CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   How to study different stationary states (http://www.cfd-online.com/Forums/fluent/35538-how-study-different-stationary-states.html)

Marta January 22, 2005 14:31

How to study different stationary states
 
I have to study an object (in the form of a brick) situated in a wind tunnel and I'm interesting to see the differences at various inclinations. Each simulations is in a steady-state. At this moment, I'm creating manually with Gambit different inclinations one at time and then exporting mesh files in Fluent to run each inclinations one at time.

I was thinking to use an appropriate journal file to create the different geometries at the different inclinations to export mesh file: is it possible? Is it a good idea? How can I do to submit to Fluent all mesh files in a single step?

Otherwise, is it possible, to do so, using dynamic meshes or another method, without creating the different inclinations with Gambit?

IMPORTANT: Take into account that I need, for each inclinations, the study of a steady-state and not an unsteady-state!

thank you, Marta

Jason January 24, 2005 09:04

Re: How to study different stationary states
 
This depends on what you are trying to take into account... If you don't care about wall effects, then I would model the control volume as a hemisphere and change the flow vector in Fluent to give different inclination angles. This way, your force and moment monitors are always setup along the body-axes. If wall effects are important, and you don't want to completely remesh the model every time, you could use a non-conformal grid interface. Split the control volume with a sphere or cylinder that has the brick in its center and the center of the sphere (or centerline of the cylinder) should be coincident with the axis of rotation of your body. When you split the volumes make sure "connected" is turned off. Then you can mesh the two volumes separately, and define the coincident faces as interfaces. Then, for each inclination, you can rotate the volume containing the brick and export that mesh. As long as your mesh is reasonable, Fluent will create a smooth transition across the non-conformal interface and you won't even know it's there from the results. Just remember that you're rotating the body, and if you're trying to get any moments out of the results, they are about the cartesian coordinate systems (which is equivalent to a wind-axis system instead of a body-axis system).

Hope this helps, and goodluck Jason

Luca January 24, 2005 10:10

Re: How to study different stationary states
 
I would define a pressure far field and apply a different velocity vector for each analysis. Luca

Jason January 24, 2005 13:32

Re: How to study different stationary states
 
I agree, but that only works if you're not interested in wall effects from the tunnel wall. That's why I gave both options.

Jason


All times are GMT -4. The time now is 18:41.