[Y..T]

I am using OpenFOAM v2412 and have a question about p_rgh divergence in

Case description
Closed-loop geometry (no inlet/outlet)
Gravity enabled (constant/g correctly defined)
Initial velocity U = 0
No heat source (temperature and density differences are at numerical-noise level)
Constant properties
Mesh passes checkMesh without issues

Observed behavior

During the simulation, p_rgh diverges, and consequently velocity U grows unrealistically

|U| reaches 10³–10⁴ m/s, and |∇p_rgh| reaches O(10⁹)

Changing

p_rgh boundary conditions (zeroGradient, fixedFluxPressure)

relaxation factors

nNonOrthogonalCorrectors
does not improve the behavior

Question
Are there any known issues, limitations, or required settings in buoyantSimpleFoam that could cause p_rgh to diverge in a closed-loop system with negligible density variation?

Any references or similar reported cases would be appreciated.

Please note that English is not my first language, and I am still learning OpenFOAM, so I am using a translation tool. Thank you for your understanding.

[VRN]

Hello Y.T.

It could be because there is no divScheme defined for the pressure field in your fvSchemes.