FSW uses a non-consumable rotating tool, which moves along the joint between two components to produce high-quality butt or lap welds. The FSW tool generally has a profiled pin or probe and a shoulder with a larger diameter than that of the pin. The pin length is similar to the required weld depth. The pin is traversed along the joint line while the shoulder is in intimate contact with the top surface of the workpiece to provide frictional heating, consolidation and to avoid expelling softened material.
The resulting joint has excellent tensile strength and fatigue and fracture properties. The tensile strength can be within a few percentage points of the parent material, with the fatigue strength very similar and fracture resistance at least as good as parent material.
TWI holds two key patents on FSW. The first was filed in 1991 and the second was filed by Norsk Hydro in 1995 and shortly afterwards assigned to TWI.
It is interesting to look at TWI's first description of FSW and to compare this with the many details that others have thought worthy of additional patent filings. Somewhat naively, TWI thought it was providing a full and comprehensive description of this new process!
TWI's first patent describes the process in Claim 1 as:
'A method of joining workpieces defining a joint region there between, the method comprising carrying out the following steps without causing relative bodily movement between the workpieces: causing a probe of material harder than the workpiece material to enter the joint region and opposed portions of the workpieces on either side of the joint region while causing relative cyclic movement between the probe and the workpieces whereby frictional heat is generated to cause the opposed portions to take up a plasticised condition; removing the probe; and allowing the plasticised portions to solidify and join the workpieces together.'
Analysis of claim 1
The first claim clearly states that the method of the patent is to join workpieces by forming a joint between them over a defined region without any relative movement between the workpieces. No mention is made of the particular configuration of the workpieces or of the orientation of the joint region, so any normal configuration obvious to the welding/joining layperson could be considered to be embodied by this, therefore butt, lap, T, scarf and other joint configurations are included, with the workpieces orientated as required, the only proviso being that the workpieces should not move relative to each other.
The claim then states that the method is carried out by causing a probe of material harder than the workpiece material to enter the joint region and opposed portions of the workpieces on either side of the joint region. That is, a probe that will deform the materials to be joined, without being substantially deformed itself.
Again, direction or orientation is not made specific, only that the workpieces oppose each other at the joint region. Therefore, the probe can approach the joint region from any direction while contacting the workpieces. This, again, is any direction that could be considered obvious to a skilled layperson, indicating that a variety of joint designs can be employed. The claim states that this occurs while causing relative cyclic movement between the probe andthe workpieces whereby frictional heat is generated to cause the opposed portions to take up a plasticised condition. That is, the probe is rotated or oscillated in a cyclic manner while it is inserted into the joint region and continues to do so during the process.
The claim then states that the probe is then removed to allow the plasticised portions to solidify and join the workpieces together. That is, the probe is removed without causing any lateral relative translation between it and the workpieces - typical of what could be called a 'spot' weld.
Other details in TWI patents
TWI's first patent discloses many of the potential uses of FSW. It mentions the use with metals, alloys, plastics and composites such as MMCs, the concepts of spot and interrupted welding, lap welding, and crack repair. Tool motions such as rotation, oscillation and reciprocation are discussed as is tool heating. Bobbin tools are shown and the use of both fixed gap and variable gap tools mentioned. The variable gap is described as being adjusted by either a cam lever or by a spring.
Claim 2 of the first patent also describes moving the probes along the joint line to make a weld. This is the most widely used implementation of the process.
The second patent shows the use of FSW in many different joint configurations. It describes the improvements that arise by making sure that adequate pressure is maintained during the operation by design of the tooling.
Key prior art
The patent offices tend look to past patent filings as their primary source of prior art so to assist them and others, TWI has collected a file of key prior art documents. This is available as a download from TWI's web site. TWI also distributed copies of these to the three main patent offices, namely European Patent Office (EPO), Japan Patent Office (JPO) and United States Patent & Trademark Office (USPTO).
By early in 1995, Dawes produced a short guide to FSW and listed many of the potential industrial applications that FSW is suitable for. These include extrusions and extruded lightweight panels, space frames, airframes, pressure vessels, fuel tanks, railway wagons and coaches and castings. He also showed a number of lap and butt joint configurations suited to FSW.
Take-up FSW by industry
TWI vigorously promoted this new joining technology to its Members and to industry at large by visits, presentations, articles, papers and research projects. By 1995 the first industrial applications began to emerge. There is some debate about what was the first successful application but it probably was the use of FSW to join extrusions and attach cover plates in the fabrication of refrigeration panels for fishing boats. The number of organizations requesting licences grew rapidly.
