GB2319977A

GB2319977A - Friction welding sheet material

Info

Publication number: GB2319977A
Application number: GB9625404A
Authority: GB
Inventors: Wayne Morris Thomas; Peter Temple-Smith
Original assignee: Lead Sheet Association
Current assignee: Lead Sheet Association
Priority date: 1996-12-06
Filing date: 1996-12-06
Publication date: 1998-06-10
Anticipated expiration: 2016-12-06
Also published as: GB9625404D0; GB2319977B

Abstract

A method of friction welding suitable for joining workpieces and forming a joint region therebetween comprises the steps of:- (i) Offering a probe 1 of material harder than the workpiece material to the workpiece surfaces in the joint region, a head region of the probe contacting one or both workpieces; (ii) Causing a rotational movement between the probe and the workpieces while urging the probe and workpieces together whereby frictional heat is generated in the joint region so as to create a plasticised region in the workpieces around the probe; (iii) Removing the probe or advancing the probe along the joint region and allowing the plasticised region to solidify and thereby join the workpieces together. The method is characterised in that the head region of the probe incorporates a plurality of protrusions 2 adapted to penetrate the workpieces. Such protrusions 11 may be crescent shaped.

Description

IMPROVED FRICTION WELDING APPARATUS Field of the Invention The present invention relates to a method and apparatus for friction welding.

It is particularly applicable, but in no way restricted to, a method and apparatus for joining lead sheets and is based on a relative rubbing movement between a probe of harder material and the two members to be joined. The present invention is also directed to an improved tool to be applied to the friction welding process.

Background to the Invention The attractive properties of lead have been put to good use in the fabrication of all types of buildings from private houses to industrial complexes. Traditionally, when it is necessary to join lead the tried and trusted technique of lead burning is utilised. This joining method cannot be considered ideal because of the skill needed and the environmental and health problems that are associated with the fumes emitted. Furthermore, lead burning is a highly skilled job and there is a general shortage of skilled operators. As a consequence, the initial cost outlay for laying lead sheet as a continuous membrane is unattractive or uneconomic when compared with felt/bitumen alternatives. The initial installation cost outlay should not be the only factor used when comparing the two roofing systems. Felt roofs generally have a life expectancy of only 10 - 15 years whereas lead roofs last considerably longer. Over an extended period of say 20 - 100 years a lead roof is likely to be considerably more cost effective. However, the initial high installation costs still tend to discourage architects and builders from specifying lead as a flat roofing continuous membrane material of choice, despite its longevity and attractive physical appearance.

If a simple and cost-effective method of joining lead sheet were available, then this could change the economic equation in favour of lead sheet as a flat roof membrane material.

Possible alternatives to lead burning are friction welding and friction stir welding. The process of friction welding has been known for many years and typically invoives causing relative movement between a pair of workpieces while they are urged together so as to generate a plasticised region, stopping the relative movement and allowing the plasticised region to solidify thereby joining the work pieces.

It has also been proposed in the past to join workpieces by use of a nonconsumable member which does not form part of the finished joints. An example of this so-called friction stir butt welding technique can be seen in WO 93110935 (The Welding Institute) and WO 95/26254 (NORSK HYDRO) in which two workpieces are urged together using various shaped tools which cause the region to be joined to plasticise and hence become joined. This technique allows sheet materials to be welded by a solid state process in either butt or lap geometries. The process operates by forcing a rotating tool with a specially shaped probe along the joint line, which causes intense plastic deformation of the immediate surrounding material. The tool is designed to prevent material escaping, and once the tool has passed a given point, a weld is made.

The leading edge of the rotating tool provides frictional heat and subsequent thermal softening in front of the tip, in effect preheating the area about to be bonded.

This effect is especially useful in allowing the passage of the tip part of the rotating tool through the material. The greater the area of the shouldered region of the rotating tool in contact with the joint, then the greater the frictional heat available.

However, increasing the diameter of the shoulder has practical limitations and tends to produce a side flash.

To date, this process has been successfully demonstrated for many aluminium alloys and also for thermoplastics. However, application of this process to lead sheet results in major difficulties because lead is relatively soft and malleable and the above techniques do not achieve satisfactory results in terms of weld acceptability.

An object of the present invention is thus to provide a method and apparatus which overcomes some or all of the above problems.

