CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Laminar or Turbulent model (http://www.cfd-online.com/Forums/fluent/106172-laminar-turbulent-model.html)

prince_pahariaa August 21, 2012 07:44

Laminar or Turbulent model
 
Hello Friends

How to model a system where flow is turbulent at one part of system and rest part flow is laminar.

mvee August 23, 2012 02:13

Hi

This is possible with UDF where you have to tell solver that up to finite length of the domain turbulence model will work and in rest of the domain another model.

I have not yet tried such UDF but it can be possible.

Best wishes
Mvee

cfd seeker August 23, 2012 06:09

Quote:

Originally Posted by prince_pahariaa (Post 377967)
Hello Friends

How to model a system where flow is turbulent at one part of system and rest part flow is laminar.

Transition is taking place from laminar to turbulent in your system or you have two different flows one laminar and one turbulent at different parts of your system? what exactly is your problem?

Richard Parker August 23, 2012 06:10

Simply solve all with a turbulent model. A turbulent model accounts for turbulence, but this does not mean you are assuming the flux is turbulent. It might as well be laminar.

cfdnewbie August 23, 2012 06:25

That would be unwise to do. The standard turbulence models (I am assuming you are doing RANS) add viscosity to the flow, even in the laminar regime. When using a additional turbulence model in the laminar region, you will likely get execessive dissipation there...

prince_pahariaa August 23, 2012 06:32

@ CFD seeker :- Transition is taking place from turbulent to laminar. My problem is like this.

I have a pipe on which 10 nozzles are mounted. This pipe system is to recirculate water in rectangular pool. Water through nozzles goes into pool and there is outlet for water inside pool.

Flow is turbulent inside the pipe and when it passes through nozzle but re- circulation inside pool is laminar. I am expecting at certain height the transverse velocity should be almost zero but i am not getting the desired result.

@ MVEE :- I will try the UDF way. Thanks.. Will let you know if it works.

@ Richard Parker :- I am not very clear what did you suggest. Please can you elaborate a little..

cfd seeker August 23, 2012 06:37

Quote:

Originally Posted by cfdnewbie (Post 378300)
That would be unwise to do. The standard turbulence models (I am assuming you are doing RANS) add viscosity to the flow, even in the laminar regime. When using a additional turbulence model in the laminar region, you will likely get execessive dissipation there...

exactly cfd newbie is spot on that turbulence models add viscosity everywhere....if the flow is transitional then we can use 2 new turbulence models incorporated in Ansys 12 onward i.e k-kl-w transition model(3 equations model) and transition k-w model(4 equations model) but in the other case if we have two separate flows one laminar and other turbulent then I guess we have to split the domain in two parts to solve laminar and turbulent portions separately

Richard Parker August 23, 2012 07:05

Perhaps it is not necessary to write an UDF. Fluent has the capability of defining a laminar zone, where turbulence equations are not solved. You just have to define the zone in the mesh and then mark is as a laminar zone in the Fluid panel of the GUI. Have a look here. (I must say that I've never done it myself, but looks promising).

About the consequences of solving laminar flows with a turbulent model whe have a good old post with a very knowledgeable insight.

prince_pahariaa August 24, 2012 00:46

Thank all of you for interest..

It has been useful :)


All times are GMT -4. The time now is 19:31.