[artymk4]

I'm new to OpenFOAM but so far I've been able to correct all my mistakes, but this time I just cannot. I'm using PIMPLE to solve propeller case with my custom CAD propeller. I get this error as soon as I run mpirun:

[0] --> FOAM FATAL ERROR: 
[0] Solution convergence criteria specified in PIMPLE.residualControl must be given as single values. Corrector loop convergence criteria, if appropriate, are specified as dictionaries in PIMPLE.<loopName>ResidualControl.
[0] 
[0]     From function bool Foam::singleRegionConvergenceControl::readResidualControls()
[0]     in file cfdTools/general/solutionControl/convergenceControl/singleRegionConvergenceControl/singleRegionConvergenceControl.C at line 73.
[0] 
FOAM parallel run exiting

I tried to change residualControl which is defined in fvSolution but nothing helped. My residualControl looks like this:

PIMPLE
{
    correctPhi          no;
    nOuterCorrectors    100;
    nCorrectors         3;
    nNonOrthogonalCorrectors 1;
    pRefCell    0;
    pRefValue   0;
    residualControl
    {
        U
        {
            tolerance   1e-05;
            relTol      0;
        }
        p
        {
            tolerance   1e-04;
            relTol      0;
        }

    }
}

I tried to replace both U and p with "(p|U|k|epsilon)", I deleted p and kept only U, I tried to delete the whole residualControl... nothing helped. Any ideas what could be the problem?

P.S. I created tet mesh with prismatic boundary layers. Mesh looks good to me, I also get "Mesh OK." after checkMesh but I have very big Max skewness of 3.11. It says OK but I've read OpenFOAM has problems with skewness above 0.9. How can I have 3.11?? Mesh non-orthogonality Max is 70.3 and average 14.9
Since I started using OpenFOAM I've been having problems whenever I had tetrahedral mesh. This time with propeller case I followed youtube tutorial but still no luck. Is there any OF solver that works with tet meshes?

[artymk4]

I fixed this problem by setting residualControl like this:

residualControl
{
    U   1e-05;
    p   1e-04;
}

So the solver finally began working. For each cycle of deltaT = approximately 1.2e-05 the Pimple solver made 100 interations. After 6 cycles the solver crashed because Courant Number became very high and deltaT very low. So it crashed after about one hour on my 4 core CPU.
This happens to me whenever I try to solve case with tetrahedral mesh. What am I doing wrong?

[Joko]

Hi artymk4.

First thanks for the info with the residualControl. I had the same problem.

But now to your problem. What did you write in your fvSolution file for nOuterCorrectors under the pimple Algorithmus?

[Yassin-K]

Hi all, I had the same problem. The problem was with OF version that I am using. OF5 uses the form:

residualControl
    {
        U
        {
            tolerance   1e-05;
            relTol      0;
        }
        p
        {
            tolerance   1e-04;
            relTol      0;
        }

    }

while OF6 uses the form:

residualControl
{
    U   1e-05;
    p   1e-04;
}

[CFD_10]

Originally Posted by Yassin-K

Hi all, I had the same problem. The problem was with OF version that I am using. OF5 uses the form:

residualControl
    {
        U
        {
            tolerance   1e-05;
            relTol      0;
        }
        p
        {
            tolerance   1e-04;
            relTol      0;
        }

    }

while OF6 uses the form:

residualControl
{
    U   1e-05;
    p   1e-04;
 }

I cannot find any tutorial in OF5 that uses that format. OF5 uses the following:

residualControl
{
    U   1e-05;
    p   1e-04;
}

which is also used by OF6.
If want to verify this, run the following grep or another tool that support pcre regex, e.g: pcregrep:

pcregrep --color -r -M '^\h*residualControl\s*{[^}]*}' $FOAM_TUTORIALS

[ttsurvase]

For OF-V5 and OF-V6, residual control needs to be given as single value. OF-v1906 and OF-v2006 asks for dictionary residual control. While OF-V5 and OF-V6 requires single value.
residualControl

{

p 0.001;

p_rgh 0.001;

U 0.001;

k 0.001;

epsilon 0.001;


omega 0.001;

nuTilda 0.001;
}
Its just an example on how to implement residual control in OF-v5 and OF-v6.

[jairoandres]

Yeah, I ran into the same issue when using an OpenFOAM.com case in OF-V8. Fixed.