A two-floor, two-bedroom submersible dwelling cast as a single seamless sphere over an inflated balloon form. Floats for daily life, submerges 1-10 m to ride out storms. Every figure below is a live, first-order engineering model you can drive yourself.
Livable space is rectilinear, so we fit the largest cube inside the internal sphere: its space diagonal equals the diameter, giving a side of 2R/√3. The curved volume outside the cube faces is dead space - and the bottom cap becomes the ballast tank. Then we stack floors inside the cube from your slab thickness and room height. Drive all four.
Submerged, the whole envelope carries external water pressure pushing inward. The failure mode is no longer "flooding" - it is buckling (implosion). A sphere turns that pressure into pure membrane compression, which concrete loves. Slide the crown depth from 1 to 10 m.
Reference geometry: 8.5 m internal Ø, 200 mm wall. Buckling uses the classical sphere formula P_cr = 1.21·E·(t/R)² with a 0.2 imperfection knockdown.
| Geometry under external pressure | Behaviour | Verdict |
|---|---|---|
| Flat panel / box wall | Bends inward, ~90 t on a 3×3 m panel at 10 m | avoid |
| Faceted icosahedron | Bending + stress concentration at every edge/corner | poor |
| Cylinder + domed ends | Membrane compression, ring-stiffener friendly | good |
| Sphere | Uniform membrane compression, lowest material for a given depth | best |
Buoyancy comes from the enclosed air volume; you submerge by flooding ballast tanks to cancel that lift, and surface by blowing them with compressed air. The golden rule is fail-safe positive buoyancy: any failure must default to floating. Fill the ballast and watch it settle down its guide wires.
A single pressure hatch on the top deck, in air when surfaced. It seats inward against its frame, so the deeper you go the harder the water clamps it shut. Dogged down, it is a structural plug, not a swinging door. No submerged access = no airlock, no decompression.
Wave energy is a surface phenomenon and decays fast with depth. Submerging to 10 m lets a unit sit calm beneath a hurricane that would batter it on the surface. The guide wires and tension moorings hold station while it is down.
The whole point is a monolithic cure: structure and liner set as one body, with no cold joints to leak or crack. That drives an inflated-balloon-form method suspended from a gantry over deep water, coated in fibre-reinforced shotcrete. Step through it.
Marine, monolithic, and formed over an inflatable. Each layer is chosen to bond to the next so the finished sphere behaves as one body.
| Option | Why | Watch-outs | Fit |
|---|---|---|---|
| PVC-coated polyester | Proven Monolithic-Dome air-form fabric; weldable seams, holds low pressure for weeks | UV degrades if left bare; single-digit-kPa working pressure | primary |
| TPU-coated nylon | Tougher, more abrasion + tear resistant, reusable more times | Costlier; seam welding needs control | reusable |
| EPDM / reinforced rubber | Cheap, very airtight | Heavy, hard to seam into a true sphere | backup |
| Component | Role | Target |
|---|---|---|
| Sulfate-resisting / marine Portland (or GGBS blend) | Binder, seawater durability | 380-420 kg/m³ |
| GGBS or fly ash replacement | Lower heat of hydration, denser matrix, chloride resistance | 30-50% of binder |
| Water / binder ratio | Low w/b = low permeability (keeps chloride + water out) | 0.36-0.40 |
| Silica fume | Densifies, cuts permeability, boosts shotcrete cohesion | 5-8% |
| Superplasticiser (PCE) | Workability at low water | to spec |
| Set retarder | Keeps successive lifts "live" so the shell cures monolithic | to spec |
| Corrosion inhibitor + hydrophobic admix | Protects any steel; reduces wicking | to spec |
| Fibre | Strength / stiffness | In seawater concrete | Where to use it |
|---|---|---|---|
| AR-glass (alkali-resistant) | Good tensile, low cost | Must be AR grade - plain E-glass is eaten by cement alkali | Primary shell mesh |
| Basalt | Higher than AR-glass, excellent | Superb chloride + alkali resistance, non-corroding | Primary mesh (marine first choice) |
| Carbon | Highest strength + stiffness; electrically + thermally conductive | Inert, non-corroding | High-stress zones + the thermal patch (see §06) |
| Steel rebar (epoxy/GFRP-clad) | Cheap, ductile | Corrodes unless well-covered or clad; avoid bare in saltwater | Only if clad; prefer non-metallic |
A poured aircrete (foamed concrete, ~400 kg/m³) liner bonded to the still-green structural shell gives a low-conductivity insulating skin and a fair interior surface, curing into the same body. It is buoyant, non-combustible, and easy to shape into built-in furniture. Keep it out of the keel thermal patch, where you want conductivity, not insulation.
This is where the carbon additive earns its place. Plain concrete insulates (~1.4 W/m·K); carbon / graphite additions raise conductivity several-fold. Cast a conductive "thermal patch" at the keel - in contact with cold deep water - and it becomes a passive heat exchanger for the fresh-water tanks, while the aircrete-lined living space stays insulated. Add deck solar heating for the warm side.
| Loop | Source | Function |
|---|---|---|
| Chilled water | Cold deep seawater via keel carbon patch | Passive cooling + cold store |
| Hot water | Deck solar thermal collector | Domestic hot water + warm store |
| Living space | Aircrete liner insulation | Buffers cabin from both |
| Bonus | Carbon fibre is conductive | Enables resistive trim-heat + corrosion sensing |
Deep-water temperature is an approximate Bermuda seasonal profile; conductive-concrete values are within the published range for carbon/graphite-modified mixes. Treat as order-of-magnitude, not design-final.