CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   how to split a surface in CFX POST (http://www.cfd-online.com/Forums/cfx/105146-how-split-surface-cfx-post.html)

mohammad July 24, 2012 02:33

how to split a surface in CFX POST
 
Dear all,
I have fnished the calculation of a blade...while meshing in the ICEM , I defined the blade surface as a single surface.

Now I want to see the force and moment on some radial section of the blade, say from R1 to R2....

How can I split the blade surface in the CFX-POST to create that smaller surface?

Thanks a lot

ghorrocks July 24, 2012 03:21

Yes, use isosurfaces of radius to define a new surface on your blade surface between your radii. Then you can do calculations on these new surfaces.

You can also set this up using a contour tool, drawing contours of radius on your surface. Then you can extract the surface between contours as a user defined surface (I think that is correct).

mohammad July 24, 2012 03:44

Quote:

Originally Posted by ghorrocks (Post 373149)
Yes, use isosurfaces of radius to define a new surface on your blade surface between your radii. Then you can do calculations on these new surfaces.

You can also set this up using a contour tool, drawing contours of radius on your surface. Then you can extract the surface between contours as a user defined surface (I think that is correct).

Dear Glenn,
First of all, thanks for your useful reply.
The second method is correct....geometrically it gives the radially extended area.
But..about the first method... I am wondering after creating those two isosurfaces, how to create a surface between two radii( Iso-surfaces) on the blade?

Thanks:)

mohammad July 24, 2012 04:41

Veryyyyyy Important and STRANGE....
 
As a test to see the accuracy of the results, I divided a surface to n smaller part using the second method of Glenn...
The results are dramatically different.... This might be because the integration algorithm and mesh size on the surface Vs. radial surfaces...

To the readers:
I will post more things as I understand the reason of this HUGE difference:)

ghorrocks July 24, 2012 19:31

The mesh on these sub-surfaces should just be the original mesh chopped up. The sum of the surface bits should equal the whole surface, and if not make a careful check you are doing this correctly.

mohammad July 24, 2012 21:40

Dear Glenn;
I in one of my files I have a blade whose surface is composed of 4 smaller surfaces. I tried with this file:


Quote:

Originally Posted by ghorrocks (Post 373337)
The mesh on these sub-surfaces should just be the original mesh chopped up.

I divided my blade radially to 100 small pieces, meaning each radial strip width equals to (1%* Blade Radius) . This method helped me to fit the newly created surfaces with high accuracy to one of those 4 surfaces which exist on blade surface. The surfaces overlap is nearly 95%.



Quote:

Originally Posted by ghorrocks (Post 373337)
.... and if not make a careful check you are doing this correctly.

then I went to "Expression" and i created expressions for the Fx, Fy, Fz; one for that surface which originally belongs to the blade and one for all newly created radial surfaces...

Amazingly the results are different.....

To my knowledge, this could JUST be because of one thing.....when we create some surfaces on "CFX_POST" , the software might work with the interpolated data instead of those original nodal values and hence the results become different. This seems correct to me, because I have highly separated flow on the blade and high mesh density near the blades surface...thus i think the gradient values of the pressure are high in that area, which make the interpolation ( and thus the values on the new surfaces) erroneous.


To investigate this matter more, I created a flat plane and tried to use "Sample " method and " Slice" method. the results were different even with the change in "#of sample points" the results of "Sample" became different....

I would be very happy if these differences can be explained to me.

Regards,

mohammad September 29, 2012 09:59

Quote:

Originally Posted by mohammad (Post 373138)
Dear all,
I have fnished the calculation of a blade...while meshing in the ICEM , I defined the blade surface as a single surface.

Now I want to see the force and moment on some radial section of the blade, say from R1 to R2....

How can I split the blade surface in the CFX-POST to create that smaller surface?

Thanks a lot

Location> Iso clip
then define the radius as the varaiable and limit the "Iso clip" with R1 ,R2


All times are GMT -4. The time now is 06:45.