Does anyone have any suggestions which would be helpful for modelling blinds
which let different amounts of light through at different angles of incident
light, an obvious example of this being a Venetian blind. The particular case
I am interested in is a woven wire blind where the light incident normal to the
blind would be transmitted better than at an angle.
At the moment, I am modelling using BRTDfunc with a tranmittance value
appropriate to the light at normal incidence, however I assume this is no
different really to modelling glass with the same transmittance value.
Thanks,
Richard
Clibborn
Hi Richard,
Attached is a .cal file which is distributed with the latest version of Radiance, I believe. It allows you to specify the spacing of a grating, and it comes up with a reasonable approximation to the angular transmission.
-Greg
grating.cal (1023 Bytes)
Greg,
Thanks for that. I'm just a little unsure about the orientation of
this "aligned with x-y plane" in the header .. is this absolute (ie. the
project x-y plane) or does it automatically align with the surface it is being
applied to. Sorry if this is a stupid question, it's just I have had trouble
before simulating fritting with mixfunc and different surface orientations.
Cheer,
Richard.
···
Hi Richard,
Attached is a .cal file which is distributed with the latest version of
Radiance, I believe. It allows you to specify the spacing of a
grating, and it comes up with a reasonable approximation to the angular
transmission.
-Greg
It means that the .cal file assumes the surface is aligned in the x-y plane. Assuming you want a vertical grid and your z-axis is up, adding the following transform to the string arguments would rotate the material so that your surface was instead in the x-z plane:
-rx -90
Since we use a right-handed coordinate system in Radiance, and positive angles correspond to counter-clockwise rotations looking down the selected axis, this rotates the positive y direction in the original orientation so it is now in the positive z direction. Since this material is symmetric front-to-back, it actually didn't matter in this case which direction we rotated it.
If you want the surface in the y-z plane instead, you would use:
-ry -90
This rotates the positive x-axis into the positive z direction. Just to be clear, your BRTDfunc would look like so:
void BRTDfunc grating
12
0 0 0
spectrans spectrans spectrans
0 0 0
grating.cal -ry -90
0
13
rrefl grefl brefl
rtrns gtrns btrns
rrefl grefl brefl
A10 A11 A12 A13
You just add a couple of arguments to the string arg count, and tack them on the end.
I hope this helps.
-Greg
···
From: Richard Clibborn <furry@ihug.com.au>
Date: February 15, 2005 10:07:55 PM PST
Greg,
Thanks for that. I'm just a little unsure about the orientation of
this "aligned with x-y plane" in the header .. is this absolute (ie. the
project x-y plane) or does it automatically align with the surface it is being
applied to. Sorry if this is a stupid question, it's just I have had trouble
before simulating fritting with mixfunc and different surface orientations.
Cheer,
Richard.