[mohammad]

Hi everybody,
I am simulating a wing span without twist and not tapered.
I want to locate nodes exactly on a specific spanwise location (say, mid-span location). This plane is to guaranty the existence of nodes at mid-span location for POST-PROCESSING.

So, it came to my mind that i can create a virtual( dummy) plane in ICEM at the mid-span and then divide the blade surface and the domain by this plane.
so at the desired location of mid-span , i have some curves and then i can associate the the nodes to these newly made curves. By this way , I will have nodes at exact location of mid-span.

but this rises the following question;" when i import the mesh into the CFX_Pre, I need to assign a boundary condition to this new plane.
What should be this boundary condition to assure no influence on the flow field?

Can i use OPENING?

Please if you know any other practical way to make sure that the nodes are EXACTLY located on mid-span let me know.

Thanks

[Far]

Originally Posted by mohammad

Hi everybody,
I am simulating a wing span without twist and not tapered.
I wanna to locate nodes exactly on a specific spanwise location (say, mid-span location). This plane is to guaranty the existence of nodes at mid-span location for POST-PROCESSING.

So, it came to my mind that i can create a virtual( dummy) plane in ICEM at the mid-span and then divide the blade surface and the domain by this plane.
so at the desired location of mid-span , i have some curves and then i can associate the the nodes to these newly made curves. By this way , I will have nodes at exact location of mid-span.

but this rises the following question;" when i import the mesh into the CFX_Pre, I need to assign a boundary condition to this new plane.
What should be this boundary condition to assure no influence on the flow field?

Can i use OPENING?

Please if you know any other practical way to make sure that the nodes are EXACTLY located on mid-span let me know.

Thanks

Well if you are simulating the blade cascade then apply symmetry condition, I have no idea about the wing.

[mohammad]

Originally Posted by Far

Well if you are simulating the blade cascade then apply symmetry condition, I have no idea about the wing.

Hi Far,
Because i am doing DES , so the symmetry BC is not valid....

Thanks a lot

[Far]

Then use slip wall, it has same effect

[ghorrocks]

Select the multiple regions in the same domain. Then in the internal surfaces are ignored, providing the meshes match.

Also be aware that the nodes are surrounded by control volumes and the values represent the average in the control volume. This means you have the average of a control volume centred at the mid point, not a point.

If you put an interface or BC at the mid plane then you are biasing the control volume to one side or the other. Then you will have not got a control volume centred on the mid point.

[mohammad]

Originally Posted by ghorrocks

Select the multiple regions in the same domain. Then in the internal surfaces are ignored, providing the meshes match.

.

Dear Glenn,
So, you mean that I must create that surface and the related association in ICEM. Then I must follow your steps in CFX-Pre. Am I right?

[ghorrocks]

You can connect the 3D regions with interfaces but try to avoid it. Usually you can get the mesh to match and that is much better. If it does not currently match you will have to remesh.

[mohammad]

Originally Posted by ghorrocks

You can connect the 3D regions with interfaces but try to avoid it. Usually you can get the mesh to match and that is much better. If it does not currently match you will have to remesh.

Thanks a million