Andrew V. Olver

The Mechanism of Rolling Contact Fatigue: An Update

Abstract A review of some of the recent work on the mechanism of rolling contact fatigue (RCF) is presented. Topics covered include the appearance and classification of RCF and the processes of strain localization, texture development, microstructural change, crack formation, crack shape and propagation, and through-fracture. It is concluded that a significant barrier to progress is the poor current understanding of the processes of running-in and of the interactions between plastic deformation, wear, lubricant chemistry and damage accumulation.

Lubricant Film Thickness in Rough Surface, Mixed Elastohydrodynamic Contact

A spacer layer imaging method has been employed to map lubricant film thickness in very thin film, rough surface, rolling elastohydrodynamic (EHD) contacts. A series of model roughnesses have been produced by depositing tiny ridges and bumps on a steel ball surface and the influence of these features on film thickness has been investigated at a range of rolling speeds. It has been shown that all the model surface features studied form speed-dependent, micro-EHD lubricating films, but the detailed shape and thickness of these films depends upon the geometry of the feature and the rolling speed. All model surface features also produce a net increase in mean film thickness, compared to the smooth surface, under operating conditions where the film thickness is less than the out-of-contact height of the surface feature studied. For a real, random, rough surface, however, mean film thickness is less than the smooth surface case. The film thickness mapping technique has also been used to measure the effective roughness of surfaces in lubricated contact. This shows that surfaces based on 2-D array of tiny circular bumps become rougher as the rolling speed and thus film thickness increases. However, real, rough surfaces appear to show a decrease of in-contact roughness with increasing rolling speed.

‘Inlet suction’, a load support mechanism in non-convergent, pocketed, hydrodynamic bearings

Abstract It is shown that a simple parallel pad bearing containing a closed pocket can support load if it operates in an ambient pressure that is appreciably in excess of the cavitation pressure of the lubricating fluid. This arises due to fluid flow driven by subambient pressures in the inlet region of the pad (‘inlet suction’). Maximum load capacity occurs when the pocket is located near the inlet to the bearing and under conditions such that cavitation is just provoked.

Wear in rolling contacts

Abstract Unexpectedly high uniform rates of wear occur from time to time in concentrated lubricated rolling contacts having low slide-roll ratios. Such wear can occur under quite mild conditions and this poses a significant practical problem, especially in gears. This phenomenon of high wear rate at low slide-roll ratios has been reproduced consistently and studied in the laboratory using a disc machine. In this paper the outcome of this study is reported. It is shown that the wear is caused by plastic deformation, fatigue cracking, ductile extrusion and fracture on a scale associated with asperity contact. The relationship of the phenomenon to delamination wear theory is discussed and ways of avoiding this type of wear in operating machinery are suggested.

Measurement of Sub-Nanometer Lubricant Films Using Ultra-Thin Film Interferometry

The ultra-thin film interferometric method of measuring the thickness of very thin films in lubricated contacts has been refined so as to be able to measure films down to 0.3 nm with a standard deviation of 0.15 nm. The main remaining source of measurement variation for films below 3 nm thick is the surface roughness of the contacting solids. This modified technique has been applied to study the film-forming properties of three fluids, hexadecane, a dilute solution of surfactant in hexadecane, and cyclohexane. Purified hexadecane shows a very slightly enhanced oil-film thickness below 1 nm. The long-chain surfactant forms a boundary film 2 nm thick. Cyclohexane behaves as though it forms a surface layer about 1 nm thick with viscosity three times the bulk fluid viscosity.

Prediction of traction in elastohydrodynamic lubrication

Abstract The prediction of traction (friction) in lubricated rolling-sliding contacts remains a challenging problem despite the development of the realistic Maxwell-Eyring-limiting shear stress model by Johnson and co-workers in the 1980s. This is largely because there is a strong coupling between the elastohydrodynamic traction and the film temperature. An added complication is that the heat conducted into the rubbing surfaces, as well as influencing traction directly, also determines the temperature in the inlet to the contact and hence the thickness of the elastohydrodynamic film. In the present paper, the traction model of Johnson et al. is combined with a heat transfer analysis of the contacting bodies as well as the film thickness regression equation. In addition, the variations in the lubricants rheological properties with temperature and pressure based upon the measurements of Muraki et al. have been included. The traction equation is expressed in dimensionless form and is solved using a simple iterative scheme, which in many cases allows estimation of the traction without the use of a computer. Closed-form equations for the friction are given for each of the traction regimes.

