CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   Multiphase Journal Bearing CFD: Modelling Air Entrainment (https://www.cfd-online.com/Forums/cfx/135447-multiphase-journal-bearing-cfd-modelling-air-entrainment.html)

eaxym May 13, 2014 14:02

Multiphase Journal Bearing CFD: Modelling Air Entrainment
 
2 Attachment(s)
Hello All,

I am trying to simulate air entrainment in a journal bearing using CFX. Oil is fed in from the feed hole and distributed into the bearing via a 180 degree rectangular groove. I have attached an image showing how a journal bearing operates. There is an inner rotating cylinder eccentric to an outer stationary cylinder. This creates a narrow converging film region which leads to generation of pressure in the fluid film. After the minimum film thickness point, the film starts to diverge and in this region the film pressure drops below ambient. In the CFD model this causes reverse flow at the bearing sides in the diverging film region.
In CFX, I specified an opening pressure boundary condition at the oil inlet at 20 kPa and an opening pressure boundary at the bearing side (outlet) at 0 kPa (ambient). I used openings because there is reverse flow occurring. When I run the simulation in single phase, it converges well. Then I turned on the CFX cavitation model, which lead to generation of oil vapour in the diverging part, and it also converges well. I used the steady state solver and homogeneous multiphase model and specified volume fraction of oil liquid equal to 1 at both the inlet and outlet.
Since I also want to model air being entrained from the bearing sides (due to reverse flow), I then specify volume fraction of air at the bearing sides (outlet) to be equal to 1. When I do this, the simulation becomes very unstable and the mass imbalance keeps fluctuating irregularly between +- 30% and it doesn't converge.

For simplicity I am only modelling a half bearing with a symmetry boundary condition being applied.

I need help on how to model air entrainment. Please could anyone suggest the right way to do it and suggest what is going wrong at the moment.

Thanks a lot!

ghorrocks May 13, 2014 19:29

Your outlet boundary is right int he middle of complicated and important flow physics. This does not sound like a good place to put a boundary to me. I would move the boundary further away from the action to a location where the conditions are much simpler.


All times are GMT -4. The time now is 02:17.