|
[Sponsors] |
Aspect ratio for CFX solver, suggested value? |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
September 25, 2007, 14:58 |
Aspect ratio for CFX solver, suggested value?
|
#1 |
Guest
Posts: n/a
|
Hi there,
I read in CFX documents (page 374, ANSYS CFX-Solver Modeling Guide) that the aspect ratio of <100 would be OK with CFX solver. Is it not too much? I was told from other sources (mainly fluent users) that aspect ratio should be maintained <5. Please correct me whenever needed. I will highly appreciate your reply. Many Thanks in advance, Kushagra |
|
September 25, 2007, 21:54 |
Re: Aspect ratio for CFX solver, suggested value?
|
#2 |
Guest
Posts: n/a
|
The smaller the better. I'd trust the CFX documentation on this one and aim for < 100. If you can do < 5, go for it. If you are really concerned, perform a mesh refinement study to see how sensitive your solution is to aspect ratio.
|
|
September 26, 2007, 00:25 |
Re: Aspect ratio for CFX solver, suggested value?
|
#3 |
Guest
Posts: n/a
|
Hi,
Aspect ratio 100 is usually OK for hex/wedge elements as long as the flow is aligned with the element faces. Aspect ratio 100 is poor for tet/pyramid elements or if not aligned with the element faces. Glenn Horrocks |
|
September 27, 2007, 15:42 |
Re: Aspect ratio for CFX solver, suggested value?
|
#4 |
Guest
Posts: n/a
|
Thanks a ton ! It helps.
Regards, Kushagra. |
|
September 27, 2007, 17:21 |
Re: Aspect ratio for CFX solver, suggested value?-
|
#5 |
Guest
Posts: n/a
|
You can generally ignore the aspect ratio limits near the wall, it is only a concern if there are high aspect ratio's in the free stream. If you try to keep the aspect ratio down near the wall you will run into conflicts with the near wall resolution requirements.
|
|
September 27, 2007, 17:28 |
Re: Aspect ratio for CFX solver, suggested value?-
|
#6 |
Guest
Posts: n/a
|
yes ! this is exactly something that was bothering me!
To resolve boundary layer, I wanted to create edge (element) prependicular to the wall of a small size (say .5 mm and increase the element size away from the wall. But to maintain an aspect ratio of 5, I could go only upto 2.5 mm maximum edge length in lateral direction. And that was causing 'unwanted' huge number of nodes. Thanks for your help ! KM |
|
September 27, 2007, 19:21 |
Re: Aspect ratio for CFX solver, suggested value?-
|
#7 |
Guest
Posts: n/a
|
Hi,
In the boundary layer the direction of flow is known - parallel to the surface. This is why hex/prism elements, aligned with the surface work well here as they are aligned with the flow direction. This means you can usually run aspect ratios in the inflation layers of much greater than 5. Glenn Horrocks |
|
September 28, 2007, 11:52 |
Re: Aspect ratio for CFX solver, suggested value?-
|
#8 |
Guest
Posts: n/a
|
Hi KM,
Don't worry about the aspect ratio next to the wall. It's only a problem when you have high aspect ratio elements in the free stream. This isn't hard to avoid with unstructured meshes, since away from the wall you are generally filling the volume with relatively orthogonal tets and, even if the tet is stretched, the orthogonality angle will be a problem long before the aspect ratio. Also, with tets you may have a few bad elements, but as long as they are isolated they often aren't a problem. -CycLone |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
DOES ASPECT RATIO PLAY A PART IN CFD RESULTS? | RIF | Main CFD Forum | 2 | January 5, 2008 14:14 |
mesh for a high aspect ratio geometry | matthias | CFX | 3 | October 20, 2006 03:55 |
aspect ratio | boe | CFX | 5 | October 20, 2006 03:13 |
aspect ratio | Joe | CFX | 3 | February 26, 2004 17:13 |
grid aspect ratio and skew | Darcy | CFX | 2 | March 10, 2001 02:52 |