Figure 1 shows the cumulative number of licences issued with time. Although over 195 have been issued, some have been terminated.
There are 172 current licence holders, as of September 2007. Their distribution by location and type of organization is shown in Table 1.
Table 1 - Current licensees
|UK & EU
|USA & Canada
|China & SEA
The take up of FSW has been highest in Japan by industrial end users, followed by UK & Europe. In the USA, National Labs, Universities and R&Dmp;D institutes have led the way, with industry following. China is just beginning to explore the advantages offered by FSW but could easily become the largest user of FSW if present trends continue.
The industrial uses of FSW have been the subject of many papers and articles. TWI's web site www.twi.co.uk is a good source of such information. FSW has been used to fabricate deck structures for ships, railway carriages, airframes and secondary structures, heat sinks, nuclear waste containment vessels, launch vehicle fuel tanks, wheels, automotive body parts, electrical distribution assemblies and small pressure vessels. It has also been employed to make large-scale blanks for subsequent machining or forming, when stock material sizes are too small for the application.
The vast majority of the industrial use has been with aluminium alloys. The process is now well developed for these and other light metals such as copper and magnesium. Development of the process for hard metals such as titanium and steels is advancing rapidly and the first major industrial deployments involving these materials are expected soon.
Patent filings by third parties
By 1996 the first of the third party patent filings began to be made. Like the issued licences, these applications grew rapidly. TWI maintains a database of all the patent applications it finds at the three main patent offices. The contents are available as a download from the TWI website.
For various reasons, TWI may only be aware of around 80 - 85% of the actual patent filings made at any one time. This is mainly because patent applications are not made public for at least 18 months from filing and, prior to 2001,not at all by USPTO. This explains the apparent slowing down of applications for 2006 in Figure 2.
TWI does not search for national filings in countries outside the three main patent offices as most 'local' filings result in a Patent Cooperation Treaty (PCT) filing or other main filing in due course.
Applicants sometimes describe FSW as 'friction churning', 'friction agitation joining', 'stir-friction', 'friction stirring joining', 'penetrating frictional agitation joining', 'friction stir bonding', 'friction mixing joining' and 'friction bonding'. This can make finding FSW patent applications difficult. Searching by International Patent Classification (B23K20/12) rather than keyword usually overcomes this problem. However, there are some FSW related patents classified under other headings either due to a classification error or by being classified under some other feature of the particular filing.
Of the total known patent applications from third parties, some 1800, about 90% are from industrial end users and about 5% each from equipment developers and academia. Table 2 below shows the geographical distribution of the origin of applicants. Japanese organisations are the most prolific. The JPO receives the majority of the patent applications but Japanese applicants also make many filings in both EPO and USPTO.
Table 2 - Where the filings come from
When the success or otherwise of these applications are examined, interesting differences begin to show as indicated in Table 3. Applicants appear to have the most success with USPTO filings no matter where they are based. The USPTO only started publishing patent applications in 2001 so it may be thought that the success rate is being overstated as, before that date, only successful patents were known and there was no information on applications that had failed, or were still in the pipeline. However, analysis of the post-2001 data has shown that the success rate in USPTO has increased, and it appears to be the office most ready to grant FSW patents. This result may be due to the longer time it takes for applications to be examined at EPO and JPO. Partial analysis of the early filing data suggests that the EPO and JPO outcomes will tend to increase towards the success rates of the USPTO.
There also appears to be a higher success rate for foreign applicants to JPO than local organisations but the sample size is quite small and this result is probably not significant.
Table 3 - The success of applicants by office
These figures are also affected by the rate at which the applications pass through the various offices. The time taken for examination and grant can vary enormously for different applications to the same office and for the same application to different offices. There is often no clear point at which an application is deemed to have failed as patent offices allow claim amendment, re-submission, continuations, divisionals and appeals. The final position on agiven application may not be known for a decade or more, so it is only practical to look at the success rates, for comparison purposes, rather than the failure rates.
Most active organisations
The top 11 organisations in terms of numbers of successfully granted patents as of September 2007 are listed in Table 4.
Table 4 Organisations holding the most patents
As was discussed previously, this table does not indicate the organisations making the most applications. It shows those applications that have been with the various offices long enough to have gone through the process.
