Platinum Metals Rev., 1966, 10, (1), 2
Wear and Friction of the Platinum Metals
Performance in Sliding Contacts
The sliding properties of the noble metals to a large extent determine the performance of separable electrical contacts, slip rings, switches and other devices in which these metals are used. Recently, certain phenomena which control wear and friction have been discovered (1). This paper summarises the fundamentals of sliding of unlubricated platinum group metals, with emphasis on these phenomena.
The Prow Formation Mechanism of Sliding
In most practical cases, the members of a sliding contact system are not identical in shape and dimension. Furthermore, action is localised generally to a small area on one surface and spread out on the other contact. In such situations, the unlubricated sliding of palladium, platinum, gold, silver, copper, and many other metals is characterised by unsymmetrical transfer; virtually all metal moves initially from only one contact member to the other, and not in both directions. In a second step, the transferred metal is removed by back-transfer to the part which contributed it, or is lost as loose debris.
These processes are readily observed with a rider-on-flat apparatus such as is shown in Fig. 1. A hemispherically-ended rider is pressed against the flat specimen attached to a turntable. A flexure plate and ring provide elastic mounting for the rider. Strain gauges on the ring and an oscillographic recorder are used to measure friction force. Contact resistance during sliding is determined with a current-limiting circuit having an open circuit d.c. potential of a few millivolts. Resistance is recorded in the same way as is friction, or is observed with an oscilloscope. The specimens, slip ring, and associated wiring are electrically isolated from the rest of the machine. A microscope positioned close to the members enables details of sliding to be seen.
At the onset of rubbing a lump of metal (called a “prow”), which comes from the flat, forms between and separates the specimens, as shown in Fig. 2. Sliding now continues at the junction between prow and flat. The prow projects against the direction of sliding and gouges the flat. Ploughed solid becomes attached to the prow so that it grows in length.
It is not certain how the prow originates, and why it is always located on the rider. A plausible explanation is that this is related to the difference in size of rubbing areas of the members. The population density of transfer particles (formed by the breaking of cold-welded asperities of the specimens) is greater on the smaller part. These particles readily weld together and to the rider to create a prow, since they are clean and tarnish-free and because of the high compressive forces which act on them. The prow is considerably harder than either contact because of the extreme degree to which it is worked in its transfer and growth. Ploughing of the flat by the prow, then, is much like a hard file scoring a softer surface.
Various processes cause the prow to break from the rider to produce loose debris or adherent back-transfer solid. The most important of these is welding of the prow to the flat, and this, incidentally, is one of the ways in which surfaces become roughened during sliding. Loose particles are sometimes roller-shaped, produced by the kneading of the prow as it passes between the surfaces.
Prows form on the smaller specimen in any arrangement of sliding members. For example, with the crossed rod geometry of slip rings, in which one contact is stationary and the other rotating, a prow is created on the stationary member; it has the smaller wearing area, and corresponds to the rider in rider-flat systems. Even with the perfectly symmetrical geometry of axially-loaded cylinders having carefully aligned flat surfaces, a prow forms on either cylinder purely by chance. And once a prow appears, geometry becomes unsymmetrical, as with the rider-flat system.
Prow formation occurs regardless of speed or type of motion—unidirectional or reciprocating, and is unrelated to the passage of current through the contact (in work to date, current levels and conditions of sliding were such as not to entail appreciable generation of heat). It has been observed at loads in the rider-flat machine from a few grammes to one kilogram. Prow formation is, therefore, a characteristic of many metals, and does not depend much on the way they are slid, provided their surfaces are clean.
Table I lists for typical metals the hardnesses of their wear debris and of the flat from which the debris comes. It is evident that the hardness of prows is considerably greater than is possible to obtain with metal by the usual forming processes such as rolling and drawing.
|Metal||Unworn Flat (kg/mm2)||Wear Debris(1) (kg/mm2)|
This leads to the unusual situation of a “full-hard” flat wearing without simultaneously damaging an annealed rider of the same metal pressed against it; the super-hardness of transfer metal dominates sliding, with the rider serving only as a holder for the prow.
Dissimilar Metals in Contact
If dissimilar metals slide, prow formation still occurs, provided that the unworn flat is not excessively harder than the unworn rider. The flat provides prow-metal which then wears the flat in the usual way. When the prow is lost, the members come together and a new prow immediately forms.
