Hello again.

I have been testing the switchable electrochromic dynamic shading solution and get some odd results. I have created four separate layers for each state and each are offset by 5 mm from the next state so they don't cause problems with DIVA by being on top of each other. The layers are defined as 80%, 50%, 20% and 2% transmission with 80% as the base state and is the inner layer while 2% is the outer layer offset by 15 mm relative to the 80% layer. I have made sure that the layers are turned on and unlocked.

I use the automated thermal control strategy and use 40 nodes in front of each person about an area of 1 m wide and 0.4 m long with 0.1 m between each node. The state intervals are set to:

- Base state (80%)

- State 1 (50%): off 1500 lx on 1600 lx

- State 2 (20%): off 1700 lx on 1800 lx

- State 3 (02%): off 1900 lx on 2000 lx

When I run the annual glare analysis and the climate based analysis it looks like no sunlight are entering the room but when I run a visualization on each state I can see that they are working as transluscent and that each state makes the room darker. Based on this I am sure that my material file is working and my geometry seems OK, but I can't seem to figure out if DIVA run some special kind of program when doing electrochromic shading. The visualization seems a bit odd in the fact that the room gets darker but the direct sunlight on the table doesn't seem to change (see uploaded pictures).

The only odd thing about the analysis is that when running gen_dc it gives me the warning:

oconv: warning - (C:\DIVA\Daylight\blank.rad): empty file

I was wondering if I should replace that particular file with something to set as the base state or if that would mess up the program? Maybe the EC layers should be on top of each other? The model and material file are uploaded in an archive .rar file.

Thanks a lot for all the help that I have received lately and I hope you can help one more time.


Tags: base, blank.rad, dynamic, electrochromic, electrochromics, empty, error, file, oconv, shading, More…state, switchable

Views: 554


Reply to This

Replies to This Discussion

Hello Mads,

There are two ways of setting up detailed dynamic shading systems in DIVA: Mechanical and Switchable. It seems like you are using the mechanical method, but let me describe the two and their uses -- maybe this will help.


Define layers containing geometry (base state, state 1, state 2, state ..., state N). In this case, if you have a window and a mechnical blind system, you would have a blank base state (blank.rad) that gets included with all of your other model geometry for the 'open' or 'unshaded' shading state. State 1 geometry (blinds down, but open) will be a second simulation, and State N's geometry will be the most closed (blinds down, fully closed). If you include something on the base state geometry, it will be replaced when State 1 is triggered. A good example of this is a rotatable fixed louver system. Even in the open or base state, there is something producing shading, but it can be modified to be more 'closed' via rotation.


Switchable is the type of system you want to use for electrochromic glazing. In this case you choose materials that get changed with each shading state. Note that in this mode the state layers are called: base glazing material, glazing state 1 material, etc. In this case, the base material is your clear electrochromic state. State 1 is your 50% material, and state 3 is your 2% material. I don't know why the lux level triggers change for these states, but you can do that if you wish. The ranges seem very narrow -- typically you would set the lower bound at something like 250 lx (where it gets too dark) and the upper bound around 3000 lx (where it is too bright). But does it make sense to make the window more transparent when less than 1900lx is achieved inside?

I strongly recommend to check out our video on the topic, which should explain some more.

DR 1: New Shading Module (Video) (A Jakubiec & J Niemasz, Solemma)


Thank you very much.

I see the problem now, and it makes sense. I have tried to use a mechanical method to model a switchable method which of cause creates problems. I apologize, I haven't paid enough attention to the menu saying "layer" when in mechanial and "material" when in switchable.

Regarding the ranges I have just been testing EC with the same ranges I used for horizontal lamellas until I got it working, but thank you very much for your expert knowledge and the presentation you uploaded. It will be a big help.



Hi no worries -- you could use the mechanical method to model switchable glazing, you would just have to make separate layers with the various materials applied and place them all on top of eachother. The switchable method just simplifies things a bit.



OK, that is actually what I have been trying to do, I just thought that it would cause an error if the dynamic shading layers were on top of each other because that is what I have experienced with dynamic lamellas I accidentally put on top of each other. Now I see should just have chosen mechanical instead of switchable then. I will rather use the simpler method now.

Can you clarify for me why the climate based analysis is running exceptionally fast when using the DGP schedule created compared with not using the DGP schedule? Just to understand the program a bit better?


Hi Mads,

Apologies for the late reply.

Can you clarify for me why the climate based analysis is running exceptionally fast when using the DGP schedule created compared with not using the DGP schedule? Just to understand the program a bit better?

There should not be a difference. Perhaps in one case you re-used the daylight coefficient file and in another you ran the full raytrace calculation?



Okay, maybe that is the case. I might just have mixed some impressions together. Thank you.

Hello again.

I am wondering how the blue shading control table is created and where I may find the numbers. I found the script for creating the DGP table and thought a similar one was needed for the shading control table, but I can't seem to find it neither the numbers.

Could you explain this for me, please?

Kind regards


Reply to Discussion


© 2020   Created by jeff niemasz.   Powered by

Badges  |  Report an Issue  |  Terms of Service