Table of contents

20 µm diameter spray-dried agglomerates made up of 210 nm zirconia particles have been individually tested by crushing between rigid plates using a modified nanoindenter. The results show that the agglomerates support considerable stress before deforming. Both the elastic and plastic deformations of the agglomerates have been observed, followed by ultimate failure through a cracking mechanism. A model has been proposed to account for the results based on van der Waals adhesion between particles within the agglomerate. This model has been realized in a macroscopic agglomerate of smooth adhesive rubber beads. The test method could distinguish variations in the properties of individual agglomerates in a batch, and could detect differences between batches made under different spray-drier conditions. Powders with weaker agglomerates were shown to give improved ceramics as a result of easier particle dispersion during mixing.

The aim of this work is to understand better the initial stage of sliding friction between a rigid asperity and an adhesive rubber surface which induces the formation of the sliding annulus, as first reported by Mindlin in 1949. Experiments were carried out by contacting a rigid axisymmetric punch with a rubber sheet submitted to an increasing biaxial extension. For three flat or spherical punches and various extension rates, we observe similar behaviour: above a critical strain, the contact area decreases by a peeling mechanism and depends mainly on the instantaneous strain. Theoretical analysis of the superficial state of stress provides that separation occurs under `mixed-mode' loading and verifies a criterion K_I + 0.15K_II = K_c with K_I, K_II the stress intensity factors and K_c a toughness not very sensitive to rupture speed.

The problems of mould release and mould fouling in the rubber industry have prompted research into the adhesion of rubber to steel mould materials. In this paper the interface formed by moulding nitrile rubber (NBR) against medium carbon steel 080M40 is examined with the aim of better understanding the factors which influence the adhesion. The level of adhesion between the steel and four test compounds was measured using the TMS rheometer. The parted surfaces were then examined using XPS. The results revealed the presence of a complex interlayer formed between NBR and the steel during moulding. The layer consists of residues some derived from the compounding ingredients and others, such as coagulants and emulsifiers, from the base polymer. It was found that the nature of the interlayer exerts a strong influence on the bonding between the two surfaces, and in particular that residues of different fatty acids affect the adhesion differently depending on their solubility in the rubber compound. The work concludes that the adhesion between NBR and medium carbon steel depends upon both the compound chosen and the polymer grade used.

This paper describes the results of five types of experiment where a poly(urethane) is adhered to a metal substrate. The adhesion has been quantified by using the `blister test' and data are given for the interface peeling energy at the same crack velocities to account for the pronounced viscoelastic character of the crack propagation process. The five types of interface studied were prepared as follows using a number of topographically characterized metal substrates. The polymer was cured against these substrates and the fracture energy measured (type 5). The experimental method allowed the interface to be re-formed without a major geometric relocation of the original juxtaposition of the prepared interface. These interfaces were ruptured and the peeling energy quantified (type 1). Alternatively, the interface was re-formed at a position different to the original juxtaposition and peeled apart (type 2). The third and fourth configuration (types 3 and 4) involved the creation and rupture of interfaces of differing combined topographies which had been re-formed following rupture. There are two extremes: a relatively smooth metal in contact with elastomers of a range of roughness and vice versa. The experimental data from the five cases show characteristically different trends and the origins of these trends are considered. The type 3 data are found to follow the predictions proposed and established elsewhere based upon a model of adhesion which incorporates surface free energy and stored elastic strain at asperity contacts. The other types of data indicate that, for these systems at least, the introduction of roughness reduces the strength of the interfaces.

The frictional properties of rubber glove surfaces modified by various treatments have been investigated in order to understand the factors which reduce the friction and give the surface appropriate qualities for practical use. Factors of importance in the reduction of dry friction were found to be an increased surface roughness and hardness. Under damp conditions the friction can increase above the dry value and here it was found useful if the treatment rendered the surface capable of quickly absorbing water. For surgeons' gloves in contact with skin a particular inner-lining treatment, a hydrogel polymer coating, gave the necessary surface roughness, hardness and water absorptivity so making the gloves easy to don and pleasant to wear. On long wearing, the hydrogel coating gradually softened as water was absorbed, the coating then acting as a `second skin'.

In this paper some recent studies on the friction of diamond are reported. These include a study of friction tracks by reflection electron microscopy (REM); analysis of wear debris using electron and x-ray diffraction, and infrared spectroscopy; friction experiments with various lubricants. These researches showed that adhesion was an important factor in the friction of diamond on diamond, that water reduces the friction of diamond on diamond, and that plastic deformation could be involved in the friction at high loads.

The nature of the `third-body layer' formed at the interface between the two elements of a friction couple in an automotive braking system has been the subject of much speculation. The development of the imaging x-ray photoelectron spectrometer is now allowing detailed studies of this layer. This investigation has examined the influence of temperature on the composition of transfer films (third-body layers) formed on grey cast iron after rubbing against asbestos-free friction materials in a friction assessment screening test (FAST). Evidence will be presented to show the strong dependence between transfer film composition and the coefficient of friction over a range of temperatures. The information obtained in this investigation may provide a partial solution to the problem of `early morning sharpness' and also poor cold friction coefficient, the latter is often referred to in the industry as `low cold mu'. These slight changes in friction coefficient are occasionally encountered by some drivers whose vehicles have a particular cast iron-friction material combination.

Recent experimental measurements using twin-axis load cells in storage vessels for bulk solids have revealed load-dependent variation of the wall friction angles measured in the presence of small normal loads on smooth silo walls. The effect is quite appreciable with smooth particles such as plastic beads and glass ballotini, but particles with rough surfaces, like some agricultural seeds, show no such effect.

