From a special issue of International Journal of Fatigue, 2003. Published by Elsevier Ltd: www.sciencedirect.com/science/journal/01421123
Friction Stir Welding (FSW) is an entirely new solid state joining process that was invented by TWI in the early 1990s. It has generated tremendous interest in industry, partly because of the relative ease with which high integrity welds can be made rapidly from one side without the need for joint preparation, filler metal or protection from the atmosphere, but also because the mechanical properties, including fatigue strength, can be exceptionally high compared with other types of weld. It has proved to be particularly valuable for joining aluminium alloys and this has been the main focus of research and development. As the process does not melt the parts being joined, materials such as series 2000 and 7000 aluminium alloy, which are difficult to weld by fusion processes due to solidification problems, are readily weldable. Single pass full penetration butt welds have been made in alloys less than 1mm to over 50mm in thickess. The process is used commercially by an ever-growing range of industries, including aerospace, shipbuilding, railway and automotive. Perhaps the most controversial of these is the use of FSW in aerospace structures. Welded joints in general can exhibit such poor fatigue properties that they have not been an option for such fatigue-critical structures. FSW has changed this.
The fatigue strength of FSW butt welds in aluminium alloys can be as high as that of the parent metal. With such potential, considerable research effort is being directed at the factors that affect the fatigue performance of FS welds and the generation of fatigue design data. There is clearly a special need to understand fully the factors that can reduce their fatigue lives, to ensure that design guidance is consistent with use of the process in an industrial environment. One potentially valuable commercial aspect of FSW is that joints can be made very rapidly. However, the lower heat input resulting from high speed welding increases the risk of incomplete bonding in the joint, leading to the small forging-type flaws known as 'kissing bonds'. This Special Issue of the journal contains three papers that address the likelihood and significance of such flaws, as well as other aspects of the fatigue performance of friction-stir butt welds in aluminium alloys under fatigue loading transverse to the joint. However, the Issue opens with a broad overview of the current status of fatigue design guidance for welded aluminium alloys and the experimental basis of that guidance. Ultimately, one aim in the current developments of FSW related to aluminium alloys is to introduce fatigue design rules for friction-stir welds.
Ericsson and Sandström compare the fatigue performance of joints made by FSW and conventional MIG and TIG arc welding. The results showed that the use of FS welding speeds up to nearly three times that used for MIG welding had no detrimental effect on either the static or fatigue strength of the joint, which was still higher than that of either of the two arc welds. The paper also provides useful insight into the degree of softening of the aluminium alloy due to welding,
The paper by James, Hattingh and Bradley presents more direct evidence of imperfections in FSWs and the influence of welding speed. Although embedded forging-type defects were observed, their presence was not related to welding speed, and indeed they did not act as fatigue crack initiation sites. Their main significance seemed to be that they could increase the effective growth rate of fatigue cracks by providing easier linking paths between adjacent cracks.
'Kissing bonds' were found to provide sites for fatigue crack initiation in the work described by Dickerson and Przydatek . However, these were surface-breaking, at the weld root, and as such would be expected to be more significant than the embedded flaws discussed above. Their presence could reduce the fatigue strength of the transverse butt welds tested, but flaws up to 0.35mm in depth were harmless. All the welds gave fatigue lives higher than the current Eurocode design curve for butt welds made from one side. From the practical viewpoint, it was noteworthy that welding conditions well outside the usual range of conditions for FSW were required to produce the flaws.
Finally, it will be noted that even these three papers show quite marked variations in the fatigue strengths of friction-stir butt welds. A great deal more research is needed to understand fully the reasons for this variation and the extent to which it needs to be considered in fatigue design rules for friction-stir welds. However, the clear potential of the process for producing high fatigue performance joints certainly provides encouragement for such research effort.