[msatrustegui]

Hi everyone,

I'm trying to simulate a solid cilinder situated inside a bigger cylinder, which is filled of air.

The inner cylinder has an energy source, so it warms and the heat is evacuated from the inner to the outer cylinder by natural convection.

Here comes the question, I'm trying to simplify the model by representing only part of the full model (only 30º out of 360º). How do i have to model the side walls?

When i define them as symmetry, i do not get the result i expect. The temperature along the tangential direction should be constant, however it gets colder at the extremes. Do i have to define the walls as periodic instead of symmetric? How do i define the periodic condition?

Thanks for your help.

PD: I've attached an image of the results i'm getting with symmetry conditions in the side walls.

[zhenglun.wei]

You may have two ways to do it:
1, define the periodic BC: you need to have some special treatment while generating the mesh for the boundary that you want to enforce the periodic BC. I'm sure that ICEM CFD has an approach to do that. However, different meshing package has difference way. You'd better check the manual.
2, use the axis-symmetric simulation: if your model is axis-symmetric, you can generate a 2D domain and use axis-symmetric simulation in FLUENT. It simplifies the problem a lot. Check FLUENT manual for: Modeling Axisymmetric Flows.

hope it helps,
Alan

[RodriguezFatz]

Hi msatrustegui,
1) Where is the heat source? At r=0 ?
3) The "red" body in your picture is the solid, and the transparent part is the air, right? So there is air at the top, bottom and side-wall of the inner cylinder?
2) Why do you simulate this in 3d? Isn't it fully cylindrical symmetric?

[msatrustegui]

Originally Posted by RodriguezFatz

Hi msatrustegui,
1) Where is the heat source? At r=0 ?
3) The "red" body in your picture is the solid, and the transparent part is the air, right? So there is air at the top, bottom and side-wall of the inner cylinder?
2) Why do you simulate this in 3d? Isn't it fully cylindrical symmetric?

Thanks for answering,

1) The heat source is constant all over the body.
2) You're right. The solid cylinder is surrounded by air.
3) I'm simulating it in 3D because this is a simplified model, the original is periodic every 30 degrees.

Any ideas to make it periodic?

The only software i have is Ansys meshing and i don't know how to define the periodic boundaries there. In Fluent the only way i found is using TUI, but it doesn't work very well..

[RodriguezFatz]

Alright. You have one problem: Normally, you should try to get a conformal mesh at the two faces that you want to become periodic. In your case, I don't see how that works, because your inner body goes to r->0. Thus, you can not just sweep the surface mesh for 30°...
Anyway you have to create a non-conformal periodic mesh. In my fluent help version this is explained under "5.4.4. Using a Non-Conformal Mesh in ANSYS FLUENT".
Try to get this running. Unfortunately, I never did that and can not help you...

[msatrustegui]

Originally Posted by RodriguezFatz

Alright. You have one problem: Normally, you should try to get a conformal mesh at the two faces that you want to become periodic. In your case, I don't see how that works, because your inner body goes to r->0. Thus, you can not just sweep the surface mesh for 30°...
Anyway you have to create a non-conformal periodic mesh. In my fluent help version this is explained under "5.4.4. Using a Non-Conformal Mesh in ANSYS FLUENT".
Try to get this running. Unfortunately, I never did that and can not help you...

I finally found a way to do it. It's explained in another thread of this forum. Here is the link:

http://www.cfd-online.com/Forums/fluent/105041-fluent-periodic-boundary-condition-solved-tutorial.html

Thanks for your time!

[RodriguezFatz]

But isn't the tutorial you mentioned for conformal meshes? Or did Ansys Meshing create a conformal mesh?

[msatrustegui]

Originally Posted by RodriguezFatz

But isn't the tutorial you mentioned for conformal meshes? Or did Ansys Meshing create a conformal mesh?

As you said, the link I posted is for conformal meshes. The thing is that i can simulate my model using conformal and also non-conformal meshes. The problem was i could not define periodic walls in anyway. However, the tutorial I linked worked very well.

[furqanrk]

Originally Posted by RodriguezFatz

Alright. You have one problem: Normally, you should try to get a conformal mesh at the two faces that you want to become periodic. In your case, I don't see how that works, because your inner body goes to r->0. Thus, you can not just sweep the surface mesh for 30°...
Anyway you have to create a non-conformal periodic mesh. In my fluent help version this is explained under "5.4.4. Using a Non-Conformal Mesh in ANSYS FLUENT".
Try to get this running. Unfortunately, I never did that and can not help you...

Dear expert.. I have to use Superficial gas velocity 0.0016 m/s. But In fluent we have to put Velocity Magnitude at VELOCITY INLET boundary conditions.( INLET VELOCITY) How can I calculate INLET GAS VELOCITY form superficial gas velocity??? ( Reactor dimensions are 20*5*50 cm W*D*L and inlet dimensions are 2.4*1.2 L*D )