"Kalwall" skylights in daylight simulation

[Brian Cassil]

We are trying to do a daylight analysis that includes some frosted glass or "kalwall" style skylights which diffuse the light into the space. We can get the VLT (Visible Light Transmittance) values easy enough. However, is there some way to accurately (or semi-accurately) account for the rays being dispersed and spread through the frosted glass? Perhaps using a SSS material? If so would there be any indication in the specs of the actual glazing that would indicate what setting I would put in my SSS material to simulate that? Or maybe it could be as simple as setting up the refraction glossiness to a certain value. But how would I determine that other than it "looks right".

[AJLynn]

I would think that an SSS material wouldn't be the most accurate thing, and that you should calculate it based on luminous flux hitting the outer surface (you remember, from Building Tech II, where you find the solar angle and do the trigonometry that projects the rectangle of the incident surface into the equivalent area perpendicular to the solar angle) at the time being considered, multiplied by the transmittance factor, to make a diffuse area light source representing the inner surface.

[Brian Cassil]

I would think that an SSS material wouldn't be the most accurate thing, and that you should calculate it based on luminous flux hitting the outer surface (you remember, from Building Tech II, where you find the solar angle and do the trigonometry that projects the rectangle of the incident surface into the equivalent area perpendicular to the solar angle) at the time being considered, multiplied by the transmittance factor, to make a diffuse area light source representing the inner surface.

 

Good idea! However rather than using the luminous flux calculation, I should be able to set up a light meter on the roof of my model to get a reading of the the intensity that way. Right? Then multiply that value by the transmittance factor and use that in my area light as you suggest. The only problem I see is that skylights usually arent flat, especially on top. Trying to calucalate the luminous flux for a dome shaped surface is something that humans just aren't supposed to do... at least not this one. But maybe I'll have to just say that this will be close enough. I'd still rather do it with a shader if that were possible.

[AJLynn]

Heh, well, if you aren't up to the math

[TheAllusionisst]

Did you try contacting Kalwall and see if they have any formulas, or charts to get you in the general "ballpark" range? Suprising what kind of info you can get from vendors, especially if they think you are an architect considering specifying their product for a project.

 

You then could place IES type lighting on inside surface of Kalwall units with the values the chart gives you to double check against your solution and reverse engineer as needed or just use the light setup.

 

Just a thought, so that if someone complains later about accuracy, you can say you used information directly from the vendor?

[Warren Mann]

You might want to tale a look at this web site: daylightmodeling

[pfbreton]

Hi!

 

here is a blog post I wrote that might help solving this problem for this task:

 

http://www.pfbreton.com/2010/02/lighting-analysis-and-translucent-glazing/

[AJLynn]

That's fantastic.

[Brian Cassil]

Thanks for all your help on this Pierre. I think I have this working accurately now although I wish there was a way to confirm that with a real guru. BTW, the idea of simulating this with a light (as discussed earlier in this thread) proved to be wildly inaccurate.

[TheAllusionisst]

.......... BTW, the idea of simulating this with a light (as discussed earlier in this thread) proved to be wildly inaccurate.

 

See, what do I know? Wait, don't answer that!