[NewtoFOAM]

Hello everyone,

Been a while since using Openfoam but I want to get back into it for an interesting problem I have at work. I thought I'd see if I could just simulate flow through a cylinder but, annoyingly, I can't even get this to work.

I have three boundaries; inlet, outlet and surface with the following simple conditions

U:

inlet
{
type fixedValue;
value uniform (0 0 1);
}
outlet
{
type zeroGradient;
}
surface
{
type fixedValue;
value uniform (0 0 0);
}

p:

inlet
{
type zeroGradient;
}

outlet
{
type fixedValue;
value uniform 0;
}

surface
{
type zeroGradient;

}

When running simpleFOAM the pressure solver falls over on the first iteration. That leads me to think that my boundaries are incorrect.....but I really don't know what I'm doing wrong. Maybe my solver set ups are wrong....but I've copied them from a simpleFOAM tutorial and seem sensible.

I'm using OpenFOAM for windows (latest). I've uploaded the case file to here. http://www.filedropper.com/ahu

If anyone could help me tell me what daft mistake I'm making I would be ever so grateful.

Thanks in advance,
Dave

[T.D.]

Hi,

with simpleFoam solver, many times you need to play with the values in "relaxationFactors" and "nNonOrthogonalCorrectors"
in system/fvSolution
Did you try this ?
Did you check mesh errors, by "checkMesh -allTopology"

Regards,
T.D.

[matejfor]

Well if your case is failing in the first pressure solution, your assumption was absolutely correct. Wrong BCs. Are you sure, that 'z' direction for velocity is the right one? (Just checking). What about your initial conditions? And as noted above, what about your mesh quality? Mainly non-orthogonality? What solver are you using for pressure? GAMG? what is the settings for it? Do you use SIMPLE or SIMPLEC? And last not least, what happens, when you run potentialFlow on the case instead of simpleFoam?

[NewtoFOAM]

Hello all,

Thank you all for your messages! I changed my GAMG solver to a PCG solver with a cheeky DIC conditioner and it seemed to work a treat! Now I can focus on my actual interesting problem.

Thank you all again for your help.

Cheers,
Dave