Archimedean Mini-House

I designed a structure to act as a proof-of-concept for small-scale, modular, dome-style houses. My design is inspired by numerous architectural works, including Biosphere 2, the Montreal Biosphere, and Vancover's Science Dome.

I am using 5 pieces of software for this project, which include the following:

1. Concepts - A vector design software that I use for 2-dimensional design and prototyping
2. Autodesk Fusion 360 - A CAD software I utilize for its extremely fast prototyping and design workflows
3. Adobe Substance Painter - A material design and application software for texturing the model prior to rendering
4. Maya- An animation and CAD software I utilize for UV editing and compositing
5. Nvidia I-ray - A rendering engine for producing final still renders

For someone following along with this guide, I will lay out the framework for the project, and show you the process I used to complete this design. You can fully overhaul the project with a variety of different variables, from changing sizes, materials, and the overall mesh shape.

My initial idea for this project was a fully modular, scalable, and environmentally friendly home, built using tessellatable components. My hope was that by using tessellatable components, the design and overall complexity of the project could be reduced, which turned out to be only half true.

I started out my research searching for the optimal shape for development, that minimized complexity whilst still having structural integrity and general usability. I landed on a truncated icosahedron (it sounds complicated but I promise it's not), which had the most efficient shapes for tesselation, and high structural stability. For my design, I chose to cut the icosahedron roughly in half, increasing the overall stability and usable space within the structure. This gave me a final shape with approximately 25 sides, with around 85% of these being tessellatable, which was around what I was looking for.

Once a general shape is decided upon, and before any designing or prototyping can begin, you need to decide on the design criteria for the project. The design criteria I decided upon include the following:

1. The design must be completely weather resistant, meaning protected against rain, snow, wind, and flooding
2. The design must be applicable in diverse environments, including areas with high and low temperatures, areas off the grid, and areas with high and low humidities.
3. The design must be environmentally friendly, and incorporate as many green technologies as possible.e
4. The design must be fully modular, meaning little to no on-site construction of core building piec.es
5. The design must provide adequate space and a habitable environment for 1-2 persons.

I’ll spend some time now going over my thought process regarding these criteria, and how I intended to solve each of them.

Firstly, weather resistance. The overall shape of the structure actually provides protection against the majority of weather-related challenges, in being fairly aerodynamic, and almost all sides being slanted. Wind is not an issue with regards to the stability of the structure, as it has a very wide base, and angled faces that provide little surface area for wind to act on. However, wind is an issue if there are gaps between the structural faces where wind can penetrate. To fix this, and water/humidity penetrating the structure, I designed pieces that couple the outer faces of the structure together, and provide sealing against any of these issues. As for flooding, my idea to combat this was simply to elevate the entire structure slightly. The design of the structure requires a concrete base, that protrudes slightly above ground level, surrounding the entire structure, and providing some flood protection. This base or foundation piece is the only part of the structure built on-site, and can easily be heightened in flood-prone areas, to provide higher levels of flood protection.

Next up, environmental challenges. I live in an area with highly erratic temperature fluctuations, so it is incredibly important to me that my design is applicable to a wide array of environments and temperatures. To start, the materials used to construct the prefabricated sections of the structure include an outer layer of reinforced concrete, a layer of plastic/rubber seal material, and an insulative layer. This insulative layer could be built out of Rockwool or any other insulation material, including spray foam. This insulation, as well as the prominence of windows, vents, and heat **capacitive materials, helps to ensure that the structure stays habitable without any outside intervention. As for off-grid locations, the modular design of the structure means that virtually all plain structural pieces can be swapped out for pieces featuring solar panels, vertical hydroponics, or windows, allowing for structures to be optimized for each deployment.

As I talked about earlier, the design is almost entirely modular, owing to the fact that the majority of the structural pieces are identical in size and shape. The only part of the structure that needs to be fabricated on-site is the base/foundation component, which needs to be built for each specific installation environment.

Finally, providing the space necessary for adequate accommodations for 1-2 persons, for an extended period of time, is the essential design requirement. My design uses an approximate livable volume of ∼225m^2 and a square footage of ∼800sqrfeet, giving an equivalent space to a small-medium size apartment. Though there isn't extremely high square footage, there is a lot of vertical space to take advantage of, allowing for stacking shelves, bunk beds, and even a lofted floor, the possibilities are endless.

I utilized the calculator below to help with some of the initial sclaing, though the majority of these values changed throughout development.

Initial Planning:

The initial plan was to build 16 tesselatable tiles to form the faces of the structure and use steel couplings to connect and seal the tiles together. In some initial testing, I found that the optimal size and mesh for the structure used a base layer of 9 slanted tiles to provide structural support and extend the livable space. These slanted tiles, and those used for the door and the capstone are not fully tesselatable, adding some complexity to the design, development, and building processes, but are necessary for the final structure. 

Planning the couplings:

The couplings need to be precisely built to mate their two tiles with high efficiency. The dimensions of each gap between two tiles vary depending on the type of tiles being connected, owing to two separate couplings, one designed for connecting pentagonal tiles to hexagonal ones, and one for connecting two hexagonal tiles. The couplings utilize pin-slot mates to connect the two tiles and spread the force load evenly. To add extra security to the connected tiles, and provide a seal against environmental variables, the coupling includes two plates which overlap the connected tiles, and use rubber rings to provide a seal. The couplings are critical for the structure's integrity and need to be designed with this in mind, and over-engineered if possible. 

