CBC.PDESys

documentation of rans

Asked by Sirditov Ivan

Hello,
where i can find the full documentation for rans, and the list of solvers parameters, or how i can orient in solvers.

Question information

Language:
English Edit question
Status:
Solved
For:
CBC.PDESys Edit question
Assignee:
No assignee Edit question
Solved by:
Mikael Mortensen
Solved:
Last query:
Last reply:
Revision history for this message
Best Mikael Mortensen (mikael-mortensen) said : 		#1

Hi,

Unfortunately there is no full documentation yet. What type of models are you interested in? I know there are some issues with the solvers right now, because the Periodic boundary conditions do not work in the dev version of dolfin. Hence the plane channel solver is broken.

Let me know if I can help you with something more specific.

Best regards

Mikael

Den Oct 9, 2012 kl. 10:31 AM skrev Sirditov Ivan:

> New question #210730 on CBC.PDESys:
> https://answers.launchpad.net/cbcpdesys/+question/210730
>
> Hello,
> where i can find the full documentation for rans, and the list of solvers parameters, or how i can orient in solvers.
>
> --
> You received this question notification because you are an answer
> contact for CBC.PDESys.

Revision history for this message
Sirditov Ivan (sirditov) said : 		#2

Thank you,
i'm interested the RANS model. And but i while not defined the type of model. I would like to understand the structure of this framework

Revision history for this message
Sirditov Ivan (sirditov) said : 		#3

Thanks Mikael Mortensen, that solved my question.

Revision history for this message
Mikael Mortensen (mikael-mortensen) said : 		#4

Hi,

I think the best way to learn the structure is to study the source.

Basically, in cbc.pdesys, each system of PDEs is contained in a PDESystem class. So each new turbulence model is an instance of the PDESystem class. The many different turbulence models all contain different PDEs and often many other derived quantities, like the turbulent viscosity. Each PDESystem (or turbulence model) is then subdivided into PDESubSystems, where each PDESubSystem contains exactly one variational form. For example, there are two ways of subdividing the k-epsilon model. One may solve a PDE for both k and epsilon individually (segregated), or simultaneously for both k and epsilon (coupled solver). The segregated approach thus contains two PDESubSystems, whereas the coupled has only one. The derived quantities are the same for both approaches.

Each turbulence model requires a Navier-Stokes solver. The linking to NS solver is taken care of in the base class for all turbulence models: TurbSolver. You can ideally choose any NS-solver to go with any turbulence model, but I have not spent much time developing the steady segregated solvers. Look at ransmodels/demo/turbulent_channel.py for how it works, choose NS solver, then choose turbulence model and finally solve problem. (Unfortunately, the channel flow does not work at the moment due to issues with periodic boundary conditions)

Hope this helps

Best regards

Mikael

Den Oct 9, 2012 kl. 11:11 AM skrev Sirditov Ivan:

> Question #210730 on CBC.PDESys changed:
> https://answers.launchpad.net/cbcpdesys/+question/210730
>
> Sirditov Ivan posted a new comment:
> Thank you,
> i'm interested the RANS model. And but i while not defined the type of model. I would like to understand the structure of this framework
>
> --
> You received this question notification because you are an answer
> contact for CBC.PDESys.