I need to compare the influence of several shading and lighting control systems on the visual comfort in some office layouts for my thesis. Currently I’m running test simulations with split blinds.
Recently I started with the reference office of the old Diva for Rhino tutorial series, named view “row1_west”. My goal was to apply a high reflection rate (70%) to the upper part of split blinds in order to get daylight deep into the room while choosing a low reflection rate to the lower part of the blinds to avoid glare by the blinds themselves (20%).
dynamic shading – setting manual, 4 ambient bounces – with shading group 1 (upper blinds open / 30°) and shading group 2 (lower blinds open / 30° / 70°).
Result: The simulation finishes much too quickly (about 35 minutes; normally it takes more than 70 minutes for EACH dgp file for this office space). This is the visual output for the base case (compared to a regular dgp output without any shading):
During the run the command line shows these lines:
Further attempts with different weather files and different shading control settings (eg shading group 1 automated glare with external sensor…) didn’t help. Always the same…
In a previous forum post Jon helped me find a bad binary file which prevented the gh glare component from working correctly ( http://diva4rhino.com/forum/topics/glare-component-issue ) . There are some exe files existing in c:\diva\daysimbinaries as well as in c:\diva\radiance\bin_64 . Actually they are not really duplicates, just the names are. File sizes differ and the files in c:\diva\daysimbinaries were all created in 2013 whereas c:\diva\radiance\bin_64 files were made in 2015. (Find the direct comparison attached as a screenshot). Now that replacing the 2013 evalglare.exe file with the 2015 version helped me solve my gh glare issue I ask myself if there could be more of these exe files causing problems.
Which file/-s is/are involved in the annual glare run? Maybe I could try to replace the old files by the newer ones either?
The reason why I suspect the annual glare run is that in some further test runs I encountered strange annual dgp values which could not be confirmed by point-in-time simulations. But I will put this into a second thread, too much stuff for a single post…
P.S. I've already changed punctuation settings in Windows. Must be something different...
I've been away from the forum for much too long, and I apologize that no one answered your question. Being that so much time has passed, this may be no longer useful to you.
It seems that for some reason the octree never builds for the image-rendering portion of your calculation. I built a kind of test version of your setup, but I didn't run into the same problems. It might have been something that is file or material-specific to your setup. In any case, I would have liked to have the Rhino and material files in hand to test this.