Description of complex glass systems (geometry+ material)

Marina -

I have not followed this thread very closely, so I don't know exactly what
you are trying to model in Radiance, but thought I would make one comment,
which may or may not be useful.

You say that the constructions you are trying to model are not in the IGDB,
and then you say that you want to model a laminate.

I just want to make sure that you know you can make a laminate in Optics,
depending on what you are using for the glass and interlayer values.

There is a detailed write-up of how to make laminates in Optics on our
Optics Knowledge Base website:

http://windows.lbl.gov/materials/optics5/CurrentVersion/Optics_Knowledge_Bas
e.htm

Click on the link for "NFRC Procedure for Creating Laminates" and you will
get to another page, and there you can click on "Laminate Procedures" for
the detailed description of the process. Here is a direct link to that PDF:
http://windows.lbl.gov/software/NFRC/Training/LaminateProcedure-Final.pdf

Robin Mitchell
Windows & Daylighting Group
Lawrence Berkeley National Laboratory

[mailto:[email protected]] On Behalf Of marina
aviles olmos

···

-----Original Message-----
From: [email protected]
Sent: Monday, May 04, 2009 8:24 AM
To: Radiance general discussion
Subject: Re: [Radiance-general] Description of complex glass systems
(geometry+ material)

Hi Marija, hi Jack

Thanks for your mails.
Some glass types I want to simulate are not in the IGDB, so I will
describe the glasses with the measurements-method,
but I am not really sure I have understood this method properly. There
are those precisions that make radiance so great, but sometimes also
difficult to learn. Could you tell me if this would be then right as
follows?

This is the hole double glazing geometry description:

#window geometry:

double_glazing_a polygon window
0
0
12 0 -0.18 0
        0 -0.18 3.34
       4.13 -0.18 3.34
       4.13 -0.18 0

Then I should measure the interior pane (a laminated float glass with an
interlayer) and separately the exterior pane (a 4 mm float glass). The
hole input data are then:

Red: 700nm -- rf1_r rb1_r rf2_r rb2_r t1_r t2_r
Green: 546nm -- rf1_g rb1_g rf2_g rb2_g t1_g t2_g
Blue: 436 -- rf1_b rb1_b rf2_b rb2_b t1_b t2_b

Then substitute these in the formel you said and this would be the
complete description of the double glazing material:

#window material:

void BRTDfunc double_glazing_a
10
if(Rdot,cr(fr(rf1_r),ft(t1_r),fr(rf2_r)),cr(fr(rb2_r),ft(t2_r),fr(rb1_r)))
if(Rdot,cr(fr(rf1_g),ft(t1_g),fr(rf2_g)),cr(fr(rb2_g),ft(t2_g),fr(rb1_g)))
if(Rdot,cr(fr(rf1_b),ft(t1_b),fr(rf2_b)),cr(fr(rb2_b),ft(t2_b),fr(rb1_b)))
ft(t1_r)*ft(t2_r)
ft(t1_g)*ft(t2_g)
ft(t1_b)*ft(t2_b)
        0 0 0
        glaze2.cal

0
9
        0 0 0
        0 0 0
        0 0 0

I also try to understand how this calculation system is related to the
physical process. So I have another question.
If I measure with a spectrometer the complete normal reflectance and
transmittance of a double glazing, the interior reflectances are already
in the total measurement. So I understand that, also when in the
physical reality the total reflectance of a double glazing could be the
input data for the simulation, the calculation system of radiance
calculates only right if you specify the interior reflectances and the
calculation take place within the BRTD function. Am I right on that?

Thanks again,

Marina

Hi Marina,

Yes you want to describe the glass geometry, that is one polygon
representing the multiple layers of the physical makeup. The material is
what should represent the behavior. If you take the output from Optics
and run it through optics2rad you will get a single BRTDfunc material
description. One thing to be careful of is that this is dependent on the
correct orientation (normal) of the glass geometry. I believe that the
normal is supposed to be pointing "into" the interior of the building.
One way to check normal orientation in radiance is to assign
(temporaryily) a "glow" material to the glass geometry and view it. The
sides that are "black" are the back sides of the geometry, the sides
that are not black are the positive normal side.

