Friction surfacing tomorrow - industrial applications of cladding

TWI Bulletin, November - December 1996

Dick Andrews, Principal Research Metallurgist, works in TWI's Electron Beam, Friction & Forge Processes Department. He is responsible for the progress of research programmes involving friction welding, friction processing, flash, MIAB and explosive welding.

Cladding by the solid phase friction surfacing technique has so far found limited application in industry, in part due to its batch status and relatively small surface area coverage capability.

Dick Andrews describes examples of cladding application and explains the process developments that will be necessary for the friction surfacing technique to achieve large scale industrial acceptance.

Surface engineering techniques used traditionally to combat wear and corrosion, and also for repair to damaged or worn components are being used increasingly to provide improved lifetime expectancy. In particular, the aerospace industry has used surfacing technologies to advantage for many years. Manufacture of value added products has enabled sufficient resources to be available for the development of a range of surfacing engineering techniques that have been successful in combating problems associated with the arduous service conditions experienced by aircraft engines and airframe components. Unfortunately many industries have not been able to adopt this type of approach.

The metal working industries who manufacture large numbers of low-cost components, compared with limited batches of expensive aerospace items, have been slower to adopt surface engineering. In fact, the range of techniques that is available has very often been considered only in an emergency capacity, and regarded as a short-term solution to a current problem. A major requirement that has been highlighted throughout the last 40 years, but only partly addressed, has been the development of methods to increase the lifetime expectancy of hot and cold work tooling. The benefits of reduced plant downtime, and improved productivity should provide lower costs.

A variety of surface engineering processes such as arc, laser, electron beam, resistance cladding and thermal spraying variants have all been used with varying degrees of success. Unfortunately some of these processes have been unfairly discredited. On occasion the tool user and the surfacing contractor have not appreciated that the tooling associated with every production process has a unique wear-couple with the manufactured component which requires a dedicated solution to reduce wear. It is possible that the cost of obtaining this solution might possibly outweigh potential cost savings. Existing tooling inadequacies are very often tolerated, rather than development of superior performance tooling, and are included in a component fabrication route.

The industrial need for cladding

Reviewing the hot working processes used for forgings, seamless tubing and extrusions has revealed considerable variation in the materials, and design of tooling used by operators, but their problems are invariably very similar. Hot work tooling differs greatly in size and application, and has to survive the most aggressive wear situations. Figure 1 shows a good example of hot work tooling known as the piercer plug, which is subjected to high forging temperatures and continuous abrasive wear during manufacture of seamless steel tubing by the rotary piercing process.

In this process large thermal and mechanical stresses are imposed simultaneously, and consequently the life of the plug is comparatively short, and the overall cost is high. Under ideal conditions the plugs will fail by gradual wear taking them outside a preset tolerance, but leaving them otherwise undamaged. The worn, undersize plugs are returned to the foundry to be recast to the required dimensions. Attempts have been made to improve the life of these plugs using available cladding techniques - the most effective method so far has been production of an adherent, high temperature, wear resistant oxide layer in a controlled atmosphere furnace. High temperature self-lubricating coatings used as bearings on space vehicles, and manufactured using a combination of thermal spraying and sintering, have been investigated as a potential replacement for the oxide layer. However, further expensive development is needed, so the oxide layer continues to be used.

This example of hot work tooling highlights the extreme conditions in which clad layers would have to operate. Therefore, in order to provide an increase in tooling life the choice of cladding method and material will need to be selected with care. Clad layers that are able to resist hot forging loads satisfactorily without deformation are likely to be brittle and fail by splitting or cracking, and it will not be possible to refurbish the tool. Alternatively clad layers which are tough enough to resist cracking are likely to be less hard, and will fail by deformation in the surface layers which will flow with each forge operation, and change the dimensions of the tooling. Other important factors that will influence tool life are time, lubrication and cooling which makes the choice of cladding method even more difficult.

Alternative cladding techniques - friction surfacing

It is important to note that the hot work tool wear scenario described is a typical example of so many applications where traditional and new surfacing technologies could and have been used to maximise tooling life. In particular a technique first recorded in 1941 ^[1] and reinvestigated in the 1980s by TWI called friction surfacing ^[2] could be used to clad new tools and repair worn tooling. Friction surfacing is a solid phase cladding technique that is a variant of the friction seam welding process. Simply, a rotating consumable rod is forced into contact under an axial pressure with the substrate plate to be clad. The rubbing action causes frictional heat to be generated such that the tip of the consumable rod becomes plasticised. Friction surfacing is accomplished by traversing the substrate plate underneath the rotating consumable, enabling transfer of a clad layer. Principles of the friction surfacing process are illustrated in Fig.2.

