[peterhess]

Hello everybody!

During my training process to use openFoam I work now on free convection simulation using chtMultiRegionFoam...

For this target I built a mesh from two regions a solid region and a fluid one.

The solid region contains a heat source programed in fvOptions file in the solid region constant dictionary.

The fluid region is a prefect gas air assumed...

No radiation (turned of). And no turbulence (laminar), to make the calculation easier...

The mesh has been generated using Ansys Mesher and have very high quality.

- Import made using: fluent3DMeshToFoam Test2.msh!

- The one region imported is splited to two regions using: splitMeshRegions -cellZones -overwrite

- The simulation is splited on two cores using: decomposePar -allRegions

The geometry contains 6 sides, three of them are symmetrical defined.

The upper, lower and back sides are constant temperature 300 K

The contact between solid and fluid is correctly defined (I suppose).

The gravity is defined in Z direction.

To run the simulation simply fire the Allrun file included. The running is in parallel on two cores.

- you can also run using: mpirun -np 2 chtMultiRegionFoam -parallel

Just the last result has been shown here to reduce the size of uploaded file!

Problem:

At the surrounding walls (upper, lower and back sides), the density of air is about 0.75 and should be theoretically 1.1881


I have no idea why is that happens.

Any idea is welcome

Regards

Peter

PS: changing to chtMultiRegionSimpleFoam brought also wrong results

[peterhess]

sorry! the file is too big for CFDOnline. I will manage it and repost...