[urosgrivc]

Hello,
I have a question and a slight problem setting my BCs the wright way.

I am simulating a brake discs thermal dissipation properties, brake disc has cooling ribs between two brake plates. figure below;

https://drive.google.com/open?id=0Bw...UtxdmdObXhzUzQ

I would like to predict ventilation losses and average HTC for different surfaces and angular velocyties.

The simulation is done with cfx 18.1.
Only the air surrounding the brake disc is simulated and 1/12 of the model is simulated (rotating domain). I will neglect the ongoing velocity and only use the angular velocity of the brake disc. Temperature of the brake discs surfaces is constant 100°C.

Everything is working nicely and convergence is good I have also made a mesh independence study and set Tbulk to constant.

The problem occurs that my results depend only on what frame type I chose for opening (rotating or stationary):
If I chose Stationary than: Ventilation loses ~30% higher, avgHTC~20% higher on some areas and 20% lower on friction surface area than values for rotating opening frame type.

What changes by changing frame type and what should I choose now?
Can someone help I can not find any direct ansver to this problem and even what changes for opening type BC by choosing frame type.

Personaly I think that I should use stationary, but I would like to be 100% about it, do you think I could improove the domain?? any suggestion will only help
I can see some huge diferences in post for velocity in stationary frame at the opening boundary. It seems like the flow gets an (omega*R) part if rotating is choosen.