Multi-scale friction modelling of lubricant-surface interaction in tribological contacts

Abstract

The study emphasises on the fundamental understanding of the tribological behaviour for a lubrication system. In a dynamic system (e.g. piston ring/liner contact), the opposing sliding surfaces are separated by a layer of lubrication film, preventing direct metal-to-metal contact. Under critical operating conditions (e.g. high applied load), the film thickness might reduce to similar order of magnitude of the surface roughness, leading to boundary friction. A multi-scale approach is required to better predict the lubrication performance in a dynamic system. The first part of this study focuses on the development of a numerical algorithm using Reynolds equation to predict fluid film formation in a lubrication system. A rough surface contact model using the Greenwood and Tripp assumption is applied to predict the frictional properties along a lubricated conjunction using a three types of commercially available SAE grade lubricant namely, SAE5W40, SAE10W40 and SAE15W40. The simulated friction force is compared to the values measured using a pin-on-disc tribometer, showing good correlation between both data sets. With acceptable level of confidence on the mathematical model, the second part of the study attempts to simulate a tribological properties of lubrication system considering biodiesel as the lubricant in the engine lubrication system. As the rough surface contact parameters are considered to be consistent in both case study, Reynolds solution is modified in order to considers the fluid film formation of biodiesel with relation to their Fatty Acid Methyl Ester (FAME) compositions. Two parameters, namely 1) average number of carbon atom (zave) and 2) average number of double bonds (ndave) for biodiesels are included when introducing the effect of FAME compositions to the classical Reynolds equation. The term χ has been introduced to the modified Reynolds equation to correlate these two parameters. The value for the term χ is determined by fitting the simulated lubrication Stribeck curve to the measured data for each of the selected biodiesels. Through this, a polynomial correlation is then obtained for the term χ as a function of the product of average number of carbon atoms and average number of double bonds for the biodiesels. The coupling of Reynolds solution and rough surface contact model give a good correlation between measured data and predicted value in determining lubrication Stribeck curve regime for typical lubricants and biodiesel.

More information

Identifiers

Catalogue record

Export record