The inclusion of an intrinsic shear resistance at zero normal load into the frictional behaviour (after Briscoe and Tabor) provides a satisfactory prediction of tangential forces required to cause gross sliding of smooth particles over smooth silo walls. This work applies the same adhesive friction concepts to the microdisplacements that occur before full tangential sliding develops.

Mindlin analysed the traction profiles for two spheres in contact assuming a constant coefficient of friction, independent of normal stress inside the contact area, with microslip occurring in those portions of the contact area that had a tangential-to-normal stress ratio exceeding the friction coefficient. The assumptions of that analysis preclude the existence of the observed load dependence of wall friction angles. In this paper we re-examine the micromechanical nature of contacts between elastic spheres incorporating tangential stress limits that contain two terms: one proportional to the normal stress at each point in the contact area, the other independent of normal stress. Two possible physical (and mathematical) interpretations of the stress-independent term are discussed, one called adhesive slip and the other stick slip. The resulting traction profiles are examined to determine implications for both microslip and gross-sliding frictional behaviour. A complete theoretical analysis requires the solution of the surface displacement integral over the contact region for a chosen form of the tangential traction profile satisfying appropriate boundary conditions for elastic displacements. Upper and lower bounds and an approximate form for the resulting tangential force-displacement curves are developed and it is shown that such models result in a load-dependent friction behaviour that is very similar to recent experimental wall friction and individual particle friction measurements.

A point-contact microscope (PCM) with a very sharp diamond tip works with loads as light as 10 nN, which allows high-resolution topographies to be obtained. This PCM can also be used to perform indentation, scratch and wear mark tests. Nanometre-level wear mark tests were performed on polycarbonate and epoxy surfaces using loads ranging from submicronewtons to micronewtons. During these tests, the scanning-scratched surfaces of polycarbonate formed projections, and no depressions or wear particles were observed. This phenomenon is novel compared with the depressed wear marks formed on the scratched surfaces of ordinary materials such as metals or ceramics. The projections on the surface involve an increase in volume and they are softer than the non-scratched surface. It seems that this distortion was caused by the frictional force. In contrast, an epoxy surface did not show any growth of projections, and wear particles were produced from the surface.

A range of ultra-high molecular weight polyethylenes have been subjected to water-lubricated, reciprocating sliding wear against a stainless steel counterface under a pressure of 10 MPa at a frequency of 2.5 Hz and a maximum speed of 0.4 m s^-1. Transfer of material to the metal counterface during sliding wear involves interlamellar shear of the polymer and results in the development of a highly oriented transfer film. The deformed surface layers of the polymer and the resultant wear debris have been examined using both optical and scanning electron microscopy, differential scanning calorimetry, infrared spectroscopy and x-ray diffraction. Significant differences have been found in the degree of crystallinity, crystallite size and orientation in the deformed surface layers of the polymer and debris compared with those of the bulk polymer. The worn surface of the polymer shows slightly increased crystallinity but the crystallinity of the debris is much higher than that of the bulk whilst the crystallite size is much reduced.

This paper examines how the durability and friction of thin films of molybdenum disulphide are affected by the surface roughness of the substrate to which they are applied. Three types of substrate, representing a variety of bearing materials (bearing steel, titanium alloy and hot-pressed silicon nitride), were prepared with surface roughnesses of nominally 0.04, 0.1, 0.2 and 0.4 µm (centre line average values). Onto these substrates were deposited 1 µm thick films of sputtered MoS₂. Pin-on-disc tests undertaken in high vacuum indicated that both film friction and film durability varied appreciably with substrate surface roughness. The most durable films were those applied to silicon nitride substrates. With this film/substrate combination, the onset of film failure was more gradual than that seen with the films applied to metal substrates, and the MoS₂ film was observed to recover its low friction properties several times before complete failure. Results from these experiments are presented and a theoretical model, based on the numerical simulation of friction between contacting rough surfaces, is introduced and is used to explain qualitatively the variation in MoS₂ film friction with the surface roughness of metallic substrates. Possible reasons for the poor agreement between the model and those results obtained with silicon nitride substrates are discussed.

Very thin films of perfluoroalkylpolyether (Rf) derivatives having benzene rings on carbon surfaces were evaluated using scanning tunnelling (STM) and frictional force (FFM) microscopes. Two kinds of surface were used as substrates: one was highly oriented, pyrolytic graphite (HOPG) as a defect-free and inactive surface, and the other was sputtered carbon as an active surface having many defects. STM was used to observe the configuration and mobility of lubricant molecules and clusters on the HOPG. The STM images reveal that the lubricant molecules in clusters are not fixed but can easily change positions. FFM was used to evaluate frictional force characteristics on the non-lubricated and lubricated sputtered carbon films. On sputtered carbon film even the lubricants that had insufficient stability on HOPG proved to be stable and able to decrease the friction. The adsorption mechanism of a lubricant on a practical sliding surface that is active and has many defects, such as sputtered carbon, is different from that on a defect-free and inactive surface such as HOPG.