Regards,

-Jack de Valpine

Marija Velickovic wrote:

Hi Marina,

1.glazing geometry

Always define glazing object as a polygon, no matter how many layers
of glass it contains in real world.
Double clear, laminated and other glazings differ only in BRTDfunc
(light transmittance and reflectance distribution function), while
real glazing width is not important here.

Note that polygon face should be oriented towards building interior.

##########
2.glazing material
Since BRTD function for double glazing is different then for single
glazing, I don't suggest using of simple glass material.
Procedure we use is next:
*Measure each glass layer transmittance and reflectance data. If you
can measure their r,g,b components it is good, if not use the same
value for all 3 components.
or
*Export from IGDB (using Optics5 I suppose), each layer separately.
And write down transmittances/reflectances you have in BRTD functions.
Note that in Optics5 front side of the glazing is towards outside, so
their front and Radiance "front" are different.

So you should have next parameters for double glazing parameters after
measures or after export:
rf1 - interior pane front reflectance (towards the room)
rb1 - interior pane back reflectance (towards the exterior pane)
rf2 - exterior pane front reflectance (towards the interior pane)
rb2 - exterior pane back reflectance (towards the exterior)
t1 - interior pane transmittance
t2- exterior pane transmittance
r,g,b - red, green and blue color components
*For clear glass layers front and back reflectance are always the same
/
void BRTDfunc double_glazing
10
if(Rdot,cr(fr(rf1_r),ft(t1_r),fr(rf2_r)),cr(fr(rb2_r),ft(t2_r),fr(rb1_r)))
if(Rdot,cr(fr(rf1_g),ft(t1_g),fr(rf2_g)),cr(fr(rb2_g),ft(t2_g),fr(rb1_g)))
if(Rdot,cr(fr(rf1_b),ft(t1_b),fr(rf2_b)),cr(fr(rb2_b),ft(t2_b),fr(rb1_b)))
ft(t1_r)*ft(t2_r)
ft(t1_g)*ft(t2_g)
ft(t1_b)*ft(t2_b)
        0 0 0
        glaze2.cal

0
9
        0 0 0
        0 0 0
        0 0 0
/

In glazing definition function
if(Rdot,cr(fr(rf1_r),ft(t1_r),fr(rf2_r)),cr(fr(rb2_r),ft(t2_r),fr(rb1_r)))
means:
-if light is incident to the front side of the glazing then
reflectance is cr(fr(rf1_r),ft(t1_r),fr(rf2_r))
-else (light is incident to the back side) reflectance is
cr(fr(rb2_r),ft(t2_r),fr(rb1_r))

Functions for double glazing reflectance and transmittance are defined
in *glaze2.cal *script which is provided with Radiance:
-cr(refl1, trans,refl2) - is reflectance distribution for double
glazing as function of light incident angle. refl1 is reflectance of
first glazing pane, refl2 of the second one, and tran1 is
transmittance of the first glazing pane
-ft(t) - transmittance distribution for single glazing as function of
light incident angle. For double glazing total transmittance is
ft(t1)*ft(t2)

Hope this helps,
Marija
De Luminae <http://www.deluminaelab.com>

------------------------------------------------------------------------

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
  
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

Hi Marina,

Looks about right to me -- does it reproduce the combined transmittance and reflectance you're measuring? Are the interfaces separated by an air gap, or are some/all layers laminated with an optical adhesive? In the latter case, the functions in glaze2.cal won't compute the correct combined transmittance and reflectances.

-Greg

···

From: marina aviles olmos <[email protected]>
Date: May 5, 2009 8:05:45 AM PDT

Hi Robin, hi All

yes I want to simulate a laminated glass in a double glass system. This
laminated glass is a prototyp of a glazing system and there is no
information about it anywhere. It is also a quite complicated system
with embeded coating and interlayer and I can get its description from
Optics5. That's why I think the way to describe this glasses are with
the spectrometer measurements.

