BSDF 3 layer glass

Hi all,
I would like to have some help in regards of a BSDF definition I’ve generated with genBSDF.
I am are modelling a triple glazing unit with horizontal lamellas within the 1st and 2nd glass panes.

The geometry of the lamellas is designed to reflect out the majority of high altitude sun (>°50), reflect onto the ceiling the sun between 45-50° and 25°, and to let lower sun in (<25°).

For the sake of debug I’ve created a simple room geometry, with a gensky definition with 45° altitude and 0° azimut (that would help me compare the bsdf definition from the viewer with the radiance model).

Despite the bsdf making sense (after checking it with the BSDF viewer) when visualizing the results in a simple test room in radiance the sun is penetrating. The bsdf is applied as material definition to a single surface (glass) having the normal pointing towards the inside of the room.

After a little debug I’ve found out that reducing the n. of glass panes to 1 (so 1 glass surface + lamellas in the model geometry) would make work as expected (lot of light reflected onto the ceiling).

It seems that the number of glass layers on the bsdf geometry built up determines the unexpected behaviour of sun penetrating when there should be no sun?

Has anyone come across this before?
Thanks in advance!

Reducing the glass layer n to 1 makes the glass not reflect anymore. Your solar penetration may be the result of multiple reflections between the lamellas and the glass elements, and is likely the more accurate simulation.

Thanks for you reply Greg.

It still does look strange since I’m working on a very specific commercial product for which we know the real life behavior (from a physical point of view).
Probably I am doing something wrong in my material definition or in setting the bsdf file.

Best Regards