The experimental validation was carried out in collaboration with Drive System Design
Ltd. (DSD) (http://www.drivesystemdesign.com/). DSD performed a
series of experiments taking a single gear and rotating it in an oil sump, while
varying the level of the oil and the RPMs. The sketch of the setup is shown in Figure 1 below.
The rotation speed is changed from 500 to 3000 RPM in increments of 500 RPM. The oil
level is varied from 40 to 80 mm, measuring from the axis of the gear.
Numerical Setup
The initial particle arrangement is obtained by directly discretizing the CAD model
provided by DSD without any simplification. This results in approximately
approximately 8.6 million particles (dx = 1e-3 m) for oil and air phases
combined.
Results
Qualitatively nanoFluidX has shown very good agreement with the
experiments. Similar flow structures between the experiment and the simulation
solution have been observed for every case.
Most notable qualitative result is the observation of the windage effect in the 80 mm
oil level case. In the 80 mm case, the gear is not immersed in the oil sump, but
rather hovers over the oil surface with several mm distance. At 3000 RPM the windage
effects (motion of the air due to the gear rotation) become prominent and the oil’s
surface is disturbed in an oscillatory manner, causing occurrence of high frequency
waves. Furthermore, small amounts of oil are carried around the domain by the
windage effects, causing visible splashes. This behavior is very well captured by
the simulation. Instantaneous time frames can be seen in Figure 3 (experimental) and
Figure 4 (nanoFluidX simulation).