CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM (https://www.cfd-online.com/Forums/openfoam/)
-   -   create new Reynolds Stress Model (RSM) (https://www.cfd-online.com/Forums/openfoam/195781-create-new-reynolds-stress-model-rsm.html)

oda November 17, 2017 01:54

create new Reynolds Stress Model (RSM)
 
Hello,

I'm beginner in OpenFOAM and want to create new Reynolds Stress Model (RSM).
That's Hanjalic and Launder model.(1998, CONTRIBUTION TOWARDS THE SECOND-MOEMNT CLOSURE MODELLING OF SEPARETING TURBULENT FLOWS)

https://www.researchgate.net/publica...port_Modelling

So, I tried method that modified Launder-Gibson-RSTM turbulence model to Hanjalic model.
I did the following two things.

1) make a character definition (aij, A eij, E etc..) in Hanjalic.C.
ex)
Code:

            volSymmTensorField aij = ((R_/k_)-(2/3)*I);
            volScalarField A2 = (aij && aij);

2) modified each term (Dij, Y etc) in Hanjalic.C.
ex)
Code:

Y = - C2_*Fe*epsilon_*(epsilon_-D)/k_
Other than that I did not do anything.So, compile of Hanjalic model is OK.
But, that model diverge soon.

So, I have a question.
1) Is that way to create new RSM corrected ?
2) Is that way to make a character definition corrected ?

I don't know whether to tell it well, please give me advice.

Please forgive me my lousy English.
Your help is highly appreciated:)

Thanks!

oda

oda December 4, 2017 05:54

please help me
 
anyone, help me:(

Ash Kot December 5, 2017 01:45

Quote:

Originally Posted by oda (Post 671862)
Hello,

I'm beginner in OpenFOAM and want to create new Reynolds Stress Model (RSM).
That's Hanjalic and Launder model.(1998, CONTRIBUTION TOWARDS THE SECOND-MOEMNT CLOSURE MODELLING OF SEPARETING TURBULENT FLOWS)

https://www.researchgate.net/publica...port_Modelling

So, I tried method that modified Launder-Gibson-RSTM turbulence model to Hanjalic model.
I did the following two things.

1) make a character definition (aij, A eij, E etc..) in Hanjalic.C.
ex)
Code:

            volSymmTensorField aij = ((R_/k_)-(2/3)*I);
            volScalarField A2 = (aij && aij);

2) modified each term (Dij, Y etc) in Hanjalic.C.
ex)
Code:

Y = - C2_*Fe*epsilon_*(epsilon_-D)/k_
Other than that I did not do anything.So, compile of Hanjalic model is OK.
But, that model diverge soon.

So, I have a question.
1) Is that way to create new RSM corrected ?
2) Is that way to make a character definition corrected ?

I don't know whether to tell it well, please give me advice.

Please forgive me my lousy English.
Your help is highly appreciated:)

Thanks!

oda

My suggestion would be first to try generating this same model using anyother propritory software.

OpenFOAM is easy to modify and develop, the issue occure when the bugs start causing divergence although the physics behind the modeling is correct.

oda December 7, 2017 03:25

modify turbulence model
 
Hi Ash!

Thanks you for your reply.
Sorry for lack of words.
I could use other software for Hanjalic and Jalirlic model.

So, I want to use that model in OpenFOAM.
Though compilation is successful, diverge quickly.

I don't know whether corrected file and how to modify is right or not.
Please give me advice.

Thanks!

oda


All times are GMT -4. The time now is 03:46.