1. Design Concept and computational Design System
Obsolete typology
Interpretation
Deconstruction
Understanding the current prisons as punitive institutions that fail to rehabilitate but imprison inmates without offering social inclusion solutions increases the risk of recidivism.
Deconstructing Alcatraz proposes an alternative to this typological lack through its own reconstruction. In a VR environment the proposal consists of a series of interventions around the very idea of reconstruction in all its aspects.
1.- Island Deconstruction
Houdini
2.- Rearrangement of Pieces
Kangaroo
3.- New connections
Heteroptera
+
Kangaroo
+
Noise & Swarm
4.- Creation of Pavilions
Mathematical Surfaces
+
Tesselation
5.- VR Experience
Unreal
OTHER TOOLS
Pufferfish
Parakeet
Human
2. Early explorations and lessons from failures (reflect on these and connect them to your next steps)
For the introspective zone, we started with some explorations in Houdini creating a movable and responsive labyrinth with stairs and cell like panels, afterwards this was reduced to the panels due to the large amount of geometry.
The connections between islands was firstly created as a multipipe of a mesh, being further developed into a swarm aggregation that allowed us to make more natural connections (tree branch like) between the islands.
Lastly for the explorations for the pavilions we tested several patterns for the tailing of the roof, finishing up with the simplest aiming a less noisy solution.
3. Computational Design Methodology (including pseudo code and geometry diagrams) 1/4
1.- Prison Reinterpretation
The beginning of our proposal occurs at the core of the project, where we propose a labyrinthine concept. Solved with Houdini.
2. Split and Select the amount of pieces
Out of a total of 21 fragments we selected one group for each aggregation (corresponding to each convict one of them). The central building being the first of the possibilities.
-N of pieces: 21
-Central building index: 0
-Selection: 1-21
3. Split and Select the amount of pieces
Use Populate Geometry to relocate the parts. By means of the “seed” parameter we control the desired aggregation.
-Seed: 0-20
4. Circle Packing & Kangaroo
Introduce the pieces inside circles and run a Collider to avoid overlapping solutions.
-Seed: 0-20
3. Computational Design Methodology (including pseudo code and geometry diagrams) 2/4
5.- Connect pieces
Use closest points to connect centers. Intersect with the pieces’ projection.
-N of connections between pieces: 3
6. Create one line bridge
Re-project back to the mesh intersections. Create a line connecting the dots.
7. Expand bridge by distance to line
Filter naked points of the pieces by distance to each end of the bridges.
Join the points with Minimum Spanning Tree.
Create a Ruled Surface with each curve.
-Distance: 0-20
8. Clean geometry & Kangaroo
Convert surface with mesh surface. Extract anchor points from the sides of each bridge. Apply a Kangaroo solver with vertical loads and edge legnts applied to the rest of the vertices.
-Seed: 0-20
3. Computational Design Methodology (including pseudo code and geometry diagrams) 3/4
9.- Clean Geometry
Join meshes and create a bounding surface mesh from where we will take its points for the field creation.
-Count of U&V: 40
10. Create field with Perlin noise
Create field with vectors from a python script.
Adjust Scale and Octaves to match size and appearance.
Use seed for similar variations.
-Points: 1681
-Scale: 75.000
-Octaves: 1
-Seed: 1
11. Starting & Ending points for Swarm
Extract bridges’ sides and divide the curve by the desired amount of starting and ending points.
-Max distance for swarm mesh: 10-40
12. Swarm
Convert surface with mesh surface. Extract anchor points from the sides of each bridge. Apply a Kangaroo solver with vertical loads applied to the rest of the vertices
-Field cohesion: 0.80
-Separation: 6.7
-Alignment: 1.4
-Cohesion: 1.6
-Speed limit: 0.3
3. Computational Design Methodology (including pseudo code and geometry diagrams) 4/4
13.- Create Math Surfaces
Use python to create 3 mathematical surfaces that will be located on the islands. One per island.
