CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FloEFD, FloWorks & FloTHERM (http://www.cfd-online.com/Forums/floefd-floworks-flotherm/)
-   -   Rotation Simulation with FloEFD (http://www.cfd-online.com/Forums/floefd-floworks-flotherm/89181-rotation-simulation-floefd.html)

ksegond June 7, 2011 07:10

Rotation Simulation with FloEFD
 
My question is related to the cooling of a rotational equipment.
When the cooling fluid enters the rotor, its circunferential component is accelerated to the rotational speed. Moving mesh would be far too complicated for my case. What could be the trick to simulate the action of the rotation in another way?

PS: I have never used FloEFD so far. I have used Star-CD, Fluent and CFX

tim_shute January 9, 2012 22:45

Hi karim
As I learnt from ur description, this is a fluid structure interaction (FSI) coupling question, which is out of EFD's reach. Since you've used Fluent, I suppose you may know how to solve it associating ANSYS and fluent. In EFD, the expediency is to simulate it and obtain the torque of the circunferential component. You could roughly estimate the performance compared with component's start torque or rated torque, even further substitute/transfer the result into SW's other modules—dynamic/motion simulation. If you cant tell rated torque of the components, make it rotating in flow simulation first and get it (average result). The reason why I say "roughly estimate" it is because the force/torque exterted onto your component driven by wind/fluid is an alternating loading based on my understandings on wind turbine principle

hope it could help
Tim

Boris_M January 17, 2012 05:03

Hi both,
Tim is not quite right, it doesn't sound like a FSI task. It's a standard rotor fluid interaction. The rotor drives the fluid and therefore accelerates it in circumferential direction. There's nothing FSI about it.
Now I don't know the type of your system as some systems require either moving mesh or sliding mesh but in a lot of cases this is more than you need. Often it is enough to use the RRF approach that FloEFD has. RRF stands for rotating reference frame. You basically define a region in which the reference frame is switched to a rotating one and therefore the fluid rotated around the rotor instead of the other way around in cases when you use moving or sliding meshes. so radial and axial blowers are usually best suited for this approach.

Regards,
Boris

kimmj January 24, 2012 23:06

Quote:

Originally Posted by ksegond (Post 310837)
My question is related to the cooling of a rotational equipment.
When the cooling fluid enters the rotor, its circunferential component is accelerated to the rotational speed. Moving mesh would be far too complicated for my case. What could be the trick to simulate the action of the rotation in another way?

PS: I have never used FloEFD so far. I have used Star-CD, Fluent and CFX


Hi all
EFD use "Mixing Planes" method.
Description is
"Similar to Frozen rotor, but domain interface is treated differently. Instead of transferring solution directly the solution is circumferentially averaged."

"Plus points are
-Relatively low computational resources
-Circumferentially averaged solution has more applicability to engineering problems than “frozen” solution.


Minus is
-Transient phenomena are not taken into account.
-Valid for cases in which the flow on the boundary of rotating subdomain is “close” to axisymmetric"


Hope to help..

P.S
by Flometrics data..


All times are GMT -4. The time now is 18:00.