An apparatus has been developed which uses a laser to measure the thickness of boundary films in lubricated rubber contacts. In earlier measurements using the same interferometric principle the active optical component was a two-piece beamsplitter. This has been replaced by a simple prism. The behaviour of surfactant solution boundary films in rubber/glass contact has been re-examined using this apparatus, and films in rubber/rubber contact have been explored in detail for the first time. Stable films in static contact have been recorded at thicknesses up to 25 nm at an average pressure of 10 kPa. At higher average pressures, up to 50 kPa, the collapse of these films has been observed, which brings into question the role of electric double layers in supporting the normal load. Introducing a surfactant solution to the region around a dry, static, loaded rubber/glass contact results in the separation of the surfaces and the formation of an equilibrium film. The thickness of the film formed is the same as by squeezing rubber and glass surfaces together under the same load when immersed in the surfactant solution. This observation indicates the presence of surface repulsive forces. The improved sensitivity of the new apparatus was demonstrated during the course of experiments. In earlier measurements it was difficult to resolve the thickness of films thinner than 4 nm. It appears that films can now be resolved down to 1 nm, which is of the molecular dimensions of surfactants.

Despite the fact that many tribological components are designed to operate with a comparatively thick film of lubricant, bearing surfaces often still deteriorate with time as hard particulate contaminants are swept through the bearing gap. These particles may arise from the external environment or be wear debris from other pairs of surfaces lubricated by the same fluid. In order to investigate this phenomenon experimentally it is necessary to develop a predictable hydrodynamic film between the test surfaces which can be contaminated by small volumes of carefully graded particles. Here we describe the use of a foil bearing to generate such films between 10 and 50 µm thick to which contaminants such as finely divided diamond can be added. As the ratio of the characteristic particle size to film thickness is varied not only do the wear rates of the solid surfaces change but examination of the wear tracks suggests that very different mechanisms of material loss come into operation. When the size ratio is low the worn surface of the cylinder is covered by what appear to be small erosion pits; these display virtually no alignment in the direction of relative sliding and it appears that the particles tumble and roll freely through the gap. Above some critical value of the ratio the appearance of the surface changes dramatically to a grooved or micro-machined surface with all the grooves aligned in the sliding direction. A relatively simple theoretical model is developed, based on what happens to a typical particle, which goes some way to explaining these observations. As well as being consistent with the observed transition from `tumbling' to `grooving', the model can also explain why increasing the hardness differential between the hard and the soft surfaces does not always lead to a reduction in the damage to the harder member of the pair.

The interdisciplinary nature of the subject of tribology has been frequently emphasized during the last quarter of a century. An interested observer might reasonably question, however, the strength of the link between fluid-film lubrication at the one end of the tribological spectrum and dry wear/surface phenomena at the other. The foundations of fundamental understanding of the former area were laid over one hundred years ago whilst one suspects that with regard to the latter fields current knowledge represents but the tip of the iceberg. In recent years there has been wide review of fluid-film lubrication and of those who gave the topic substance. A repetition would serve no particular purpose. Rather it is proposed to consider the subject in the context of the thin lubricating films to be found in the internal combustion engine which powers our motor vehicles. This engine can be described as a device to generate lubricant films of the order of a micrometre in thickness. A consideration of the current understanding of the tribological machine elements involved from the perspective of fluid-film lubrication is revealing, both with regard to the development of this subject area and its growing links to other tribological disciplines reflecting surface-contact phenomena.

Steady load performance data are presented for a sterntube bearing lined with an elastic material, which is contained in a rigid housing and subject to angular misalignment. The analysis is thus applicable to oil-lubricated sterntube bearings in which reinforced resin liners are used as an alternative to conventional white metal lining. The hydrodynamic analysis method used satisfies flow continuity throughout the cavitation zone. A rigorous elastic model, accounting for inter-element and element-to-housing forces was developed. Results showed that a simplified approach neglecting element interaction was of acceptable accuracy. An effective oil viscosity based on a simple thermal balance calculation was used. A relaxation method was employed for the combined hydrodynamic and elastic deflection solutions, with over-relaxation applied to the former. The application of under-relaxation to the elastic deflection solution was necessary in order to obtain convergence. The most significant results are those for minimum film thickness which is shown to be reduced by liner elasticity in all cases. This reduction is shown to be a direct function of both specific bearing pressure and misalignment angle. The rigid bearing assumption is confirmed to be acceptable for white-metal-lined sterntube bearings.

Described in this paper is a new approach to the analysis of the performance of a tilting-pad journal bearing under static load. This method is based on the fact that the oil-film thickness on a pad can be expressed as a linear combination of the journal centre coordinates and the tilting angle of the pad. The coefficients in these linear combinations are known functions of the bearing geometry, whereas the journal centre coordinates and the tilting angles of all the pads together form a complete set of unknowns for the bearing system. Once these are found, useful indicators of bearing performance such as power loss, journal eccentricity and minimum film thickness can be easily deduced. In a similar way the oil-film pressure on each pad can be expressed as a linear combination of these unknowns with coefficients that are obtainable from a Fourier series solution of the Reynolds equation. The oil-film temperature on each pad is allowed to vary circumferentially, and the oil viscosity is assumed to be a simple function of temperature derived from empirical data. By integrating the normal and shear stress functions over the surface of a pad the oil-film force can be obtained as a linear combination of the set of unknowns also. Then, by applying the equilibrium conditions for the bearing system, a complete set of algebraic equations is established from which the unknowns can be uniquely determined. As the Reynolds equation is temperature dependent and the oil-film temperature is not known initially, an iterative scheme must therefore be devised. In the present approach, however, this is the only iteration that is needed, representing a great improvement to existing schemes which contain multiply nested iteration loops. A computer program has been written and tested for internal consistency. Comparisons with published work, both theoretical and experimental, have been made and are found to be satisfactory.