The Effects of Three-Dimensional Model Surface Roughness Features on Lubricant Film Thickness in EHL Contacts

The Spacer Layer Imaging method has been used to investigate the influence of three-dimensional roughness features on the thickness and shape of elastohydrodynamic (EHL) films. An array of near-hemispherical bumps was employed to represent asperities. A micro-EHL film developed at the bumps whose orientation depended on that of the inlet boundary at the location at which the bump had entered the contact. Rolling-sliding conditions induced a micro-EHL film with a classical horseshoe shape at the bumps. The flow of lubricant around the bumps appeared to differ between thin and thick films.

The residual stress distribution in a plastically deformed model asperity

Abstract When rough metallic surfaces come into contact, plastic deformation may occur locally, even at the lightest loads. This plastic deformation is thought to be an important element in a wide range of contact failure mechanisms, including fatigue and nearly all forms of wear. In this paper a simple model of asperity plastic deformation is presented. The model is based on slip line field theory and is used to calculate residual and full-load stress distributions at fully plastic asperity contacts for normal and moderate tangential loads. Measurements of surface residual stress were carried out using two different techniques on a range of plastic contacts of various materials and geometries. The results show agreement with the main predictions of the theory.

Compression of a Single Transverse Ridge in a Circular Elastohydrodynamic Contact

A detailed experimental study has been made of the behavior of a 100 nm high transversely oriented ridge in an elastohydrodynamic (EHD) contact. Ultra-thin film interferometry has been used to measure film profiles accurately over a very wide range of lubricant film thicknesses, from a few nanometers up to nearly one micron. This enables the recovery of the amplitude of the inlet perturbation geometry with increasing EHD film thickness to be quantified and compared with numerical predictions. In pure rolling under very thin film conditions, corresponding to a smooth surface EHD film thickness of 10 nm, the surfaces near the ridge were squashed down, leading to a constriction in the film of only about 9 percent of the height of the un-deformed ridge. As the EHD film thickness increased, this deformation recovered until the ridge constriction regained about 90 percent of its original height at film thicknesses of about 1 μm. However this relatively rapid recovery only occurred in pure rolling and is attributed to the local perturbation of film convergence which the ridge generates while in the inlet region. This propagates through the contact at the mean speed of the surfaces and-in pure rolling-acts to diminish the effect of local squeeze. When sliding was present, the ridge remained almost fully deformed even when the mean film thickness was as much as twice the height of the original ridge. In this case, the ridge travels through the contact at a different speed from the mean of the two surfaces. The consequent decoupling of the ridge and the convergence perturbation results in a large local pressure due to squeeze which acts to inhibit recovery of the ridge. The general trend of the behavior of the lubricated ridge is shown to be in good agreement with earlier theoretical results.

The Effect of a Friction Modifier Additive on Micropitting

Previous studies have shown that many anti-wear additives have a tendency to aggravate micropitting damage in hard steels subjected to rolling sliding contact at low lambda values. This is apparently because anti-wear additives suppress the gradual smoothing of the rough surfaces that takes place when a pure base stock is used under mild conditions. However, in practice, it is common to combine anti-wear and friction modifier additives in formulated oils, a combination that leads to reduced boundary friction. In the work described in this article, we added a common friction modifier agent, a molybdenum bis-diethylhexyl dithio-carbamate (MoDTC), to an oil containing an anti-wear additive, a secondary zinc dialkyl-dithio-phosphate (ZDDP) in a mineral base-stock. We examined micropitting behavior using a disc tester. The oil with both MoDTC and ZDDP showed initial micropitting that gradually disappeared with continued running, whereas ZDDP alone led to continuing severe damage. Analysis of the counter-discs after the test showed the presence of MoS2 deposits on the asperity crests. Reduced boundary friction was confirmed using a tribometer test. It is speculated that the improved micropitting behavior resulted from the effect of a reduction in local tensile stress due to reduced asperity friction. This may have reduced the opening of the surface cracks and inhibited their extension.