Big difference between vertical blinds with BSDFs and static geometry

grasshopper honeybee[+]
1

Hi,

I try to simulate vertical blinds with the three phase method. The simulation case is an office room with one big south oriented window. The shading system are interior vertical blinds.
I generated BSDFs for four angles: 0, 45, 90, 135 degrees. The image below presents an overview of the hourly illuminance values for an office room on a clear, sunny day from 8 in the morning till 6 in the afternoon.

image
image.png1601×641 211 KB

BSDFs:
The image above shows the results, but I would expect other results for the angles 45 and 135.

Because of the strange results for these two angles, I also gathered illuminance values by modelling the blinds for these angles as static geometries. So I runned the model two times to collect hourly illuminance values for each of the two angles. The pattern in these images seems more realistic for me.

image
image.png1540×265 95.5 KB

My question is: Does anyone recognizes this problem of does could anyone tell me where I go wrong in the three-phase method??

Any advice is very welcome!

Kind regards,
Kim Bodde

Verticalblinds_three_phase.gh (1000.2 KB)

Bsdfs.rar (622.4 KB)

2
  1. If you generated your BSDF (xml files) by using LBL Window (Complexshade: venetian type) keep in mind that it uses radiosity approach and does not represent high reflective slats material and complex geometries.
  2. Due to the shortcoming of 3phase method, which represents sun into big sky patches, 3phase has some very high picks.

image

3

Just keep in mind, for using 5phase method, you need a BSDF with proxy geometry or a variable resolution BSDF material.(!!!Advanced users only)

4

Hi Kim,

First of all, I would suggest before running an annual simulation try to understand the behavior of your BSDFs. It has 2 ways. One what @sarith did in here for you. The other is using BSDF viewer.
So, for your vertical slats 45° the BSDF makes sens to me but the way you apply it in your model might be influential as you got some comments in here.

45_Bot
45_Bot.PNG1176×694 216 KB
45_Right
45_Right.PNG1171×694 202 KB
45_Left
45_Left.PNG1176×694 218 KB
45_Top
45_Top.PNG1177×696 217 KB

Again I suggest you to run an point in time image based (rendering) by using the same model setup only for 2 hours (one winter day, morning time and one afternoon) and see the behavior of your shading.
you can use the HB+ example files.

Best,
Fazel