simulating fritted glazing systems

Hello:

I am interested in people's comments on how best to simulate frit glazing systems from a performance and appearance standpoint. We are interested in comparing the appearance based on available performance data of a variety of fritted glazing systems on the exterior of a building system under daylight conditions. The fundamental challenge it seems to me is how to use the data that is readily available (from software such as LBL's Optics5 as well as from manufacturers themselves) to make an appropriate material in Radiance and whether the readily available data is in fact sufficient to describe the material behavior.

All the frit systems that we are looking at use the same basic insulating glass with the following makeup:

    * (exterior) 6mm clear glass with low-e coating on the second surface
    * (airspace) 13.2mm
    * (interior) 6mm clear glass

A Radiance material description (brtdfunc) for this makeup can be generated by the LBL Optics5 program. Note that Optics5 outputs material descriptions for the glass primitive as well as interior and exterior descriptions based on the BRTDfunc. In this case I believe that the "front" version of the BRTDfunc represents the exterior and is the appropriate description to use for this study.

void glass ve1-2m_glass
0
3 0.712 0.791 0.719

void BRTDfunc ve1-2m_front
10
      0.111 0.114 0.116
      0.653 0.726 0.659
      0 0 0
      .
0
9 0 0 0 0 0 0 0 0 0

void BRTDfunc ve1-2m_back
10
      0.112 0.125 0.122
      0.653 0.726 0.659
      0 0 0
      .
0
9 0 0 0 0 0 0 0 0 0

The information that we do know is as follows:

    * measured photopic transmittance, and exterior and interior
      reflectance of base insulating glass system (70%, 11% and 12%)
    * radiance material description of base insulating glass system
    * measured photopic transmittance, and exterior and interior
      reflectance of selected composite frit glass systems (ie the
      composition of the frit silkscreen [white] onto the 2nd or 3rd
      surface of the base insulating glass system, for example 45%,
      22%, and 24% based on a 50% frit on the 2nd surface) , note that
      as far as I can tell it is not possible to build up such as system
      in a program like Optics5 (unless additional baseline measured
      data can be obtained from the manufacturer)

Is it sufficient to figure out the percentage variations for the transmittance and reflectance values between the original glass and the composite frit glass and then modify the brtdfunc material description accordingly? I know that this will not completely account for the change in color of the glass due to the white silkscreen on the 2nd or 3rd surface. But is this a reasonable approximation to start with.

Another thing to consider is the 'modeling' of the actual frit pattern on the glass. My thinking is that this is only important for cases where the glass is very close to the viewer. Typically a lot of the frit patterns (such as dot and lines) tend to disappear the the farther away from the building you are. Now in the cases where the view is close to the glass, what is a good way to build a composite material in Radiance (a mixfunc is probably a good starting point)? We know the data for the base insulating glass but what is a good way to figure out the reflectance of the frit silkscreen as seen through the glass?

I will look forward to your suggestions and comments.

Regards,

-Jack de Valpine