It is now recognized that micro-elastohydrodynamic lubrication contributes substantially to the formation of effective lubricating films in synovial joints. In this paper we examine the influence of a single blemish on the bearing surfaces upon film formulation. We also investigate the influence of roughness, amplitude, wavelength and phase angle upon minimum film thickness and report that whereas phase angle has little effect, large amplitude and small wavelength roughness leads to smaller film thicknesses. The overall conclusion is, nevertheless, that micro-elastohydrodynamic lubrication is an exceedingly effective mechanism of lubrication under all the circumstances examined. The influence of conjunction kinematics, ranging from pure rolling to the translation of either the rough or smooth surface, is also considered.

The ability of grease to generate elastohydrodynamic (EHD) films has been studied and interpreted in terms of the response of grease structure to contact conditions. For most liquid lubricants, EHD film thickness depends upon the rheological properties of the fluid in the contact inlet. This mechanism is reasonably well understood for fluids that are Newtonian in behaviour. However, greases have complex bulk structures, the scale of which is larger than the dimensions of a contact inlet. The EHD performance of grease must therefore depend upon the response of grease structure to the high-shear inlet conditions. In this study, EHD film thickness has been measured for a range of greases using optical interferometry. In parallel work, grease structural changes have been monitored directly in the inlet of EHD contacts using reflection-absorption infrared spectroscopy. By using these two approaches in combination, changes in composition and loss of bulk grease structure have been observed and correlated with the ability of greases to form films under EHD conditions.

A generalized analysis for the squeeze-film lubrication of a rigid ellipsoid approaching a plane is presented for both isoviscous and piezoviscous lubricants. It is shown that the relationships for load and time of approach reduce to well known expressions when a sphere lubricated by an isoviscous fluid is considered. It is shown that for a given effective radius a sphere sinks more quickly through the lubricant than an ellipsoid. The rate of sinkage is retarded significantly by piezoviscous behaviour of the lubricant and theoretically infinite pressures are readily attained. A truncated series solution is developed to yield expressions for squeeze-film velocities and times of approach in piezoviscous lubricants. The results indicate that squeeze-film times are unlikely to be significant when spherical wear specimens are initially placed upon counterfaces, in the presence of lubricants, prior to sliding wear or boundary lubrication experiments, whereas it may be necessary to consider them in any complete analysis of the skidding phenomenon in ball and roller bearings.

Modern technology places increasing demands on equipment performance via a combination of design, cost and performance parameters such as miniaturization, a long life, and maintenance-free units which are sealed for life and can withstand extreme temperatures and have zero failures. At the leading edge of design in advanced technology such as magnetic data storage, semiconductor processing, aerospace, vacuum, electronics and automotives, problems can arise because of lubricant degradation when conventional fluids are at the limit of their performance. Evaporation, polymerization, carbonization, etc of only a small volume of oil may be the cause of a catastrophic and expensive failure of critical equipment. The identification of lubricants and functional fluids likely to be available by the turn of the century is probably the first stage in the design process. Work in Europe and the USA suggests that perfluoropolyethers and their derivatives, whilst currently falling short of anticipated requirements, represent today the best starting point for the development of advanced liquid lubricants and lubricant additives. This paper will cover this expanding family of lubricants and greases and the wide combination of high-performance properties they offer.

Rolling-bearing fatigue tests showed that under marginal film lubrication conditions (lambda, lambda = 1.2), a commercially available sulphur-phosphorus extreme-pressure (EP) additive package used in blended lubricating oil reduced the bearing fatigue life by at least a factor of 4.5 compared with the base oil alone. The Weibull slope was found to be significantly higher than normally observed for material tests made under identical conditions. Post-test examination showed a remarkable difference of topography of the raceway contact surfaces between those run with the same oil with or without the EP additive package. It is suggested that the reduction in bearing life is primarily due to the chemical reactivity of the EP additives in blended oil. The action of lubricant oxidation by the addition of EP additives is also discussed.

The lubricating effect of oils and lubricants has been investigated in local contact of bearing systems using models and real units under different types of non-stationary conditions taking into account their dynamic, kinematic and temperature characteristics. The results obtained show that the antiwear and antiscuff resistance of bearing systems under non-stationary conditions is explained by the discovered mechanism of lubrication by oils and lubricants through complicated physico-chemical processes of transformations of friction surfaces.

Abrasive wear is likely to occur whenever a hard asperity or a trapped hard particle is dragged across a softer surface, and it has been estimated that this form of wear contributes to as many as half of the wear problems that are met in industry. Such damaging hard particles may be external contaminants, products of corrosion or even the debris from previous wear events. During the life of a component, damage caused by individual asperity or particle interactions builds up and, at each stage of its life, the worn surface is the result of many such superimposed wear events. The practical, quantitative prediction of wear rates depends on having both a satisfactory understanding of individual interactions and a suitable procedure for combining these when subsequent contacts are made on a surface whose topography and material properties may have been much changed from their initial states. The paper includes some details of an analytical model for the interaction of a representative asperity and the worn surface which can both predict the frictional force and the balance between ploughing, when material is displaced but not lost from the surface, and micromachining or cutting, when actual detachment occurs. Experiments to investigate the validity of the model have been carried out on a novel wear rig which provides very precise control over the position of the asperity and the counterface. This facility, together with that of on-board profilometry, means that it is possible to carry out wear experiments on areas of the surface whose previous deformation history is well known; in this way it is possible to follow the development of a worn surface in a controlled manner as the damage from individual wear events accumulates. Experimental data on the development of such a surface, produced by repeated parallel abrasion, are compared with the predictions of the model.