For example, Showa have filed more applications than Boeing (166 to 131) and Honda have filed more than Mazda (80 to 78).
FSW, a disruptive technology for the manufacturing industry, has the effect of spawning many patent applications from early adopters of the process. They are often filed to protect the competitive advantage gained by being an early adopter for a given application.
For Hitachi, the benefits and promises proved so great that they adopted a tactic of filing a wide variety of patent applications covering one particular product type so as to try and 'picket-fence' FSW for their market. They consider FSW as a differentiating technology that allows them to create a global patent portfolio so as to gain new orders and expand global business.
Of course, this high level of filing by early adopters led to others concluding that they too needed to make filings to protect their business position. In TWI's view, many of the granted patents have considerable overlap with each other and assertion of rights might prove problematic. Also, strong prior art exists in many cases. It appears that most third party patent filings are defensive.
Main themes in patent filings
TWI monitors third party patent activity as part of its exploitation strategy. This provides some opportunity for spotting potential infringement and allows the progress of the technology to be studied. Over the 10-year course of this observation, common themes have become apparent in many of the patent filings. The main ones appear to be:
- Products & applications
- Tool shape & configuration
- Process variation
- Spot welding - FSSW
TWI attempts to classify the patent applications it finds into these themes. However, some filings deal with more than one theme so they are counted in what appears to be the main theme of the filing.
The next two tables show the filings and the granted patents broken down by theme and office.
Table 5 - Main Themes in patent application filings
|Tool shape & config.
|* also includes PCT applications
Nearly half the granted patents deal with a specific product or industrial application, as might be expected. Process variation claims make up another quarter of the grants, with tool shape, configuration, apparatus and equipment and FSSW making up the remainder. The only significant difference between the main patent offices is that process variation patents make up over 40% of the grants in JPO, whereas elsewhere the figure is much lower.
Table 6 - Main Themes in granted patents
|Tool shape & config.
Early adoption by manufacturers and 'picket-fence' style patenting has led to a high level of filing activity relating to specific products, including patent filings for materials or specific configurations applicable to the manufacture of the product. It may seem that applying an innovative process to a known product is an obvious step, especially when the benefits and applicability of that process are well known. This, however, has not stopped the flowof patent applications.
Many of these filings are related to joint design or a particular arrangement of materials, where the FSW process leads to the requirement for modified designs when compared to joints made using conventional fusion techniques.
A common theme in this type of patent filing is the division by product, for example whether the application relates to the manufacture of a train, an aeroplane or a heat sink. This does not provide much information concerning the general pace of technological development, just the areas into which its use has emerged. However, many of these applications can be seen to deal with generic issues that may be representative of solutions across the broader field evenif they appear to deal with the issues of a particular product. Other themes of interest include:
One of the characteristics of FSW, due to the forging/extruding nature of the process, is a slight workpiece thinning around the joint region. Material flows to take up any small gaps between the workpieces and some is used to form the surface flash. The simplest way to compensate for this, if needed, (by thickening the local region) may seem obvious but it has not prevented many patents being granted on this subject.
Hitachi has filed many patent applications relating to the joining of extrusions using FSW, in particular for improving the visual properties of the finished joint. This has been achieved by a number of methods but mostly by altered joint design - localised thickening specifically for the purpose of gap filling using protrusions of various shapes, overlapping regions, mating interfaces, force supporting regions, additional plating and similar. Other methods include adjusted welding parameters (welding in a particular direction, plunging a particular amount), tool design (including use of bobbin tooling), procedure ordering (such as welding the lower plate face first from the insidefollowed by welding of a covering plate) and joint finishing/smoothing techniques.
Dissimilar material joining
Using FSW to join dissimilar materials has attracted some interest. Most patents have described different tool orientations, material lay-ups and minor operational modifications. Applications that are more interesting have appeared for the creation of such structures as transition joints (such as aluminium to steel) where the softer material is forced into holes or re-entrant features in the harder material.
Heat treatment/material specific
Almost all FSW is currently carried out on aluminium alloys, as these are the most readily joined by the process. However, interest in transferring the well-known benefits of FSW to other materials is growing. Although many filings for joining 'other' materials specify types of tooling and equipment, filings have been made that purely cover a configuration of workpieces, tooling, clamping or even a variation of parameters to deal with specific material types.
Early in the development of FSW it was clear that the process could be applied to a wide variety of joint types, thickness and configurations with minor alterations to typical process parameters. This did not stop many filings appearing for methods of making common joints, although several inventive concepts for joining lapped workpieces and making fillet/corner joints have appeared.
