Week 9 (Part1): Scripting Preliminary Designs
On Thursday the 4th a few of us went to a workshop in the SquareHouse with Alessandra to pump out some final designs. I love using Grasshopper for Rhino and so jumped on the bandwagon straight away. I was in charge of scripting the forms, while other tried to do material testing or script some additional content for the design.
DESIGN ITERATION 1
The first design was created to signify stress through form. The beginning dome would be smaller and more confined and as you walk through the transitional corridor, you enter a large open space with light peering through.
I created the script so that the pavilion was parametric. A few problems that had come across in the earlier weeks was that some scripts were only slightly parametric and so it made it very difficult to change and
The occulus in the large dome was scripted by Nariddh and was scripted parametrically based on the latitude and longitude of the Sun in relation to the centre of that dome.
Some issues that we came across included the height of the domes. If the radius of the dome was 2m and it was created from a perfect sphere, then that would meant that the maximum height at the centre of the dome would be 2m. This became a running issue throughout the rest of the design iterations we made throughout the day.
The pavilion was made based off two physical points in Rhino. These points determined the centre of spheres which were cut in half by the group plane. Using those physical points a line was drawn between them and evaluated based on its length. An angled plane was then put on this evaluation point which then cut in to the domes, creating angled openings for the transition corridor to be placed.
Lines were generated using divide curve to create enough U and V curves to generate a smooth organic transition space between the two domes.
From there, the small entrance dome and transition space was subdivided to create the framework structure that would hold up the pavilion. The larger occulus dome was harder to divide because we wanted it not to be divided from the top of the dome like the small dome, but from the centre of the occulus. The required some playing around, but eventually I was successful in generating the vertical divisions.
To achieve this, I found the centre of the occulus by dividing the curve in 4 sections then drawing a line between two opposing points. Using the middle of that line, a normal plane was placed and arrayed in a circular fashion. Circle by base plane was then used to create large surfaces that would intersect with the dome. And lastly, the division were created by identifying where the circular arrayed surfaces and the dome intersected. Being a parametric script, it was easy to refine the amount of divisions based on what was thought to be more structurally stable and efficient. Nariddh had scripted the horizontal divisions of the large dome, however by the time we switched designs there was no point in combining the scripts.
DESIGN ITERATION 2
Design iteration 2 was proposed when concerns about how AR,VR and HoloLens could fit in design iteration 1. We needed to be able to separate areas where different activities would occur.
I liked this design a lot more in comparison to the first due to its shape and the entrance that draws you in. Although it could still be developed I think it’s a good starting point for a promising form. Similar to the problems in design 1, the domes would be too short if they were perfectly domed and so I scaled them up in the z direction to give some height, although this looks not as aesthetically pleasing.
This design was generate off physical points in rhino which enabled me to create a spline along them and move their geometries in the z direction at different height values to accomplish the flow of the walls.
The domes were created the same way as iteration one, with one sphere based on a physical central point in Rhino and sliced on the ground plane. These domes were then stretched up to allow for extra height. The green geometry above represents spheres that were stretched to create doorways for the domes. These were booleaned from the main dome shelters to create the entrances and exits.
They were then subdivided using the same method as the dome shelters in iteration 1.
The tapered walls were created from physical points in Rhino which were then moved in the z direction at different values. This is not the most efficient way of scripting this, however it gave me more control over which point was at exactly which height and at what location.
The same process was repeated on the remaining two walls. Unfortunately, I was unable to make Friday’s workshop session and so Dom took the reigns and picked up where I left off for Friday. He took this iteration and developed it further.