In this paper, the frictional temperature field and its relationship to the transition of wear mechanisms of steel 52100 during dry sliding were studied by wear testing, SEM observations of worn surfaces and particles. Computer simulation of the temperature field in a surface layer during sliding wear was based on a mathematical model of frictional temperature field which itself was based on the law of heat conduction. The results of computer simulation for thermometric data, which were real-time recorded by a thermal video system, show that there was an ideal fit between the measured data and calculated curve. On the basis of analysing the relationship between the transition from mild wear to severe wear and the temperature field in surface layers during sliding of the steel, a new concept is put forward that the lower heat effect corresponds to mild wear and the higher heat effect corresponds to severe wear. The value of T_b corresponding to the transition of wear mechanisms of steel 52100 is about 200 °C.

The combined effects of erosion and corrosion can lead to a number of material wastage scenarios in industrial environments. In some cases, for example at high temperatures in fluidized bed environments, the wear may be corrosion dominated where chipping of an oxide scale is the predominant process, while in more aggressive environments such as gas turbine conditions, formation of a protective scale may never occur. Thus, pre-oxidation to form a protective scale or coating of a substrate may be effective in reducing material loss in corrosion-dominated environments since particle energies may be low enough to prevent significant deformation of the underlying surfaces. A computer model has been developed to simulate the erosion of a preformed scale or a coating on a metal substrate. This model uses computer graphics to demonstrate the sequential stages of erosion-corrosion. By incorporating established erosion and corrosion algorithms to calculate the wastage rate, the surface topography and extent of wear can be shown after exposure to erosion-corrosion conditions of varying severity. Thus, one can extrapolate from this technique whether erosion of oxide or metal is likely to be dominant wastage mechanism. This model was developed to describe the erosion-corrosion of a preformed oxide scale on a metal and comparisons are made between the erosion results for pre-oxidized surfaces and the graphics after computer simulation of erosion-corrosion. However, since the factors which determine adhesion of a preformed scale may be similar in many respects to those of a coating, the model may equally describe the erosion-corrosion behaviour of a coated metal. Hence, the paper shows the stages of the model development and compares the experimental results with those of the model.

So-called adhesive wear is caused by microscopic fractures on sliding surfaces. An approach is described in which the wear process is broken down into several elemental processes, i.e. formation of real contact points, action of forces on real contact points, distribution of stresses and strains, accumulation of damage in the subsurface, initiation and propagation of cracks, removal of particles leading to wear and modification of microgeometry so as to form a feedback loop. This paper discusses updated theories on selected elemental processes.

In the sliding wear of reaction-bonded silicon nitride (RBSN) a glassy phase was formed at the wear interface when the applied stress was below a threshold value. Under the heat generated at the contact between the two surfaces the glass liquefied and acted as a lubricant so that wear and friction were reduced dramatically below the threshold value. When the load was increased beyond the threshold value the wear and friction increased dramatically, with microfracture of the surface the dominant form of wear.

An investigation has been carried out into the effects of thin titanium nitride coatings, produced by several PVD and CVD processes, on the sliding friction and wear behaviour of 321 stainless steel in carbon dioxide under reciprocating sliding conditions at temperatures to 500°C. At 20°C and 300°C there was a very considerable reduction in both the coefficient of friction and the wear rate for the coated surfaces compared with the uncoated steel during the early stages of like-on-like sliding. However, after some time, which depended on the type of coating and the temperature, the friction and wear rate of the coated surfaces increased significantly. In some cases, this was accompanied by gross coating failure, followed by a severe rate of wear, similar to that observed for the uncoated steel. The general mechanism of coating failure involved thinning, leading to cracking and fragmentation, which produced areas of exposed substrate. The wear resistance of the coated surfaces was improved at 500°C. Although thermal softening resulted in deformation of the substrate through the coating, oxidation of the exposed alloy prevented severe damage resulting from metal-metal contact. Possible explanations for this behaviour are discussed in relation to the recorded tribological data and morphological features of the wear scars.

Mechanical equipment used in BP operations is subjected to extremely adverse conditions, often in remote locations. Downtime due to wear and component failure can be very costly and the need to choose materials based upon an understanding of wear resistance is required. Selection is often based on the economics of component manufacture and on physical properties of materials such as hardness and toughness. It is often assumed that selecting materials on the basis of suitable physical properties will automatically lead to their possessing wear-mitigating properties but, in practice, this is not always the case. To help reduce wear of critical components, BP have developed a suite of laboratory test methods to provide basic insight into the wear performance of new materials as an aid to the material selection process. In this paper, wear results are presented from a series of pin-on-disc and slurry impingement erosion tests performed on stellites and ceramics, materials already in use in applications requiring superior wear properties. Through an understanding of wear mechanisms, simple laboratory test methods can be used to simulate the hostile environmental conditions experienced in service and enable materials to be ranked according to their resistance to abrasive, erosive and adhesive forms of wear. Consideration of wear mechanisms in pin-on-disc studies has identified factors such as transitions from mild oxidative to severe adhesive wear in metallic samples and thermally induced spalling of ceramic samples which leads to order-of-magnitude increases in wear rates as sample loading is increased. Profilometry of slurry erosion scars reflects differences in the erosion mechanism for ductile and brittle materials.

