CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   stopping and re-starting moving mesh for static lift coefficients (https://www.cfd-online.com/Forums/cfx/63796-stopping-re-starting-moving-mesh-static-lift-coefficients.html)

haconk April 20, 2009 05:19

stopping and re-starting moving mesh for static lift coefficients
 
I'd like to calculate the static lift, moment and drag coefficients for various angles of incidence.

I created one model with a moving mesh but how do I make that transient model a stationary one within the run-time environment?

Or do I need to create individual models and do separate runs for each angle of incidence?

ghorrocks April 20, 2009 18:42

Hi,

Using moving mesh to do various AOA is a very complicated way of doing it. Just change the angle of incidence at the inlet boundary. If it is a simple steady case then use a simple steady model.

Glenn Horrocks

haconk April 21, 2009 06:18

Thanks Glenn,

I have a moving anyway for the transient runs, besides the model actually changes with various AOA (actively controlled), that's why I'd like to use the moving mesh method.

Just wondered whether I could obtain the stationary results by stopping and restarting the run and if so, how? Couldn't find anything in the manual re this.

ghorrocks April 21, 2009 07:16

Hi,

You can't use moving mesh in steady state runs. Also I don't think you can move a mesh to a position and use that mesh for a steady state run.

You would only bother doing a moving mesh simulation of what you are doing if the flow timescales are similar to the wing motion timescales. If the wing is much slower than the air it can be accurately modelled as a series of steady state models at the various positions.

Glenn Horrocks

Tristan April 27, 2009 16:05

We have done some moving mesh simulations to get "steady-state" port discharge coefficients as the ports are closed. Rather than generate a sequence of meshes and run them steady-state one at a time, we did a moving mesh transient simulation where the ports closed very slowly. In order to make sure our results were "steady-state", we did a 2nd run where it took twice as long to close the ports. A comparison of the two runs showed no difference between them until the ports were nearly closed.

I doubt that this was the most efficient way to solve the problem but it sounds like it might be an option for you.

Tristan


All times are GMT -4. The time now is 13:38.