Courant number explodes
 

Hello Foamers!

i am trying to simulate the bifurcation that is depicted in the bifurcation.png file. There is one inlet in the right of the geometry and three outlets at the right.

I use icoFoam to solve the flow problem. the deltaT is 1e-4. The courant number at first is very small but is increased with a small step after each iteration. After 10000 iterations the courant number is getting too big so the simulation stops.

The mesh is ok after checkMesh.

I calculate the Dp in order to have the desired mean velocity in the pipe before the bifurcation and then i divide with density.

0/p File

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

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

internalField uniform 0;

boundaryField
{
inlet
{
type fixedValue;
value uniform 0.012957446;
}

outlet_internal
{
type fixedValue;
value uniform 0;


}

outlet_externalbig
{
type fixedValue;
value uniform 0;

}

outlet_externalsmall
{
type fixedValue;
value uniform 0;

}

walls
{
type zeroGradient;
}

}



0/U File


object U;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

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

internalField uniform (0 0 0);

boundaryField
{
inlet
{
type pressureInletVelocity;
value uniform (0 0 0);
}

outlet_internal
{
type zeroGradient;
}

outlet_externalbig
{
type zeroGradient;
}

outlet_externalsmall
{
type zeroGradient;
}

walls
{
type fixedValue;
value uniform (0 0 0);
}


}


transportproperties

transportModel Newtonian;

nu nu [ 0 2 -1 0 0 0 0 ] 0.000002057;




Can anyone help me to solve the problem ? I stuck over 15 days

Hey
Why are you running a transient simulation for this case?

I know that Poiseuille flow is treated as steady state.

But i have developed a code based on transient solver and i need to evaluate the results of my case with experimental data

Ahh okay. Did you try running adjustable time stepping? Also, try pimplefoam solver : use higher outerCorrectors and relax the solution

If you know the desired velocity at the inlet, just impose that, rather than the pressure at the inlet (it seems you use that to find the pressure drop?).

Also, as a suggestion, I would not recommend icoFoam. You could use pimpleFoam, and have better control on the solution process (outer correctors, adaptive time-step, under-relaxation, ...).