A review of previous work concerning the influence of water on the wear of alumina, silicon carbide, silicon nitride (or sialon) and zirconia (or partially stabilized zirconia, PSZ) has revealed a very confused picture: wear rates can either be increased or decreased by the presence of water. In an attempt to rationalize this behaviour, experiments are described in which the above ceramics are worn against a bonded diamond counterface in various environments - air, water, a synthetic diester (di-2-ethylhexyl sebacate) and a diester containing dissolved water. All of the observed trends in wear can be explained in terms of the competing effects of water-accelerated crack growth leading to increased microfracture, and tribochemical reactions leading to reduced microfracture. The relative importance of these two processes depends on the specific ceramic/environment combination and on the counterface roughness.

Transitions in wear rate and mechanism are a common feature of the erosive and abrasive wear of ceramics and other brittle materials. In such cases, a relatively small change in applied conditions, such as impact velocity or angle in the case of erosion, or of applied load in abrasion, results in a significant change in the mechanism of wear. Such a transition often represents a change in the extent or nature of fracture, leading to a concomitant change in the rate of material removal. Transitions in wear mechanism can also be associated with a change in the shape, hardness or size of the abrasive particles. In this paper, theoretical models for the onset of fracture and plastic flow in a material being eroded or abraded, and also in the abrasive or erosive particles, are used to define regimes over which a single wear mechanism will be dominant. A method is outlined by which these regimes can be illustrated graphically on plots of particle size against impact velocity or applied load, and it is suggested that in this way `maps' showing wear rates and mechanisms might be developed for the abrasive and erosive wear of brittle materials.

A model has been developed to predict the erosive wear behaviour of elastomers under conditions of glancing impact by small hard particles. Previous work has shown the erosive wear mechanism of elastomers under these conditions to be similar in nature to that of abrasive wear by a sharp blade. The model presented here was developed from the model of Southern and Thomas for sliding abrasion, by combining their treatment of the growth of surface cracks with a model for particle impact in which the force - displacement relationship for an idealized flat-ended punch on a semi-infinite elastic solid was assumed. In this way an expression for the erosive wear rate was developed, and compared with experimental measurements of wear rate for natural rubber, styrene - butadiene rubber and a highly crosslinked polybutadiene rubber. Good qualitative agreement was found between the predictions of the model and the experimental measurements. The variation of erosion rate with impact velocity, impact angle, particle size, elastic modulus of the material, coefficient of friction and fatigue properties were all well accounted for. Quantitative agreement was less good, and the effects of erosive particle shape could not be accounted for. The reasons for these discrepancies are discussed.

A confocal scanning optical microscope has been used to characterize the surface topography of alumina ceramic wear specimens. The system that was used is briefly described, and the performance of the instrument compared with that of a conventional stylus profilometer. The results of an examination of alumina ceramic wear surfaces is then presented. This includes quantitative measurements on the size and shape of delamination pits in a moderately worn wear surface.

Fretting involves contact between surfaces undergoing small cyclic relative tangential motion. The resultant wear and initiation of fatigue cracks are strongly influenced by the nature of the surface. Shot-peening is an established surface treatment which produces both surface roughening and the development of compressive stresses in the surface. If the material is susceptible to work hardening it also produces surface hardening, since local plastic deformation occurs. Experiments have been devised to separate these three effects in a study on the behaviour of an Al-4Cu-1Mg age-hardening alloy. The results on the fretting - fatigue behaviour of the alloy show that the residual compressive stress has the greatest effect on retarding the propagation of fatigue cracks initiated by fretting. The present investigation is designed to assess the effect of the roughening produced by different levels of shot-peening on the fatigue and fretting fatigue of an age-hardened aluminium alloy after removal of the residual compressive surface stress. On removal of the stress by heat treatment, the surface damage resulting from shot-peening is more damaging in plain fatigue than the increase in roughness would suggest, due to the presence of incipient cracks. These are less damaging in fretting fatigue because they are nearly parallel to the surface and are closed up by pressure from the fretting pad.

The performance of polymeric bearings for metrological applications requiring nanometre-level precision is examined. An initial assessment of lubricated and dry polyacetal and PTFE/Pb composite bearings using a modified stylus measuring instrument is presented. Results of these tests indicate that smooth and repeatable motion of this precision is possible. Reducing the thickness of PTFE in the latter stick - slip behaviour and a change in the friction coefficient from approximately 0.31 to 0.09. To further investigate these and other bearings, a dedicated test rig based on capacitance gauging has been produced. Being constructed almost entirely from the low-expansivity glass ceramic `Zerodur', it has a low thermal susceptibility of 300-600 nm K^-1 enabling the monitoring of long-term effects such as creep and wear. The precision gauging has subnanometre precision over a bandwidth of 1 kHz introducing the possibility of measuring transient effects such as film delamination and stick - slip. Initial results, indicating nanometre-level performance with PTFE thin film bearings, are presented.

Height and wavelength parameters of surface roughness are varied to show that three types of surface contact can exist in relation to tribology. These are (i) predominantly plastic asperity contact with gaps in the contact interface, (ii) the same contact geometry but with predominantly elastic asperity contact and (iii) almost total surface contact under an elastic state of stress (this state approaches the theoretical smooth surface conditions). These contact states are confirmed by rough surface contact techniques of a ball-on-flat geometry. Results from simple boundary-lubricated sliding tests are also presented to show that the frictional and tribological behaviour differs significantly for each of these contact states. The fundamental implications for tribology are discussed in relation to this work.

