Glazing transparency/reflectivity study

radiance-general
1

I'm considering doing a study comparing the transparency (as viewed from outside) of two glazing types on a building modeled in radiance. The study would render several viewpoints (and view angles looking up the facade) for the multiple glass types under a variety of sky conditions (sunny, cloudy, etc). The primary goal of the study would be to as follows:

1. Demonstrate how exterior transparency changes with different glass types (with different coatings and exterior reflectances)
2. Demonstrate how exterior transparency changes with angle of incidence
3. Demonstrate how exterior transparency changes based on ratio of interior to exterior illuminance (or more accurately, luminances).

I'm only interested in this as a comparative study: does glass type A look clearer than glass type B when viewed at X angle and under Y sky condition, and Z interior/exterior illuminance ratio.

My question to the list is:
1. Comments on the usefulness of this study?
2. Advantage of using Optics 5 rad file (BRTDfunc) over glass primitive? My understanding is it more accurately captures the color of the glass and reflectivity... anything else?

Thanks,
Galen

···

____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

2

Hi Galen,

I won't comment on the usefulness of your proposed study, other than to say it sounds interesting to me... I do have a couple of suggestions, however.

The output of Optics5 should not be used directly, but should be passed through the "optics2rad" script to get the desired angular behavior from the selected glazing. In the case of multi-layer glazing systems, you may be better off using the "glaze" script to define your material. This script does a much better job modeling the multiple reflections that occur in a double-glazed system.

Best,
-Greg

···

From: Galen Burrell <[email protected]>
Date: January 14, 2009 2:39:40 PM PST

I'm considering doing a study comparing the transparency (as viewed from outside) of two glazing types on a building modeled in radiance. The study would render several viewpoints (and view angles looking up the facade) for the multiple glass types under a variety of sky conditions (sunny, cloudy, etc). The primary goal of the study would be to as follows:

1. Demonstrate how exterior transparency changes with different glass types (with different coatings and exterior reflectances)
2. Demonstrate how exterior transparency changes with angle of incidence
3. Demonstrate how exterior transparency changes based on ratio of interior to exterior illuminance (or more accurately, luminances).

I'm only interested in this as a comparative study: does glass type A look clearer than glass type B when viewed at X angle and under Y sky condition, and Z interior/exterior illuminance ratio.

My question to the list is:
1. Comments on the usefulness of this study?
2. Advantage of using Optics 5 rad file (BRTDfunc) over glass primitive? My understanding is it more accurately captures the color of the glass and reflectivity... anything else?

Thanks,
Galen

3

Hi Galen,

You should probably consider the glaze script to produce a material description for your glazing. I think that this will you get the most sophisticated description for a glass material. You can incorporate data that is output from Optics for single monolithic glass types to build up double pane insulating units if needed. A coating can be applied to a selected layer of the make-up. Note, that using the simple BRTDfunc description output from Optics (unless this has been updated in the past few years) does not account for variation due to incident angle. There is actually a way to set up the glaze script so it accepts a file of glass types as input rather than using the rather minimal inputs that come with script.

Let me know if you have any questions and I will do the best I can to help out.

-Jack

Galen Burrell wrote:

···

I'm considering doing a study comparing the transparency (as viewed from outside) of two glazing types on a building modeled in radiance. The study would render several viewpoints (and view angles looking up the facade) for the multiple glass types under a variety of sky conditions (sunny, cloudy, etc). The primary goal of the study would be to as follows:

1. Demonstrate how exterior transparency changes with different glass types (with different coatings and exterior reflectances)
2. Demonstrate how exterior transparency changes with angle of incidence
3. Demonstrate how exterior transparency changes based on ratio of interior to exterior illuminance (or more accurately, luminances).

I'm only interested in this as a comparative study: does glass type A look clearer than glass type B when viewed at X angle and under Y sky condition, and Z interior/exterior illuminance ratio.

My question to the list is:
1. Comments on the usefulness of this study?
2. Advantage of using Optics 5 rad file (BRTDfunc) over glass primitive? My understanding is it more accurately captures the color of the glass and reflectivity... anything else?