14.- Create Pattern to Remap Surfaces
Create pattern to remap surfaces.
15.- Refine geometries
Create 2 layers (surface + 3d pattern). Add architectonic elements.
16.- Place Pavilions on Platforms
These platforms will serve as a transition between the irregularities of the topographic mesh and the delicacy of the roofs that are subtly placed on them.
4. Form Finding Process (step-by-step images using a fixed camera angle)
1. – Houdini iterations.
4. – Pack pieces in circles.
7. – Swarm with Perlin Noise field
2. – Fragmentation.
5. – Kangaroo to avoid overlapping of pieces
8. – Locate Math – Roofs
3. – Fragmentation.
6. – Connect islands with mesh
9. – Locate Platforms
5. Form Iterations (9 variations displaying parameters for each)
Iteration 1
Count of islands: 3
Population seed: 8
Connections: Max width: 24m
Perlin noise: -Scale: 100 -Octaves: 1 -Seed:1
Swarm: -Field cohesion: 0.80 -Separation: 6.7 -Alignment: 1.4 -Cohesion: 1.6 -Speed limit: 1.2
Iteration 2
Count of islands: 7
Population seed: 8
Connections: Max width: 24m
Perlin noise: -Scale:100 -Octaves:1 -Seed:2
Swarm: -Field cohesion: 0.01 -Separation: 6.7 -Alignment: 1.4 -Cohesion: 10 -Speed limit: 10
Iteration 3
Count of islands: 4
Population seed: 6
Connections: Max width: 10m
Perlin noise: -Scale:50 -Octaves:2 -Seed:5
Swarm: -Field cohesion: 1 -Separation: 1 -Alignment: 1.4 -Cohesion: 0.6 -Speed limit: 6
Iteration 4
Count of islands: 15
Population seed: 2
Connections: Max width: 34m
Perlin noise: -Scale: 90 -Octaves: 2 -Seed:2
Swarm: -Field cohesion: 6 -Separation: 4 -Alignment: 1.4 -Cohesion: 0.6 -Speed limit: 10
Iteration 5
Count of islands: 21
Population seed: 9
Connections: Max width: 24m
Perlin noise: -Scale: 80 -Octaves: 2 -Seed: 3
Swarm: -Field cohesion: 0.85 -Separation: 5 -Alignment: 5 -Cohesion: 2 -Speed limit: 2
Iteration 6
Count of islands: 5
Population seed: 6
Connections: Max width: 40m
Perlin noise: -Scale: 30 -Octaves: 1 -Seed:6
Swarm: -Field cohesion: 0.85 -Separation: 5 -Alignment: 1.4 -Cohesion: 2 -Speed limit: 8.0
Iteration 7
Count of islands: 10
Population seed: 3
Connections: Max width: 30m
Perlin noise: -Scale: 30 -Octaves: 1 -Seed:2
Swarm: -Field cohesion: 6 -Separation: 2 -Alignment: 2 -Cohesion: 2 -Speed limit: 8
Iteration 8
Count of islands: 3
Population seed: 8
Connections: Max width: 30m
Perlin noise: -Scale: 80 -Octaves: 2 -Seed: 3
Swarm: -Field cohesion: 0.85 -Separation: 5 -Alignment: 5 -Cohesion: 2 -Speed limit: 2
Iteration 9
Count of islands: 6
Population seed: 4
Connections: Max width: 24m
Perlin noise: -Scale: 90 -Octaves: 2 -Seed: 4
Swarm: -Field cohesion: 0.10 -Separation: 3.0 -Alignment: 3.0 -Cohesion: 0.1 -Speed limit: 2.0
6. One comprehensive aerial view
7. One layout view or plan demonstrating the project’s relationship with the context
8. One comprehensive human-eye view image of the final geometry with human figures
Redemption
Rehabilitation
Salvation
9. Screen recording to showcase the user experience in Unreal Engine using gameplay.
