CFD Online Discussion Forums - gamma-ReTheta turbulence model for predicting transitional flows

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- OpenFOAM Programming & Development (https://www.cfd-online.com/Forums/openfoam-programming-development/)

- - gamma-ReTheta turbulence model for predicting transitional flows (https://www.cfd-online.com/Forums/openfoam-programming-development/85382-gamma-retheta-turbulence-model-predicting-transitional-flows.html)

Airfoil FX-W-343.

Attachment 7470

Comparison for the results with and withou transition model in the boudnary layer.

Attachment 7471

Attachment 7472

Airfoil FX-77-W-343.

Comparison for the results given by OpenFOAM and ANSYS CFX.

Attachment 7473

Attachment 7474

Airfoil FX-77-W-400.

Attachment 7475

Comparison of results of OpenFOAM and ANSYS CFX. Both with transition model in the boundary layer applied.

Attachment 7478

Attachment 7479

Airfoil FB-3500-0875

Attachment 7480

Comparison of OpenFOAM, ANSYS CFX, XFOIL and Wind Tunnel measurements.

Attachment 7481

Attachment 7482

Airfoil FB-3500-1750

Attachment 7483

Comparison of OpenFOAM, ANSYS CFX, XFOIL and Wind Tunnel measurements.

Attachment 7484

Attachment 7485

hi jsm,
first of all can you comment on the results obtained for NACA0012 airfoil with blunt trailing edge. The results look reasonable in the linear region with OF, CFX versus the literature.
:confused: But what happens in the maximum Cl region with CFX results the lift drops dramatically. I guess you did not run that long enough to get a converged solution or their is any other explanation.
On the other why with OF you overpredicted the maximum Cl and OF does not seem to predict the leading edge type stall observed in the literature.
I have the following queries:
--> what bc's you used
--> what turbulence level, is this same for both CFX and OF
--> did you try unsteady in the non linear region

really looking forward to your reply.

taxalian.:confused:

Hello, José,

thanks for sharing your results with us. Especially the NACA0012 looks great! I didn't think the implementation would perform so well. It does seem to have some problems in the non-linear region, though, so probably there still are some improvements left.

A few questions:

- which correlations did you use?
- did you try the suggestion of SULUKSNA et. al. 2009, i.e. keeping lambda = 0 in gammaReThetatSST::ReThetat()?
- do you have any detailed information about the CFX implementation and simulations?

Thank you very much again!

Greetings,
Felix.

Running the transition model in parallel

Dear all,

Has anyone of you tried to run the transition model given by Felix L. in parallel mode? I have tried it but it starts to diverge.
I know it sounds strange but this is how it is. I just one run one case in single mode and it works, then I do the decomposition and run it in parallel the force coefficients of the airfoil I am computing start to diverge.
I have tried to use another compiled model and it works (the k-omegaSST corrected, which only one number is added, no new equation is added as it happens with the transition model).

If any of you has something to say about this, please let me know.

Thanks,

José

Hello Felix, Jose and All,

Do you have your test cases available? Both system and constant folder? Mesh file?

I would like to test here and compare with my data simulated during my PhD thesis about aircraft icing. At that time, I did some validation with ERCOFTAC, transonic airfoil, NACA airfoil (two types) and non-isothermal flat plate cases in order to check the results of the transition (prediction and development) models of the boundary layer codes. Therefore, I am interested to see or download any case you may have to compare with my results. Further, I intend to use OpenFoam in aircraft icing cases in order to calculate local Cf and St around it.

You can download my thesis from here:
http://www.teses.usp.br/teses/dispon...-082825/en.php

Regards,

Guilherme

Hello,

I need help. I want to compile your model in openfoam 2.01. But i got some error.

Code:

gammaReThetatSST.C: In member function ‘void Foam::incompressible::RASModels::gammaReThetatSST::ReThetat(Foam::volScalarField&) const’:

gammaReThetatSST.C:244: error: no match for ‘operator[]’ in ‘Foam::incompressible::turbulenceModel::nu() const()[cellI]’

gammaReThetatSST.C: In member function ‘virtual void Foam::incompressible::RASModels::gammaReThetatSST::correct()’:

gammaReThetatSST.C:726: error: ‘omega0_’ was not declared in this scope

gammaReThetatSST.C:749: error: ‘k0_’ was not declared in this scope

In file included from gammaReThetatSST.H:109,

                 from gammaReThetatSST.C:26:

/opt/openfoam201/src/turbulenceModels/incompressible/RAS/lnInclude/RASModel.H: In static member function ‘static Foam::autoPtr<Foam::incompressible::RASModel> Foam::incompressible::RASModel::adddictionaryConstructorToTable<RASModelType>::New(const Foam::volVectorField&, const Foam::surfaceScalarField&, Foam::transportModel&, const Foam::word&) [with RASModelType = Foam::incompressible::RASModels::gammaReThetatSST]’:

