A paradox in tribosimulation: ball bearing grease friction

Tribometers are primarily used to measure contact friction between two bodies in relative motion, lubricated or dry. In recent times ball on disc contact test rigs have become very popular. They allow pure rolling motion, rolling with sliding, and pure sliding motion, and seem the perfect tool to simulate ball bearing lubrication under laboratory circumstances.

In this presentation lubricating grease friction obtained with a ball on disc rig is compared to friction in a full ball bearing for a series of greases. The greases chosen purposely show very different traction behaviour. The ball on disc rig is a so-called mini traction machine. The bearing is an axially loaded deep groove ball bearing.

For full film lubrication, friction coefficients obtained for greases with high friction coefficient correlate well for both rigs. A qualitative ranking of friction values is the same for both devices. This is less so for very low friction greases, having small differences in friction. Moreover, the values of the coefficient of friction between the two devices differ by one order of magnitude. This apparent inconsistency is attributed to various mechanisms, which affect both film thickness and friction, ball spin in the bearing being the most important one. The two devices must be seen as different tribosystems, yielding different results. Despite the large differences, which lead to complementary use of the two devices, agreement exists in a qualitative ranking of grease friction.