genBSDF and genblinds

asked 2017-01-06 22:37:34 -0500

basdav
107 ●6

updated 2017-08-13 16:40:50 -0500

Hi Radiance experts, I'd appreaciate your help regarding the following questions:

I want to assess the performance of a window-system which is formed by daylighting-systems placed in horizontal position (parallel one above the other and certain distance in between). To test, I used genblinds and genBSDF to generate the XML file of such configuration, here the first question regarding this procedure:

1) I wonder if the correct way to indicate the tilt-position of the original daylighting-system (horizontal-slats) its by the specifications in the genblinds command (tilt 0°, 20° etc), and by indicating an horizontal position in the BSDF mat definition (window-BSDF file):

void BSDF system
6 0 system.xml 0 1 0 .
0
0

Note: the final window-assemby's position is vertical (in the final window-BSDF file: 0 0 1 .)

Hoping that this approach is correct so far, then I used the resulting XML file to obtain illuminance and to generate images, while the results seem to be as expected, I have two more questions regarding this procedure:

2) I'm assuming that in this case the calculation of the interreflections occurring between the slats would be different than when using a solid material of uniform reflectance, since I guess, the transmission/reflection-front/back data included in the XML file of the assembly (incident-outgoing rays) would be taken into account for this calculation. Then, I wonder if the results may be at some extent inaccurate according to the following:

a) the following sentence in the 'Three Phase Method Tutorial' (pag. 4, 2.2 Transmission Matrix (BTDF):
'Currently, Radiance only uses the front transmission data, thus front and back reflection and back transmission are ignored by Radiance'.

b) the XML file of the original redirecting-daylighting-system was generated with a goniophotometer, where no back transmission and front and back reflection data was considered, (although the resulting BSDF-XML file contains all such data).

I hope I have explained my questions clearly,

I thank you in advance.

Chantal.

edit retag flag offensive close merge delete

Comments

I am confused by your query. Could you clarify in your question what is horizontal and what is vertical? Maybe you could offer the actual commands you executed and the polygon(s) you used for your BSDF window. Be sure to show the options you are giving to genBSDF, which matter in this case.

GregWard ( 2017-01-07 09:47:29 -0500 )edit

Hi Basdav, I'm also confused by your question. Maybe you could include a sketch to illustrate your system and describe in detail how you're planning to simulate the system.

I can offer a response to your question in 2a: The statement regarding Radiance only using the transmission front data is only true for matrix multiplaction using rfluxmtx and dctimestep. The BSDF material (which didn't exist at the time the tutorial was written) uses front and back, transmission and reflection.

Andyrew ( 2017-01-07 18:53:38 -0500 )edit

Thank you for your answers, I rephrase below in a new 'Answer' space. Regarding the sketch, I'm so afraid that my sketching skills would make things worse, so I really hope that the new explanation can be more useful to describe the issue.

basdav ( 2017-01-08 08:46:56 -0500 )edit

add a comment

2 Answers

Sort by » oldest newest most voted

answered 2017-01-08 23:00:32 -0500

GregWard

1203 ●6 ●5 http://www.anyhere.com/gward

Hi Chantal,

You are right to be concerned about the reflections of the original XML file. If your measurements do not include reflection on both sides as well as transmission, then there will be no interreflection between the slats in your subsequent genBSDF simulation. I think this will lead to an underestimate of the transmitted (as well as reflected) components.

You also need to be very careful with your xform command, which currently puts the genblinds output into the positive-Z domain you should avoid. Your maximum Z value, which you can check with getbbox, should be zero (within some tolerance). Based on my test, your transform should be "xform -rz -90 -rx -90 -t 0 0 -1.414" rather than what you have. You should also double-check that your slats are slanting in the desired direction. You can use objview to look at the Radiance scene file you are giving to genBSDF.

Finally, your BSDF specification should have its "up vector" set to "0 0 1" for any slat angle that is less than 90°, since that will orient the measured distribution into the room, which is the positive-Z direction.

edit flag offensive delete link

Comments

I managed to use bsdf2rad for the visualizations as well, also useful to see the difference in the distribution when using the XML with the missing data, Thank you so much for your help! Chantal

basdav ( 2017-01-09 14:13:30 -0500 )edit

add a comment

answered 2017-01-07 21:23:57 -0500

basdav
107 ●6

updated 2017-08-13 16:42:05 -0500

Thank you so much, and sorry for the confusing explanation. In short, I want to create a system of horizontal-blinds where the material is a redirecting-daylighting-system (BSDF), I rephrase:

I used genblinds to create a shading system (horizontal blinds) but instead of using a uniform reflectance material (void plastic white), I'm using the BSDF of a daylighting-redirecting system (redirecting-system.xml - which was generated by a gonio-photometer). Since the position of the blinds is horizontal, I'm assuming that first I need to place the material (XML file of the redirecting-system) in an horizontal position (in the window-BSDF file: 0 1 0 .), then the tilt of the blinds (0°, 20°, 45°), would be specified by the genblind parameters' (this is just a test, so its pretty much like in the tutorial):

genblinds cfs blinds 2 400 4 4 45 |xform -rz -90 -rx -90 -t 0 0 -.939693  > sys/blindsRS45.rad

the genblinds RAD file (blindsRS45.rad) looks like this:

# xform -rz -90 -rx -90 -t 0 0 -.939693
# genblinds cfs blinds 2 400 4 4 45

void BSDF cfs
6 0 xml/redirecting-system.xml 0 1 0 .
0
0

Then I used such file to generate an XML file using genBSDF:

genBSDF +f +b -c 500 -geom inch -dim 199 201 1.5 2.5 -.939693 0 sys/blindsRS45.rad > sys/blindsRS45.xml

Then I used the resulting XML (blindsRS45.xml) file to create a window-system where the horizontal blinds (using the redirecting-system as material) are placed behind a vertical glass pane (in this case such glass material was included in the same window-BSDF file but I know I can also add it in Window-7), thus in the final RAD file (mywindowsystem.rad) the XML position is vertical: 0 0 1

##mywindowsystem.rad

void BSDF fenestration
6 0 sys/blindsRS45.xml 0 0 1 .
0
0

fenestration polygon window
0
0
12
    10.000000  -0.010000  2.800000

My questions are, first if this approach is correct to simulate horizontal blinds using BSDF material, and then about the calculation of the interreflections occurring between the horizontal-slats, since I'm not using a diffuse uniform material (plastic) but a BSDF material, then I guess such calculation would be done using the transmission-reflection/back-front data of the XML file (incident and outgoing rays). I wonder if my understanding of this process is correct, then if the results might be at some extent inaccurate according to:

a) the following sentence in the 'Three Phase Method Tutorial' (pag. 4, 2.2 Transmission Matrix (BTDF):
'Currently, Radiance only uses the front transmission data, thus front and back reflection and back transmission are ignored by Radiance'.

b) the XML file of the original redirecting-daylighting-system was generated with a goniophotometer, where no back transmission and front and back reflection data was considered. (although the resulting BSDF-XML file contains all such data).

Hoping that my question is better described now, I thank you in advance.

Ch.