The glass geometry from rhinoceros should not be solid?

Hi, I have a question about geometry when I call Radiance from Rhinoceros(grasshopper).
The glass geometry is solid(a thin hexahedron). If Radiance calculates glass transmission for each surface, it would be simulated as there are two glasses.

Should I remodel the glass geometry into single mesh?
I couldn’t find this kind of basic help, Is there any place for beginners q&a?

Thank you.

Assuming all the glass surfaces are oriented outwards (i.e., vertices are listed counter-clockwise as seen from the front), then you may use the “dielectric” type rather than glass to get the same behavior. The calculation is slightly more expensive, since it involves multiple internal reflections, but the result will be correct.

Another option if it is not too difficult is to create two materials, one made of glass and the other completely transparent. Use the glass material for the inside-facing polygon and the completely transparent material for the rest of the glazing surfaces, e.g.:

void trans invisible
7 1 1 1 0 0 1 1

void glass glass_mat
3 .92 .92 .92

You can also replace the modifier of the extra surfaces with “void”, which is even faster than using the invisible material, above.


Hi Greg,

Thank you for your perfect advice.
I understand what I should do.

Hi Akito, with grasshopper you can also explode the solid brep(assuming it’s box shape) into 6 surfaces and keep the largest of them. In that case you keep one but it’s random whether it’s the interior or exterior pointing pane. For 3/5 phase simulations the normal is important but afaik not on a 2phase simulation/daysim.

Let me know if you need any grasshopper definition to find the largest pane.

Best, Mathias

Hi Mathias, sorry for my late late reply. Thank you for your practical advice. I think I can code finding
largest surface. But can I ask what you mean by 3/5 phase simulations and 2phase simulation?