The Friction Facts Hub Tester measures the friction created by the hub; specifically, the friction created by the rotating bearings, under load, within the hub.
The Hub Tester was designed to simulate real world conditions, while achieving a high level of measurable accuracy and precision. The following considerations were incorporated into the design.
Traditionally, hub friction has been analyzed by placing an assembled wheel on a drum, applying load, and bringing the drum up to speed. The total work required to turn the drum is then measured. The work contributions stemming from drum bearing friction, tire rolling resistance, and aerodynamic drag of the wheel and drum are part of the total system losses. Thus, subtracting these components from the total measured work gives the work done to turn the hub, ie, the hub friction. If the contributing factors are kept at a constant, it is theoretically possible to single out the friction of the hub.
However, the major drawback to this type of tester is that the frictional losses produced by the contributing factors are substantially greater than the friction created by the hub. For example, if friction due to the tire rolling resistance and aerodynamic drag of the wheel and drum amount to 50 watts, yet the friction of the hub is 3 watts, the hub wattage is only 6% of total system losses. The hub friction becomes diluted within the overall losses. When a larger full scale torque transducer is used, the relative accuracy required to effectively measure the smaller friction in the hub is reduced.
Because of the accuracy limitations seen with the drum-style test method, a novel design concept was used in the creation of the Friction Facts Hub Tester. This design lends itself to a much higher level of measurement resolution and thereby accuracy when analyzing the hub friction. By using a ‘bearing within a bearing’ design, the Friction Facts Hub Tester measures the friction created only by the hub and dedicated fixture bearings, and is not affected by any friction contributions of tire rolling resistance nor wheel aerodynamic drag. Ultimately, this design allows the use a lower full scale torque transducer, effectively increasing accuracy by an order of magnitude or greater when compared to a drum-style tester.
With the Friction Facts Hub Tester design, the wheel is non-rotating, with a load applied, while the fixture bearings both support and rotate together with the rotating hub axle. This design functions in the opposite manner to a traditional rotating wheel and fixed axle, yet the force diagrams on the hub are identical. Instead of a spinning wheel, the axle is spinning. When the axle is rotated, the torque transducer is measuring the friction of both the hub bearings and the fixture bearings, with no frictional contributions from the tire, aerodynamics, drum, etc.
The known fixture bearing friction is subtracted from the raw friction measurement, leaving only pure hub friction. The ceramic fixture bearings are calibrated, and calibration curves created over a range of both radial and radial-axial loads. Using these defined curves, the fixture bearing friction can be treated as a constant.
The accuracy of the Hub Tester is +/- 0.05 watts.
Loading Yoke: The loading of the hub is achieved by the use of a yoke bar centered across the top of the wheel. Each end of the yoke is attached to large-displacement springs and a ratchet system, with load cells in series. Each of the ratchets can be tightened to create a specific load on the yoke. In turn, the load is transferred to the wheel. The force on each side of the yoke is measured by one of two load cells. Each load cell is summed to get the total loading on the wheel.
Creating a compressive load on the wheel is crucial to the design. Hanging a weight from the bottom of the wheel, creating a tensile load, would have been a simple and easy-to-use design. A hanging weight would create the same general vertical loading on the hub as with the yoke design. However, the load transfer at the spoke/hub attachment point would be different than real-world. A hanging weight would create a tensile condition on the wheel, rather than compressive condition. It is possible that the hub bearing friction is affected by the loading characteristics of the spokes onto the hub. Because of this, the yoke is used to simulate the real-world compressive nature of the load, thus achieving a similar spoke-loading situation. In other words, the yoke simulates the weight of the rider on the wheel pushing against the ground.
To achieve angled loading, the point at which the ratchets are anchored to the floor can be adjusted laterally. The angle of the ratchets/springs relative to vertical is reported as the angled load applied. Angled loading is used to simulate any non-vertical riding orientation, such as sprinting.
Testing of Stand-Alone Hubs (no wheel): Stand alone hubs can be tested. For this application, two straps are placed over each end of the hub (one per side, directly over the location of the hub bearings), and weight is hung on the straps. The straps transfer the vertical force onto the hub. While this method can and is used, it is not the preferred method. This method does not simulate the real-world transfer of load at the spoke/hub attachment points, as described in the above section. The preferred method of testing is to test the fully assembled wheel.
Radial Ceramic Fixture Bearings: The fixture bearings used to support the axle are much larger in diameter than the respective hub bearings. The large bearing diameter decreases the relative friction contribution of the fixture bearings. The seals were removed from the fixture bearings and the grease replaced with low viscosity bearing oil to further reduce the friction in the fixture bearings.
Thrust Ceramic Fixture Bearing: When the wheel is subjected to pure radial force loads, the radial ceramic fixture bearings, which are ceramic deep-groove radial bearings by design, are operating at their most efficient level (as radial bearings in general are designed to operate in radial loading situations). As the wheel load is varied from pure radial to a combination of both radial and axial, (the above referenced angled loading), not only are the hub bearings subjected to increased axial load, the fixture bearings are also subjected to the same axial loading. The tester is designed to always minimize the friction contribution from the fixture bearings. With this in mind, thrust bearings are incorporated into the tester during angled loading to remove the axial load from the radial fixture bearings. As the radial fixture bearings operate most efficiently with pure radial loads, similarly, the thrust fixture bearings operate most efficiently with pure axial loading. When the two radial bearings and the single thrust bearing are used in conjunction with one another, the radial bearings support the radial load and the thrust bearing supports the axial load, allowing all three bearings to operate at maximum efficiency, thereby minimizing the frictional losses due to the fixture bearing assembly.
Meanwhile, the hub bearings, regardless of design, are fully subjected to all vectors of the load situation, whether pure radial or a combination of radial and axial loading is applied. This is the intent of the angled loading to subject the hub bearings to multiple angles of loading and analyze hub efficiency at various angles.
Axle Adapters: The outside diameters of the hub axles are dimensionally different by manufacturer and model, and since the inside diameter of the fixture bearings is much larger than the outside diameters of the axles, axle adapters were fabricated to seat and center the axles at the rotational axis of the fixture bearings. Multiple adapters had to be machined at 1/1000 in gradients to accommodate the slightly different axle diameters.
Quick Release Force: In order to accurately simulate the compressive side-loads applied to the hub by the quick release skewer, the main shaft of the tester is a threaded rod-design. Nuts are located on either side of the hub axle, and tightened to create the compressive force. A torque wrench is used when tightening the nuts to achieve the desired force, and also repeatability of compressive force between hub tests.
Fixture parallelism: The hub bearings and fixture bearings must exhibit parallelism. If they are not parallel, an undesirable angular loading situation could be introduced within the bearings, creating erroneous friction. Fixture parallelism is ensured by using a connecting rod at the top of the fixture and struts on both fixture plates.