Thanks,
Galen

____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses
  ------------------------------------------------------------------------

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
  
--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

4

Hi Galen,

A couple of subtleties to watch out for.

   1. the normals for glass geometry must be uniformly oriented and
      point to the interior of the space
   2. the header spacing if very specific and relies on tabs only as I
      recall
   3. I think that you will also find that the appearance in the images
      will be sensitive to the ambient parameters that are used

-Jack

Jack de Valpine wrote:

···

Hi Galen,

No problem.

Regarding the header

    * Rc<r,g,b> - these are the rgb components for the reflectance on
      the "coated" side of the monolithic glass
    * Rg<r,g,b> - these are the rgb components for the reflectance on
      the non-coated side of the monolithic glass
    * T<r,g,b> - these are the rgb components for the transmittance
      for the monolithic glass
    * Part - this is a boolean value
          o 0 means this is a regular glass
          o 1 means this is a frit - if this is the case the "coated"
            side reflectance needs to represent the reflectance of the
            frit at 100% coverage which should be possible to extract
            from the glazing manufacturer

The RGB values are extracted from the file generated for a "radiance" export from Optics. This should be for a single pane monolithic glass description (not a multi-layer make-up). The file will have three Radiance material definitions, the two BRDTfunc definitions are the ones to pay attention to. The makeup is something like the following:

    void BRTDfunc some.glass.front
    10

        Rfr Rfg Rfb
        Tr Tg Tb
        0 0 0
        .

    9 0 0 0 0 0 0 0 0 0

    void BRTDfunc some.glass.back
    10

        Rbr Rbg Rbb
        Tr Tg Tb
        0 0 0
        .

    9 0 0 0 0 0 0 0 0 0

So you need to take the two rgb triples for reflectance and one for transmittance. So an entry in the data file could look like this (you will need to figure which represents the coated side if needed):

    Surface Tr Tg Tb Rcr Rcg Rcb Rgr Rgg Rgb Part
    <name> Tr Tg Tb Rfr Rfg Rfb Rbr Rbg Rbb 0
      
A file for a double pane insulating unit might look like the following:

    Surface Tr Tg Tb Rcr Rcg Rcb Rgr Rgg Rgb Part
    green-lowe Tr Tg Tb Rfr Rfg Rfb Rbr Rbg Rbb 0
      
The glaze script contains a default "clear" glass. I seem to recall that this is required for the script to output a material description (eg it must have a "clear" glass description"). If this is not suitable the script requires some modification, which I think I can offer some guidance on so that you can include your own "clear" glass description.

I would suggest before you do anything else you should test out the script as is with the built in values. This will let you see how it works. You will assign different materials to different layers to build up the composite unit.

Try it out and let me know if you have more questions.

-Jack

Galen Burrell wrote:

Jack,
Thanks for the input and offer to answer a few questions. Much appreciated.
I see the header information the glaze script is looking for the following when importing a data file:

Surface Tr Tg Tb Rcr Rcg Rcb Rgr Rgg Rgb Part

Questions:
1. Are the RGB values for T, Rc, Rg easily found in Optics 5? I selected 'Color - Transmittance' under Tools>Select Spectrum, and tried all the different wavelength functions, but I'm not seeing RGB anywhere.
Am I missing something?
2. Between Rc and Rg, which one refers to outside surface?
3. What does 'Part' refer to in the header above?
4. For a double pane IGU with a low-e coating on green glass (surface 2), would the data file look something like this?

Surface Tr Tg Tb Rcr Rcg Rcb Rgr Rgg Rgb Part
1 Trgb, Rrgb, Part information for green glass
2 Trgb, Rrgb, Part information for green glass + LowE coating (or would it be just the LowE coating?)
3 Trgb, Rrgb, Part information for clear glass
4 Trgb, Rrgb, Part information for clear glass

Thanks for your time and assistance - much appreciated.

