Simulations of rheological properties of lubricants under operational conditions

Abstract

Understanding the microscopic behaviour of lubricants and their additives under operational conditions is critical to developing new lubricants and improving existing ones. Tribology researchers have been applying a quite diverse range of experimental techniques, that, while providing excellent insights into the macroscopic rheological properties of such systems, are not able to capture the microscopic behaviour fully, or they are limited in the range of conditions they can test. In this context, computational techniques have come forward as a tool to investigate microscopic and macroscopic behaviour at scales and conditions which are very difficult to access or probe experimentally. This thesis focuses on various computational techniques such as equilibrium molecular dynamics, non-equilibrium molecular dynamics and density functional theory to tackle a range of important technical and industrial problems in the field of rheology. In particular, we start by investigating a set of 11 ideal mixtures of two industrially relevant synthetic esters via equilibrium and non-equilibrium molecular dynamics and compare the temperature dependence of static and dynamic properties of these systems against experimental values. We showed that computational methods can reproduce the experimental trends as well as provide a workflow that can be applied to a wide range of lubricants. Next, we moved to non-equilibrium molecular dynamics simulations of a more complex set of non-ideal mixtures of three hydrocarbons to investigate the pressure dependence up to 4 GPa, a range which is hard to reach experimentally. We compared our simulated densities and viscosities against known and commonly employed empirical equations and we showed that we can obtain reliable estimates of the densities and viscosities, while also highlighting the limitations of current models at extreme conditions. We further moved towards even more realistic systems by confining a hydrocarbon-based lubricant between two iron oxide slabs as the interactions at the surface–lubricant interface is relevant in a range of technological applications. We employed nonequilibrium molecular dynamics simulations with reactive and non-reactive force fields to assess the effect of confinement on rheological properties. We showed that by increasing the film thickness we approach the viscosity value of the bulk fluid and that, at the conditions studied, no reactions are happening at the lubricant-surface interface. Finally, we shifted our focus to the interactions between additives and hematite surfaces as they are widely present in a variety of industrial applications. Particularly we focus on the adsorption of Zinc dialkyldithiophosphates as it is the first step that leads to the formation of protective films (tribofilms) which are key in the anti-wear property of this class of lubricant additives. We investigated the changes in electronic structures and geometry that could increase the reactivity towards possible decomposition pathways. The work presented in this thesis shows the applicability of computational techniques in studying the rheological properties of lubricants and related systems. The techniques described here showed that it is possible to accurately describe realistic systems at operating conditions and beyond experimental limitations. In future, this could aid the design of new lubricants or additives as well as improving existing ones.

More information

Identifiers

ORCID for Chris Skylaris: orcid.org/0000-0003-0258-3433

Catalogue record

Export record