[Reineke]

Hello everybody,
I am investigating a simple CHT test case in 2D: Slow (laminar & compressible) air flow around solid copper wires which are comprising a constant volumetric heat source (fvOptions, scalarSemiImplicitSource).

An apparently converged, stationary solution is obtained using (sufficiently long) transient chtMultiRegionFoam simulations under constant & homogeneous (T, U) influx conditions.

All resulting fields look qualitatively as expected (see images attached), but the overall energy balance is violated: In the stationary limit, the enthalpy flux difference of the air differs from the overall energy source of the wires by factor 3.7. The kinetic energy change can be neglected (checked). The grid is not optimised but seems sufficiently fine for this test.

What's going wrong here? Thanks for any aid I receive in advance.
Best, Reineke.


-----
Self-reported solution:

Due to quite low Reynolds number and small-scale dimension, conductive axial heat flux must be considered in the present case. Especially in the influx section, the temperature field exhibits a destinct gradient, which causes conductive heat flux across the influx section. Taking this heat flux into account, the overall energy balance of converged solutions is kept.

Thanks for considering this post anyway,

Best, Reineke.