CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Siemens (
-   -   axisymmetric model issue (

prc February 3, 2007 15:40

axisymmetric model issue

I have a funny problem creating an axisymmetric model. For some reason, the centerline of my axisymmetric model, which is on the y-axis appears as a wall in the model. This area near at the centerline is so small that I cant enforce a symmetry condition on this face in prostar. The model converges and the velocity profile does not show a boundary line on this wall.However, a y+ value is associated with this centerline wall. How do I convert this centerline wall (which really is a symmetry line) into a symmetry boundary condition so that the code knows it as a symmetry line and doesnt give me a y+ value when I plot the solution. I am a new starcd user and would appreciate your help, if you let me know the command to do this.

Thanks PRC

PS: I created the axisymmetric model by extruding a shell by 1 degree in angular direction.

prc February 4, 2007 10:47

Re: axisymmetric model issue
Hi: Could anyone tell me what are all the explicit commands prostar gui executes when you try to create a 5 degree, 1 layer thick axisymmetric model under create 2d grid category. Thanks PRC PS: I would like change the angular extent to 1 degree and impose symmetry conditions on either angular face and the centerline

Anton Lyaskin February 5, 2007 06:36

Re: axisymmetric model issue
1) The command is 2D3D, AXISYMMETRIC (actually you can see it yourself in the output window when you press "Generate 3-D mesh from 2-D mesh") 2) You can't change the extent from 5 deg to 1 deg, at least if you're going with 2D3D. For trully axisymmetric flow it actually doesn't mater, which sector do you take. 3) The correct boundary condition for axisymmetric flow is Cyclic, not Symmetry, and it is automatically defined by 2D3D 4) If you want to have a wall at the centerline (I still don't understand, what do you want there - wall, i.e. no-slip boundary, or symmetry?) then you have to make your initial 2D mesh is such a way, that min Y > 0, not = 0. In this case after 2D3D you'll have a small face there. And boundary conditions can be aplied only to FACES, not to LINES! With min Y = 0 you'll get prismatic cells at the centerline, they don't have faces where you can put boundary conditions.

prc February 5, 2007 19:09

Re: axisymmetric model issue
Thanks for the info. I was able to resolve the problem. I was able to extrude in angular direction by 1 degree. Had to create a polar coordinate system and then do a vcextrude command. My model is a 2d axisymmetric model. So I eliminated the cyclic conditions and imposed symmetry conditions on either angular face. For some reason the centerline of the model (which is on the axis) was giving me problems. For example, I was seeing a y+ value at the centerline when there is no wall there. Not sure why though. I was able to resolve it by moving the centerline slightly from centerline (eg. 1e-10 as opposed to having it at 0). Thnaks for taking the time to respond. PRC

Anton Lyaskin February 7, 2007 07:42

Re: axisymmetric model issue
Oh, that's the second part. If you make your model with VCEX command, then at the symmetry axis you'll get not the proper prisms, but just cells with one collapsed (i.e. zero area) face. And by default it will be treated as wall

All times are GMT -4. The time now is 15:02.