In the cases of gold riders sliding on much harder palladium, platinum or rhodium flats, there was a tendency early in the runs for the riders to smear on to the flats. This tendency increased in going progressively to harder flat metals. Specimen hardnesses (kg/mm2) with a Knoop indenter at 250 g were: gold, 63; palladium, 123;platinum, 176; and rhodium, 504. With the palladium and platinum flats, the dual processes of (1) transfer of flat-metal directly to the gold rider (prow formation) and (2) back-transfer from the flat of smeared gold predominated after an initial period of smearing of rider-metal to flat. With the rhodium flat, there was negligible transfer to the gold rider, hence, no prow formation. Figs. 3 and 4 illustrate these details in sliding at 500 g on a one-inch diameter track. In Fig. 3, a gold-platinum system, gold transferred and back-transferred between rider and flat for 110 revolutions with growth of a leading edge on the rider. This growth was not a prow, for the rider and flat were continuously in contact. Eventually, the back-transfer gold became so worked, with resultant increase in its hardness, that it began to plough the flat. Prow formation ensued. The figure clearly reveals the projection of gold and a platinum prow underneath the rider (the photomicrograph does not show the initial platinum prow, but one formed in subsequent sliding). All wear debris was platinum.
In Fig. 4 are gold on rhodium specimens after 500 revolutions. The back-transferred gold (not a prow) projects ahead of the rider. There was no loss of gold by the system—only net transfer of it to rhodium. Apparently, gold cannot be severely enough worked to form a prow that is capable of ploughing rhodium. Rhodium itself, as pointed out later, does not wear by prow formation.
Rider Wear Mechanism
If sliding is repetitive on the same track the prow formation process eventually ceases and is replaced by “rider-wear”, a mechanism in which both rider and flat wear. This is due to the gradual accumulation of sufficient back-transfer prows on the flat to increase its hardness at all places to the level attainable by extreme work hardening. When the surface of the flat reaches the hardness of prows on the rider, ploughing of the flat ceases. A wear land appears on the rider and a groove grows in the flat. This transition occurs within 500 revolutions at 500 g on a one-inch diameter track with palladium, gold, silver and copper. Platinum requires lengthier sliding at these conditions for the transition. The greater the load, the more quickly will the transition come.
Rhodium and other very hard metals do not wear by prow formation. From the onset of sliding, both rhodium rider and rhodium flat lose material. A possible explanation for the inability of transfer particles of rhodium to adhere to each other and to the rider to form a prow is based on the theory of asperity-bridge breaking when elastic stresses are released on removal of load (2). The metal is so hard that the ratio of elastic to plastic deformation on loading is much greater than for softer metals such as palladium and platinum. Consequently, there will be a greater tendency for springing apart of compressed transfer particles than with particles of metals that readily form prows. Explanations for the absence of prow formation in certain base metal systems are based on the oxidation of transfer particles; oxide or oxide-covered particles have little tendency to accrete.
Friction and Contact Resistance
Friction and contact resistance largely depend on the metal which constitutes the prow-flat interface, the degree to which interface metal has been worked, on the presence of loose debris and its shape, and on roughness of the flat. The Fig. 5 curves, showing kinetic friction coefficient versus sliding distance for various combinations of gold and palladium specimens, illustrate some of these points. First, sliding friction is determined by the metal of the larger surface. A gold rider on a palladium flat slides with the same friction as an all-palladium system. In both cases a palladium prow is formed, and friction is that of palladium. Conversely, the palladium rider-gold flat combination behaves identically to the gold-gold system. Secondly, the curves in Fig. 5 have two distinct regions, one of rising and one of falling friction. The rising region corresponds to the part of the run when the track is rapidly widening. Here, friction is largely due to ploughing, which is proportional to the cube of track width (3). In the falling part of the curve, the influence of progressive prow-transfer to the flat is seen. The ploughing contribution to friction decreases as the flat hardens by progressive prow-transfer to the flat.
Examination of friction behaviour over short time spans, from microseconds to seconds, reveals other expected phenomena. The amplitude of stick-slip swings in friction increases as the track roughens, owing to elasticity of the apparatus and to interference with motion by the back-transfer metal. Roller-shaped wear debris causes a momentary drop in friction, by a factor of 10 to 20, as the particles move like miniature bearings between the surfaces.
