Hi all,

I have performed a simulation for a rotating piece of equipment in a stationary frame of reference. For this I have defined an angular velocity on the rotating wall. This gave me a good solution.

Now I would like to perform the same calculations in a rotating frame of reference. Is there a way to use the results of the calculations in the stationary frame of reference as an initial guess for this new calculation? Pressure, temperature and additional varaibles are OK but of course the velocity components are incorrect.

Thanx, Pascale

Hi Pascale,

You did this in CFX-5?

Robin

That is a YES.

Pascale

Hi Pascale,

Wall boundary conditions cannot enforce flow normal to the wall, regardless of the expression used, only tangential velocities can be specified. Unless your geometry is made entirely of swept surfaces, your initial analysis will be incorrect.

Regards, Robin

Hi Pascale,

I do not understand your concept of solving something which is moving in a stationary frame of reference. Do you mean that you specified the fluid domain as stationary, and then in the def file you added a parameter for the wall - angular speed ? Is that possible? I wonder what the solution represents.

Anyway, I am myself puzzled about the way CFX treats rotation. Please have a look at my recent post Re: Rotation effects. Your opinion will be appreciated.

Martin

Hi Martin,

That is exactly what I did. I have a cilinder that is rotating in a rectangular space. So the space is stationary and the cilinder just has an angular velocity on the outer wall.

I will read your other postings.........

Pascale