Indentation cracking behaviour and hardness of (100), (110) and (111) surfaces of single crystals of MgO have been studied using a Vickers diamond pyramidal indenter and a 0.4 mm diameter tungsten carbide spherical indenter. Also, the plastically deformed zone around the indentations has been investigated with the technique of cathodoluminescence. It is shown that the crystal orientation has a very marked influence on the indentation cracking patterns. In the case of the Vickers indentations on (100) and (110) surfaces, radial cracks, four and two in number respectively, form for indenter loads as low as 50 gf. On the other hand, no such cracking is observed when indentations are made on a (111) surface even for indenter loads of up to 0.5 kgf. These differences have been explained in terms of the material flow and dislocation interactions on the various slip systems. The hardness is also found to be dependent upon both the indenter geometry and the indenter crystallographic plane. The Vickers hardness of all three crystal planes decreases with increasing indenter load in the load range of 0.025 to 1 kgf, whereas the Meyer's hardness (for the spherical indenter) increases with increasing load in the range of 0.5 to 13 kgf for the (111) and (110) surfaces, but shows a monotonic decrease with increasing load for the (100) surface. It has been suggested that this is related to the indentation cracking. The cathodoluminescence studies in the scanning electron microscope have been shown to be very successful in revealing the distribution of dislocations around the indentations. This information has been valuable in the interpretation of the cracking patterns. Finally it is concluded that the cathodoluminescence technique is a sensitive and non-destructive one for detecting contact-induced plasticity and cracking of ceramics.

The sliding of a hard surface over a relatively soft one is considered. For rough surfaces a model is given which shows how wear can result from the tearing off of asperities on the soft surface. Such wear is akin to the adhesive wear process of earlier theories. For very rough surfaces wear is modelled as an abrasive process with material removed by chip formation. From the resulting theory it is predicted that when wear occurs as a result of asperity tearing then lubrication reduces both friction and wear. On the other hand in the abrasive range lubrication is predicted to increase both the friction and wear. For smooth, well lubricated surfaces the model used assumes that wear results from the repeated straining of the soft surface by the passage of hard asperities. In this way predictions are made of the influence of surface roughness and lubrication on wear. Metallographic evidence supporting the three proposed wear models is given, together with comparisons between experimental and theoretical wear results.

Contact mechanics is a well established subject, often associated with the work of Boussinesq and Hertz, but larger in scope. This paper discusses new areas where contact mechanics could contribute extensively if a certain number of questions were answered. Some domains just lack data to include in existing models; for instance information on the thermo-mechanical properties of thin coatings is lacking, and an effort in that direction would open up the design of coated machine elements to contact mechanics. Similar deficiencies are noted in fracture mechanics. In other areas, transfer of contact mechanics methods to industrial problems is hindered by changes in material properties under working conditions. Flow stress, toughness, etc vary during most sliding or rolling operations, and contact analyses should include these variations. A strong cooperation between mathematicians, mechanical engineers and material scientists can extend the field of application of contact mechanics in the design of machine elements.

By using an enhanced system to measure surfaces in three dimensions it has been possible to compare topographic data from the same location during the bedding-in of automotive power train tribological surfaces. In addition, it has been demonstrated that AlSn12Si4Cu2 crankpin bearings have the ability to polish asperities from nodular cast iron crankshafts. There is also evidence of erosion of the graphite from the surface of the cast iron leaving shallow dimples which in conventional 2D analysis would have been interpreted as scores in the surface. The limiting minimum oil-film thickness can be related to a characteristic height of the asperities (lambda ratio) at which the `bearing area' is 0.13% of the projected area (3sigma).

In order to generate the optimum surface, without negatively influencing the desired properties of the subsurface material, a myriad of surface selection techniques are available to choose from. This paper provides an overview of these methods, by broadly classifying the surface and overlay coating treatments currently used, according to the fabrication styles of atomic deposition, bulk deposition and direct surface modification. Insight into the research, development and utilization of these concepts, so as to address the needs of surface engineers today and in the future, is offered.

Wear resistance of cemented carbide tools with TiN and (Ti, Al)N coatings in lathe turning of steel was investigated. WC-15%TiC-6%Co and WC-3%Co cemented carbides were used in experiments. The thickness of coatings in all cases was 5 µm. Cutting speed in the tests ranged from 70 to 220 m min^-1, from 0.1 to 0.2 mm rev^-1 and depth from 0.25 to 0.5 mm. All coated inserts outperformed uncoated ones. In the case of WC-15%TiC-6%Co inserts (Ti,Al)N outperformed TiN, but for WC-3%Co inserts TiN coatings demonstrated better wear resistance than (Ti,Al)N. The method of arc vapour deposition was selected for experiments with titanium and Ti-Al alloy cathodes. Ti-Al cathodes were made using electron beam remelting of a powder metallic mixture in vacuum. The compositions of Ti-Al cathodes were 70Ti-30Al (wt%) and 50Ti-50Al. Al/Ti ratios (at.%) in (Ti,Al)N coatings, based upon energy dispersive x-ray microprobe measurements, were 0.1 and 0.3 respectively. From x-ray diffraction data all coatings were found to have an NaCl structure. The preferred orientation for TiN coating was (111); for both (Ti, Al)N coatings it was (200). Lattice parameters calculated from TiN(200) reflections via Al/Ti ratio were as follows: a = 0.4219 nm (for TiN), a = 0.4215 nm and a = 0.4191 nm. The stresses in coatings were determined assuming the plane stress state and an equality of stresses in substrate and coating. It was demonstrated that growth of Al content in the coating led to increased stresses. Investigations of surface topography were performed using scanning electron microscopy. The main roughness parameters (asperity height, average radius of separate asperities, angle of obliquity etc) of (Ti,Al)N coating were higher in comparison with TiN because of a great number of microdrops on the coating surface. The turning tests demonstrated that there was only a slight difference in wear resistance between the two (Ti,Al)N coatings. Probably, the principal reason for improvement of tribological properties was in stronger links between atoms in the (Ti, Al)N structure. The effect of microdrops was harmful, but in oxidizing conditions synthesis of Al₂O₃ on the surface was a favourable factor.

