Irradiance map

radiance-general
1

Hi,

I am trying to calculate an irradiance map on the ground level in front of a
concaved building facade for an instant time at a specific location. Here is
the problems that I have:

1. If I use the glass type material for the building facade with -i
option in rpict, there will be no facade in the final result. I have checked
the manual page that I know it is because of the Boolean calculation will
ignore the transparent surface in the final result. But I do need the facade
to be seen in the final result, how can I solve it.

2. I found in the result that if I use the glass type material, no
matter how low I set the transmissivity values, there will always no
'hotspot'/convergent irradiance on the ground level. It seems that the
program just ignore the reflection process on the glass surface.

3. If I change the facade material to mirror type, I can find an
obvious 'hotspot' on the ground level as well as the facade in the final
result. So I just wonder is it proper to modify the facade material into a
mirror type if I am only interested in the exterior irradiance map of a
building. Besides, if mirror type material can be used in my situation, how
can I set the RGB reflectance values for mirror? For example, if the facade
glass has a 40% transmittance, is it equivalent to set RGB values into 0.6
0.6 0.6 for a mirror material?

Many Thanks,

Jiajie

2

Hi Jiajie,

I put some responses inline...

From: "Jiajie Zhu" <[email protected]>
Subject: [Radiance-general] Irradiance map
Date: August 10, 2015 9:31:13 AM PDT

Hi,

I am trying to calculate an irradiance map on the ground level in front of a concaved building facade for an instant time at a specific location. Here is the problems that I have:

1. If I use the glass type material for the building facade with –i option in rpict, there will be no facade in the final result. I have checked the manual page that I know it is because of the Boolean calculation will ignore the transparent surface in the final result. But I do need the facade to be seen in the final result, how can I solve it.

The -i option is a hack to show the irradiance on surfaces, but glass surfaces appear as transparent to facilitate viewing through windows, etc. The results will still be correct within the limits of the calculation.

2. I found in the result that if I use the glass type material, no matter how low I set the transmissivity values, there will always no ‘hotspot’/convergent irradiance on the ground level. It seems that the program just ignore the reflection process on the glass surface.

Radiance is a conventional light-backwards ray-tracer, so needs special facilities to find caustics. These are the "mirror" type you mention below, and photon-mapping, which is incorporated in the latest HEAD.

3. If I change the facade material to mirror type, I can find an obvious ‘hotspot’ on the ground level as well as the facade in the final result. So I just wonder is it proper to modify the facade material into a mirror type if I am only interested in the exterior irradiance map of a building. Besides, if mirror type material can be used in my situation, how can I set the RGB reflectance values for mirror? For example, if the facade glass has a 40% transmittance, is it equivalent to set RGB values into 0.6 0.6 0.6 for a mirror material?

You need to know the reflectance of your glass, not the transmittance, in order to use the mirror type appropriately. Then, you can give the original glass as the "alternate" type. Most uncoated glass surfaces have a specular reflectance of between 4 and 8% depending on their transparency.

···

Many Thanks,

Jiajie

3

There is a cool Greg Trick you can use to render irradiance on the glass material, which I summarize here:

http://www.rumblestrip.org/2003/03/19/rendering-falsecolor-on-glass-with-vwrays/

As for your quest to find the hotspot, if mirror material gave you a result in a reasonable timeframe, I'd say that may be the most expedient way to get what you're looking for, with the caveat that simply using the inverse of the glazing's transmittance for the reflectance is not correct. Most glazing manufacturers will provide the actual reflectance, which is what you should use.

If you want to use the glass material you will need to adjust some of your rendering parameters to better account for the specular reflections. Can you share what you used, and perhaps a sample rendering of the façade?

···

On 8/10/15, 10:31 AM, "Jiajie Zhu" <[email protected]<mailto:[email protected]>> wrote:

Hi,
I am trying to calculate an irradiance map on the ground level in front of a concaved building facade for an instant time at a specific location. Here is the problems that I have:

1. If I use the glass type material for the building facade with -i option in rpict, there will be no facade in the final result. I have checked the manual page that I know it is because of the Boolean calculation will ignore the transparent surface in the final result. But I do need the facade to be seen in the final result, how can I solve it.

