Quick question about MFR
 

Hi, CFX guys

I am doing a impeller ratation tutorial using MFR method, and was wondering if I need to set a rotational speed for balde wall and shaft wall boundaries.

In the tutoral, only the impeller domain speed was defined. I want to confirm that I do not need to define the rotation velocity for all parts inside this domain.

Thanks!

Rob

Re: Quick question about MFR
 

Hi Rob,

For rotor/stator simulations then usually you have stationary walls and rotating walls. Walls in the stationary domain are assumed to be stationary, and walls in the rotating domain are assumed to rotate. You can make a wall on the rotating domain stationary by setting it to "Counter-rotating wall". You should check in CFX-Post that the walls are rotating correctly by looking at the velocity and velocity in stationary frame variables.

Glenn Horrocks

Re: Quick question about MFR
 

Hi, Genn,

Thanks. The blades should rotate with the domain, so I understand that I do not need to set a rotational velocity for the blades.

I did a simulation example, one phase problem and the impeller was put in a rectangular tank. In the Post, my results showed that a row of velocity vectors are perpendicular to the shaft. Do you have experience that what may cause such a problem?

Regards!

Rob

Re: Quick question about MFR
 

Hi,

What variable were you looking at? "Velocity" or "Velocity in stationary frame"? The velocity variable is relative to the local frame of reference, and vel in stn frame is relative to the absolute (or stationary) frame of reference.

Thus walls in a rotating domain which are stationary in the absolute frame will show a velocity at the wall, but will show no velocity in stn frame.

Glenn Horrocks