Table II lists the average coefficients of kinetic friction (μk) of palladium, platinum, rhodium and several other metals at 500 g after 400 to 500 revolutions on a one-inch diameter track, together with wear measurements after 500 revolutions.
|Metal and hardness (kg/mm2 at 250 g)||Weight changes (1) (mg)||Volume wear rider plus flat (cm3/cm/g)||μk, avg. 400–500 revs|
|Copper (136)||Copper (79)||−2.4||−14.6||0.96 · 10−9||1.14|
|Gold (63)||Gold (68)||−2.3||−32.7||0.91 · 10−9||1.87|
|Palladium (139)||Palladium (123)||−1.1||−18.0||0.80 · 10−9||1.06|
|Platinum (162)||Platinum (176)||−0.2||−33.8||0.79 · 10−9||0.74|
|Rhodium (371)||Rhodium (504)||−0.5||−0.2||0.03 · 10−9||0.39|
|Silver (81)||Silver (58)|
|a2||+0.3||− 2.0||0.40 · 10−9||1.18|
|b||−2.8||− 0.5||0.14 · 10−9||1.09|
|Silver (81)||Silver (58)||−2.5||− 2.8||0.25 · 10−9||1.04|
|Gold (63)||Platinum (176)||+0.6||−19.0||0.43 · 10−9||0.81|
|Gold (63)||Rhodium (504)||−0.8||+ 0.8||0||0.95|
|Rhodium (504)||Gold (68)||+0.5||−13.6||0.34 · 10−9||1.49|
Contact resistance is related to prow formation and its associated phenomena, just as is friction. Variations in contact resistance are caused by stick-slip, roller debris formation, work hardening and, with dissimilar contact metals, changing composition of the sliding interface. For example, progressive work hardening during sliding increases contact resistance because the area of contact diminishes at constant load; also, worked metal has a slightly greater bulk resistivity than unworked metal. In runs with gold riders on palladium flats, contact resistance at 100 g rose from about 15 milliohms at the start to about 30 milliohms after 500 revolutions on a one-inch diameter track. This contact resistance is approximately the same as was obtained with an all-palladium system after a similar period of sliding. In stick-slip motion, electrical opens often occur on breakaway when there is momentary separation of rider and flat. Roller debris between the contacts approximately doubles contact resistance; rollers create two constriction resistances in series.
Some Practical Consequences of Prow Formation
All but the hardest noble metals and their alloys have a pronounced tendency to slide by prow formation, because they are not normally covered by oxide or other inorganic films. Consequently, their transfer particles tend to accrete to form prows. Numerous lubricants, effective on base metals, did not prevent prow formation in experiments with a typical noble metal (4).
The contacts in many electrical devices, such as the separable connector used in electronic equipment, do not slide large distances during their lifetimes. Only prow formation is then operative; the transition to rider-wear is not attained. Since prow formation involves preferential wear of one member, the non-wearing member need not be made with as much noble metal as the wearing contact.
The growth of prows requires that the specimens be free to move apart in order to accommodate them. If there is a constraint on the members, as with spring-loaded contacts, the actual load will increase during prow formation, and with this the transfer and wear rates become even greater. If clearances are small, as in sleeve bearing geometry, prows can produce gross seizure of the device.
aM. Antler, “Wear, Friction and Electrical Noise Phenomena at Severe Sliding Conditions”, A.S.L.E. Trans., 1962, 5, 297 – 307 ;
bM. Antler, “Processes of Metal Transfer and Wear”, Wear, 1964, 7, 181 – 203 ;
cM. Cocks, “Interaction of Sliding Metal Surfaces”, J. Appl. Phys., 1962, 33, 2152 – 2161
- 2J. S. MacFarlane and D. Tabor, “Adhesion of Solids and the Effect of Surface Films”, Proc. Roy. Soc., 1950, A202, 224, 244
- 3F. P. Bowden and D. Tabor, “The Friction and Lubrication of Solids”, Rev. ed ., 1954, Chap. 5; Oxford, Clarendon Press
- 4M. Antler, “The Lubrication of Gold”, Wear, 1963, 6, 44 – 65