Galen

Message: 3
Date: Wed, 14 Jan 2009 18:08:02 -0500
From: Jack de Valpine <[email protected]>
Subject: Re: [Radiance-general] Glazing transparency/reflectivity
        study
To: Radiance general discussion <[email protected]>
Message-ID: <[email protected]>
Content-Type: text/plain; charset="iso-8859-1"

Hi Galen,

You should probably consider the glaze script to produce a material
description for your glazing. I think that this will you get the most
sophisticated description for a glass material. You can incorporate data
that is output from Optics for single monolithic glass types to build up
double pane insulating units if needed. A coating can be applied to a
selected layer of the make-up. Note, that using the simple BRTDfunc
description output from Optics (unless this has been updated in the past
few years) does not account for variation due to incident angle. There
is actually a way to set up the glaze script so it accepts a file of
glass types as input rather than using the rather minimal inputs that
come with script.

Let me know if you have any questions and I will do the best I can to
help out.

-Jack

Galen Burrell wrote:
  

I'm considering doing a study comparing the transparency (as viewed
from outside) of two glazing types on a building modeled in
radiance. The study would render several viewpoints (and view angles
looking up the facade) for the multiple glass types under a variety of
sky conditions (sunny, cloudy, etc). The primary goal of the study
would be to as follows:

1. Demonstrate how exterior transparency changes with different glass
types (with different coatings and exterior reflectances)
2. Demonstrate how exterior transparency changes with angle of incidence
3. Demonstrate how exterior transparency changes based on ratio of
interior to exterior illuminance (or more accurately, luminances).

I'm only interested in this as a comparative study: does glass type A
look clearer than glass type B when viewed at X angle and under Y sky
condition, and Z interior/exterior illuminance ratio.

My question to the list is:
1. Comments on the usefulness of this study?
2. Advantage of using Optics 5 rad file (BRTDfunc) over glass
primitive? My understanding is it more accurately captures the color
of the glass and reflectivity... anything else?

Thanks,
Galen

____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

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

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

--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://radiance-online.org/pipermail/radiance-general/attachments/20090114/1769d356/attachment-0001.html

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

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

End of Radiance-general Digest, Vol 59, Issue 8
***********************************************

--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction
  ------------------------------------------------------------------------

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
  
--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

5

Thanks Jack.

All seems clear and I've got some results that I'm pretty happy with. One thing I'm still a little unclear about is how you would apply a low-e film to a non-clear glass (i.e. tinted green, blue, etc) if the script requires one surface to be "clear" for each pane.

For example:
VE2-2M (http://www.viracon.com/makeupsheet.php?id=3840) has the VE-2M coating applied to Evergreen glass on the #2 surface. As it is now, I would input:
Surface 1 - Clear
Surface 2 - VE2-2M on evergreen (from optics 5 data)
Surface 3 - Clear
Surface 4 - Clear

Is there an easy hack to make Surface 1 evergreen instead of clear?

Thanks,
Galen

···

-----Original Message-----
Date: Thu, 15 Jan 2009 21:07:06 -0500
From: Jack de Valpine <[email protected]>
Subject: [Radiance-general] Re: Radiance-general Digest, Vol 59, Issue
        8
To: Galen Burrell <[email protected]>
Cc: Radiance general discussion <[email protected]>
Message-ID: <[email protected]>
Content-Type: text/plain; charset="iso-8859-1"

Hi Galen,

No problem.

Regarding the header

    * Rc<r,g,b> - these are the rgb components for the reflectance on
      the "coated" side of the monolithic glass
    * Rg<r,g,b> - these are the rgb components for the reflectance on
      the non-coated side of the monolithic glass
    * T<r,g,b> - these are the rgb components for the transmittance for
      the monolithic glass
    * Part - this is a boolean value
          o 0 means this is a regular glass
          o 1 means this is a frit - if this is the case the "coated"
            side reflectance needs to represent the reflectance of the
            frit at 100% coverage which should be possible to extract
            from the glazing manufacturer

The RGB values are extracted from the file generated for a "radiance"
export from Optics. This should be for a single pane monolithic glass
description (not a multi-layer make-up). The file will have three
Radiance material definitions, the two BRDTfunc definitions are the ones
to pay attention to. The makeup is something like the following:

    void BRTDfunc some.glass.front
    10

        Rfr Rfg Rfb
        Tr Tg Tb
        0 0 0
        .