Summary of the Invention According to a first aspect of the present invention there is provided a method of friction welding suitable for joining workpieces and forming a joint region there between comprising the steps of: (i) Offering a probe of material harder than the workpiece material to the workpiece surfaces in the joint region, a head region of the probe contacting one or both workpieces; (ii) Causing a rotational movement between the probe and the workpieces while urging the probe and workpieces together whereby frictional heat is generated in the joint region so as to create a plasticised region in the workpieces around the probe; (iii) Removing the probe or advancing the probe along the joint region and allowing the plasticised region to solidify and thereby join the workpieces together; characterised in that the head region of the probe incorporates a plurality of protrusions adapted to penetrate the workpieces.

Moving from a single point probe to a multiprong whisk type tool causes plasticising material movement without causing melting of the low-melting lead, enabling a fault free joint to be made.

Preferably, when the workpieces are joined by a lap joint the protrusions are so sized and shaped that they fully penetrate the workpiece nearest the probe but only partially penetrate the other workpiece.

Not only does this achieve the optimum strength and appearance of the joint but it also ensures that the lower workpiece remains intact.

Preferably the protrusions are symmetrically displaced around the probe head about the rotational axis of the probe. This leads to smooth operation and prevents the tool from tending to drag to one side or the other.

Preferably the protrusions are substantially crescent-shaped in cross-sectional profile.

In a particularly preferred embodiment the concave surface of the crescentshaped protrusions is directed substantially towards but displaced from the axis of rotation of the probe.

Preferably the probe head incorporates two protrusions.

Alternatively and for some applications the probe head may incorporate three or four protrusions.

Preferably the protrusions take the form of slightly tapering cylinders, the taper narrowing towards the tip of the protrusion. This facilitates penetration and ease of operation generally.

In a particularly preferred embodiment the head region of the probe is substantially convex in shape, the edge of the head i.e. the portion furthest from the rotational axis being in effect chamfered away from the workpiece. This allows the convex surface to the probe to cause a slight undercut in the lead. It is important that this convex surface is to the correct depth to ensure sufficient heat is generated in the lead to allow welding to occur.

Preferably the probe is slightly displaced, in use, from the plane perpendicular to the workpieces. This displacement is typically 3 from the plane perpendicular to the workpieces, the inclination being away from the general forward direction of travel of the probe. Angles that vary from 3 can be used and this will depend in part on the rotational speed of the probe and the speed of forward movement.

One advantage of this method is that it incorporates all the benefits of friction welding whilst providing excellent results on lead sheet for the first time. But the method is not restricted to lead. It applies equally well to other low-melting or soft materials such as thermoplastics.

The present invention also encompasses any probe suitable for use in a method of friction welding as described above and incorporating any of the features herein described.

Brief Description of the Drawings The invention will be further described, by way of example, with reference to the accompanying drawings in which: Figure 1 shows a schematic side elevation of the welding apparatus employed in the present invention; Figure 2 shows diagramatically the depth to which the probe penetrates the workpieces; Figures 3, 4 and 5 show the head profiles of a number of different probes applicable to the present invention; Figure 6 shows a typical plot of tool rotational speed vs traverse speed; Figure 7 shows the results of typical peel tests; Figure 8 shows microstructure of successful weld; Figure 9 shows an overlapped butt joint.

Description of the Preferred Embodiments Embodiments of the present invention are described below by way of example only. These examples represent the best ways of putting the invention into practice that are currently known to the Applicant although they are not the only ways in which this could be achieved.

Referring to Figure 1, this illustrates a probe 1 made of hardened steel which incorporates two prongs or protrusions 2 in a head region 3 of the probe. Probes of this generally type have been developed by The Welding Institute and others for friction welding aluminium and its alloys. However, tool design is critical for use with lead sheet and the multi-pronged tool is most important for effective weld formation.

In this method as illustrated, a pair of lead sheets 5, 6 are overlaid one on another in order to form a lap joint, the lower sheet 5 being placed on a firm substrate 7 (not shown). The probe has a relatively narrow central or body region 4 which is adapted to fit into a rotating engine or motor of some description (not shown).

The probe head/prong dimensions are determined specifically for the thickness of sheet to be joined. In use, the prongs must fully penetrate the uppermost sheet and only partially penetrate the lowermost sheet 5, whilst the head of the probe causes a slight undercut into the lead sheet 6 by distance (t). To facilitate this, the head of the probe is preferably slightly convex in shape as shown in the various profiles in Figures 3, 4 and 5. In effect, the circumferential edge of head 3 is slightly chamfered away or radiused off. In use, the bottom face of the head thus makes good contact with the top sheet 6, as evidenced by a visibly polished zone along the centre of the joint region.