Planning the tiles:

Apart from special tiles, 5 types of tile need to be designed, they include:

1. The standard hexagonal tile
2. The standard pentagonal tile
3. A left-sloped hexagonal tile
4. A right-sloped hexagonal tile
5. A foundation layer pentagonal tile

All tiles of these five types will have the same relative dimensions, materials, and coupling design, but will vary in shape. Each of these tile types can be modified to include additional features such as solar panels or green tiles, whilst keeping the general tile dimensions.

In addition to the tesselatable tiles, two additional tiles need to be produced, the capstone tile, and the door tile, each of which provides a unique challenge. Firstly, the capstone tile must be produced with an altered design, using a sloped surface to ensure load limits aren't exceeded with heavy snowfall, allowing for snow to slide off the capstone section. Secondly, a doorway tile needs to be built to connect to the entire structure and allow for entry. The door tile should be pretty simple, essentially a regular tile with a hole cut for a door.

The attached pictures consist of the following:

1. Side profile of a "green" and solar panel tile
2. First rendition of a coupling design (shockingly close to the final design)
3. Front facing views of a pentagon window, hexagon window, and a "green" tile
4. Front facing views of a half hexagon and window tiles, and a hexagonal door tile (none of these were ever used)
5. my first ideas for the structural tile system

The design of the structure, and the fact that it incorporates two oddly shaped tiles at various angles, meant that I needed to get creative with how the tiles would connect. I began prototyping the connection components by using two couplings, one for hexagon-to-hexagonal connections, and one for hexagon-to-pentagon connections. It took a lot of trigonometry(and guessing), but I was finally able to come up with a design that allowed for the pieces to align properly, but it had one massive flaw. The way my couplings were designed, they did not share a single axis, with each being tilted in 3d space to some degree. On top of that, Each coupling has a seal component that overlaps the back of each tile to provide security and stability to the structure. This backing plate and the odd angles of each connector meant that the point at which three connectors met up was (at least to me) impossible to calculate. I ended up simply cutting the backing plates off far enough away from the meet-up points that the issue was negligible.

I ended up creating a rough mock-up of the entire structure with simple shapes, which helped me calculate the necessary angles, and ensure that the structure went together perfectly. Additionally, I was able to use simple colours to figure out which tile types I wanted where, for the final design.

I initially planned to use a foundation layer of tiles, and simply bury the portion of those tiles that extended below the structure, but after some calculations with my mock-up, I was able to come up with some simple tile designs that fit perfectly into the structure whilst only extending into the foundation as far as necessary.

This prototyping, aided by the extremely quick nature of F360, saved me a lot of time down the road, as I made upwards of 80 coupling versions before landing on one that finally worked.

I began my final building process by creating two initial tiles, one pentagon base, and one hexagon base to serve as templates for the rest of the development. Each had the same relative dimensions, sticking with a side length of 2 meters. I added a hollowed out section to the inside face of each tile to allow for insulation space, and chamfered the inner edges to work better with my coupling design.

Once I had my tiles created, I used them as references to design my two coupling versions, scaling all my previous work up. I had to make some changes due to unforeseen collisions between coupling types, but they were mostly minor cosmetic tweaks. Finally, I duplicated my base tiles and created versions for my capstone, door, and foundation tiles, as well as specialty tiles for windows, solar panels, and green walls. Becuase of the tiles being polygon based, I utilized a lot of mirroring, patterning, and revolving to create the features on each tile.

Once I had all my components designed, I began the arduous process of adding joints to each piece and assembling the final structure. I ended up altering a copy of my hexagon couplings to work with my foundation tiles, which made my life a lot easier when building the foundation. Speaking of the foundation, that was the final assembly step. I used a sketch in the assembly to extrude a foundation that held all the tiles properly and overlapped the bottom portion of my foundation tiles.

My plan for rendering the final version of the house, was to use Nvidia Iray in Adobe Substance painter however, Fusion 360 exports its files with UVs that are unable be textured properly, and therefore require editing. My software of choice for editing the UVs was Autodesk Maya, an animation and design software I am quite comfortable with. BUT, I had a major issue when attempting to import my files into Maya, and any other program for that matter. Fusion for some reason was exporting broken or corrupted versions of my files whenever I attempted to export as .FBX. I spent quite a few hours trying to work out what was wrong, and I eventually stumbled across the solution. If you are having the same issue as me, go through the document you wish to export and unling all external files that are referenced in your design. I found that unlinking then exporting produced files that worked in Maya and Substance Painter.

In may, after importing the file as an FBX, and reorienting the components to align with the global coordinates, I had to delete all the UVs for my design, and manually recreate each, for over 300 bodies. To recreate a UV, select the body you wish to edit, open the UV editor, and click create UV. It seems like a rather simple process, but after 30minuters of clicking, it can get quite tiring.

With all my UVs edited, I was able to re-export as an FBX file, and import into Substance painter. Whilst importing, I selected a document resolution of 2k, a resolution I continued to use throughout the rest of the texturing and rendering process. After importing, I added textures to each of the different material groups created by F360, and baked all the texture and normal maps for the document. After baking, I modified a few of the textures, and rendered through Iray.

Honestly, my final renders were kind of lacklustre, and I'm hoping to redo them in the future.

Thanks soo much for reading, Have an amazing day:)