|
[Sponsors] |
December 21, 2003, 19:12 |
CFX, 2 dimensional flow
|
#1 |
Guest
Posts: n/a
|
Hi!
I have an assignment to simulate flow in two dimensions. Is that possible with CFX 5.5.1, and how do you do it? Thx! |
|
December 22, 2003, 06:18 |
Re: CFX, 2 dimensional flow
|
#2 |
Guest
Posts: n/a
|
Hi Beno, Yes it is possible, but it depends upon your geometry as to how best to approach the problem. If the geometry is nice and simple (assuming you do not have lots of meshing experience), I'd use a structured mesh which can easily be defined as 1 cell deep. On the boundaries of the 3rd dimension you have to specify symetry planes. Now if your geometry is more complex then you may well have to use the unstructured mesher. But that will give you 3D cells. So to get a truely 2D mesh you need to create a 3D mesh, define GGI boundaries rather than a symetry planes on your 3rd dimension, and then invoke an enviroment variable to tell the solver to extrude one of the GGI surface mesh's through to the other boundary.
Unfortunately, I can't remember which enviroment variable it is, plus I'm not 100% sure that the GGI is the correct definition, it may be a periodic pair ? sorry. But the long and short of it is that you can do 2D runs. May be some one else can clarify the last points for you. Have you done a search on the CFX comunity site, as I'm sure there is help there either in an FAQ or just a regular querrie. Good Luck Bob |
|
December 22, 2003, 09:43 |
Re: CFX, 2 dimensional flow
|
#3 |
Guest
Posts: n/a
|
Hi Beno, the enviroment variable is: CFX5_2D_MESH set this to be TRUE.
To return to 3D meshing you will need to rempove this variable. The boundary condition you need to apply is a periodic pair. This should then create a 2D mesh when the def file is created. Hope this is a little more clear now ? Bob |
|
December 25, 2003, 03:01 |
Re: CFX, 2 dimensional flow
|
#4 |
Guest
Posts: n/a
|
If it's truely 2D (no flow in the third direction) you don't need the GGI. Simply use the symmetry planes.
If there's rotational flow (2D axysymmetric) then you can still do this without a GGI. Use the paver option for meshing and pave quads over your 2D face. Then sweep these elements 1 cell thick in the Z direction. This gives hexes which match up on the high and low Z faces and can be set as symmetry planes. Jeff |
|
December 30, 2003, 17:00 |
Re: CFX, 2 dimensional flow
|
#5 |
Guest
Posts: n/a
|
CFX is point based. Aren't two cells required for the thickness? (i.e. points in the middle)
With one cell thick, there are only points on the symmetry planes. Not sure. -Ron |
|
January 4, 2004, 18:26 |
Re: CFX, 2 dimensional flow
|
#6 |
Guest
Posts: n/a
|
Hi Ron,
The 2D simulations work fine with just one cell in the thickness. Done it many times. Glenn |
|
January 5, 2004, 11:29 |
Re: CFX, 2 dimensional flow
|
#7 |
Guest
Posts: n/a
|
Thanks Glenn!
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Flow Analysis tilting Pad thrust bearing using CFX | julien | CFX | 5 | April 9, 2014 09:38 |
[HELP] Slip flow boundary condtion in CFX | jeffwmb | CFX | 20 | March 13, 2013 17:21 |
Ansys 11.0 CFX - solving electric potentials and multiphase flow | cfd_multiphyiscs | CFX | 2 | March 10, 2010 14:43 |
Different flow pattern between OpenFOAM and CFX | AirS | OpenFOAM | 0 | January 12, 2010 08:08 |
demo free flow blunt body in cfx ansys 11 | jan | CFX | 1 | July 31, 2007 20:44 |