    9 0 0 0 0 0 0 0 0 0

    void BRTDfunc some.glass.back
    10

        Rbr Rbg Rbb
        Tr Tg Tb
        0 0 0
        .

    9 0 0 0 0 0 0 0 0 0

So you need to take the two rgb triples for reflectance and one for
transmittance. So an entry in the data file could look like this (you
will need to figure which represents the coated side if needed):

    Surface Tr Tg Tb Rcr Rcg Rcb Rgr Rgg Rgb Part
    <name> Tr Tg Tb Rfr Rfg Rfb Rbr Rbg Rbb 0

A file for a double pane insulating unit might look like the following:

    Surface Tr Tg Tb Rcr Rcg Rcb Rgr Rgg Rgb Part
    green-lowe Tr Tg Tb Rfr Rfg Rfb Rbr Rbg Rbb 0

The glaze script contains a default "clear" glass. I seem to recall that
this is required for the script to output a material description (eg it
must have a "clear" glass description"). If this is not suitable the
script requires some modification, which I think I can offer some
guidance on so that you can include your own "clear" glass description.

I would suggest before you do anything else you should test out the
script as is with the built in values. This will let you see how it
works. You will assign different materials to different layers to build
up the composite unit.

Try it out and let me know if you have more questions.

-Jack

____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

6

I think you just specify surface 1 as evergreen and surface 2 as VE2-2M. The effect is the same as making them the same surface, I think.

-Greg

···

From: Galen Burrell <[email protected]>
Date: January 15, 2009 7:54:09 PM PST

Thanks Jack.

All seems clear and I've got some results that I'm pretty happy with. One thing I'm still a little unclear about is how you would apply a low-e film to a non-clear glass (i.e. tinted green, blue, etc) if the script requires one surface to be "clear" for each pane.

For example:
VE2-2M (http://www.viracon.com/makeupsheet.php?id=3840) has the VE-2M coating applied to Evergreen glass on the #2 surface. As it is now, I would input:
Surface 1 - Clear
Surface 2 - VE2-2M on evergreen (from optics 5 data)
Surface 3 - Clear
Surface 4 - Clear

Is there an easy hack to make Surface 1 evergreen instead of clear?

Thanks,
Galen

7

Hi Galen,

It has been a while, but I think that the color/tint comes from the rgb values for the coated glass. So the color would be conveyed by "VE2-2M on Evergreen" monolithic export from Optics 5. So this should account for the color of the coated glass. However, I think that I have to defer to Greg on this for confirmation.

Note for Viracon:

    <type><color>-<transmittance>

    so
    type = VE = Low-E
    color = 2 = Green
    transmittance = 2M = ~70%

    if you want "Evergreen" you need to look at color 8.... so VE8-2M

One way to sanity check is to run a single layer output from glaze and specify s1 as clear and s2 as VE2-2M on evergreen. The last three lines of the material description should give you:

    Rext_r Rext_g Rext_b
    Rint_r Rint_g Rint_b
    Tr Tg Tb

The rgb components should be the same as those exported from Optics5 (even though you are specifying "clear" for one layer).

Hope this helps.

-Jack

Galen Burrell wrote:

···

Thanks Jack.

All seems clear and I've got some results that I'm pretty happy with. One thing I'm still a little unclear about is how you would apply a low-e film to a non-clear glass (i.e. tinted green, blue, etc) if the script requires one surface to be "clear" for each pane.

For example:
VE2-2M (http://www.viracon.com/makeupsheet.php?id=3840) has the VE-2M coating applied to Evergreen glass on the #2 surface. As it is now, I would input:
Surface 1 - Clear
Surface 2 - VE2-2M on evergreen (from optics 5 data)
Surface 3 - Clear
Surface 4 - Clear

Is there an easy hack to make Surface 1 evergreen instead of clear?

