CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] Problems about meshing the swept wing with cylinder end cap (http://www.cfd-online.com/Forums/ansys-meshing/76035-problems-about-meshing-swept-wing-cylinder-end-cap.html)

kepeng May 12, 2010 07:05

Problems about meshing the swept wing with cylinder end cap
 
2 Attachment(s)
Hi, everyone, I have some prolems about meshing the swept wing with cylinder end cap, wish to get your help, thanks.

As shown in the attachment, i first meshed the 2D airfoil then translate to 3D, , then delete the block inside the wing.

but there are two problems,
1) the grid inside the tip face of the wing, where are triagles.
2) the quality of whole mesh is very poor, somes is below 0.0.

could someone give me some advice to improve such mesh?

Thanks a lot.

PSYMN May 24, 2010 13:29

Suggestions for now...
 
2 Attachment(s)
Hey KePeng,

I took a quick look at your model and noted several areas that could be improved...

First, your far field. It is a strange shape that leads to poor quality... Does it need to be that shape or was it somewhat arbitrary?

Second, I see what you are doing by creating construction Geometry between the wing tip and the far field, but this isn't necessary for ICEM CFD. You don't need to subdivide the volume with geometry. In your case, you also projected the edges at the far field to this construction geometry... Totally not required and it actually reduces your mesh quality with an unnecessary constraint.

Third, Near the wing tip, you project the end of the blocking to the perimeter curve. With ICEM CFD hexa, you do not need to respect the surface boundaries. You could move those front corner verts out more onto the wing tip to get better angles with less distortion.

For instance, this is what you have, not the bad elements near the tip (and all the way out to the FF)...
Attachment 3508

To fix this, I split the edge (top and bottom) and associated the front half to surface. Then I used "Move Vertex" to move these out a bit to the "real front corner" of the wing. This both minimized (averaged out) the distortion and removed the profile curvature.
Attachment 3509


You also didn't create an OGrid inside your wing... This is acceptable if you square off the tip a bit, I probably would have gone another way with that though and either put the CGrid partially inside the wing (collapse three blocks) or wrapped the wing entirely with an Ogrid and just collapsed the back edge. This would have resulted in pyramids though... I am guessing that you can handle the wedge elements... How about Pyramid elements? What is your solver?, some don't mind a bit, in which case, I could get rid of all the prisms (Wedges) in your model in exchange for a few pyramids.

Get back to me on these questions and I can come up with the optimal blocking strategy for you.

kepeng May 25, 2010 09:57

1 Attachment(s)
Thanks for you advices very much!

As for the problems you mentioned, i give the reply as follows,
1) the far field has been changed.
2) all the construction geometry has been deleted
3) i will try later.

i tried again and failed, as shown in the attachment.
currently, i don't know how to mesh without triangles inside of the 2D airfoil.
I tried the O-grid but there still some triangles inside.

btw, I develop a code to manipulate the Fluent cas and dat file, but current the code can only deal with single element type, so the simplest mesh is better. However, Unstructured grid could be generated easliy, but the precision of flow simulation result is unacceptable.

Waiting for your suggestions more, thanks a lot.

kepeng June 15, 2010 05:21

Haha, got it!
 
2 Attachment(s)
Finally i got an usable mesh using ICEM, As show in the picture.
Thanks for your advice.

But the quality needs to be improved.


All times are GMT -4. The time now is 18:04.