[wildfire230]

Hi all,

I am running into a problem which I do not quite understand related to how exactly continuity errors are calculated. I am using the icoFoam solver. When I run the solver, I consistently have time step continuity errors that are on the order of 1e-16, and if I also have the code print max(fvc::div(phi)), this consistently gives values on the order of 1e-9 or less, except the first few time steps. However, when I view my results in paraview and calculate the divergence of the velocity field, it is 7-8 orders of magnitude larger. Also, when I have the code print max(fvc::div(U)), it is also 7-8 orders of magnitude larger than max(fvc::div(phi)).

I assume that this error is somehow introduced through an interpolation scheme somewhere along the way? However, it's very important for my results to have zero divergence in the velocity field.

I would greatly appreciate any help with this issue.

EDIT: I guess the question boils down to: why are fvc::div(U) and fvc::div(phi) different by 8 orders of magnitude???

Thanks.

[wildfire230]

Sorry for the bump, does anyone have any ideas on why fvc::div(U) and fvc::div(phi) could be different by 8 orders of magnitude?

Thanks.