At the start of the welding procedure, the rotating probe may be introduced slowly into the lead sheet, giving time for plasticising to begin. Alternatively, a suitable hole can be drilled in one or both sheets or a run-on (and run-off) tab provided.

Having entered the rotating probe into the two workpieces, the probe is arranged such that its axis of rotation is angled away from the plane perpendicular to the workpieces by angle "a" This arrangement is clearly shown in Figure 1. The probe is then progressed in the direction of the forward arrow at a speed V whilst maintaining a substantially constant downward force F. As the probe moves forward the plasticised matter behind the probe sets and forms a joint between the two workpieces in a joint region which follows the path taken by the probe.

A large series of trials was undertaken to establish the optimum parameters for joining. The principle variables investigated were improvement to tool design, rotation speed, travel speed and angle of tilt of the tool. Essential parameters such as tool rotation speed and travel speed were measured for each run. Initially, welds were supported by a steel substrate. Later welds used substrates which were more typical of building practice, namely wooden roof boards of 18 mm thick marine WWP ply covered by a layer of commercial geotextile. The use of tarred building paper over wood and concrete were also briefly investigated.

Appropriate instrumentation and data retrieval software were used to measure and record the downforce, traverse force, rotational speed and current consumed during tool rotation and traverse.

All welds were examined visually to assess the quality and surface appearance. Selected visually satisfactory welds were further examined by X-ray in order to detect buried volumetric defects. Welds which were still considered satisfactory were subjected to a peel test to determine the relative strength of the weld. Peel tests were considered satisfactory when failure occurred through the parent material rather than through the weld itself.

Tool Design: The design of the tool is, of course, of paramount importance in friction stir welding. The tool must be able to plunge into the lead at the start of the weld, provide sufficient frictional heat to soften the material, and allow the material to pass from the front of the weld to the rear without leaving any cavities. Whilst a number of design concepts which have been investigated for lead and other materials, trials in this work have established that a whisk type tool, as shown in Figures 1, 3, 4 and 5 gave the best results. In this design, the separation of the whisk probes leads to a fairly wide weld, ensuring a large area of penetration into the lower material. Sketches of the various tools used are shown in Figure 5.

The most reliable welds were obtained with tool 14(a). Typical dimensions for this tool for use with 1.8 mm thick lead sheet are shown in Figure 4. These dimensions are given by way of example only. However, satisfactory results could also be obtained with the other multi-pronged tools shown in Figure 5 by carefully controlling the operating parameters (see below). The single pronged tools in Figure 5 did not give such good results.

Substrate The substrate under the sheets has a significant influence on the quality and nature of the final weld. Steel is one option that was investigated but this is uncharacteristically hard and is also a good conductor of heat. However, reliable results could be obtained when welding over steel. Further tests were carried out using the more realistic geotextile underlay laid over wood. This is the sort of material that would be found in practice under a flat lead roof.

This had two effects on welding. First of all, the geotextile is a fairly efficient insulator, preventing heat flow from the lead into the steel; secondly, the material is very compliant, and can effectively reduce the downforce on the tool if the same registered tool position is used. To provide suitable welding conditions the geotextile materiel will be temporarily compressed as the rotating tool moves along the weld seam.

The effect of heat conduction is a significant factor in the difference between process conditions for steel and wood substrates. Process trends indicated that a relative significant increase in the traverse rate would be necessary when welding is carried out on top of geotextile/wood substrate materials.

Typical results are shown in Table 1. These experiments were conducted using a vertical milling machine to weld two 1.8mm + 0.09mm lead sheets in a lap weld.

The welding conditions used to produce test welds 389 - 439 were used to produce welds for tensile test specimens T1-T4 (see tables 3 and 4). Although these conditions used to produce the welds were not optimised, the test results gave tensile values close to that measured from similar parent material tensile results.

Probe Angle Tests were carried out with a tilt on the tool, generally of 3 . This produced a very significant increase in quality, and enabled much higher rotational and travel speeds to be used. In fact, the upper limit on rotational speed was dictated by the partial fusion of the geotextile layer, and its subsequent fusion to the lead. Since the ultimate use of this process will be outside, tolerance to water would be a distinct advantage. Some trials were therefore undertaken where the lead sheet was completely underwater (welds 1419-1449), and yet this did not seem to adversely influence the quality of the weld.