/opt/openfoam201/src/turbulenceModels/incompressible/RAS/lnInclude/RASModel.H:138:   instantiated from ‘Foam::incompressible::RASModel::adddictionaryConstructorToTable<RASModelType>::adddictionaryConstructorToTable(const Foam::word&) [with RASModelType = Foam::incompressible::RASModels::gammaReThetatSST]’

gammaReThetatSST.C:44:   instantiated from here

/opt/openfoam201/src/turbulenceModels/incompressible/RAS/lnInclude/RASModel.H:126: error: no matching function for call to ‘Foam::incompressible::RASModels::gammaReThetatSST::gammaReThetatSST(const Foam::GeometricField<Foam::Vector<double>, Foam::fvPatchField, Foam::volMesh>&, const Foam::GeometricField<double, Foam::fvsPatchField, Foam::surfaceMesh>&, Foam::transportModel&, const Foam::word&)’

gammaReThetatSST.C:341: note: candidates are: Foam::incompressible::RASModels::gammaReThetatSST::gammaReThetatSST(const Foam::volVectorField&, const Foam::surfaceScalarField&, Foam::transportModel&)

gammaReThetatSST.H:128: note:                 Foam::incompressible::RASModels::gammaReThetatSST::gammaReThetatSST(const Foam::incompressible::RASModels::gammaReThetatSST&)

make: *** [Make/linux64GccDPOpt/gammaReThetatSST.o] Error 1

Could you help me please?

Hello,

sorry about being slow with things lately - I caught the chance of having a cold and thus staying at home today to make the turbulence model compatible with OF 2.0.x. Please bear in mind that the code still is NOT validated!

Changelog:

Code:

2011-09-06 : - ensured compatibility with OpenFOAM 2.0.x

Greetings,
Felix

Hello Felix,

thanks for your work. It was possible to compile so that i'm going to test it in the next days.

Greetings,
Christian

Dear Felix,

thanks for the good job and the effort to implement this turbolence model in openFoam. Exactly what i was looking for. I owe you a beer.

I will test and validate it in the next few days and report back.

Cheers, mate, a cold beer's always appreciated! ;)

Greetings,
Felix

Hello, FelixL

I am using version 1.6-ext and I am a new user. Do you think I can apply this transitional model to komega SST model in version 1.6-ext?
In version 1.6-ext, there are both Komega SST low Rey model and Komega SST model. Which one we need to apply this transitional model?

please help me!

regards

jia

I have run the code using the T3A testcase, but when I apply the model in airfoil, it blows up. Can somebody give me an testcase of airfoil?

Hi Foamers!

First of all, thank you Felix for sharing the code, saved me (and likely many others) lots of effort in implementing it.

I just wanted to post my experience with using the model with OpenFOAM-2.2.2 for future reference if anyone else tries to compile under this version.

List of changes I had to make to make it work (I started off from the most recent version that you posted in this thread):
1. divDevRhoReff method had to be added (simply copied over from standard kOmegaSST)
2. in the correct() method there was a problem with some naming convention at the line:

Code:

omega_.boundaryField().updateCoeffs();

Closer look at kOmegaSST.C revealed that replacing the line before that as follows solves the issue:

Code:

    //volScalarField G("RASModel::G", nut_*2*S2); // replaced this

    volScalarField G(GName(), nut_*2*S2);

I also took the liberty of updating the test case you posted on page 1 of this thread to match the recent OF conventions of schemes.

Once again thanks a lot for posting this.

Peace,

Artur

Hi there!
Has anyone of you tried Felix implementation on an airfoil case and had a look on the C_f distribution? Especially transition onset points and length?

I am trying to get a solution with the S809 airfoil right now. It is kind of difficult to get a well converged solution I think. My C_l is not too bad with a 5% error compared to experimental results. But if I look at my pressure distribution the onset point on the upper side has significantly moved towards the leading edge.
I also made some simulations with Ansys CFX for comparison with the exact same mesh and I get good and reasonable results respecting the pressure distribution.

Anyone who faces the same complications?

About the Missing Correlations in Gamma-Re_Theta Turbulence Model

Thanks a lot for sharing your code Artur. We know that there are two proprietary correlations in the original gamma-Re_theta turbulence model proposed by Meter et. al which are unknown to the public and implemented in the commercial code like Fluent. Different researchers (e.g. Sørensen) have attempted to propose their own correlations in the past. May I know which correlations are used in the default code?

I have attempted to use your code for analyzing NACA 4415 and found that though Cd values are closer to the experimental results, but Cl values are worse than kOmegaSST!

Thanks.

Hi,

I was not the one who created the code, I merely upgraded it to work with OF version 2.2.2. Hopefully the original Author (Felix) will be able to answer your questions.

Sorry,

A