Unsteady laminar flow around cylinder with CS

  • Alicia
  • Topic Author
  • Offline
  • Junior Boarder
  • Junior Boarder
More
11 years 3 months ago #4108 by Alicia
Hi all,

I'm trying to run a very simple case with code saturne, based on a tutorial made for Fluent : the unsteaday laminar flow around a circular cylinder.
I used all the parameters given in the Fluent tutorial (at this address : https://confluence.cornell.edu/display/SIMULATION/FLUENT+-+Unsteady+Flow+Past+a+Cylinder+-+Problem+Specification ), I tried as much as I could to reproduce the mesh (using GMSH instead of Gambit, and having a 3D mesh instead of a 2D one). The calculation is running fine, but the results I obtain are not what I should get. It seems that I can't get the vortex shedding in the wake of the cylinder, I only obtain a large recirculation zone.
I also tried the Fluent tutorial for a steady laminar flow around a circular cylinder, with the same mesh, and the flow I obtained was ok (in accordance with the results given in the Fluent tutorial).
I can't figure out why I can't get the vortex shedding downstream of the cylinder.

I attach the Saturne case and the routine I used for the initialisation (usiniv.F). I can't attach my mesh file because it's too big. I just attach a screenshot.
I hope someone can help me on this one...

Alicia

Attachment case.gz not found


Attachments:

Please Log in or Create an account to join the conversation.

More
11 years 3 months ago #4109 by Claus
Have you considered which turbulence model you're utilizing?
Heres a bit of info on the models in CS

cfd.mace.manchester.ac.uk/twiki/bin/view.../SatTurbulenceModels

Also: Is this what we're talking about? A small for-fun study I made a while back using CS:

www.caelinux.org/wiki/images/b/b9/Velocit.avi.tar.gz

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version

Please Log in or Create an account to join the conversation.

  • Alicia
  • Topic Author
  • Offline
  • Junior Boarder
  • Junior Boarder
More
11 years 3 months ago #4111 by Alicia
The flow is laminar, so I'm not using any turbulence model.
Basically it's the same study as yours, but with a circular cylinder instead of a boat shape. What was the Reynolds of your study ? Mine is 150. Which turbulence model did you use ?
What was your time step, and how many iterations did you run to have the little animation ?

Please Log in or Create an account to join the conversation.

More
11 years 3 months ago #4112 by Claus
Alicia wrote:

The flow is laminar, so I'm not using any turbulence model.
Basically it's the same study as yours, but with a circular cylinder instead of a boat shape. What was the Reynolds of your study ? Mine is 150. Which turbulence model did you use ?
What was your time step, and how many iterations did you run to have the little animation ?


Sorry I cannot remember the R-number, but you have:

Re = ρ u dh / μ

= u dh / ν

So just take the physical properties for air (it's a bullet! not a boat :) )

dh = <hydraulic_diameter>0.02
u = <dirichlet name="velocity_V">10.0</dirichlet>


<turbulence model="k-epsilon">
<time_step_ref>0.01</time_step_ref>
<iterations>600</iterations>

That's what I could come up with - Im not sure how much resemblance this simulation really have with the real world, but it shows you that vortex shredding should be obtainable.

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version

Please Log in or Create an account to join the conversation.

  • Alicia
  • Topic Author
  • Offline
  • Junior Boarder
  • Junior Boarder
More
11 years 3 months ago #4113 by Alicia
Hi Claus,

Thanks for the answers. Looks like your calculation was for a turbulent flow (Re~1e4), mine is laminar. But I should get the vortex shedding anyway... Did you have a specific initialization (so there is something to cause the vortex shedding) ?
I really don't understand why I can't obtain it... :dry:

Please Log in or Create an account to join the conversation.

More
11 years 3 months ago #4119 by Claus
Alicia wrote:

Hi Claus,

Thanks for the answers. Looks like your calculation was for a turbulent flow (Re~1e4), mine is laminar. But I should get the vortex shedding anyway... Did you have a specific initialization (so there is something to cause the vortex shedding) ?
I really don't understand why I can't obtain it... :dry:


I don't know enough about CS to properly diagnose the problem, but how is convergence (in 'listings' file) ?

Do you run the simulation long enough for the vortex shredding to appear? The ones in mine don't appear until 20 (video) seconds into the simulation.

If you don't figure it out, you can try code-saturne.info/products/code-saturne/forums - they are usually very helpful - or wait for David Monfort to poke in his head on this forum.

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version

Please Log in or Create an account to join the conversation.

Moderators: catux
Time to create page: 0.119 seconds
Powered by Kunena Forum