
[Sponsors] 
August 11, 2014, 03:15 
Airfoil lift and drag using kklomega turbulence model

#1 
New Member
Karl Hylen
Join Date: Dec 2013
Posts: 3
Rep Power: 4 
Hi FOAM:ers,
I'm trying to get nice lift/drag prediction on 2D airfoils at low AoA:s (05 degrees) and a Re of 10^6, using the simpleFoam solver. I've used some lowRe turbulence models, especially SST komega, and some different meshes, but am consequently getting to high drag, an overprediction of 100150%. From the threads here on this subject I understand that this is normal for fully turbulent modells, and that a transitional model may be needed. Thus I have now turned to the kklomega model, which I believe is the only transitional model implemented in OpenFOAM. The problem is that I can't even get this model to run. In the first iteration, the solver ends with a floating point exception. I would very much appriciate help on making this case run. These are my BC:s, Kt: inlet fixedValue 3.4e4 outlet inletOutlet inletValue 3.4e4 airfoil fixedValue 1e11 Kl: inlet fixedValue 1e11 outlet inletOutlet inletValue 1e11 airfoil fixedValue 1e11 omega: inlet fixedValue 0.25 outlet inletOutlet 0.25 airfoil omegaWallFunction nut: inlet/outlet calculated airfoil nutLowReWallFunction p: inlet zeroGradient outlet fixedValue 0 airfoil zeroGradient U: inlet fixedValue (15.11 0 0) outlet inletOutlet inletValue (15.11 0 0) airfoil fixedValue (0 0 0) These are my numerical schemes, divSchemes { default none; div(phi,U) bounded Gauss limitedLinearV 1; div(phi,kt) bounded Gauss limitedLinear 1; div(phi,kl) bounded Gauss limitedLinear 1; div(phi,omega) bounded Gauss limitedLinear 1; div((nuEff*dev(T(grad(U))))) Gauss linear; } I've been using this wonderful mesher by Mads here on the forum, http://hvirvel.dk/airfoilmesher/ Should probably also mention that I'm using OpenFOAM 2.3.0. I would very much appriciate help on making this case run. Regards, Karl EDIT: After reading the article a bit closer, I changed omega to zeroGradient at the airfoil. Not the problem though. EDIT2: The error message might be helpful. It suggests a division by zero in the kklomega::correct function, if I understand it correctly. I cannot see what might be zero at the start though... Code:
smoothSolver: Solving for Ux, Initial residual = 0.020463, Final residual = 0.00138025, No Iterations 1 smoothSolver: Solving for Uy, Initial residual = 0.0726345, Final residual = 0.00486526, No Iterations 1 GAMG: Solving for p, Initial residual = 1, Final residual = 0.00940194, No Iterations 2 time step continuity errors : sum local = 0.00295125, global = 1.54808e06, cumulative = 1.54808e06 #0 Foam::error::printStack(Foam::Ostream&) at ??:? #1 Foam::sigFpe::sigHandler(int) at ??:? #2 in "/usr/lib/libc.so.6" #3 Foam::divide(Foam::Field<double>&, Foam::UList<double> const&, Foam::UList<double> const&) at ??:? #4 void Foam::divide<Foam::fvPatchField, Foam::volMesh>(Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh>&, Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> const&, Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> const&) at ??:? #5 Foam::tmp<Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> > Foam::operator/<Foam::fvPatchField, Foam::volMesh>(Foam::tmp<Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> > const&, Foam::tmp<Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> > const&) at ??:? #6 Foam::incompressible::RASModels::kkLOmega::correct() at ??:? #7 in "/opt/OpenFOAM/OpenFOAM2.3.0/platforms/linux64GccDPOpt/bin/simpleFoam" #8 __libc_start_main in "/usr/lib/libc.so.6" #9 in "/opt/OpenFOAM/OpenFOAM2.3.0/platforms/linux64GccDPOpt/bin/simpleFoam" Last edited by hylleman; August 12, 2014 at 04:18. 

August 15, 2014, 02:13 

#2 
New Member
Karl Hylen
Join Date: Dec 2013
Posts: 3
Rep Power: 4 
I managed to get this working, after a lot of fiddeling. Unfortunatly, I can't say exactly what the problem was since I changed a lot (schemes, bcs, relaxation factors...). Now I get really nice force prediction on the airfoil, so the fiddeling was worth it


August 20, 2014, 05:32 

#3 
New Member
Michael D.
Join Date: Jun 2014
Posts: 14
Rep Power: 4 
Hi Karl,
I'm also interested in trying the kklomega model, and I've run into the same problem. Could you post the setup you used to get it working? Thanks in advance! 

