CompressibleInterFoam crashing with negative temperature values
 

Hello Dear Foamers,

I have been trying for quite a time to solve the compressible flow of helium and water.
my boundry conditipons are as follows,
FoamFile
{
version 2.0;
format ascii;
class volVectorField;
location "0";
object U;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions [0 1 -1 0 0 0 0];

internalField uniform (0 0 0);

boundaryField
{
outlet-top
{
type zeroGradient;
}
outlet-front
{
type zeroGradient;
}
inlet-gas
{
type flowRateInletVelocity;
massFlowRate constant 1.1e-09;
rhoInlet 1.63;
value uniform (0 0 0);
}
inlet-liquid
{
type flowRateInletVelocity;
volumetricFlowRate constant 4.1e-12;
value uniform (0 0 0);
}
front
{
type wedge;
}
back
{
type wedge;
}
fixedWalls
{
type fixedValue;
value uniform (0 0 0);
}
}


// ************************************************** *********************** //

object T;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions [0 0 0 1 0 0 0];

internalField uniform 300;

boundaryField
{
inlet-gas
{
type fixedValue;
value uniform 300;
}

inlet-liquid
{
type fixedValue;
value uniform 300;
}

fixedWalls
{
type zeroGradient;
}

outlet-top
{
type zeroGradient;
}

outlet-front
{
type zeroGradient;
}

front
{
type wedge;
}

back
{
type wedge;
}

}

object p_rgh;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions [1 -1 -2 0 0 0 0];

internalField uniform 1e5;

boundaryField
{

inlet-liquid
{
type zeroGradient;
}

inlet-gas
{
type zeroGradient;
}

fixedWalls
{
type fixedFluxPressure;
}

outlet-top
{
type uniformFixedValue;
uniformValue tableFile;
uniformValueCoeffs
{
fileName "pout";
outOfBounds clamp;
interpolationScheme linear;
};
value $internalField;
}

outlet-front
{
type uniformFixedValue;
uniformValue tableFile;
uniformValueCoeffs
{
fileName "pout";
outOfBounds clamp;
interpolationScheme linear;
};
value $internalField;
}

back
{
type wedge;
}

front
{
type wedge;
}

}

object p;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions [1 -1 -2 0 0 0 0];

internalField uniform 1e5;

boundaryField
{
inlet-liquid
{
type calculated;
value $internalField;
}

inlet-gas
{
type calculated;
value $internalField;
}

fixedWalls
{
type calculated;
value $internalField;
}

outlet-top
{
type calculated;
value $internalField;
}

outlet-front
{
type calculated;
value $internalField;
}

front
{
type wedge;
}

back
{
type wedge;
}


}

I get to have the problem of case crashing with
MAXIMUM NUMBER OF ITERATIONS EXCEEDS when temperature goes to negative.
I have tried to change the pMin value and find out that it is very sensitive to the solution.
I am very much stuck.
I have opened a new tread as I have seen there are only few already exists and are quite old.

Hi Rizwan,

Two suggestion:

1.Use the buoyantPressure boundary condition for pressure at walls.

2.Use the upwind scheme for div(Phi,T)

Hello rapierrz

I am using of3.01 which has this alternative of pressure boundary condition to buoyantPressure you suggested as "fixedFluxPressure" which I am using. and Currently I have tried with the Upwind scheme for Temperature as well.
I have tried to use the CompressiblemultiphaseInterFoam for 2 fluids to see that it effect or no.
It currently solved the temperature problem but now When I alpha.* sums increase to something 34 which is very strange even with limited schemes as well like vanLeer01.
It has been few weeks And I have used whatever possible known to me in schemes boundary conditions.
One thing that struck to me is setting the pMin in thermoPhysical properties as chaniging these values effect the run time before crashing.
I would really appreciate if you or some one can guide me through this problem.

Thanks in advance

Hi,

I'm facing same problem too, my case is a cht.
Have you found a solution?

Hello,

Yes i have found the solution it was just some tweeking the problem settings.
If you can explain in more detail the problem might be able to suggest something because the solution i found is problem dependent.
Regards
Rizwan

Thx for the answer, well let's begin.
My case is a conjugate heat transfer with 3 regions: two fluids and one solid.
You can see the attached picture:

• White region is air, inlet blue arrow is inlet, outlet blue arrow is outlet.
• Green region is air, inlet red arrow is inlet, outlet red arrow is outlet.
• at the inlet side of the green region a scalarSemiImplicitSourceCoeffs has been applied with 5000W (absolute).
• case is symmetryc, from the picture the front surfaces are the symmetric Patches

Here follows BC for the green region, that's where the solver blows up.
U
p_rgh
T
fvSchemes
fvSolution
From a rough calculation, air at the green region should reach 2400K, but it gose over 5550K!!!
Then when it cools down pressure start to oscillate and when T becomes negative, it blows up...

this is the output from checkMesh for green region
What is strange is that I already ran the same domain over a coarse mesh, even with more errors in the mesh, but it ran and I "succefully" (ok let's take with care) 2600K and the solution ran with no problem at all...

Just adding some plots about min & max for p T and U

I think I solved my problem by setting:
I saw it, by checking direction of the flow at the inlet side with paraview.
If it is set to yes, U vectors (magnitude) are not perpendicular to the inlet patch and, as my heat source is at the inlet side, combination of the 2 factors raises this problem...I don't think that this is the solution, but 'till now it worked for me.

Maybe this setting is useful when you have a long pipe before entering the domain and you don't know the profile....I supposed :confused:

Regards

Hello Foamers
I too have run across the problem described here. After a small time (0.2 seconds) my simulation breaks down and i get negative T. Also the max courant numbers, both interface and normal become bigger than one, while the mean is 100 to 1000 times smaller. I use compressibleInterFoam to simulate a hot gas (700K) entering from the ground, in an one block mesh, and the outlet is at one of the corners at top. It is an adaptation from a case I was using with interFoam that worked. Here are my files:

blockMesh

topoSet (for inlet and outlet)
[CODE]

actions
(
{
name jetCells;
type cellSet;
action new;
source cylinderToCell;
sourceInfo
{
p1 (5 5 -0.01);
p2 (5 5 0.26);
radius 1;
}


}

{
name f2;
type faceSet;
action new;
source cellToFace;
sourceInfo
{
set jetCells;
option all;
}
}
{
name f2;
type faceSet;
action subset;
source boxToFace;
sourceInfo
{
box (4 4 -0.1) (6 6 0.1);

}
}



{
name f1;
type faceSet;
action new;
source boxToFace;
sourceInfo
{
box (7.9 7.9 14.9) (10 10 15.1);
//boxes ((0 0 0) (1 1 1) (10 10 10)(11 11 11));
}
}
);



U

T

p_rgh

p

alpha.gas (we have gas and air)

thermophysicalProperties

thermophysicalProperties.air

thermophysicalProperties.gas

transportProperties

and I am calculating this as a laminar flow.

fvSolution

fvSchemes

controlDict

So, where am I going wrong with this? Any help would be greatly appreciated.

Best Regards.