On the Chaotic Behavior of Grease Lubrication in Rolling Bearings

Abstract

Grease-lubricated bearings often show a sharply fluctuating temperature signal. At the start of the bearing operation the temperature rises. This is generally ascribed to grease churning. This phase takes approximately 2–5 h. After that the temperature usually reaches a quasi-steady-state value. Quite often, this situation is not maintained and the temperature signal typically shows “events,” characterized by periods where the temperature rises significantly and falls back to the “steady-state” value after some duration in time. The time intervals between the events and their duration vary significantly throughout a single bearing test. Moreover, this signal varies markedly for individual bearings tested under the same conditions with identical grease. This article gives an analysis of the temperature and relative film breakdown time series signals from various tests using the concepts of nonlinear dynamics and chaos. The analysis demonstrates that the embedding dimension of the system for the group of bearings tested is consistently 5, showing that a five-parameter nonlinear model provides an adequate description. The maximum Lyapunov exponent lies in the narrow range 1.14 < λmax < 1.21, while another exponent lies close to zero. It may thus be concluded that grease lubrication here exhibits “deterministic chaotic” behavior, implying that the initial filling conditions for these bearings will be crucial for the life of the grease in the bearing. The minimum Lyapunov exponent is negative, indicating a dissipative mechanism in the dynamics. The article shows that the film thickness in a grease-lubricated bearing cannot be described using conventional starvation theory only and that it is fluctuating in time in a chaotic manner. This is most important for the development of film thickness and (grease) life models in grease-lubricated bearings.