In this phase of the work, tests were conducted under various conditions to establish the range of conditions for which satisfactory welds could be made. During this phase, tests were carried out to show that it was possible to achieve a travel speed Im/min (16.6mm/sec) with a satisfactory weld quality. The results are shown in Figure 6, and all data relates to tool 14a, used with a 3 tilt, on a geotextile over wood substrate. At low travel speed and high rotation speeds, it is difficult to prevent surface voids. Moreover, such conditions also result in melting of the geotextile, and its adherence to the lead, which is unacceptable for most applications. Satisfactory welds can be made at travel speeds below 16.7mm/sec, if a suitably slow tool rotation speed ( < 400 rev/min) is used, but there is no advantage to operating in this area. It is believed that travel speeds beyond 18mm/sec (the fastest investigated) may be possible. Machine capabilities prevented such speeds being used.

The data obtained are summarised in Table 2. In this part of the programme, 75 welds were made, each witnessed by representatives of the Lead Sheet Association. All welds were made with tool 14a, and using an angle of 3 . Initial trials (welds 1 - 7) were not entirely satisfactory, but a closer examination revealed that the wooden substrate was uncharacteristically soft, and the problem was resolved by replacing the wood with a harder piece. Welds 8 - 51 were all made with the same condition (travel speed 16.6mm/min, rotational speed 950 revimin), and gave good results. Radiographic examination of sample welds produced from this group gave no indication of any sub-surface defect. The temperature of the tool and the lead were monitored using a contact thermometer, and the results are indicated in Table 2.

In some cases the lead was either cooled or preheated, so simulate the extremes of weather conditions. Perhaps surprisingly, the temperatures after welding were little different to those measured on other welds.

Welds 52 and 53 used a tar paper underlay instead of the geotextile cloth.

The welds appeared to be good, although there was some melting of the tar.

The effect of soft wood was explored in welds 54-70. In general this resulted in an intermittent void, although this was not always particularly serious. Welds 57-60 employed very fast (30mm/sec) or slow (12mm/sec) travel speeds, but the results were similar. However, some melting of the geotextiie was observed in the slower speed welds, as expected. Better results were obtained in welds 62-70, where a faster rotational speed (1100 rev/min) was used. Occasional melting of the geotextile demonstrated that this would be the upper limit of rotation speed.

The final series of welds (70-75) were made onto a concrete substrate, with generally acceptable results.

In almost all cases, satisfactory results were obtained in pull tests, i.e. the failure occurred in the parent material rather than in the weld. The only exceptions to this were in a few welds made on the soft wood substrate, when the presence of surface voids indicated poor weld quality and when welding lead sheet which varied in thickeness and did not have a consistent thickness, typically that found with a cast material.

In the majority of peel tests, the parent metal failure occurred in the top lapped plate. Many examples where the parent metal failure occurred in the lower lapped plate were also observed. This aspect of parent metal failure location indicated that the effect of the friction stir welding technique was fairly well balanced, i.e. there was no significant difference between the top plate and the effect of the rotating shoulder and the bottom plate and the effect of the rotating probe within the lower plate, as shown in Figure 7.

Metallographic Examination An examination of the microstructure was made on a few selected welds (1149 and 141g). The weld nuggets were invariably fine grained, showing that recrystallisation had occurred. The welds were defect free, except for occasional strings of inclusions, believed to be lead oxide. A microsection taken from an underwater weld is shown in Figure 8.

Instrumentation Data A linear slide was modified and suitable load cells fitted to monitor axial and traverse force. An electronic device was fitted to capture current consumed during machine running under no load and under optimised welding conditions. Ten welds using a 25mm dia (type 14a) rotating tool at 3 tilt have been carried out. A typical instrumentation data sheet is attached under Appendix 2. To show the interrelationship and sequence of process events, the instrumentation data are included on one chart. Summary of Data

Average maximum downforce 188.31kg Average maximum traverse force 38.29kg Rotating head - average maximum power drawn from the mains during 2,855.6W welding Rotating head - average maximum power drawn from the mains under 2,729W no load conditions Rotating head - power drawn from the mains by welding operation 126.9W Traverse table - average maximum total power drawn from the mains 314W during welding Traverse table - average maximum power drawn from the mains under 272W no load conditions Traverse table - power drawn from the mains for welding operation 42W The latter current consumed data are those measured during friction stir welding trials on TWl's milling machine designated 857, and are representative of the transmission system and machine characteristics associated with this machine.