I am still not sure anyway if I describe the glasses propertly.
Does anybody knows if this is right like this:

#window geometry:

double_glazing_a polygon window
0
12 0 -0.18 0
        0 -0.18 3.34
       4.13 -0.18 3.34
       4.13 -0.18 0

#window material:

void BRTDfunc double_glazing_a
10
if(Rdot,cr(fr(rf1_r),ft(t1_r),fr(rf2_r)),cr(fr(rb2_r),ft(t2_r),fr(rb1_r)))
if(Rdot,cr(fr(rf1_g),ft(t1_g),fr(rf2_g)),cr(fr(rb2_g),ft(t2_g),fr(rb1_g)))
if(Rdot,cr(fr(rf1_b),ft(t1_b),fr(rf2_b)),cr(fr(rb2_b),ft(t2_b),fr(rb1_b)))
ft(t1_r)*ft(t2_r)
ft(t1_g)*ft(t2_g)
ft(t1_b)*ft(t2_b)
        0 0 0
        glaze2.cal

0
9
        0 0 0

where I just have to substitute the next measured data:

Red: 700nm -- rf1_r rb1_r rf2_r rb2_r t1_r t2_r
Green: 546nm -- rf1_g rb1_g rf2_g rb2_g t1_g t2_g
Blue: 436 -- rf1_b rb1_b rf2_b rb2_b t1_b t2_b

Thanks again,

Marina

Hi Marina,

I think you had best treat this as a two-glazing system, and use glaze2.cal similar to how you suggested in your previous e-mail. You'll need to measure the transmittance and front/back reflectances of each piece of glazing (laminated and float), and provide these as the inputs to your BRTDfunc model. You could even use the glazing script as is if your float glass is close enough to a standard clear glazing by providing your own definition of the other glass as input a la:

  http://www.radiance-online.org/pipermail/radiance-general/2009-January/005665.html

I hope this helps.
-Greg

···

From: marina aviles olmos <[email protected]>
Date: May 5, 2009 11:02:00 AM PDT

Hi Greg,

first of all, there is a mistake in my previous mail. I wanted to say
< and I cannot get its description from Optics5.>

It is the latter case:
outside the laminate (glass pane + optical adhesive coating +... +
interlayer + ... + adhesive coating + glass) then the air gap and the
inside that is a float glass.
(It is a switchable glass in a double glazing)

How can I then describe this glasses?

(I have the possibility to measure the double glazing or the different
layers separately)
Just a naive idea that could helps:

If the combined R and T are the ones resulting from the hole double
glazing which could be measured in the spectrometer if I measure the
complete double glass and not the separated layers:

Total Reflectance front
Total Reflectance back
Total Transmitance

Could be with this data a way to describe them?

Marina

Hi Marina,

That is correct. The "glass" type is only suitable for single-pane glazings, and simple ones at that.

Best,
-Greg

···

From: marina aviles olmos <[email protected]>
Date: May 6, 2009 10:04:20 AM PDT

Hi Greg,

thanks.

What I actually try to do in a part of my current research is a
comparative study of the daylight efficiency of different glass systems
and some of this glasses are switchable, so thanks the link.

From your answer, I guess that the glaze2.cal is not the exact script to

this ones although the results will be quite accurate.

The method that Jack describes has some points that I can understand at
the moment, so I have tried (for the moment) the first method:

void BRTDfunc double_glazing_a
10
if(Rdot,cr(fr(0.075),ft(0.831),fr(0.153)),cr(fr(0.188),ft(0.158),fr(0.075)))
if(Rdot,cr(fr(0.075),ft(0.899),fr(0.134)),cr(fr(0.150),ft(0.215),fr(0.082)))
if(Rdot,cr(fr(0.083),ft(0.88),fr(0.141)),cr(fr(0.150),ft(0.195),fr(0.083)))
ft(0.831)*ft(0.158)
ft(0.899)*ft(0.215)
ft(0.88)*ft(0.195)
0 0 0
glaze2.cal

0
9
0 0 0

and the results looks like right, but just to be sure:

the descrition of:

mod glass id
0
3 rtn gtn btn

is not included in this cases of double glazing. Is it right?
and, it is only included, I guess after your mails, only in monolitic
not double glazings, isn't it?

Marina