I was comparing performance of glass vs trans materials and come across a rather unusual phenomenon. For my test case I used a simple box with lid at the top as glass/trans material. I placed my analysis planes above and below the glass surface to define the analysis points. I defined rest of the box surfaces with a Reflectance of 0.01 to minimize any internal reflections.
It seems that the analysis points are affected by the normal direction of the analysis surface. However I am very sure the analysis planes themselves are not part of the simulation. For the test case mentioned below I have removed the glass surface as it doesn’t really play any role. Pls refer to the below images to see what I mean. Yellow colour represents backface.
I also tried simulating hourly daylight, same phenomenon. I changed different scales - meter, millimeter, changed dist from base surface, grid size, etc. same!
Additionally, if I change the box interior surface reflectance from 0.01 to say 0.5 (or something significant), depending on the no. of bounces I DO get the reflected values recorded on the analysis points.
(0.9 box surface reflactance, 5 bounces)