Stochastic model for lubricating film thickness in rolling contact

On-line monitoring plays an important role in lifetime management of machinery. An especially interesting element of machinery is oil – as a lubricant or hydraulic fluid. The present paper introduces a model that has been developed for evaluating film thickness in rolling contact based on parameters such as viscosity, water content of lubricant, temperature, pressure and rolling versus sliding ratio. After combining direct or indirect information from different sources, statements can be made concerning machinery and the condition of its components. These statements are used in maintenance and replacements decisions, in order to avoid failures leading to unexpected unavailability and to optimise overall costs. Here, a mathematical modelling framework for combining various condition monitoring measurements to predict or estimate the lubricating oil film thickness is presented. The approach is based on stochastic filtering principles.