Translucent material modeling with Radiance

[T_Charles]

Dear Radiance users and experts,

As an engineer in a the building industry, I use Radiance to model different facades options in terms of daylighting. I also use Daysim for annual calculations to estimate annual illuminance exposure and other metrics (DA, etc.).

I'm currently working on a translucent glazing option and below is my understanding of the several possibilities I have to model it with Radiance.
Please feel free to correct me.

The easiest way is to use trans but I understand that this does not allow one to model an angle-dependant transmittance and reflectance.
To overcome this limitation, I could use transfunc or transdata but the angle-dependant transmittance/reflectance only applies to the specular components and not the diffuse ones.
I understand that BRTDfunc suffers from the same limitations even if it allows the user specifying more distribution functions (mirror, reflectance, transmittance lobes).

From my understanding then only a BSDF material allows a complete material description. This can be achieve with Window and (or?) the genBSDF routine.

Moreover, on the use of transdata/transfunc, I understand from the Radiance documentation that the only modified value is the transmitted specular component so the last one called tspec. Is that correct ?
I've carried out some calculation with a trans and transdata material following the process described in Reinhart's paper (Development and validation of a Radiance model for a translucent panel).

trans:
void trans south_panel
0
0
7 0.40446 0.40446 0.40446 0.08 0 0.435635 0

transdata
void transdata south_panel
4 noop refl.dat rang.cal rang
0
6 0.40446 0.40446 0.40446 0.08 0.435635 1

Reading the paper, I though that the values contained into the refl.dat file where multipliers of the transmitted specular fraction (tspec) set to 1 (last parameter of the transdata). Then I ran a simulation with a sunny sky with sun (simple square box with a window).
In the case of the trans material, with no transmitted specular component, I was expecting to see a perfect diffuser which was the case.
In the transdata case, I expected to see a patch of light on the floor as the transmitted specular fraction was not zero (from my understanding at least). Instead, the transdata material seemed to behave as a trans material with no transmitted specular component.
Am I missing something ?
These results really confuse me.

Thank you all for your support.

Regards,

Thibault

42-43 Gloucester Crescent
London
NW1 7PE
T +44 (0) 20 7267 5161
F +44 (0) 20 7482 0329
maxfordham.com

Max Fordham LLP is a Limited Liability Partnership. Registered in England and Wales Number OC300026
Registered office 42-43 Gloucester Crescent, London, NW1 7PE

[Greg_Ward]

Hi Thibault,

It seems like no one responded to this. First, you might want to read through Peter Apian-Bennewitz's recent review of the various Radiance types, which he has posted at:

  http://www.pab.eu/docs/Review_of_simulating_four_classes_of_window_materials_for_daylighting_with_non-standard_BSDF_using_simulation_program_Radiance.pdf

Peter does an excellent job of explaining the different types, what they do, and when they are useful -- something that has been missing from the documentation and literature for decades, now.

What I think you will learn as it applies to your case is that without an interreflection calculation, the transdata and transfunc types will not give you a solar patch on the interior of your test space no matter what your material parameters. The BSTDfunc type would, however.

Cheers,
-Greg

···

From: [email protected]
Date: November 14, 2013 11:01:49 AM PST

Dear Radiance users and experts,

As an engineer in a the building industry, I use Radiance to model different facades options in terms of daylighting. I also use Daysim for annual calculations to estimate annual illuminance exposure and other metrics (DA, etc.).

I'm currently working on a translucent glazing option and below is my understanding of the several possibilities I have to model it with Radiance.
Please feel free to correct me.

The easiest way is to use trans but I understand that this does not allow one to model an angle-dependant transmittance and reflectance.
To overcome this limitation, I could use transfunc or transdata but the angle-dependant transmittance/reflectance only applies to the specular components and not the diffuse ones.
I understand that BRTDfunc suffers from the same limitations even if it allows the user specifying more distribution functions (mirror, reflectance, transmittance lobes).
From my understanding then only a BSDF material allows a complete material description. This can be achieve with Window and (or?) the genBSDF routine.

Moreover, on the use of transdata/transfunc, I understand from the Radiance documentation that the only modified value is the transmitted specular component so the last one called tspec. Is that correct ?
I've carried out some calculation with a trans and transdata material following the process described in Reinhart's paper (Development and validation of a Radiance model for a translucent panel).

trans:
void trans south_panel
0
0
7 0.40446 0.40446 0.40446 0.08 0 0.435635 0

transdata
void transdata south_panel
4 noop refl.dat rang.cal rang
0
6 0.40446 0.40446 0.40446 0.08 0.435635 1

Reading the paper, I though that the values contained into the refl.dat file where multipliers of the transmitted specular fraction (tspec) set to 1 (last parameter of the transdata). Then I ran a simulation with a sunny sky with sun (simple square box with a window).
In the case of the trans material, with no transmitted specular component, I was expecting to see a perfect diffuser which was the case.
In the transdata case, I expected to see a patch of light on the floor as the transmitted specular fraction was not zero (from my understanding at least). Instead, the transdata material seemed to behave as a trans material with no transmitted specular component.
Am I missing something ?
These results really confuse me.

Thank you all for your support.

Regards,

Thibault

42-43 Gloucester Crescent
London
NW1 7PE
T +44 (0) 20 7267 5161
F +44 (0) 20 7482 0329
maxfordham.com

Max Fordham LLP is a Limited Liability Partnership. Registered in England and Wales Number OC300026
Registered office 42-43 Gloucester Crescent, London, NW1 7PE

http://www.radiance-online.org/mailman/listinfo/radiance-general