Modelling and measurement of wear particle flow in a dual oil filter system for condition monitoring

Abstract

Wear debris is an indicator of the health of machinery, and the availability of accurate methods for characterising debris is important. In this work, a dual filter model for a gear oil system is used in conjunction with operational data to indicate three different system operating states. The quantity of wear particles in gear oil is analysed with respect to system running conditions. It is shown that the model fits the data in terms of startup “particle burst” phenomenon, quasi-stationary conditions during operation, and clean-up filtration when placed out of operation.In order to establish boundary condition for particle burst phenomenon, the release of wear particles from a pleated mesh filter is measured in a test rig and included in the model. The findings show that a dual filter model, with startup phenomenon included, can describe trends in the wear particle flow observed in the gear oil. Using this model it is possible to draw conclusions on the filtration system performance and wear generation in the gears. Limitations of the proposed model are the lack of ability to describe noise and random burst spikes attributed to measurement error distributions. Trending of gear wear particle generation is made possible by model parameter estimation and identification of an unintended lack of filter change. The model may also be used to optimise system and filtration performance, and to enable continuous condition monitoring.