Research and development programmes ^[3] have established friction surfacing as an alternative cladding technique to the traditional methods. Deposition of clad layers is achieved in the solid phase with no melting occurring, or required, to achieve a metallurgical bond. There is no major dilution between the clad layer and substrate allowing dissimilar material combinations to be joined that are normally considered to be difficult using conventional techniques. As dilution can occur when using fusion techniques relatively thick layers have to be used to ensure that the exterior surface of the clad layer has suffered no physical property degradation. Also the thinner clad layers permissible by friction surfacing c an allow considerable cost savings where expensive cladding materials are necessary, such as the high temperature locations within an aero-engine that experience localised wear. Friction surfaced deposits are characterised by a fine grained hot forged microstructure.

Friction surfacing - current status

Process development continued in the UK, at TWI within the Core Research Programme funded by TWI members and the DTI, and also at Portsmouth Polytechnic, to understand fully and extend the application of friction surfacing. Industrial use of friction surfacing is limited to one company, FricTec Ltd (Portsmouth, UK). A purpose-built machine, shown in Fig.3, was manufactured for FricTec by Blacks Equipment Ltd (Doncaster, UK). One commercial application of this machine is to clad the cutting edges of guillotine and knife blades with various wear-resistant materials, in batch process mode. Friction surfacing is a batch process, due to flash generation from the rotating consumable rod, and the need to use short consumable rod lengths which limits application to relatively small surface areas and components.

Excessive consumable flash build-up limits the time available for surfacing, and also the area that can be clad. During the friction surfacing cycle the rotating consumable is supported by a bush located within a support steady (to prevent it from bending when subjected to side loading). The flash collar, which is generated continuously through the surfacing cycle, eventually impinges on to the bush, and prevents further deposition. Location of the steady, bush and the flash collar interaction is shown in Fig.4. These process deficiencies will limit the friction surfacing process to batch production, and hinder progress into full-scale industrial large area applications. Additionally, deposits produced by conventional rotary friction surfacing suffer from small non-bonded regions at the deposit edges.

Identification of essential future process developments

To make using friction surfacing more attractive in production, and elevate it to industrial status, TWI saw that a series of interlinked developments should be pursued. Rotary friction surfacing required a purpose-built continuous consumable rod feed system, complemented by a suitable method of either removing, or restraining the collar of undeposited material known as consumable flash, to assist with improved deposit quality and deposition rate.

Additionally, development of alternative friction surfacing motions from the current rotary method, were considered necessary to maximise the deposit bond area, and help to reduce or eliminate non-bonded regions at the edges of deposits, thus improving quality. Potential methods of addressing these problems were filed as patents by TWI ^[4] , Rolls-Royce ^[5] and AS&T. ^[6]

'Close' fitting bush

The consumable flash collar generated during a friction surfacing cycle impinges on the support steady illustrated in Fig.4. This flash is continually increasing in height and volume, and pushes the steady upwards, leaving it typically 50mm above the surface of a deposit at the end of a surfacing cycle. Obviously this is not the best situation, as an increasing length of consumable rod is exposed and also unsupported. Side loading on the consumable rod caused by the traversing motion of the substrate can cause bending or fracture, particularly when a long unsupported length is exposed at the end of the surfacing cycle.

Extensive modifications were made to one of TWI's hydraulically actuated friction welding machines enabling it to operate in the surfacing mode. Instead of the steady being permitted to travel freely upwards during a surfacing cycle it was possible to lock it at variable 'stand-off' positions above a substrate. A wear-resistant bush was mounted on to the steady as shown in Fig.5 Friction surfacing trials were carried out with varying stand-off distances, to constrain the hot softened flash, and force it to flow laterally. A range of bush materials such as cast iron, Stellite 12 and hot work tool steels were assessed with the result that nitrided H13 hot work tool steel proved to be the most wear resistant.

a) with (smooth texture);
b) without (rough texture) a 'close' fitting bush in 6082 aluminium alloy.

It was possible to use this bush with relatively ductile materials such as 6082 aluminium alloy, which resulted in the production of fine textured surface deposits of uniform width and thickness. This feature was combined with a deposit/substrate bond width increase of ~15%, effectively reducing the size of the defects at the edges of a deposit. A comparison of the surface of deposits made with and without a bush is shown in Fig.6.