November 29, 2014, 23:20 

#4  
Senior Member
Khamlaj
Join Date: Nov 2010
Location: United States
Posts: 166
Rep Power: 7 
Quote:
Could you please provide us more explications since I am working on the same problem? 

February 5, 2015, 08:52 

#5 
New Member
Anonymous
Join Date: Dec 2014
Posts: 2
Rep Power: 0 
I get the same instant error
Code:
smoothSolver: Solving for Ux, Initial residual = 0.020463, Final residual = 0.00138025, No Iterations 1 smoothSolver: Solving for Uy, Initial residual = 0.0726345, Final residual = 0.00486526, No Iterations 1 GAMG: Solving for p, Initial residual = 1, Final residual = 0.00940194, No Iterations 2 time step continuity errors : sum local = 0.00295125, global = 1.54808e06, cumulative = 1.54808e06 #0 Foam::error::printStack(Foam::Ostream&) at ??:? #1 Foam::sigFpe::sigHandler(int) at ??:? #2 in "/usr/lib/libc.so.6" #3 Foam::divide(Foam::Field<double>&, Foam::UList<double> const&, Foam::UList<double> const&) at ??:? #4 void Foam::divide<Foam::fvPatchField, Foam::volMesh>(Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh>&, Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> const&, Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> const&) at ??:? #5 Foam::tmp<Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> > Foam::operator/<Foam::fvPatchField, Foam::volMesh>(Foam::tmp<Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> > const&, Foam::tmp<Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> > const&) at ??:? #6 Foam::incompressible::RASModels::kkLOmega::correct() at ??:? #7 in "/opt/OpenFOAM/OpenFOAM2.3.0/platforms/linux64GccDPOpt/bin/simpleFoam" #8 __libc_start_main in "/usr/lib/libc.so.6" #9 in "/opt/OpenFOAM/OpenFOAM2.3.0/platforms/linux64GccDPOpt/bin/simpleFoam" 

February 5, 2015, 12:01 

#6 
Senior Member
Alexey Matveichev
Join Date: Aug 2011
Location: Nancy, France
Posts: 1,419
Rep Power: 25 
Hi,
If you take a look at correct method of kkLOmega, programming there always is rather defensive, like this: Code:
const volScalarField lambdaT(sqrt(kT)/(omega_ + omegaMin_)); Code:
const volScalarField nutl ( min ( C11_*fTaul(lambdaEff, ktL)*omega*sqr(lambdaEff) * sqrt(ktL)*lambdaEff/nu() + C12_*BetaTS(ReOmega)*ReOmega*sqr(y_)*omega , 0.5*(kl_ + ktL)/sqrt(S2) ) ); Code:
const volScalarField S2(2.0*magSqr(symm(gradU))); Update. Though if you have got diverging solution and something that should be positive has become negative, then all these omega_ + omegaMin_ can become zero. 

June 17, 2016, 15:10 
New model

#7 
New Member
Alberto
Join Date: Sep 2013
Posts: 21
Rep Power: 5 
After 8 years, there is a new version (or new model) of the kklomega model.
There are a few problems with the kklomega model in the farfield. One of them is the growth of Laminar Kinetic energy when separation occurs. Lopez and Walters have a paper (have not been published yet) correcting this issue: Maurin Lopez. D. K. Walters. “A recommended correction to the kklomega transition sensitive eddyviscosity model”. Journal of Fluid Engineering. This correction has to be made to the 2008 kklomega model from now on. Now, Lopez and Walters also developed a new transitional model (komegav2) as an alternative to the kklomega one. This new model has more capabilities (it is more reliable) than the kklomega model, especially in the farfield computations. Fortunately the paper for this new model is already publish. Maurin Lopez. D. K. Walters. “Prediction of transitional and fully turbulent free shear flows using an alternative to the laminar kinetic energy approach”. Journal of Turbulence, Vol 17, Iss. 3, 2016. If you see the papers, you will immediately see how the kklomega model is not good for free shear flows, and how the new model corrects all those issues. From now on, kklomega users have to start using the new komegav2 model. 

Tags 
airfoil, drag, kklomega, lift, simplefoam 
Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
SA Turbulence Model and drag prediction  Alan  FLUENT  2  January 11, 2016 09:33 
wrong SU2 calculation for lift and drag coefficient for NAC4421  mechy  SU2  5  June 24, 2014 16:20 
Airfoil lift and drag understimated  Kio  OpenFOAM PreProcessing  14  September 13, 2013 06:41 
Problem with restart solution in shape_optimization.py  robyTKD  SU2 Shape Design  21  May 29, 2013 09:26 
Water subcooled boiling  Attesz  CFX  7  January 5, 2013 04:32 