CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   axisymmetry set up for multiple axes of symmetry (http://www.cfd-online.com/Forums/fluent/31980-axisymmetry-set-up-multiple-axes-symmetry.html)

varghese August 16, 2003 10:06

axisymmetry set up for multiple axes of symmetry
 
Hi all!

I have a geometry with three different axes of symmetry. Is it now possible to define this problem as axisymmetric in Fluent.

Thanks for any hint in advance.

Greetings Varghese

ap August 16, 2003 12:24

Re: axisymmetry set up for multiple axes of symmet
 
If all your domain is symmetric relatively to three axes, you can choose one of them and use the axisymmetric solver. Remember that FLUENT consider the x-axis as axis of symmetry.

If only a part of your domain is symmetric relatively to an axis, you can't use the asisymmetric solver for your case.

Also, you have to consider flow behaviour before using an axisymmetric model. Is the flow really symmetric? If not, you're imposing a non-physical condition.

Hope this helps

Hi

ap

varghese August 17, 2003 09:55

Re: axisymmetry set up for multiple axes of symmet
 
Dear ap

I think I must clarify my problem little further that you can help me.

In the calculational domain I have one major axis of symmetry at the centre of the domain and two minor axes of symmetry at two inlets. The axes of symmetry are however distinct. When the major axis turns 180 the minor axes would have revolved through 360.

I was thinking of multiple frame of references(MRF). But am not sure whether I can keep all the three domains stationary in that case.

Hoping your further reflection on the problem

Greetings Varghese

ap August 17, 2003 15:02

Re: axisymmetry set up for multiple axes of symmet
 
I've some difficulties to visualize your domain because I don't know what you're modeling and what form it has. So it's difficult to give proper answers :)

Specifically, I'm not sure if your domain is a solid of rotation around the main axis of symmetry.

Only in this case you can use the axisymmetric solver.

Keep me up to date and e-mail me if you want.

Hi

ap


varghese August 18, 2003 04:40

Re: axisymmetry set up for multiple axes of symmet
 
Dear ap

The domain is a large cylindrical tank, to which two inlets are open. The inlets themselves are radial jets into the tank, but not placed axially with the main tank. They shift themselves away from the central axis and this is the main problem to model it in a single axisymmetric scheme. There are now a central axis of symmertry for the large tank and another two axes of symmetry centred on the two inlets. But all the three axes remain in the same plane, hence opening possibility of creating an axisymmetric model.

The question is whether Fluent allows one to have three stationary axes of symmetry for the same calculational domain.

Greetings Varghese

ap August 18, 2003 08:08

Re: axisymmetry set up for multiple axes of symmet
 
If I'm not wrong, your geometry is something like this:


|
-----
| | |
-| | |-
| | |
-----
|
<- axis



This model can't be solved using the axi solver, because of the presence of radial inlets. If you use the axi solver, representing your domain like:


|
|--
| |
| |-
| |
|--
|
<- axis



you'd obtain a circular inlet all along the cilinder circumference.

You have two choices:

1) Use a 3D model which represent the whole domain. It's the best solution, but it's time consuming and computationally expensive.

I think you can do this for sure if you're going to do a steady simulation. If you want to do an unsteady calculation the computational time will grow a lot.

2) Use a 2D model (not Axi)

If you decide for the 2D model, you can use the Axis boundary condition to represent the main axis of symmetry.


|
|--
| |
| |-
| |
|--
|
<- axis



Honestly I don't see any significant advantage (reduction of computational time) imposing the symmetry condition at inlets.

Hope this helps

Keep me up to date

Hi

ap

varghese August 18, 2003 10:24

Re: axisymmetry set up for multiple axes of symmet
 
Dear ap

Thanks for the effort you make to help me. Sorry for causing confusion once again.

The geometry infact look like this: The two inlets are not from the side, but from the top.

|------||--||--|

|------------|

major axis

It is a transient calculation, hence can be really expensive, if I donot make an approximation. As you may observe in the domain, the axes of the inlets donot coincide with that of the major axis. It is precisely the problem for the axisymmetric reductiion.

(I really wonder how you manage to draw precisely. It was tough time for me to draw at least in this crude way.)

Expecting further insights

Greetings

Varghese

ap August 18, 2003 11:08

Re: axisymmetry set up for multiple axes of symmet
 
So, your domain is like


inlets
| <->
||| ||
| ||| || |
| | |
| | |
|_________|________|
|
<- Major axis


This can't be modeled using the axi-symmetric solver because of the presence of the two inlets.

You can try a 2D model, in which you can represent your whole domain, without considering symmetry.

I don't know what model you're going to use, but you shouldn't have problem with grid dimension with this approach.

Anyway, the best solution is always the 3D model...but it's really expensive for an unsteady calculation.

P.S.

To draw better, just insert your pre-formatted text between the HTML tags



PRE

put here your pre-formatted text and drawings

/PRE



PRE and /PRE has to be between <>


Keep me up to date

Hi

ap

varghese August 19, 2003 04:16

Re: axisymmetry set up for multiple axes of symmet
 
Thank you very much Mr.ap

You spent considerable time for me. Thank you very much.

I shall try the 2d model for my problem. But what if I use MRF? Is it really possible to use Multiple frame of reference when all the individual domains are stationary? These are the remaining questions.

If you find some answers, please inform me.

Greetings Varghese

ap August 19, 2003 05:43

Re: axisymmetry set up for multiple axes of symmet
 
I never personally used multiple Multiple Frame of Reference, but I think that in your case it's easier the direct approach of a 2D model.

Anyway, I'll look for more info and tell you if I find something interesting.

I'd like to know how big is your tank, just to extimate the mesh size.

It's a pleasure to help other people which works in CFD :)

Hi

ap

varghese August 19, 2003 06:34

Re: axisymmetry set up for multiple axes of symmet
 
The tank is 12000m3 in size. So the 3d calculation can be extremely expensive. That is why I am looking for closer 2d approximations.

Thanks a lot

Varghese

ap August 19, 2003 06:47

Re: axisymmetry set up for multiple axes of symmet
 
That's a really big system. What about trying to use a scaled model?

Hi

ap

varghese August 19, 2003 07:42

Re: axisymmetry set up for multiple axes of symmet
 
If axisymmetric simplification is not possible I will have to think in that direction.

Greetings Varghese


All times are GMT -4. The time now is 04:53.