Conclusions An extensive programme of work has been undertaken to further investigate friction stir welding of lead sheet. The following conclusions have been reached: 1. It is possible to make high quality welds in 1.8mm thick rolled lead sheet manufactured to the appropriate standard using the friction stir process.

2. Tool design and inclinations are both critical. Providing the correct tool design and inclination are used, good welds can be made over a wide range of conditions.

3. Lead can be satisfactorily welded in the wet using friction stir welding, with no apparent loss of quality.

4. The process has been shown to be very repeatable.

It will be apprnciated that the whisk design of probe enables plasticised material to flow around and through the prongs. The degree of whisking can be increased if necessary by having three (tools lOs, 11a, 12a, 13a) or four (7a, 8a) prongs as illustrated in Figure 5. As well as carefully controlling the depth of each prong, the prong profile is also important. Crescent-shaped prongs as shown in Figure 4 have proved to be particularly effective. The prongs are effectively in the form of curved blades, the leading edge 10 of which cuts into the metal as the probe rotates. The thicker region 11 forces its way through the metal causing plastic deformation. This is repeated many times a minute as the probe rotates.

This example is just one type of blade shape that could be used. Elongate diamonds, needle point and curved blades can be used in straight or curved configurations as required. The prongs can be straight sided, tapered, threaded or ridged as required. Generally, however, the probes tend to be slightly tapered away from the tip region as illustrated.

The material of the probe is obviously harder than the workpieces. Hot work steel, high speed steel or other durable metals can be used as well as ceramic and other synthetic materials. Preferably the surface of the probe head which comes into contact with the workpiece is highly polished.

A further important advantage of this method is that the lead surface requires no preparation. Furthermore, various different types of joint can be formed as well as lap joints. A butt joint can be formed by butting together two sheets and overlaying the joint with a strip and forming two welds side by side. This is illustrated in Figure 9.

In summary the present invention provides:1. A method of joining lead sheets without the need for a naked flame as required in conventional joining methods.

2. A tool design which allows fast linear weld speeds in a lap joint of lead sheet without weld flaws or inclusion being a faster joining process than any conventional alternatives.

3. A method of joining lead sheet which can be applied in all weather conditions.

4. A method of jointing lead sheet which offers greater protection to the building fabric, by the avoidance of hot working, meeting a recent imposed workplace restriction on the use of naked flames whilst working on roof structures.

5. A method of joining lead sheet which enhances the Health and Safety of the workplace and thereby protecting the public at large.

6. A quality joining method which enhances the fundamental qualities of lead sheet as a weather protecting membrane in Building and Construction.

Tool Rotational Welding Tool Specimen Base Material Tool No Speed Traverse Speed Angle General Comments No mm rev//min mm/s degrees 38g 14a 200 0.6 Geotextile on wood V Very good weld surface. Good - best one yet 39g 14a 200 0.6 Geotextile on wood V Very good weld surface. Pull test good 40g 14a 200 0.6 Geotextile on wood V Very good weld surface. X-ray taken + photo 41g 14a 200 0.6 Geotextile on wood V Very good weld surface appearance 42g 14a 200 0.6 Geotextile on wood V Good result although too deep 43g 14a 200 0.6 Geotextile on wood V Very good result Table 2 WITNESSED WELDS

FRICTION STIR WELDING EFP DEPARTMENT TWI DATA SHEET EFP/F5.95 Project No 2 2 0 7 2 6 Machine Tooling Vertical milling machine No 857 Operator P Temple-Smith Additional Information Base material for trials. Wood & geotextile - concrete - bitumen paper Machine 8 5 7 Material Details 1.8mm #0.1mm lead sheet lap welds. (These welds were witnessed by LSA representatives) Tool Rotational Welding Tool Tool Lead Specimen Tool Base Material Speed Traverse Speed Angle Temp Temp X-ray Pull Test General Comments No No min rev/min mm/sec degrees C C Poor surface appearance - void on 1 14a 950 16.6 Geotextile on wood 3 - - # surface 2 14a 900 16.6 Geotextile on wood 3 - - Poor Poor surface appearance 3 14a 850 16.6 Geotextile on wood 3 - - Reasonable Poor surface appearance 4 14a 850 16.6 Geotextile on wood 3 - - Reasonable Poor surface appearance 5 14a 850 16.6 Geotextile on wood 3 - - Reasonable Poor surface appearance 6 14a 850 16.6 Geotextile on wood 3 - - Reasonable Poor surface appearance Poor surface appearance, wood 7 14a 850 16.6 Geotextile on wood 3 - - Reasonable appears spongy Comment - Plywood supplied by LSA appears to have defects in grain, also has soft spongy effect during welding Change of wood backing good 8 14a 850 16.6 Geotextile on wood 3 - - Good weld 9 14a

