Hi Hakan,
you're right. It was the fvSolution file.I've compared with other files from the tutorials.
It is just a matter of the syntax. I had to move the "solver" and delete some " ; ".
Just like this (an other file, made for pimpleDyMFoam):
solvers
{
p
{
solver PCG;
preconditioner DIC;
tolerance 1e-06;
relTol 0.05;
}
pFinal
{
solver PCG;
preconditioner DIC;
tolerance 1e-06;
relTol 0;
}
U
{
solver PBiCG;
preconditioner DILU;
tolerance 1e-05;
relTol 0;
}
k
{
solver PBiCG;
preconditioner DILU;
tolerance 1e-05;
relTol 0;
}
epsilon
{
solver PBiCG;
preconditioner DILU;
smoother DILU;
minIter 1;
maxIter 3;
tolerance 1e-07;
relTol 0;
}
B
{
solver PBiCG;
preconditioner DILU;
tolerance 1e-05;
relTol 0;
}
nuTilda
{
solver PBiCG;
preconditioner DILU;
tolerance 1e-05;
relTol 0;
}
}
PIMPLE
{
nOuterCorrectors 2;
nCorrectors 1;
nNonOrthogonalCorrectors 2;
}
relaxationFactors
{
p 0.3;
U 0.75;
k 0.5;
epsilon 0.5;
omega 0.5;
}
// ************************************************************************* //
The strange thing about this file is, that only p shows bad divergence behaviour. EDIT: I meant convergence. Hakans next post adjusts my presumption.
Have a look:
Courant Number mean: 0.00926673 max: 0.500046 velocity magnitude: 40.8202
deltaT = 1.94298e-06
GGI pair (INTERFACE_R, INTERFACE) : 0.2921 0.292085 Diff = 1.50259e-05 or 0.0051441 %
GGI pair (ROTOR_EINTRITT, EINTRITT_INTERFACE) : 0.2921 0.292 Diff = -0.000100211 or 0.0343071 %
Time = 5000.009005
Initializing the GGI interpolator between master/shadow patches: INTERFACE_R/INTERFACE
Initializing the GGI interpolator between master/shadow patches: ROTOR_EINTRITT/EINTRITT_INTERFACE
DILUPBiCG: Solving for Ux, Initial residual = 8.07566e-05, Final residual = 6.08241e-08, No Iterations 1
DILUPBiCG: Solving for Uy, Initial residual = 8.27376e-05, Final residual = 4.53439e-08, No Iterations 1
DILUPBiCG: Solving for Uz, Initial residual = 5.062e-06, Final residual = 5.062e-06, No Iterations 0
DICPCG: Solving for p, Initial residual = 0.0142832, Final residual = 0.000619964, No Iterations 6
DICPCG: Solving for p, Initial residual = 0.00150792, Final residual = 7.42823e-05, No Iterations 259
DICPCG: Solving for p, Initial residual = 0.00057985, Final residual = 2.86792e-05, No Iterations 121
time step continuity errors : sum local = 2.36639e-09, global = -2.03075e-10, cumulative = -7.55483e-07
DILUPBiCG: Solving for epsilon, Initial residual = 7.88549e-06, Final residual = 7.33085e-09, No Iterations 1
DILUPBiCG: Solving for k, Initial residual = 2.40367e-05, Final residual = 8.46692e-09, No Iterations 1
DILUPBiCG: Solving for Ux, Initial residual = 2.88001e-05, Final residual = 3.82796e-08, No Iterations 1
DILUPBiCG: Solving for Uy, Initial residual = 2.94872e-05, Final residual = 6.60261e-08, No Iterations 1
DILUPBiCG: Solving for Uz, Initial residual = 1.9312e-06, Final residual = 1.9312e-06, No Iterations 0
DICPCG: Solving for p, Initial residual = 0.00409901, Final residual = 0.00017685, No Iterations 6
DICPCG: Solving for p, Initial residual = 0.000444541, Final residual = 2.17006e-05, No Iterations 260
DICPCG: Solving for p, Initial residual = 0.000225137, Final residual = 9.72926e-07, No Iterations 371
time step continuity errors : sum local = 8.81938e-11, global = 7.52353e-13, cumulative = -7.55482e-07
DILUPBiCG: Solving for epsilon, Initial residual = 4.02741e-06, Final residual = 2.43883e-09, No Iterations 1
DILUPBiCG: Solving for k, Initial residual = 1.21342e-05, Final residual = 1.76598e-09, No Iterations 1
ExecutionTime = 340714 s ClockTime = 341807 s
So U, epsilon and k are computed well done.. but there are so many iterations for p! And it doesn't become better. My understanding of mathematics is now limited. What can I do to make it better? My tutor told me to use pimpleDyMFoam.(First I wanted to use transientSimpleDyMFoam=) I suspect 2 weeks will pass until simulation ends. no good perspective. (I could learn a bit of latex in this time..but I think there is some mistake or bad setup for p.)
Best regards,
Christian |