Crack repair was directly cited in TWI's first FSW patents and developments have recently been the subject of further filings.
Low heat input components
One of the most useful characteristics of FSW is its low heat input relative to other welding processes. This has been used widely in the joining of components with heat sensitive regions, such as in the manufacture of automotive suspension arms with rubber bushings or the packaging of electronic components.
Tool shape & configurations
Although tools could be considered part of the apparatus/equipment theme, a clear delineation can be seen between active tooling having some form of mechanical actuation or control (other than being rotated and plunged), and passive tooling, where the tool features are static.
The earliest filing to cover modified tool designs and shapes, apart from the tools introduced in the first two TWI-controlled patents, was another filing from TWI.  This covered a wide range of tool designs including textures, shapes, projections, profiles, grooves and scrolls on pins and shoulders as well as many non-tool related features such as consolidating rollers and underwater FSW. The main aim of this filing was to place these ideas in the public domain, and TWI did not proceed with the application once it was made public.
One of the most revolutionary tool designs to appear following this was the MX TrifluteTM. This tool design enabled a threefold increase in joining speed, in part enabled by the tapered pin shape and action of the tool features. TWI filed a patent for this design in 1998,  but again chose not to progress it as placing this design in the public domain should benefit the process at large. Since then, no outstanding patents have been granted on basic tool design, although there are some patents covering basic design features (even tool size ratios), many incremental improvements and similar applications that, in some cases, appear to overlap.
Common themes covered by tool design patents include:
The first successful FSW tools had a concave shoulder that required the tool to be slightly tilted. Altering the shape of the shoulder to be convex allows for variation in plunge depth and material thickness. This concept was presented in the TWI filings, but was combined with other features in later filings. The patent by Colligan  shows a typical example of this type of tool. One of the primary advantages of this type of tool is the variable effective diameter of the shoulder, meaning that variations in penetration and shoulder contact zone can be madeduring welding.
Bobbin tools were described in the first TWI patent but have been the subject of many subsequent filings. A bobbin tool has a second shoulder connected to the end of the pin, providing clamping and consolidation to the underside of the workpiece. This lessens the amount of clamping required and removes the need to provide support during welding. Many of the filings deal with developments of such tools with variable force, position control and independent shoulder rotation.
Scrolled or grooved shoulders
Adding a scroll or groove to the shoulder  allowed the tool to be operated with no tilt angle, allowing better 3D control. Developments on this, incorporating modified tool profiles, contours, pins and shapes have been mentioned in many filings.
FSW can be easily carried out on a modified milling machine. However, industrial end users would probably need bespoke apparatus and welding equipment suppliers were interested in the process from the start.
Many filings have been made for general purpose FSW equipment. These filings tended to introduce common-sense engineering practice combined with elements of control specific to FSW, such as separately controlled shoulders and pins, spindle configurations, supporting members, clamping and temperature control.
Developments of particular interest include:
Retractable pin tool (RPT)
An important development was that of RPT technology, where the position and length of the FSW pin is adjusted to compensate for workpiece variations and help deal with the exit hole left at the end of an FSW joint.
The first filings of this type were in the US in 1996 by Boeing and Rockwell, although filings by European and Japanese companies were not far behind. This has been one of the most popular subjects for patent application filings. The following table illustrates some of these:
Table 7 - Selected RPT & adjustable pin patent filings
Suffix 'A' following the patent number denotes an application, while 'B' indicates a granted patent. Many of the early filings above were only made in the territory of the company responsible and it is probable that no single company can claim to have an absolute right to RPT due to the number of overlapping filings. Although seemingly similar filings have been made for RPT variants, new filings tend to introduce new uses and modifications on the theme,which appear to differentiate them enough to warrant patent grants.
|US 5697544 B
|US 5713507 B
|US 5718366 B
|JP 3771972 B
|EP 1021270 B; US 6264088 B
|US 5893507 B
|JP 3045698 B
|US 6199745 B; EP 1105246 A
|US 6247633 B
|EP 1286805 A
|EP 1230062 B; US 6722556 B
|JP 2001259863 A
|US 6497255 B
|JP 3471313 B
|JP 2003305576 A
|US 6758382 B
|US 2005045695 A; EP 1510280 A
|JP 2005288465 A
|JP 2005319484 A
|JP 2006000858 A
|US 2006163316 A
Although in its broadest sense cooling or heating can equally be covered as a modified method of carrying out FSW, the different methods have mostly been specified around an actual device. These include tool cooling (internal waterflow, external sprays, gas jets, heat sinks, conductive tool bodies), workpiece cooling (including underwater welding, as stated earlier) and combined workpiece/tool heating (using water, gas, incoherent & coherent radiation, arc, flame and induction heating).
