A Numerical and Experimental Approach of the Flange-Roller End Contact
Résumé
Commonly, EHD lubricated contacts are modelled by a line, a circular or an elliptical contact. However, when the geometries of the contacting solids cannot be locally assimilated to parabolic curves, another approach is required. The study pertains to the flange-roller end contact in tapered roller bearing. The roller end has a toric shape and the flange is conical. Consequently, the contact has an unusual shape and modified behaviours.
The investigation was focused on the solids geometry influence on film thickness, pressure and friction forces. The results are presented, together with an analysis of the mechanisms involved in this causality relationship. The flange-roller end contact is unusual in several ways. A first analysis shows that the contacting tracks are narrow, which predicts a starved contact. However the fully flooded contact will be studied here, as there is a lack of knowledge on the topic even under fully flooded conditions. Moreover, the flange is conical and the roller-end has a toric shape. Consequently, a dedicated description of the rigid bodies is required as the parabolic assumption is no more valid. And finally, the flange-roller end contact has a complex kinematic. Indeed, the solids surfaces kinematic is composed of a rolling-sliding component and a spinning component.
In order to study the flange-roller end contact, the Tribogyr test-rig (Dormois et al.1)) was developed in the 90s. It enables to reproduce a single flange-roller end contact with its specific kinematic and geometry. Moreover, it’s a full scale experiment, as the curvature radii of the specimens are similar to those of the actual large size tapered roller bearing. With its rigid aluminium structure, the contact load is also similar to the one of the actual bearing.
Two specimens are used in the experiment. The first one is a glass disc, which represents the conical flange. Indeed, the difference between a conical and a flat shape is a slight conformity change along one direction. The other specimen is a steel cylinder which is machined at its contacting end like a roller end. Fig.3 describes the toric shape. They both spin around their geometrical axis. These geometrical axes have an angle λ between them.
The glass disc allows to visualize the lubricant during the test. Through white light interferometry measurement2), the film thickness can be measured all over the contact area during the experiment. Besides, the forces can be measured on the two specimens, allowing friction coefficients measurement. The torques in the contact plane can also be measured in order to obtain the torque generated by the contact.
To study the EHD contact with the specifics stated above, a dedicated model was developed. It was based on the model of Habchi et al.3) and further developments were led. The initial model solves the EHD problem, computing the solution to the Reynolds equation, the solids strain and the load balance. Moreover, the model takes into account the non-Newtonian and the thermal effects. Indeed, the kinematic generates high shear stress in the lubricant, which causes shear thinning and shear heating. Now, the model also includes the geometry specifics of the torus on plane contact. As no analytical description of the gap between the solid exists, the toric shape was designed with a CAD tool. The surface height was then used as the gap between the two solids.
The numerical model and the experiment allow to develop the flange-roller end contact understanding. Experiments led on the Tribogyr test-rig allow to reproduce faithfully the physical mechanisms of the flange roller-end contact. Moreover, the experimental results allow an assessment of the numerical model by comparing the two approaches results. Based on the model evaluation, it is possible to use the numerical tool to develop an insight analysis of the flange-roller end contact.
References
[1]Dormois H., Fillot N., Vergne P., Dalmaz G., Querry M., Ioannides E. and Morales-Espejel G. E., “First traction results of high spinning large-size circular EHD contacts from a new test rig: Tribogyr,” Tribology Transactions, 52, 2, 2009, 171-179.
[2]Molimard J., Querry M. and Vergne P., “New tools for the experimental study of EHD and limit lubrications,” Proc. 25th Leeds-Lyon Symposium, 1999, 717-726.
[3]Habchi W., Eyheramendy D., Bair S., Vergne P. and Morales-Espejel G.E., “Thermal elastohydrodynamic lubrication of point contacts using a Newtonian/generalized Newtonian lubricant,” Tribology Letters, 30, 2008, 41-52.