CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   CFX Post- Force function problems? (http://www.cfd-online.com/Forums/cfx/109196-cfx-post-force-function-problems.html)

watttt November 12, 2012 05:47

CFX Post- Force function problems?
 
Hi there,

I am currently trying to simulate a simple prolate ellipsoid (http://en.wikipedia.org/wiki/Ellipsoid) in a fluid flow. But I am currently having problems showing that the simulation is independent of the position of the outlet. I have set the outlet as an average static pressure of 0 and can see form a pressure plot on the sym plane that the pressure has stabilised behind the object. But whenever I lengthen/shorten the outlet the drag force (calculated using the force function) changes! To make matters worse it doesn’t even change in a pattern, the force just randomly jumps about by 10% both up and down. Has anyone ever seen this before? The mesh I am using is consistent between all runs and looks pretty reasonable. This makes me think its a problem with the force function? I could be wrong through. Its driving me nuts!

Thanks

murx November 12, 2012 09:57

Hi,

unfortunately I do not have a solution for your problem. But have you tried to calculate the forces manually by integrating the inertial and viscous forces around the elipsoid?
e.g. areaInt_x(Pressure)@elipsoid + areaInt(Wall Shear X)@elipsoid
Does the resulting value show the same behaviour as the force functions?

Maybe this can give you a hint where your problem originates from.
I have some similiar problem where the manual integration always differs from the force functions.

ghorrocks November 12, 2012 17:35

Can you post an image of your body and the outlet position? Also include some mesh details.

watttt November 14, 2012 09:01

4 Attachment(s)
Quote:

Originally Posted by murx (Post 391653)
Hi,

unfortunately I do not have a solution for your problem. But have you tried to calculate the forces manually by integrating the inertial and viscous forces around the elipsoid?
e.g. areaInt_x(Pressure)@elipsoid + areaInt(Wall Shear X)@elipsoid
Does the resulting value show the same behaviour as the force functions?

Maybe this can give you a hint where your problem originates from.
I have some similiar problem where the manual integration always differs from the force functions.


Interestingly this manual calculation does differ to that force function, but it still shows the same problems I was having before.


Quote:

Originally Posted by ghorrocks (Post 391722)
Can you post an image of your body and the outlet position? Also include some mesh details.

I have attached some pictures of the mesh

Defauly body spacing=0.7m
default face=0.035-0.7
face on body=0.03-0.031 with 50deg angular resolution
infaltion= 5 layers max height 0.01m
Max yplus=17

Inlet: 3m/s Medium Turbulence
Outlet: Average Static Pressure over whole outlet 0 Pa Relative


Looking at the forces given in the .out file it also varies between outlet lengths. So maybe it is something to do with my mesh.

ghorrocks November 14, 2012 17:45

Your mesh does not have inflation layers. Alternately the inflation layers are so small that the transition from the inflation layers is terrible. Yes, the problem is your mesh.

You need to use inflation layers for any flow which generates a significant boundary layer, and the transition from the inflation layers to the bulk mesh need to have roughly the same volume elements on both sides. You might also need to refine the wake area a bit.


All times are GMT -4. The time now is 12:37.