Discussion Forum

Hello,

My problem contains both radiation types: a glass-like solid cylinder with an absorption coefficient is
surrounded by a transparent noble gas within an cylindrical enclosure that has a known temperature
distribution. Every item of the problem is rotationally symmetric, however, I use 3D in order to use
COMSOLs Radiation capabilities.

In the model setting I have used for 'Add Physics'
# Heat Transfer with Surface-to-Surface Radiation (ht)
but have activated the Physical-Model-boxes for both
- Surface-to-Surface Radiation and
- Radiation in Participating Media (no scattering, only absorption)

When disabeling 'Radiation in Participating Media' the results seem reasonable, i.e. the temperature
on top of the cylinder is rotationally symmetric and somewhere inbetween the lowest and highest value
of the boundary temperatures.

However, when including Radiation in Participating Media, the temperature has cold spots along the
periphery every 90 degree. In case of low absorption, the temperature on top of the cylinder has even colder spots than the coldest value at the boundary: totally strange!

Changing parameters (discret: S4 to S6 to S8) has slightly improved the result but not solved the problem.
No effect by finer grid or changing from linear to quadratic discretisation.

Can anybody recommend to me some modification of my settings??
Does anybody know about a demo-example? The ones in the Model library are single problem tasks.

Thanks in advance,
Klaus Boettcher