Toward Improved Understanding of Bearing Lubrication: Role of Cage Geometry and Operating Speed

<p dir="ltr">The lubricant flow dynamics in rolling element bearings (REBs) are governed by a complex interplay of viscous forces and flow phenomena, including recirculation, starvation, cavitation, surface wetting, aeration, and capillary effects, which are strongly influenced by the cage geometry and operating conditions. This thesis examines how variations in cage design and rotational speed affect these lubrication phenomena. Bearing cages, often designed for specific applications, play a crucial role in governing both the kinematics of the rolling elements and the flow of lubricant within the bearing chamber. Specifically, the design and shape of the bearing cage can affect: <i>(i) </i>its dynamic motion and interaction with the rolling elements, <i>(ii) </i>the overall flow of oil in the region between the cage and raceways, as well as <i>(iii) </i>the distribution of oil inside the cage pocket. Additionally, bearing operating speed can substantially modify lubrication behavior, particularly under high-speed conditions, which necessitates a systematic investigation. In order to study these different phenomena, unique experimental setups were designed as well as appropriate numerical and computational models were developed to evaluate the effects of cage designs and rotational speeds on the mechanisms governing lubricant flow and distribution in REBs.</p><p dir="ltr">The first experimental platform developed in this work was the Counter-Rotating Angular Contact Ball Bearing Test Rig (CRACTR), which was designed to rotate both raceways simultaneously in opposite directions while keeping the cage stationary. This unique configuration allows for detailed observation of lubricant behavior in the cage reference frame and was primarily utilized to investigate the effects of various cage designs at moderate operating speeds. CRACTR was used to evaluate cage whirl, ball-cage contact forces, and lubrication phenomena for different cage geometries using transparent acrylic cages and high-speed imaging. Bubble Image Velocimetry (BIV) was employed to capture qualitative and quantitative oil flow data, which were also used to validate a single-phase Computational Fluid Dynamics (CFD) model. The test rig was further enhanced with ultraviolet lights to visualize oil-air distribution within different cage pocket designs. These investigations revealed the occurrence of oil-air striations along the ball surface due to surface tension effects, which were also corroborated using multiphase CFD models.</p><p dir="ltr">The next major focus of this work was to investigate lubrication behavior under high-speed operating conditions and to highlight the specific differences from previously studied low and moderate-speed regimes. A new high-speed bearing test setup was designed and developed to enable detailed visualization of lubricant behavior at elevated speeds. High-speed imaging revealed the occurrence of the Aerodynamic Leidenfrost Effect (ALE), wherein oil droplets levitate on an entrained air film, preventing their adhesion to bearing surfaces. Among the examined lubricant properties, extensional viscosity was identified as a key contributor for the difference in ALE across different lubricants. To complement the experimental observations and gain deeper mechanistic insights, a numerical ALE model was also developed by coupling the discretized Reynolds and elasticity equations to predict the interaction between air-film pressure and droplet deformation. The model was subsequently used to derive curve-fit expressions for estimating critical lubrication parameters under ALE conditions.</p><p dir="ltr">This thesis places particular emphasis on the development of innovative experimental techniques and advanced computational and numerical methods tailored to study various fascinating aspects of lubricant flow within REBs. Future research can build upon these findings by exploring additional lubrication regimes in REBs, bearing configurations, and modeling enhancements to further advance understanding and optimization of tribological performance in a wide range of engineering applications.</p>