New numerical results, based upon the concept of solidification, are produced which match the intriguing dimple observed by Kaneta in elastohydrodynamic lubrication of point contacts under pure sliding conditions. It is shown that Kaneta's dimple is consistent with transport conditions of a solidified lubricant in the high-pressure region of the conjunction. The difference of velocity between the average velocity of the solidified core and the mean entraining velocity is referred to as 'slip'. The value of slip is shown to be related to the stress at the wall, which is deduced from the limiting shear stress.

The development of understanding of the phenomenon of elastohydrodynamic lubrication (EHL) throughout the twentieth century is reviewed. The development of solutions for both line and point contacts is considered for both fully flooded and starved conditions. Particular attention is given to the introduction of non-Newtonian rheological models and numerical methods. Progress in the analysis of impact and general non-steady-state conditions is reviewed, together with the consideration of rough surface and micro-EHL. Attention is drawn to the need to consider realistic models of real surfaces and real fluids in future studies.

The numerical developments described in Part 1 are extended to incorporate the consideration of non-Newtonian effects, and fully coupled solutions of the thermal non-Newtonian elastohydrodynamic lubrication point contact problem are presented. The numerical findings are compared with a series of experimental results performed by Cann and Spikes with a steel-sapphire point contact in pure sliding conditions. Fair agreement with the experimental results is obtained in terms of surface temperature. Comparison of the results under identical conditions but with a steel-steel contact is also undertaken and shows that a steel-sapphire contact represents a good simulator of a steel-steel contact in terms of temperature and film thickness. The influence of the viscosity relationship is also considered and reveals that the Yasutomi et al. equation leads to only a marginal difference compared with the Roelands equation for the cases considered.

Solutions of the thermal elastohydrodynamically lubricated point contact problems are presented for both low and high Peclet number conditions. The surface temperatures are calculated using the full expression of the moving heat source equation given by Carslaw and Jaeger. Of interest is the potential of the method to predict solutions for pure sliding conditions in which one surface velocity is zero. The energy equation is treated as a two-dimensional problem by approximating the variation in temperature across the film with a quadratic profile. To solve the strongly non-linear set of equations which govern the problem a relaxation scheme is proposed which takes advantage of the weak coupling between the Reynolds equation and energy equation in the high-pressure region of the contact. The iterative process corresponds to a two-stage process in which pressure and temperature are relaxed independently. The numerical scheme provides a stable and fast convergence and has the added advantage of allowing the isothermal relaxation schemes for pressure previously developed to be reused.

The micro-elastohydrodynamic lubrication of a single transverse ridge is revisited using an experimental technique, which combines an optical interferometry technique and a high-speed color video camera. The purpose of this study is to augment prior experimental analyses, by providing a complete and detailed history of the ridge associated with changes in film thickness as it passes through a high-pressure conjunction. An enhanced experimental procedure has been developed to enable an automatic analysis of the interferograms. In particular, the methodology allows abrupt changes in film thickness and rapid variations of interference orders to be taken into account. The observations presented in this paper exhibit interesting and fascinating features that have not been previously reported. In particular, it is observed that under rolling/sliding conditions the ridge undergoes further deformations as it proceeds to the exit to the contact. In addition, there appears to be an important contribution of pressure flow to the transport of lubricant and, contrary to current understanding, entrapped lubricant is seen to accompany the ridge as it passes through the contact, therefore appearing not to move at the entraining velocity.

A multigrid multi-integration method has been used to solve the elastohydrodynamic lubrication ( EHL) point contact problem over a large range of loads. Solutions obtained with the multigrid method are compared with those computed with an effective influence Newton method. Good agreement has been obtained, which validates the results obtained by both of these independent methods. Smooth surface problems have been used to test the multigrid method, but an example that takes into account a wavy surface has demonstrated the robustness and the large potential of the multigrid method to analyse EHL problems with three-dimensional surface roughness.

No abstract

A direct comparison between experimental and numerical results for the passage of an array of 3D flat-top, square shaped surface features through an EHL point contact is presented. Results for pure rolling conditions show that the features’ deformation in the high-pressure region is governed by their ability to entrap lubricant both underneath and in the grooves during their passage through the inlet zone. Film perturbations associated with each defect occur as locally enhanced regions of lubricant and film thickness micro-constrictions. Under sliding conditions the features sustain further deformations as they traverse the high-pressure conjunction and meet the highly viscous lubricant entrapped in the grooves, which moves at a different velocity. Lubricant is also seen to accumulate just in front or behind the features depending on the slide-to-roll ratio. Overall, the results highlight the importance of understanding the effects of the defects structure and the lubricant rheology on the film thickness to unravel the effects of real roughness patterns.