Radiance and Optics

Hi,

We are working on simulations involving complex glazing assemblies and
day lighting. We use LBL Window4.1, and LBL Optics5 to construct and
simulate these assemblies. We now want to use the assemblies we have
created in Radiance. We see that Optics5 exports "Radiance" files. Has
anyone used this method to simulate glazing systems? Any info would be
appreciated.

Thanks,

Robert Westling

Interesting. I didn't know about those programs. Seems
to be a large collection of glazing definitions. When you
export to radiance in Optics5, you get something like this:

···

On 4 Jun 2002, at 12:09, Westling, Robert wrote:

We are working on simulations involving complex glazing
assemblies and day lighting. We use LBL Window4.1, and LBL
Optics5 to construct and simulate these assemblies. We now
want to use the assemblies we have created in Radiance. We
see that Optics5 exports "Radiance" files. Has anyone used
this method to simulate glazing systems? Any info would be
appreciated.

======================
void glass GRAY_6_glass
0
0
3 0.492 0.490 0.515

void BRTDfunc GRAY_6_front
10
      0.049 0.049 0.051
      0.451 0.449 0.472
      0 0 0
      .
0
9 0 0 0 0 0 0 0 0 0

void BRTDfunc GRAY_6_back
10
      0.050 0.050 0.053
      0.451 0.449 0.472
      0 0 0
      .
0
9 0 0 0 0 0 0 0 0 0

So it appears to be defining a highly accurate material
definition using BRTF. Could some of the gurus comment?
This could be a useful database for the novice Radiance
user with a skimpy materials library (me).

         ====================
Rob Guglielmetti <[email protected]>
    http://home.earthlink.net/~rpg777

Hi Rob,

These definitions are good and bad. they allow you (or Optics5) to define both the specular reflectance and specular transmittance of the glasses, but they don't allow for any angular variation in either. you may be better off just sticking with the glass primitive, using transmissivities calculated from the transmittances delivered by Optics5.

Phil.

···

Interesting. I didn't know about those programs. Seems
to be a large collection of glazing definitions. When you
export to radiance in Optics5, you get something like this:

======================
void glass GRAY_6_glass
0
3 0.492 0.490 0.515

void BRTDfunc GRAY_6_front
10
      0.049 0.049 0.051
      0.451 0.449 0.472
      0 0 0
      .
0
9 0 0 0 0 0 0 0 0 0

void BRTDfunc GRAY_6_back
10
      0.050 0.050 0.053
      0.451 0.449 0.472
      0 0 0
      .
0
9 0 0 0 0 0 0 0 0 0

So it appears to be defining a highly accurate material
definition using BRTF. Could some of the gurus comment?
This could be a useful database for the novice Radiance
user with a skimpy materials library (me).

         ====================
Rob Guglielmetti <[email protected]>
    http://home.earthlink.net/~rpg777

Dear Robert,

I 'm using frequently Optics software to get data for radiance window
modelling. You can also use WINDOW 4.1 instead. I think that the BRTDfunc
description of the glazing gives values for transmittance (and reflectance)
instead of transmissivity (glass material). Naturally there are two
descriptions for front and back surfaces of the window since properties are
quite different. The description in your example deals only with direct
normal transmittance hence it does not take into account the transmittance
angular behavior of the glazing. This can be treated as follows: by using
for example Window 4.1 you can calculate the transmittance in various
incidence angles. By using a polynomiac interpolation you can create a
function of transmittance versus angle of incidence (i.e. Rdot). In fact
this fucntion is a normalized function because each value is diveded by the
normal transmittance. Thus is your BRTDfunc description you should add
glazing.cal function with modified transmittance and reflectance
accordingly.
I hope that this might help.

Aris Tsangrassoulis

Dear Robert,
the difference between different glazing types is actually the difference between the fitting functions (transmittance versus angle of incidence, reflectance versus angle of incidence). Therefore if you change these functions in glazing.cal you can model "anything?".
I'm sure that you are aware that we are dealing with the whole system (i.e. triple glazing). An alternative way is to modify the winxmit.cal file but this has the disadvantage that you have to translate transmissimity to transmitance in the glass type description. I havn;t tested but I suspect that the formula that relates the above variables is not valid for oblique angles of incidence (it -of course- depended on glazing type as well).

Regards,
Aris Tsangrassoulis

Thanks Aris, and the others who replied to this one. Looks like I'll shy away from
these more complex materials until I have a real handle on them. I'll look at
glazing.cal too, so I understand what's going on there before I start using it.

I don't recall seeing any tutorials or info on applying the BRTDfunc in "Rendering
with Radiance". Is there a good tutorial available?

···

On 12 Jun 2002 at 17:09, Aris Tsangrassoulis wrote:

Dear Robert,
the difference between different glazing types is actually the
difference between the fitting functions (transmittance versus angle
of incidence, reflectance versus angle of incidence). Therefore if you
change these functions in glazing.cal you can model "anything?". I'm
sure that you are aware that we are dealing with the whole system
(i.e. triple glazing). An alternative way is to modify the winxmit.cal
file but this has the disadvantage that you have to translate
transmissimity to transmitance in the glass type description. I havn;t
tested but I suspect that the formula that relates the above variables
is not valid for oblique angles of incidence (it -of course- depended
on glazing type as well).

=================================
       Rob Guglielmetti
     [email protected]
http://home.earthlink.net/~rpg777