MMC deposits from powder mixtures by friction surfacing

TWI Bulletin, January/February 1993

David Abson gained his BA degree in Natural Sciences at Cambridge University in 1964, and subsequently obtained his master's degree and PhD at Sheffield University. After working in Montreal, Dayton (Ohio), and the University of Wales in Swansea, he joined TWI in 1976. He is a Principal Research Metallurgist in the Materials Department. Work at TWI has included carrying out many failure investigations. Research activity has covered investigations of the relationship between microstructure and toughness in ferritic weld metal, non-metallic inclusions and their influence on microstructural development, and hydrogen-induced cracking. More recent research interests have included deposition of hardfacing, and wear testing.

A new development in friction surfacing means metal matrix composites can now be produced from powder contained in a tube, as David Abson reports.

Friction surfacing, using a consumable in the form of a solid bar, is now well established as a surfacing technique. Among the materials which can be deposited by this process are tool steels, stainless steels and hardfacing materials. The process involves rotating a consumable bar, with one extremity pressed hard against a plate, disk, or cylinder of the substrate material; heat is generated at the tip of the consumable, producing a plasticised layer. Lateral movement of the substrate, relative to the rotating consumable, deposits this softened material on to the substrate ( Fig.1). ^[1] There is no melting of the substrate material, and therefore no dilution of the substrate into the deposit. One layer is generally sufficient to achieve the required deposit composition.

The process has been extended to deposit metal matrix composites (MMCs) by inserting hard particles into one or more holes or slots machined within the consumable bar. The material of the bar therefore becomes the matrix, with the hard particles distributed throughout it. ^[2]

An additional development is the production of MMCs, from powder contained within a metal or alloy tube. ^[3] The powder may be simply a metal or alloy powder, or mixtures of hard particles and metal or alloy powders. Sufficient heat is generated during processing that the powder particles weld to one another, and a sound MMC deposit is produced.

Consumable

The consumable consists of a tube, typically 20mm diameter, which is filled with powder and sealed at one end with a plug and at the other with a bar which is held in the chuck of a converted milling machine; the bar and the end plug are joined to the tube by TIG welding. The powder is introduced in quantities which are sufficient to add approximately 20mm to the powder depth, and each increment is packed down before more powder is added. The configuration of a typical consumable is shown in Fig.2.

Processing conditions

Typical rotation speeds are in the range 345-690rpm, feed rates 0.4-1.0 mm/sec and traverse rates 1.0-4.0 mm/sec. The rotating consumable bar is lowered until it comes into contact with the substrate; the end plug is allowed to soften and form a plasticised layer, before lateral movement of the substrate is started.

Deposits

Initial trials were carried out on 304L (austenitic) stainless steel powder, and mixtures of this powder with chromium boride. ^[3] One of the early deposits, produced from 304L stainless steel powder with no chromium boride added, is shown in Fig.3, and a micrograph of one of the chromium boride plus stainless steel deposits is shown in Fig.4.

In this early work, the maximum volume fraction of chromium boride achieved was approximately 10%. An unexpected finding was that the containment tube was not deposited, but rather it softened and folded back, forming a flash.

Subsequent studies have concentrated on an aluminium alloy, with silicon carbide as the hard particle, contained within an aluminium alloy tube. A powder comprising a mixture, in approximately equal proportions, of 2618 aluminium alloy and finely dispersed silicon carbide (with a particle size of less than 35µm) is available as a by product of the production of MMCs.

This powder mixture was used as supplied, and was also diluted by mixing with further 2618 aluminium alloy powder, giving approximate silicon carbide proportions of 25, 30, 40 and 50%.

Sound deposits have been produced from each of these, and from the 2618 alloy powder alone. Some of these deposits are illustrated in Fig.5-7. It is clear that uniform dispersions of SiC have been achieved in this system. Also, there was apparently no Al ₄C ₃ formed at the SiC/aluminium alloy interfaces, as would have been expected if the matrix material had been subjected to a fusion process.

The surface appearance of the friction surfacing deposits is not always as smooth as desired. However, this does not necessarily detract from the merits of the deposits, and it is likely that improved process control and optimisation of processing parameters will lead to improvements in surface appearance.

As always with this process, there is poor adhesion near the edges of the deposit, and the edges themselves are uneven. As the remainder of the interface is free from defects, it is confidently expected that a deposit laid down in a groove of appropriate configuration and width will have a sound interface across the whole width of the deposit.

Developments

The process is considered to have potential for exploitation, and so a patent application has been filed. ^[4] Among likely applications of the process are joining, i.e. butt welding, of MMCs, deposition of long life cutting edges and surfaces, and deposition of wear-resistant inserts in engineering components such as pistons. The process has the potential to be easily adapted to deposit a wide range of wear-resistant and other materials, and it may also be used for producing thin layers of MMCs.

The low thermal conductivity of aggregates of individual particles will probably allow deposition of powders or powder mixtures of materials which cannot be deposited readily by other means. Further investigation is planned to deposit such materials, to produce multiple layer and multiple-pass deposits, to explore the processing conditions required for filling grooves, and to make the process a continuous one.

Readers with a potential application for this new technology are invited to contact David Abson or Philip Threadgill in the Materials Department or Wayne Thomas or Dave Nicholas in the Forge and Resistance Processes Department.

References