Molecular modelling of lubricant-solid interactions that control friction and wear

Abstract

Lubricants are used widely in industry to reduce the effects of friction and wear. Lubricant formulations for engine applications are typically composed of a base oil and various additives, each having a different function. This complexity means that their properties are not well understood, and it is difficult to study lubricants while they are in operation in an engine. The rheology of the base oils, which make up the majority of the lubricant formulation, is particularly important as the lubricants are exposed to a wide range of pressures and shear rates. Recent high-pressure experiments at Edinburgh indicate that model base oils – squalane and a poly-α-olefin (PAO) mixture – are no longer hydrostatic media above roughly 1 GPa. Diamond-anvil cell measurements show that the hydrostatic limit for these base oils is in the range of 0.74 to 1.24 GPa, as they appear to undergo solidification. X-ray diffraction shows no Bragg peaks, suggesting that a glassy structure, and not a crystalline structure, is formed. In this work, the structure, dynamics, and thermodynamics of model base oils were studied using molecular-dynamics (MD) simulations. The equation of state, self-diffusion coefficient, viscos ity, and radial distribution functions were calculated over a wide range of pressures, from 0.0001 to 10 GPa. The results show that molecular diffusion is essentially arrested above about 0.1 GPa, which supports the hypothesis that the samples are kinetically trapped in metastable amorphous-solid states. The shear viscosity is immeasurably large at very high pressures, but the coefficient governing its increase from ambient pressure is in good agreement with the available literature data. Subtle changes in the short-range real-space correlations are related to a collapse of the molecular conformations with increasing pressure, while the evolution of the static structure factor shows excellent correlation with the available X-ray diffraction data. Next, oleamide, an industrially relevant organic friction modifier (OFM), was introduced to the base oil systems, where they formed clusters both with and without the presence of water. Taking a step towards engine conditions, these base oil and OFM systems were placed under confinement between iron oxide surfaces. It was observed that the oleamide would adsorb onto the surfaces and form a layer. If water was present, it would out-compete the oleamide for surface coverage, even being able to remove pre-existing oleamide from the iron oxide. The oleamide would then form a layer on top of the water. The free energy of adsorption for oleamide onto iron oxide surfaces was computed using umbrella sampling, and agreed well with experimental measurements. Computing association constants for oleamide showed that, at the experimental conditions and concentrations, it would adsorb onto the surfaces as monomers rather than in a cluster. The association constants were also in excellent agreement with experimental measurements. The final part of this work involved adding shear to the confined base oil/OFM systems to simulate the movement of the engine during operation. Initial frictional measurements using pure base oil systems correlated well with literature data from both experiments and simulations, and the effects of the OFM and water contamination on friction were then explored. Oleamide was found to reduce friction in most cases, being particularly effective for thin lubricating films under relatively low-shear conditions (< 107 s −1 ). Additionally, water was shown to have a larger effect on reducing friction than oleamide under a wide range of conditions. These effects showed good qualitative agreement with experimental measurements. High-pressure simulations showed some evidence that the base oils will undergo the same pressure-induced solidification that was observed in the initial bulk simulations. This work has made steps towards building up a comprehensive picture of how typical base oils and organic friction modifiers interact under engine conditions, starting from simple bulk systems and gradually adding in complexity to get closer to engine conditions. The simulations have been extensively compared and validated with experimental measurements, showing consistently strong agreement. The simulation methods described here are applicable to other lubricant components to continue elucidating how these complex formulations work.