Tool Rotational Welding Tool Tool Lead Specimen Tool Base Material Speed Traverse Speed Angle Temp Temp X-ray Pull Test General Commets No No mm rev/min mm/sec degrees C C 13 14a 950 16.6 Geotextile on wood 3 - - Good Good surface appearance 14 14a 950 16.6 Geotextile on wood 3 - - Good Good surface appearance 15 14a 950 16.6 Geotextile on wood 3 - - # Good Good surface appearance 16 14a 950 16.6 Geotextile on wood 3 - - Good Good surface appearance 17 14a 950 16.6 Geotextile on wood 3 - - Good Good surface appearance 18 14a 950 16.6 Geotextile on wood 3 - - Good Good surface appearance 19 14a 950 16.6 Geotextile on wood 3 - - Good Good surface appearance 20 14a 950 16.6 Geotextile on wood 3 106 65 Good Good surface appearance 21 14a 950 16.6 Geotextile on wood 3 - 67 Good Good surface appearance 22 14a 950 16.6 Geotextile on wood 3 106 65 Good Good surface appearance 23 14a 950 16.6 Geotextile on wood 3 117 65 Good Good surface appearance 24 14a 950 16.6 Geotextile on wood 3 124 67 Good Good surface appearance Good surface appearance although 25 14a 950 16.6 Geotextile on wood 3 137 88 # slight geotextile melt 26 14a 950 16.6 Geotextile on wood 3 133 86 Good Good surface appearance Lead cooled to outside temp. 5.4 C 27 14a 950 16.6 Geotextile on wood 3 121 73 Good - good 28 14a 950 16.6 Geotextile on wood 3 128 85 Good Geotextile melt at start Geotextile melt at start. Good 29 14a 950 16.6 Geotextile on wood 3 148 82 Good surface appearance 30 14a 950 16.6 Geotextile on wood 3 150 82 Good Good surface appearance 31 14a 950 16.6 Geotextile on wood 3 136 77 Good Good surface appearance 32 14a 950 16.6 Geotextile on wood 3 143 84 Good Good surface appearance 33 14a 950 16.6 Geotextile on wood 3 146 82 Good Good surface appearance 34 14a 950 16.6 Geotextile on wood 3 124 75 Good Good surface appearance 35 14a 950 16.6 Geotextile on wood 3 144 81 # Good Good surface appearance

Tool Rotational Welding Tool Tool Lead Specimen Tool Base Material Speed Traverse Speed Angle Temp Temp X-ray Pull Test General Comments No No mm rev/min mm/sec degrees C C 36 14a 950 16.6 Geotextile on wood 3 148 86 Good Good surface appearance Lead pre-heated. Some geotextile 37 14a 950 16.6 Geotextile on wood 3 148 98 Good melt 38 14a 950 16.6 Geotextile on wood 3 140 83 Good Good surface appearance 39 14a 950 16.6 Geotextile on wood 3 149 85 Good Good surface appearance 40 14a 950 16.6 Geotextile on wood 3 136 74 Good Good surface appearance 41 14a 950 16.6 Geotextile on wood 3 121 80 Good Good surface appearance Lead pre-heated 80 C. Good 42 14a 950 16.6 Geotextile 3 112 70 Good surface appearance Lead cooled to -50 C. Good surface 43 14a 950 16.6 Geotextile on wood 3 78 32 Good appearance 44 14a 950 16.6 Geotextile on wood 3 83 63 Good Good surface appearance 45 14a 950 16.6 Geotextile on wood 3 118 78 Good Good surface appearance 46 14a 950 16.6 Geotextile on wood 3 125 80 Good Good surface appearance 47 14a 950 16.6 Geotextile on wood 3 - - # Good surface appearance 48 14a 950 16.6 Geotextile on wood 3 139 88 Good Good surface appearance 49 14a 950 16.6 Geotextile on wood 3 136 80 Good Good surface appearance Good surface appearance. Slight 50 14a 950 16.6 Geotextile on wood 3 158 90 melt geotextile 51 14a 950 16.6 Geotextile on wood 3 155 88 Good Good surface appearance Good surface appearance. Some tar 52 14a 950 16.6 Building paper on wood 3 124 84 melt from paper Good surface appearance. Some tar 53 14a 950 16.6 Building paper on wood 3 147 82 melt 54 14a 950 16.6 Geotextile on wood 3 77 65 Good Use of soft plywood reasonable Table 2 (continued)

