[piotr.mecht]

I prepared some simulation setups for buoyantPimpleFoam which worked perfectly fine whith every scenario so far on my home PC, which runs OpenFOAM v1906 Windows10-native on six core 9th gen Intel processor (I used variable number of cores). I moved my simulation files on some offhand PC at my workplace, so I could do other stuff at home. This machine runs Openfoam 7 with Linux Mint on some older 4 core Intel Core 2 Quad Q8200 processor. I had issues with correct docker setup (i'll work on this when i'll find some more time for this), so i switched to foundation version for easy installation and usage.


My simulation quickly goes unstable after first iterations on the second machine with the same files that worked on the first machine. When I reduced time step vale the max Courant number grows even quicker.
I've seen this symptom before, while using icoPolynomial density model with pressure-work included to energy equation (sensibleEnthalpy). I disabled this with:

dpdt off;

in thermophysicalProperties dictionary and this worked well on my home machine. Is this input unrecognized with foundation version maybe?


What shoud I do to expect the same output?

Update:
I've managed to install Docker on Linux machine later taht day and the simulation in v1906 works like on my home computer now, so it's definitely .org/.com build dependent. Despite this, I'm curious about differences to my input, that would possibly produce the same results. I'm not convinced to use OF commads in separate container as it's harder to navigate between e.g. the text editor and other stuff.

[jherb]

I see one major difference in the solvers between the Foundation and the ESI (OpenFOAM-plus) development lines:


In the pEqn.H of the ESI version, the following lines have been added (assuming, that you have a boundary condition on at least one patch with a fixed pressure

thermo.correctRho(psi*p - psip0);

It would be really interesting, what happens, if you remove this line form the solver, recompile it and test it:
https://develop.openfoam.com/Develop...am/pEqn.H#L100