[ozzythewise]

Hello all,

I have a fairly straight-forward problem that for some reason I am having difficulties with. I am trying to model flow in a 2D pipe that makes a sharp "U" shape with simpleFoam, but after a few iterations it crashes and I can't decipher the error. I've reproduced the error below.

I've checked all my BCs and they seem to be as they want as this is usually the problem with this type of error. I've included my case below as well. If I had to hazard a guess it would be a problem with my turbulence modelling BCs (using kEpsilon model), as my values for epsilon and k become unstable after an iteration.

If anyone has any thoughts on this it would be very appreciated. I'm pretty stuck with ways to fix this at the moment.

Thanks!

Link to case: http://www.speedyshare.com/zEfdF/NozzleHT.tar.gz

Sorry that the link isn't on the forum, was too large to upload

ERROR MESSAGE

Time = 6

DILUPBiCG:  Solving for Ux, Initial residual = 0.99859, Final residual = 219004, No Iterations 1001
DILUPBiCG:  Solving for Uy, Initial residual = 3.45633e-05, Final residual = 9.03368e-06, No Iterations 3
DICPCG:  Solving for p, Initial residual = 1, Final residual = 5.01613e-07, No Iterations 12
time step continuity errors : sum local = 8.09745e+121, global = -5.25148e+107, cumulative = -5.25148e+107
#0  Foam::error::printStack(Foam::Ostream&) in "/opt/openfoam211/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#1  Foam::sigFpe::sigHandler(int) in "/opt/openfoam211/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#2   in "/lib/x86_64-linux-gnu/libc.so.6"
#3  Foam::PBiCG::solve(Foam::Field<double>&, Foam::Field<double> const&, unsigned char) const in "/opt/openfoam211/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#4  Foam::fvMatrix<double>::solve(Foam::dictionary const&) in "/opt/openfoam211/platforms/linux64GccDPOpt/lib/libfiniteVolume.so"
#5  Foam::fvMatrix<double>::solve() in "/opt/openfoam211/platforms/linux64GccDPOpt/lib/libincompressibleRASModels.so"
#6  Foam::incompressible::RASModels::kEpsilon::correct() in "/opt/openfoam211/platforms/linux64GccDPOpt/lib/libincompressibleRASModels.so"
#7  
 in "/opt/openfoam211/platforms/linux64GccDPOpt/bin/simpleFoam"
#8  __libc_start_main in "/lib/x86_64-linux-gnu/libc.so.6"
#9  
 in "/opt/openfoam211/platforms/linux64GccDPOpt/bin/simpleFoam"
Floating point exception (core dumped)

[Antimony]

Hi,

You might want to try adding in some relaxation factors for the solver in fvSolution file. I have noticed that that helps in making things a little better (the number of iterations hit by the epsilon and p solvers was 1001 and that is never a good thing). Also, you might want to put the relTol at a non-zero value say 0.1 or something. So far I have been able to run your simulation for around 60 steps and haven't had any blow up. So maybe that is the fix you require. The relaxation parameters I used are below:

relaxationFactors
{
fields
{
p 0.3;
}
equations
{
U 0.5;
k 0.5;
"epsilon|omega" 0.5;
}
}

Hope this helps.

Regards,

Antimony

[ozzythewise]

yep, that totally fixed it. Thanks a million!

[ozzythewise]

I wanted to follow-up on this problem actually. I was able to solve the flow without an issue, but what I'm trying to do now is use scalarTransportFoam to find temperature distributions throughout the flow field (steady-state temperature). I am having a similar issue where the temperature field is unstable, but cannot get it to stabilize (including playing with relaxation factors and relTol). I have included my solution below. If anyone has any ideas for this that would be more than appreciated.

Link to case: http://www.speedyshare.com/5VVY8/NozzleScalar.tar.gz

[Antimony]

Hi,

I just did a quick check on your mesh using the checkMesh utility and it seems that there are a lot of highly skewed elements (300 odd) and probably that is affecting the quality of the results.

You might want to remesh and see if you can get a better one.

Regards,

Antimony