2. I found in the result that if I use the glass type material, no matter how low I set the transmissivity values, there will always no 'hotspot'/convergent irradiance on the ground level. It seems that the program just ignore the reflection process on the glass surface.

3. If I change the facade material to mirror type, I can find an obvious 'hotspot' on the ground level as well as the facade in the final result. So I just wonder is it proper to modify the facade material into a mirror type if I am only interested in the exterior irradiance map of a building. Besides, if mirror type material can be used in my situation, how can I set the RGB reflectance values for mirror? For example, if the facade glass has a 40% transmittance, is it equivalent to set RGB values into 0.6 0.6 0.6 for a mirror material?
Many Thanks,
Jiajie

4

Could the lack of hotspot reflection also be related to the -st specular threshold parameter? By default this is -st 0.15 if not otherwise defined, so a glass definition with specular reflectance less than 0.15 could be ignored. I suspect -st 0.01 is more desirable for most typical simulations unless there are many parallel specular planes where it's desirable to leave out more specular component calculations.

···

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

5

ambient rays (this is what you do when using glass for your facade and plastic for the surrounding) will not traced along any specular reflection towards a light source (here sun), whatever (uhh.. not sure this word is allowed to be used any more...) specular settings you are using for the glass material. E.g. also rtrace -I facing a glass will get 0 contribution from sun reflection in the glass.
So either you use mirror material, photon-mapping or a modified version of Greg's cool tip:
You calculate the intersection points and render there a 180 degree-fish-eye image instead and calculate from the image the illuminance or irradiance at that intersection point and transfer this back to an image. The main difference is that for an image view rays are sent out and the specular reflection towards a light source is taken into account. But be aware that secondary effects are not covered with this special trick (e.g. if your reflection hits a white surface (there it is covered) and then lights an adjacent surface (this is not covered).
When using mirror you have to take care that you miss the sky contribution from the mirror reflection, when using it normal. A mixing between mirror and glass might help, also regarding the "appearance".

good luck!
Jan

···

Am 8/10/15 um 7:15 PM schrieb Guglielmetti, Robert:

There is a cool Greg Trick you can use to render irradiance on the glass material, which I summarize here:

http://www.rumblestrip.org/2003/03/19/rendering-falsecolor-on-glass-with-vwrays/

As for your quest to find the hotspot, if mirror material gave you a result in a reasonable timeframe, I'd say that may be the most expedient way to get what you're looking for, with the caveat that simply using the inverse of the glazing's transmittance for the reflectance is not correct. Most glazing manufacturers will provide the actual reflectance, which is what you should use.

If you want to use the glass material you will need to adjust some of your rendering parameters to better account for the specular reflections. Can you share what you used, and perhaps a sample rendering of the fa�ade?

On 8/10/15, 10:31 AM, "Jiajie Zhu" <[email protected]<mailto:[email protected]>> wrote:

Hi,
I am trying to calculate an irradiance map on the ground level in front of a concaved building facade for an instant time at a specific location. Here is the problems that I have:

1. If I use the glass type material for the building facade with -i option in rpict, there will be no facade in the final result. I have checked the manual page that I know it is because of the Boolean calculation will ignore the transparent surface in the final result. But I do need the facade to be seen in the final result, how can I solve it.

2. I found in the result that if I use the glass type material, no matter how low I set the transmissivity values, there will always no 'hotspot'/convergent irradiance on the ground level. It seems that the program just ignore the reflection process on the glass surface.

3. If I change the facade material to mirror type, I can find an obvious 'hotspot' on the ground level as well as the facade in the final result. So I just wonder is it proper to modify the facade material into a mirror type if I am only interested in the exterior irradiance map of a building. Besides, if mirror type material can be used in my situation, how can I set the RGB reflectance values for mirror? For example, if the facade glass has a 40% transmittance, is it equivalent to set RGB values into 0.6 0.6 0.6 for a mirror material?
Many Thanks,
Jiajie

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

--
Dr.-Ing. Jan Wienold
Ecole Polytechnique F�d�rale de Lausanne (EPFL)
EPFL ENAC IA LIPID

LE 1 111 (Office)
Phone +41 21 69 30849