Tool Rotational Welding Tool Tool Lead Specimen Tool Base Material Speed Traverse Speed Angle Temp Temp X-ray Pull Test General Comments No No mm rev/min mm/sec degrees C C Soft plywood soaked in water.

55 14a 950 16.6 Geotextile on wood 3 96 69 Good Void on surface 56 14a 950 16.6 Geotextile on wood 3 86 76 Poor Poor result using soft plywood 57 14a 950 30 Geotextile on wood 3 100 66 Poor Poor void on surface. Soft plywood 58 14a 950 12 Geotextile on wood 3 10 77 Poor Poor void visible. Soft plywood Slight void on surface. Soft 59 14a 950 12 Geotextile wood 3 117 87 Good plywood 60 14a 950 12 Geotextile on wood 3 136 97 Good Geotextile melt. Soft plywood 61 14a 950 16.6 Geotextile on wood 3 108 81 Good Void on surface. Soft plywood Good surface appearance. Soft 62 14a 1100 16.6 Geotextile on wood 3 111 79 Good plywood Slight geotextile melt. Soft 63 14a 1100 16.6 Geotextile on wood 3 144 87 Good plywood Good surface appearance. Soft 64 14a 1100 16.6 Geotextile on wood 3 138 95 Good plywood 65 14a 1100 16.6 Geotextile on wood 3 144 97 Good Slight void at start. Soft wood 66 14a 1100 16.6 Geotextile on wood 3 149 94 Good Slight void at start. Soft wood 67 14a 1100 16.6 Geotextile on wood 3 149 94 Poor result Good surface appearance. Some 68 14a 1100 16.6 Geotextile on wood 3 131 88 Good geotextile melt 69 14a 1100 16.6 Geotextile on wood 3 112 82 Good Increase in heel depth. Soft wood 70 14a 1100 16.6 Geotextile on wood 3 106 79 Good Good surface appearance Reasonable surface appearance 71 14a 1100 16.6 Geotextile on concrete 3 - Reasonable concrete. Not flat 72 14a 1100 16.6 Geotextile on concrete 3 88 77 Reasonable Good surface appearance 73 14a 950 16.6 Geotextile on concrete 3 74 63.6 Reasonable Poor void to surface

Tool Rotational Welding Tool Tool Lead Specimen Tool Base Material Speed Traverse Speed Angle Temp Temp X-ray Pull Test General Comments No No mm rev/min mm/sec degrees C C 74 14a 950 16.6 Geotextile on concrete 3 134 94 Poor Good surface appearance 75 14a 950 16.6 Geotextile on concrete 3 128 100 Good surface appearance

NAMES The TEST HOUSE NAMASuCSSb8 No. P Certificate of Test Page 1 ofi TWI, Abington, Cambs.

-enceNo T51075 MI No.: N/A rNo.: Specification: Clients own. 6 3 2 notion: Four off friction stir lap welds in Pb sheet (Pre machined tensiles) 4 ity: Project No. 220726 W M Thomas. 7 methods: Procedure: TPOl-l, BS EN10002-1:1990 7 Inspection Authority: N/A 3 5 3 SILE TEST . Test machine calibrated to Grade 1.0 requirements of BS EN 10002-2:1992 1 Dimensions Max Stress 5 Size I CSA I GL Load I Stress Load Stress El RA 3 ziryiposition Mark mm mm2 mm kN Nlmm2 kN N/mm~ % % 3 :s weld tensiles 3 ;ile T1 1 50.05 x 1.82 91.09 1.49 16.3 .ile T2 2 50.02 x 1.80 90.03 1.60 17.8 ;ile T3 3 ~ 50.01 x 1.79 89.51 1.47 16.4 while T4 4 -4 49.98 x 1.78 88.96 1.71 19.2 I :ification: lents: All samples fractured in the parent material immediately adjacent to the weld. The minimum calibrated range of the test ne is 2.00kN. Valves quoted which are below this figure should, therefore,, be treated with discretion.

