[svens]

Hello

I am an absolutely OF beginner and to learn the ropes I'm trying to create and solve a simple 2D laminar channel flow with heated walls.

-What solver fits best for my problem? I successfully calculated laminar flow in same channel geometry without heat flux with the icoFoam solver. Is there a way to modify icoFoam to use it also for my new case with heat transfer?

-I tried to solve my problem with the buoyantFoam solver. At some point the Courant-Number exploded which led to an abort of calculation. Is there a problem using buoyantFoam for a laminar case?

-Is there a case comparable to mine which I can use as assistance to get a little bit easier in the problem?

Thanks so much - svens

[Bedotto]

Hi Svens

For your first point, you should go on the OpenFoam Wiki. There is an example about how to implement the heat equation in a solver with icoFoam:
http://openfoamwiki.net/index.php/Ho...ure_to_icoFoam

For the problem concerning your courant number, sorry I've no idea (I'm a beginner like you )

However if you want to compare your results with a DNS , you should have a look on this site:
http://murasun.me.noda.tus.ac.jp/turbulence/

Regards

Quentin

[sven]

I also faced the same problem as svens. I used the manual from the link posted. It worked, but I still have some questions on this manual:

1. What is the equation that is implemented in step 4? This is not the normal energy equation, is it?

2. What is the DT specified in the transport properties (step 5)?

Thanks for your help.

[svens]

Thanks Quentin for your reply - it seems to work.

Unfortunately I've got the same questions like sven to get a fully understanding what I did. I would be really grateful about some further explanations.

Another question which occurred is the lack of an implementation for the Thermophysical Model which is performed in Chapter7 of the OF Users Guide.

I guess it's necessary to create such a model for the complete definition of this case, isn't it?

Thanks a lot for your help & greetings.
svens

couple of hours later:

Now I am pretty sure that DT labels the thermal diffusion also know as alpha [=k/(roh*c_p)].

With DT my fluid properties are completely defined and I guess the above mentioned thermophysical model isn't necessary anymore.

But I still have some huge problems to understand how the energy equation is implemented in my new icoFoam solver.

fvScalarMatrix TEqn
(
fvm::ddt(T)
+ fvm::div(phi, T)
- fvm::laplacian(DT, T)
);
TEqn.solve();

About some further explanatinons of the implemtation I would be really grateful.

Thanks a lot

[Bedotto]

fvScalarMatrix TEqn
(
fvm::ddt(T)
+ fvm::div(phi, T)
- fvm::laplacian(DT, T)
);
TEqn.solve();

In the tutorial these lines are written in the file .C, more precisely in the time loop. The term fvm means a numerical scheme used to discretize the derivative operation. There's a good example on the wiki which shows the 'traduction' of a equation written in OpenFoam. Here's the link:
http://en.wikipedia.org/wiki/OpenFOAM

Tell if I'm wrong.

Regards
Quentin

[Bedotto]

By the way, I would like to add to this equation the term nu_t/Pr_t where nu_t is the SGS model term and Pr_t is the turbulent Prandt. But I don't know how to "call" the nu_t variable. Does someone know how to implement this term?

Regards

Quentin

[santos]

Hi,

I think you may call it with:

sgsModel->nuSgs()

Regards,
Jose Santos

[Bedotto]

I've checked it, it works.
Thanks Santos!