CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Post processing of rotational moving reference frame (http://www.cfd-online.com/Forums/fluent/115054-post-processing-rotational-moving-reference-frame.html)

Brathmann March 22, 2013 10:37

Post processing of rotational moving reference frame
 
5 Attachment(s)
Hello, cluerich and clueless readers,

as a beginner to Ansys-FLUENT v14 I try to set up a flow problem with a propeller in a closed hollow cylindrical volume. In the attached pictures you can see the geometry which is one third of the real geometry and the mesh I use for the moment as displayed in FLUENT.
The problem I'm facing for now is that the results I obtained so far are not displayed identically in FLUENT-Results and the Ansys postprocessor CFD-post. You can see the different visualisation by comparing the picture CFDpost to FLUENTresults. I believe to have set all the options identically pretty much, as is hopefully shown in the dialogues next to them. The outer planes are the periodic boundaries and the middle plane is defined by the same 3 points in both cases. Also in CFD-post the streamlines look like there is no real interaction between the two fluidzones (see streamlines.jpg) but as if they pass uneffectedly through the other zone and even the airfoil.

For the setup I chose to use a rotating reference frame and periodic boundaries to set up the problem roughly according to the Ansys 12 Tutorial for Multiple Moving Reference Frame .
The slice with the airfoilshape in it is defined as a rotationally moving frame with the airfoil shaped wall as moving wall at a velocity of 0m/s relative to adjacent cell zone. The contact to the steady fluid volume was set as an interface by default. The right vertical faces are defined as periodic boundaries with the left vertical faces as their shadow walls and the model is a viscous k-epsilon-model. All other settings are basically default.

Do you have any idea why the results are displayed differently? May it be a problem of too little convergence? Or of a mesh too coarse around the contact inface? Or is something wrong in the setup that prevents the mass transfer between the moving and the steady fluid volumes?

Since I have to document my work anyway, I'm thinking to make it a tutorial and post it here, so any hints you give may impove this already in advance.

I will be happy to provide further information if necessary.
Thanks already
Brathmann

Brathmann March 22, 2013 11:39

Streamlines.jpg
 
1 Attachment(s)
Here is the picture of the so far obtained streamlines mentioned in the initial post.


All times are GMT -4. The time now is 14:25.