Friction Stir Welding, Kobe, Japan, 27-28 September 2001.
  26. Godet M: 'The third-body approach: A mechanical view of wear'. Wear, Vol. No. 100, 1984, pp.437-452.

For more information please email: