CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Can CFX-5.7 solve axisymmetric problem in 2-D? (http://www.cfd-online.com/Forums/cfx/21753-can-cfx-5-7-solve-axisymmetric-problem-2-d.html)

Xueying October 25, 2005 18:24

Can CFX-5.7 solve axisymmetric problem in 2-D?
 
If it can, how to asign the boundary condition in the axisymmetric plane (z-r plane)? I saw that the existing boundary types in CFX5.7 only include inlet, outlet, opening, wall, and symmetric. Obviously, axisymmetric face is different from symmetric face.

Thanks!

Xueying

Rui October 26, 2005 06:42

Re: Can CFX-5.7 solve axisymmetric problem in 2-D?
 
> Obviously, axisymmetric face is different from symmetric face
[list]Why ????

Korsh Mik October 27, 2005 06:10

Re: Can CFX-5.7 solve axisymmetric problem in 2-D?
 
Hi Rui,

Do you think it is the same: 1. Cilinder 2. Infinite block

You said that in both cases one should use symmetric face. But it is well known that the solution, for example, of the heat transfer and temperature profile is different for the cilinder and the block. For cilinder we use log function, while in flat block case - linear.


longbow October 27, 2005 08:24

Re: Can CFX-5.7 solve axisymmetric problem in 2-D?
 
Periodic interface should be used on axisymmetric face.

Rui October 27, 2005 09:34

Re: Can CFX-5.7 solve axisymmetric problem in 2-D?
 
Hi,

In CFX (where you cannot do a true 2D analysis) to simulate an infinite block you would have two parallel faces where you would apply symmetry BC, while to simulate a cylinder you would have a slice of that cylinder (let's say ~5 degrees) and would apply symmetry BC to the two slice faces (which are not parallel).

In a symmetry BC, the normal velocity component is set to zero (Un=0), and the variables gradients normal to the boundary are also set to zero (d_phi/d_n = 0). This is valid for an axissymetric problem.

You may also think that a cylinder is made up of 72 slices of 5 degrees, and if the problem is axyssimetric, all the 72 slices will have the same flow, temperature, etc patterns. Thus there is symmetry at each of the slices faces.

Take at look at the Documentation, Solver Modelling, Advice on Flow Modelling, Modelling 2D Problems. Take also a look at Tutorial 13 (Reacting Flow in a Mixing Tube), where the tube is modelled as an axisymmetric section.

Regards,

Rui

Rui October 27, 2005 09:52

Re: Can CFX-5.7 solve axisymmetric problem in 2-D?
 
Hi,

The Documentation says: "For axisymmetric 2D geometries, apply symmetry conditions to the high-theta and low-theta planes unless there is swirl anticipated in the flow, in which case 1:1 periodic connections should be applied instead. Do not use GGI periodic connections; doing so will hurt accuracy. The extrusion rotation angle for axisymmetric geometries should be small (eg. 1 to 5 degrees)."

Korsh Mik October 27, 2005 09:53

Re: Can CFX-5.7 solve axisymmetric problem in 2-D?
 
Yes, I agree, if we mesh the 5degrees slice with symmetry on both sides - it is Ok! But it is pity we have no possibility to mesh 2d picture and solve the cilindrical problem.

Rui October 27, 2005 09:57

Re: Can CFX-5.7 solve axisymmetric problem in 2-D?
 
Hi,

CFX "just" has a 3D solver.


All times are GMT -4. The time now is 15:30.