Thanks,
Galen

-----Original Message-----
Date: Thu, 15 Jan 2009 21:07:06 -0500
From: Jack de Valpine <[email protected]>
Subject: [Radiance-general] Re: Radiance-general Digest, Vol 59, Issue
        8
To: Galen Burrell <[email protected]>
Cc: Radiance general discussion <[email protected]>
Message-ID: <[email protected]>
Content-Type: text/plain; charset="iso-8859-1"

Hi Galen,

No problem.

Regarding the header

    * Rc<r,g,b> - these are the rgb components for the reflectance on
      the "coated" side of the monolithic glass
    * Rg<r,g,b> - these are the rgb components for the reflectance on
      the non-coated side of the monolithic glass
    * T<r,g,b> - these are the rgb components for the transmittance for
      the monolithic glass
    * Part - this is a boolean value
          o 0 means this is a regular glass
          o 1 means this is a frit - if this is the case the "coated"
            side reflectance needs to represent the reflectance of the
            frit at 100% coverage which should be possible to extract
            from the glazing manufacturer

The RGB values are extracted from the file generated for a "radiance"
export from Optics. This should be for a single pane monolithic glass
description (not a multi-layer make-up). The file will have three
Radiance material definitions, the two BRDTfunc definitions are the ones
to pay attention to. The makeup is something like the following:

    void BRTDfunc some.glass.front
    10

        Rfr Rfg Rfb
        Tr Tg Tb
        0 0 0
        .

    9 0 0 0 0 0 0 0 0 0

    void BRTDfunc some.glass.back
    10

        Rbr Rbg Rbb
        Tr Tg Tb
        0 0 0
        .

    9 0 0 0 0 0 0 0 0 0

So you need to take the two rgb triples for reflectance and one for
transmittance. So an entry in the data file could look like this (you
will need to figure which represents the coated side if needed):

    Surface Tr Tg Tb Rcr Rcg Rcb Rgr Rgg Rgb Part
    <name> Tr Tg Tb Rfr Rfg Rfb Rbr Rbg Rbb 0

A file for a double pane insulating unit might look like the following:

    Surface Tr Tg Tb Rcr Rcg Rcb Rgr Rgg Rgb Part
    green-lowe Tr Tg Tb Rfr Rfg Rfb Rbr Rbg Rbb 0

The glaze script contains a default "clear" glass. I seem to recall that
this is required for the script to output a material description (eg it
must have a "clear" glass description"). If this is not suitable the
script requires some modification, which I think I can offer some
guidance on so that you can include your own "clear" glass description.

I would suggest before you do anything else you should test out the
script as is with the built in values. This will let you see how it
works. You will assign different materials to different layers to build
up the composite unit.

Try it out and let me know if you have more questions.

-Jack

____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

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

--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

8

Yes, well, unfortunately Greg doesn't know what he's talking about, because Galen was right when he said that the glaze script demands one surface be clear glass. Your suggestion is better, and should work, since the glaze script only uses the transmittance and reflectance values for it's models.

Creating a model that properly accounts for the deepening hue experienced at off-angle transmission through colored glazing would require hand-modeling, since the glaze script only considers a surface coating. I don't know the construction of the Evergreen glass, so I'm not sure what's appropriate.

-Greg

···

From: Jack de Valpine <[email protected]>
Date: January 16, 2009 6:51:56 AM PST

Hi Galen,

It has been a while, but I think that the color/tint comes from the rgb values for the coated glass. So the color would be conveyed by "VE2-2M on Evergreen" monolithic export from Optics 5. So this should account for the color of the coated glass. However, I think that I have to defer to Greg on this for confirmation.

Note for Viracon:
<type><color>-<transmittance>

so
type = VE = Low-E
color = 2 = Green
transmittance = 2M = ~70%

if you want "Evergreen" you need to look at color 8.... so VE8-2M

One way to sanity check is to run a single layer output from glaze and specify s1 as clear and s2 as VE2-2M on evergreen. The last three lines of the material description should give you:
Rext_r Rext_g Rext_b
Rint_r Rint_g Rint_b
Tr Tg Tb
The rgb components should be the same as those exported from Optics5 (even though you are specifying "clear" for one layer).

Hope this helps.

-Jack