The use of temperature control is not generally necessary when welding many typical aluminium alloys. Cooling can improve surface finish, especially when joining higher temperature materials. Heating can also be useful in prolonging tool life and increasing traverse speed when joining higher temperature materials. How temperature is controlled, and to what degree, is heavily dependent on the material being joined and the tool material being used.
The development of FSW on materials other than aluminium has led to corresponding progress in the research of suitable tooling materials.
FSW requires the use of a tool material much harder than the workpiece at the softening temperature of the workpiece. The tools undergo exposure to high forces at this temperature and they should not break or wear.
It would seem obvious that a suitable candidate tool material for a given workpiece is one that is harder, stronger, resistant to high temperatures, formable with decent high-performance properties. Nonetheless, various filings have been made, with some grants, for the use of known materials to make FSW tools. These include PCBN, W, W-Re, SiN, SiC and a range of other engineering ceramics and refractory metals.
Many elements of process variation have been covered in preceding sections, including temperature control and RPT. RPT is just one method of dealing with the exit hole left at the end of a FSW joint. Another, closely related to RPT,is by forcing plasticised material back into the exit hole using a metal annulus surrounding the pin in pumping action. This requires quite complex and expensive equipment. A far easier method of dealing with the hole is to provide run-off plates or to design a component in such a way that the exit hole is formed in a non-critical or easily removable region. Various examples of this are provided in the patent literature, both for basic shapes (plates, cylinders)and actual products.
Other interesting process variations include:
As mentioned earlier, the forces created during the FSW process are quite large and require extensive clamping and reacting. Several patent applications address methods for reducing them, including the use of bobbin tools, use of C-frames for spot joints (FSSW) and other methods. One of the most interesting is the use of a second probe or pin contra-rotating with the normal probe. The action of the two probes cancels each other out, resulting in much reduced forces. Patents describing the application of variable forces to the workpieces to alter the residual stresses after joining have also been filed.
Combining FSW with other processes has attracted interest, even though this would seem to be one of the more obvious ideas. The low heat input of FSW has led to its application in joining heat sensitive components. Patents have been applied for using FSW in conjunction with composite forming processes that are equally sensitive to temperature gradients. There are also cases when the heat provided by FSW has been used positively, particularly in the curing of adhesive or other coatings between the workpieces These can provide additional sealing and corrosion protection. This is of particular interest in the aerospace sector and has been the subject of several filings, some successful.
Filings were first made in Europe late in 1999 for methods and apparatus relating to spot joining, in particular for creating hole-free joints by forcing material down under retraction of the pin, followed by methods of joining aluminium billets in early 2000. In the wake of legislation over environmental concerns, the benefits of FSSW for joining aluminium in automotive structures began to emerge at the same time. This saw the first automotive FSSW applications, initially claiming the actual process of forming a spot joint (disclosed in the first FSW patent by TWI) before being argued back by the patent offices to specific types of machinery, such as robots with C-frameheads.
The actual use of robots for FSW had started many years earlier with implementations using parallel kinematics-type robots, such as the Tricept, and even the less rigid articulated arm robots. However, the addition of a force containing C-frame head, near identical to that used for other joining operations such as riveting and clinching, introduced FSSW as a viable, low-cost replacement for the energy intensive resistance spot welding of aluminium,overcoming the problem of high forces and articulated arm robots.
Since then, companies in the automotive field have filed more than 182 FSSW patent applications.
FSW has been successfully developed for the lighter metals and alloys such as Al, Cu and Mg. The main thrust of much current research work is developing tools and procedures for FSW of steels, titanium and higher temperature alloys. All these materials can be joined using FSW and FSSW but the tools are not yet robust enough to be employed in industrial production. A number of promising developments are expected to mature in the next year or so.
Sources of information
The TWI web site www.twi.co.uk contains extensive information about TWI, the FSW process, licences, prior art and published reviews and papers. The current list of known filings of patent applications, and the list of key prior art, is available as downloads, although TWI cannot vouch for the completeness of these lists.
For more information, please email firstname.lastname@example.org