S - iir'zo +'' H : . . V b (7 Tn Tes. House (Carnonce Lo HD^-.D~ Ha. Afin:riD^ Aong:o Cm Te. l e' 0 ~1~ rcZ fuzzO Drreczo-s D!' r~ , z ?E' se: steaet ^en e e Tme ~ -rn -

{A{AS T1^ T=T LW o TESTINC eSna8 No OWE Certificate of Test Page 1 of I TWI, Abington Hall, Abington, Cambridge. tce No.: T51083 MI No.: N/A No.: To follow Specification. Client's own. 6 3 vtion: Three off pre-machined tensile specimens in Pb sheet. 2 Project No. 220726 W M Thomas. 7 methods: Procedure: TPOla-l, BS EN10002-1:1990 7 Inspection Authority: N/A 3 5 LE TEST(S) Test machine calibrated to Grade 1.0 requirements of BS EN 10002-2:1992 3 Dimensions Max Stress 1 Size I CSA GL Load Stress Load Stress El RA 5 y/Position Mark mm mm I N/rnrn2 kN N/mrn % % B 7 1 50.20 x 1.79 89.86 50 1.59 18 82.8 ~ ) 1 1.65 19 80.7 ) 2 50.18 x176 88.32 50 1.65 19 80.7 3 50.18 x 1.77 88.82 50 1.62 18 84.0 I ~ ~~ ~ ~= = ~ 1 I ication: l nuts: The tensile test machine used to perform these tests is only calibrated in the range 2-60kN.

F . . . . . . . . oar PXfyamo&commat;"t . - > . cate':XP Ptr . 30 js .

.. . + . . ... ......

-3-0 -C-37 --I Ao.-.:so Camn=?e 6Aa -elec . G C - - - - - -:e

Claims

Claims 1. A method of friction welding suitable for joining workpieces and forming a joint region therebetween comprising the steps of: (i) Offering a probe of material harder than the workpiece material to the workpiece surfaces in the joint region, a head region of the probe contacting one or both workpieces; (ii) Causing a rotational movement between the probe and the workpieces while urging the probe and workpieces together whereby frictional heat is generated in the joint region so as to create a plasticised region in the workpieces around the probe; (iii) Removing the probe or advancing the probe along the joint region and allowing the plasticised region to solidify and thereby join the workpieces together; characterised in that the head region of the probe incorporates a plurality of protrusions adapted to penetrate the workpieces.
2. A method according to Claim 1 wherein when the workpieces are joined by a lap joint the protrusions are so sized and shaped that they fully penetrate the workpiece nearest the probe but only partially penetrate the other workpiece.
3. A method according to any preceding Claim wherein the protrusions are symmetrically displaced around the probe head about the rotational axis of the probe.
4. A method according to any preceding Claim wherein the protrusions are substantially crescent-shaped in cross-sectional profile.
5. A method according to Claim 4 wherein the concave surfaces of the crescentshaped protrusions are directed substantially towards but displaced from the axis of rotation of the probe.
6. A method according to any preceding Claim wherein the probe head incorporates two protrusions.
7. A method according to any of Claims 1 to 5 wherein the probe head incorporates three or four protrusions.
8. A method according to any preceding Claim wherein the protrusions take the form of slightly tapering cylinders, the taper narrowing towards the tip of the protrusion.
9. A method according to any preceding Claim wherein the head region of the probe is substantially convex in shape, the edge of the head i.e. the portion furthest from the rotational axis being in effect chamfered away from the workpiece.
10. A method according to any preceding Claim wherein the probe is slightly displaced, in use, from the plane perpendicular to the workpieces.
11. A method according to Claim 10 wherein the probe is displaced in use 3 + 20 from the plane perpendicular to the workpieces, the inclination being away from the general forward direction of travel of the probe.
12. A method according to any preceding Claim wherein the probe is mechanically propelled across the workpieces.
13. A method according to Claim 12 wherein the probe is propelled along a track.
14. A method according to Claim 13 wherein the probe is propelled along a geared track
15. A method substantially as herein described with reference to and as illustrated in any combination of the accompanying drawings.