The role of surface engineering in the war against wear has been well recognized in recent years, not only because wear and friction are surface-related phenomena, but also because of the growing commercial maturity of a wide range of cost-effective surface technologies. In the present paper, the basic principles of surface engineering and its impact on tribology are illustrated principally through the technology of thermochemical processing of ferrous materials. The success of surface engineering to combat wear is demonstrated by the current application of such surface technologies as thermochemical treatments and ceramic coatings to a variety of engineering components. However, surface engineering is not simply about using one of the available surface technologies to modify the surface properties. Surface engineering is essentially about design. Hence, under situations where components are subjected to complex external loadings, more than one surface technology needs to be employed in order to produce the required combination of properties to combat the loadings. This philosophy of designer surfaces is discussed in the paper, and illustrated through detailed theoretical and experimental studies on the tribological behaviour of plasma-nitrided and combined plasma-nitrided and PVD TiN-coated low-alloy steel.

Cutting tests are very costly when expensive materials are to be cut. Therefore model tests are required which allow a preselection of materials or coatings to increase the lifetime of cutting tools. The wear of indexible inserts for milling a soft magnetic highly ductile nickel - iron alloy is simulated by means of a modified pin-on-disc system. For the simulation a detailed systems analysis of the tribological processes during cutting and model wear testing is necessary. Cutting tests revealed that flank wear is the wear-determining process. Therefore the flank of an original indexible insert was pressed against the flat surface of a rotating disc which was made of the nickel - iron alloy. By reversing the direction of rotation in comparison to the cutting process only sliding took place. The flank wear - distance curves of the model wear tests resembled the wear curves of the cutting tests. In both tests nickel - iron was transferred to the flank of the indexible inserts. By physical vapour deposition (PVD) of TiN, friction, adhesive transfer and wear were considerably reduced. These tests give valid information on the tribological behaviour of uncoated and coated indexible inserts for milling highly ductile nickel - iron alloys, but real lifetimes of the tools can only be determined by milling tests.

The cam and follower of the automobile valve train has proved to be the most difficult component of the internal combustion engine to lubricate. It has been recognized that the extremely thin films which occur during part of the operating cycle will inevitably mean some substantial surface interaction between the components. For this reason the mechanism of lubrication is often described as boundary in nature. During the last decade, however, the importance of a satisfactory design approach incorporating the principles of elastohydrodynamic lubrication has been established. Such design analysis is based upon a quasistatic approach and clearly identifies critical points at which the entraining action of lubricant into the contact ceases to operate and no lubricant film is predicted. In the present paper a transient analysis is described in which the local variation of normal surface velocity may be considered. The protective action of the squeeze-film lubrication mechanism at the critical points is demonstrated enabling an improved design sensitivity.

With the increasing data densities demanded of magnetic recording media, and in particular of floppy diskettes, wear and its effects on the durability of the media is becoming increasingly important. A consequence of the higher recording density is reduced magnetic layer thickness, and hence mechanical contact problems are more significant. In this investigation the wear and durability of diskettes of nominally the same formulation have been studied as a function of process burnishing time and state of lubrication. The study was conducted in commercial disk drives and in two forms of simulation test apparatus. Durability was found to increase with burnishing time to a maximum value and then decrease with further burnishing. Durability was also found to increase substantially with surface lubrication, but to be relatively unaffected by internal lubrication. A mechanism for the wear of flexible magnetic recording media is proposed consisting of three phases: initial plastic flow of the surface region, fatigue induced delaminative wear, and finally avalanche breakdown due to three-body action.

Wear mechanisms occurring at the video head were examined for a Mn-Zn ferrite single-crystal head in sliding contact with iron oxide formulation video tape. Of particular interest was the effect on wear due to the inclusion of a head cleaning agent (HCA) in the tape formulation. Wear tests were performed in situ using specially formulated iron oxide video tape samples containing a range of levels of two types of HCA: Al₂O₃ and Cr₂O₃. Head performance was monitored by measuring the degradation of the RF signal during wear, and wear to the head was measured using Knoop diamond indentation. The resulting wear surfaces were examined using optical and scanning electron microscopy. The tape was examined in cross section using transmission electron microscopy. Results showed that the dominant mechanism of wear was abrasive, characterized by the ploughing of isolated asperities on the tape surface through the ferrite head, and damage to the head gap by three-body abrasion. Evidence is presented which suggests that the isolated asperities are a result of poor dispersion of tape particles. The addition of an HCA to the tape formulation reduced the abrasion damage. Wear increased with increasing HCA content, but the total signal degradation did not vary significantly with HCA level for either HCA. Substantial differences in wear were observed between the two types of HCA, and evidence is presented which suggests that the differences are a product of the relative quality of dispersion of the two HCA materials.

Research on lubrication is as active today as it has ever been during the 20th century, yet it is well over a hundred years since Osborne Reynolds exposed the fundamental nature of fluid-film lubrication. Since 1886 the effective film thickness in lubricated machine components has been reduced by a few orders of magnitude and the reasons for this `thinning film' are reviewed on the occasion of the 10th anniversary of the Institute of Physics Tribology Group and the 25th anniversary of the publication of the Jost Report.