Similar improvements were achieved when producing mild steel deposits again with surface improvement. Heavy wear of the working face of the bush was caused by flash impingement. Additional development will be required to identify improved materials and to optimise bush geometry.

Restraint 'shoe'

Further work to influence the shape and bond integrity of a deposit involved design and manufacture of a constraint 'shoe' shown in Fig.7. This shoe acted rather like the die in a hot extrusion press, because its purpose was to influence the shape of the hot softened friction surfaced deposit. However, to maintain, and use effectively, the finite amount of heat generated during a friction surfacing cycle, auxiliary cartridge heaters were fitted to the shoe, to enable preheat temperatures of up to 400 deg.C to be obtained. Despite these refinements to the basic shoe, seizure of metal occurred within the body when attempting to extrude softened flash from a 6082 aluminium alloy consumable rod. It proved impossible to prevent the combined bulk quench effect of the shoe body, mounting fixture, consumable rod and substrate from cancelling the finite amount of heat generated during the surfacing cycle. In further development of a shoe use of a suitable, ceramic or cermet material would help prevent excess loss of heat, and would also not suffer excessive wear. Trials indicated that a higher axial force than normally used for friction surfacing was necessary to generate additional consumable/substrate heating.

Flash cutting

In a large scale application, a continuous consumable rod feed system would need to be complemented by an appropriate method of removing the continuously forming flash collar. Four simple types of flash cutter were designed and built. Initially a cemented carbide parting lathe tool was positioned closely to the consumable flash. The tool was clamped securely to the support steady which is illustrated in Fig.4. Other variants of this type of cutter were built and tested, and an alternative cutter location is shown in Fig.5. Finally a hollow milling cutter with interchangeable cemented carbide teeth was devised such that the rotating consumable passed through the centre. An improved performance was achieved using this device, but it rapidly became clogged with swarf. When depositing aluminium alloys, or mild steel, the performance of these cutters was disappointing, with the flash being produced at a faster rate than the tools could cut it. These results showed that a more advanced flash cutting device would be needed, and a design study was carried out that suggested that the softened flash should first be knurled or deformed, and the resulting protrusions removed using a tungsten carbide cobalt chip cutting tool. Considerable effort will be needed with design and testing of the tool before transfer to large scale industrial application can be made.

Continuous rod feed systems

Concept designs were developed for the continuous feed of consumable rods by AS&T, Rolls-Royce and TWI. Some of these ideas were patented, and therefore drawings are not included. Each system involved multi-collets which alternately grip and release a consumable rod in a sequence that generates a forward motion. Due to the complexity of fabrication of these devices, prototype development will be essential, although there is no reason to doubt that they will work.

Motions

As an alternative to conventional rotary friction surfacing, and to improve deposit quality, work was concentrated on the design and manufacture of a device that could provide an orbital motion to a consumable rod, a nd be interfaced with TWI's friction welding machines. The lack of bond defect at the edges of a deposit produced when rotary friction surfacing, when compared with those areas of the substrate exposed to the centre of the consumable, is due to both velocity profile across the consumable rod, and the short time the deposit/substrate edges are exposed to the consumable. Additionally, the axial force profile decays from the centre of a consumable rod to its outside diameter, resulting in insufficient welding force at the edges o f a deposit to permit a solid phase weld being made. A more uniform velocity profile, heat generation, and material flow was thought to be achievable using orbital motion. The orbital motion device built at TWI is shown in Fig.8.

When tested it was discovered that TWI's friction welding machines were unable to develop sufficient surface velocity to test fully the orbital device. Excessive vibration occurred during trials, and no major improvement in coating quality was determined. A small multi-motion friction welding machine was designed and built, which in future developments will enable a more thorough assessment of the benefits of friction surfacing with orbital or linear motions.

Conclusions

Investigations into the necessary steps required to improve quality and deposition rate of friction surfacing for large-scale industrial applications indicated that considerable investment and development will be needed. However, each topic addressed by TWI provided benchmark information for the next stage.

An improvement in deposition rate brought about by adopting the topics previously described, will reduce material wastage, and increase bond width, resulting in cost savings of ~30% when compared with current technology.

Success with the continuous consumable rod feed system and flash cutting devices will increase commercial acceptance and use, particularly for large areas, of friction surfacing in the future.

References