This article presents analytical friction prediction models applicable to lubricants in point contacts under thermal elastohydrodynamic lubrication (TEHL). The types of models used consider the heat generated and its penetration into the bulk of the contacting solids. Therefore, the increase in temperature is determined, which causes important variations in the operating conditions. The article sets out the hypotheses assumed, the theoretical calculation procedures and the ensuing equations for determining the friction coefficient under TEHL. An experimental stage is performed on a mini-traction-machine, which allows the measurement of friction coefficient in ball–disc contacts under a wide range of control parameters involved in TEHL, such as lubricant bath temperature, load, average velocity, slide-to-roll ratio, and contacting materials. The experimental results for different lubricants are compared to those given by the models, and show the proposed models to be accurate for predicting the friction coefficient.

Surface texturing has proved to be a very useful tool for expanding the behaviour under hydrodynamic and elastohydrodynamic regimes instead of mixed or boundary lubrication regimes, and therefore for reducing the friction coefficient under high-load low-speed conditions. This article presents the texturing of different copper test-samples using photolithography and chemical etching to measure the friction coefficient using a point contact machine. The effects of texture size, texturing density, the initial roughness of the samples and the operating conditions have all been studied. Some combinations of texturing density and texture size achieve up to 30% reduction in the friction coefficient. Taking into account experimental data, artificial neural networks are used as a tool for both predicting and optimising the friction coefficient on the textured surface for any given operating condition.

Under operating conditions which are unfavourable for lubrication, such as high load and low velocity, the use of textured surfaces significantly promotes the formation of a thick lubricant film and an improvement of the friction coefficient. This paper relates to the manufacture of textures using a photolithography and chemical etching process. Different surface geometries, texturing densities and depths were designed to analyse the influence of these parameters. The friction coefficient was measured in a ball-on-disc tribometer under different lubrication regimes, and the results have been used to develop an artificial neural network with texturing optimisation potential.

No abstract

This paper analyses the applicability of artificial neural networks for predicting the lubricated friction coefficient. We will consider their use as faster and simpler alternatives to simulations based on theoretical behaviour equations. The development of several different artificial neural networks is presented. They have been trained through tribological tests on a mini‐traction‐machine, which furnishes the friction coefficient in point contacts. Once the training has been completed the networks are applied as tools for predicting the results in different operating conditions. Their advantages and disadvantages are analysed compared with conventional simulation tools. Copyright © 2013 John Wiley & Sons, Ltd.