Group Product R&D Director
SKF Engineering & Research Centre
Utrecht, The Netherlands
Edited by Kenneth Korane
Any golfer knows that a ball with dimples has vastly superior aerodynamic characteristics versus one that's completely smooth. Researchers at SKF have found that dimpled ball and roller bearings produce similar benefits in terms of lubricant flow.
This may result in a new generation of bearings that perform better than today's versions in less-than-optimum lubrication conditions.
Until recently, internal surface design of bearings had a narrow focus. With few exceptions most efforts went into creating smoother surfaces, and most theory was based on smooth-film calculations. Variations from smooth surfaces were approached strictly on a trial-and-error basis.
In the late 1990s, engineers began to take a closer look at surface effects because with advances in steel chemistry and bearing manufacturing, traditional metal fatigue was no longer the main cause of bearing failure.
Instead, surface-originated failures became more prominent. A large portion of these are attributed to the lack of a fully separating lubricant film that permits metal-tometal contact. Causes for an inadequate separating film include high operating temperatures, low speeds, and severe lubrication starvation which, in turn, trigger failure.
Researchers now have a good understanding about how surface conditions such as roughness, topography, and the direction of microgrooves all affect bearing lubrication. This is leading to the design of bearing surfaces that optimize performance and extend service life.
Magnifying the surface finish of a bearing raceway many thousands of times shows that machined surfaces are never perfectly smooth. Generally the raceway of a roller bearing has deviations from true smoothness with an average height of 0.1 microns. A typical contact between ring and roller in a cylindrical roller bearing is 100 microns wide. The lubricant that separates the roller from the ring has a typical thickness of a few times the roughness height for a well-separated condition. Or, on another scale, if the contact area were the size of a soccer field turned on its side (100 meters), the lubricant film would be about the height of two soccer balls.
The lubricant separates and protects the surfaces in most cases. However if, for example, the speed or viscosity is low or the contact is severely starved, film thickness will fall below 0.1 micron. The running surfaces will touch and metallic contact starts to destroy the surfaces. This reduces the life of a bearing.
SKF has made detailed calculations of lubricant film thickness and surface roughness in the contact area. The calculations also predict the "deformed" surface roughness, because under extremely high pressures in the contact area, surface asperities deform somewhat before contact occurs.
Two major developments have been made in this respect. First is the ability to predict the so-called "lift-off," the point when the deformed-roughness surfaces separate and ride on the lubricant film. Next, a method has been developed to calculate lubricant transport in the contact zone as well as how introducing special surface features, such as surface pits, can redistribute the lubricant.
To determine the best surface finish for an application, researchers calculated the real contact area as a function of speed. Two phenomena compete when a bearing transitions from high to low speed: both the film thickness and the surface-roughness height are decreasing, the latter due to high-pressure deformation. The decrease in film thickness, however, predominates.
Therefore the real area of contact — the area of asperities sticking through the film — increases when the bearing slows from high to low speed. The speed at which metal-to-metal contact starts to become significant (the lift-off speed) obviously should be as low as possible, and is one of the main criteria for designing an optimally performing surface.
A bearing running under starved lubrication conditions requires a low lift-off speed and, in addition, the surface structure should be designed to enhance lubricant transport towards the inlet of the contact.
Researchers at SKF have investigated many different surface-finish alternatives with this goal in mind. To date, dimpled, golf-ball-like surfaces provide the most-efficient lubrication transport. The pits on the surface act as lubricant reservoirs, and rolling action transports oil or grease into the inlet of the contact. Once a lubricantfilled pit enters the contact, localized high pressure slightly compresses the pit.
It releases part of its content in the inlet and subsequently pushes the inlet meniscus upstream, which reduces the level of starvation for a short time. The surface should be designed such that the next pit follows before the inlet meniscus reverts to its original position. The result is a well-defined array of pits. Extensive computer simulations show that a surface structure with 1-micron-deep pits can triple the lubricant-film thickness. Further work is underway to map the relationship of pit size, shape, and distribution as a function of separation for different application conditions.
These theories and calculations have been verified by extensive testing at the SKF Engineering & Research Centre in the Netherlands. In parallel with the theoretical and experimental work on the lift-off and separation mechanisms have been investigations into the manufacturing techniques to produce such surfaces. The delivery of optimally designed surfaces for specific applications will be part of SKF's new generation of bearings currently under design. The resulting products will require less running-in and feature low friction and longer life.