Week 4: Optimizing the location of the Pavilion
I wanted to research if there was a way to optimize the position of the pavilion. As discussed by the group, we are using the hololens in majority of the pavilion (stage 1–4). Nissie pointed out that the hololens doesn’t work in direct sunlight, so the pavilion position would need to be adjusted to find the most optimal location to address this requirement
As shown last week, I conducted a solar analysis in grasshopper utilizing ladybug. I feel like it was a really successful and now will become really useful.
The images show the sun during the month of may (which is when we thought the pavilion would be on show at UNSW). The blue demonstrates that the area of the site reaches between 1–0 hours of sunlight per day. The red and oranges shows between 2–7 hours of the day, sun reaches that area (light orange being the most amount of hours). These images demonstrate that the requirements are possible to be meet. The hololens/start could be towards the eastern side of the site and the end in the center.
OPTIMAL SCRIPT
Now I have established a rough positioning for the site, I wanted to create a script that would determine the perfect orientation and positioning. The script would have to consider the amount of daylight hours,trees and the size of the site to position the pavilion in the optimal location.
I decided to look into:
- Optimo for Dynamo
- Galapagos for Grasshopper
I started to research Optimo for dynamo and watched a few tutorial on it. http://dynamobim.org/optimo/ This article explains what optimo does and links you to a few tutorials.
I ran into some issues with the tutorials but these were easily fixed once I spoke to Alexis. I believe optimo is an extremely useful tool, but whilst watching the tutorials, I realized it would actually be easier to create an optimal script in grasshopper, as the solar analysis has already been completed in there.
I watched a few tutorials on Galapagos and it seemed to be much easier than optimo (maybe because I have more experience with grasshopper).
I wanted to create a script that would move and rotate the pavilion to receive the smallest amount of daylight hour possible (only because there is a tiny area with no day lighting hours). I figured this would be the easiest way to set up the script.
For first test, the rotation was set far to high, and it was rotating it off the site. I reduced the rotation and ran it again.
After 20 minutes, Galapagos optimized numerous positions. The script needs some work which I will continue to develop next week. I was able to view a few solutions before my computer crashed. This might be because each solution Galapagos makes, the solar analysis has to run again. As is, the analysis takes between 5 minutes — 1 hours (depends on how many polygons in the iteration).
This was one of the solutions optimized, and was the best one after 20 minutes. As said earlier, this script still needs some work as every parameter wasn’t working, but this is the general idea.
As shown in the image, half of the pavilion receives 0–1 hours of light per day. I believe the Hololens team could work with this and create their experience around the above image if this pavilion was selected.
I ran the simulation again on a different pavilion and my computer crashed again (maybe I need a stronger RAM/Graphics Card)
Again, this the best solution with a half working script.
If I can fix the script to run at its full potential and also get this running without crashing, I believe it will be an extremely useful tool as it will influence the design of the pavilion and the VR/AR experience.
