Field dossier · Floating architecture · Pneumatic envelopes
A geodesic frame plus air-supported ETFE foil cushions turns one 40-foot container into nearly a thousand cubic metres of protected living volume. This is how you order it, what it costs, how long it survives salt air — and how much sunshine the membrane steals from your solar.
Terminology
All three of your names are used in the wild. Here's which one gets you taken seriously when you email a fabricator.
The common industry shorthand — exactly what architects write on drawings. Works perfectly in a request-for-quote.
Technically the most precise. "Pneumatic" is the key word: the cushions aren't structural fabric under tension — they're held in shape by a continuous low-pressure air supply (typically 200–600 Pa, about 0.03–0.09 psi).
A good description, though engineers reserve "gridshell" for specific double-curved lattice structures. On a dome you'd say space frame or geodesic frame.
The formal name in procurement documents. Famous precedent: the Eden Project biomes (2001), Allianz Arena's 2,874 cushions (2005), Beijing's Water Cube (2008). If they can wrap a stadium, they can wrap your dome.
Why this concept works
The trick is that the envelope is air. The foil ships folded; the volume is manufactured on site by a fan the size of a suitcase.
That's a ≈12× volume ratio — and the envelope itself barely registers. A 50 ft dome at 4V frequency gives ~160 triangles of 3.7–4.5 m edge, an ideal span for cushions. All ~170 cushions, folded, occupy roughly 1 m³ and weigh ≈500 kg. The container space is consumed by the frame, floor deck and interior — not the skin.
Fabrication & ordering
To your core question: cushions are ordered, not built in the field. They're patterned, CNC-cut and welded in a specialist shop, shipped flat, then clipped into the frame and inflated in hours. Here's the pipeline.
Every triangle is measured from a CAD model of the erected frame (or a laser survey of it). A 4V dome of this size has ~160 faces but only a dozen or so unique geometries — the rest are repeats, which keeps price down.
Each triangular face is flattened into a 2D cutting pattern with a "compensation" allowance (typically 2–5%), so that when the cushion is inflated to its design pressure (say 300 Pa) the foil domes to exactly the intended crown without over-stretching. This is the specialist's real secret sauce.
Foil (usually 200–300 µm ETFE film) is cut on a plotter table, and layers are welded together around the perimeter and at internal rope-lines with high-frequency (RF) welders. The edge gets a folded hem or keder that locks into the aluminium profile.
Every strut carries an aluminium clamping/extrusion profile. This is the most expensive line item per metre — so fewer, larger cushions means a cheaper envelope.
Finished cushions fold to a few centimetres thick. Your entire skin — 2 or 3 layers per triangle — arrives on a pallet, plus a crate of profiles and the air unit(s).
On site, each cushion's edge is pressed into the profile and capped. A small air-supply unit (redundant fans + filter + desiccant) keeps every cushion at ~200–600 Pa permanently. Commissioning takes days, not weeks.
Money
Planning-grade ranges in USD (2024–25), for the ≈365 m² envelope of a Ø15.2 m dome. Get three quotes; one-off geometry always carries a premium.
| Route | What you get | Typical range | Per m² of envelope |
|---|---|---|---|
| DIY / self-weld | Raw film, open aluminium profiles, used fans, all labour yours. Highest risk, best story. | $60k – $150k | $165 – $410 |
| Fabricator supply | Factory-welded cushions to your DXFs + extrusions + air unit. You install and commission. | $150k – $350k | $410 – $960 |
| Turnkey specialist | Engineering, patterning, fabrication, installation, commissioning, warranty. | $300k – $600k+ | $820 – $1,650+ |
Two levers cut cost dramatically: drop triangle count (each cushion needs a full perimeter of expensive profile) and mix in opaque bays (see §04). For context, an ETFE envelope is still usually 30–60% cheaper than curved glazing of similar span, at ~1% of the weight — which is why it dominates large transparent roofs. Ongoing costs are modest: fan power runs roughly 1,000–3,500 kWh/year, filters are cheap, and the membrane itself needs almost nothing.
Optics & insulation
Every cushion is a stack of 1–5 foil layers. Layer count is your insulation dial; printing is your shade dial; and you can mix strategies bay-by-bay across the dome.
Each added foil layer costs ~6% of light and buys real insulation. For a lived-in seastead, 2 layers is the bare minimum and 3 is the comfort sweet spot. Air gaps are typically 200–600 mm.
| Layers | U-value W/m²K | Light in | Verdict |
|---|---|---|---|
| 1 | ~5.6 | 94% | Shelter only, unheated |
| 2 | ~2.8 | 87% | The standard roof |
| 3 | ~1.9 | 80% | Year-round comfort ✓ |
| 4 | ~1.5 | 74% | Cool climates |
| 5 | ~1.2 | 68% | Diminishing returns |
≈94% visible transmission per foil — clearer than glass, and it passes most UV. Plants adore it; your upholstery will fade. Budget shades anyway.
White or silver dots screen-printed on a layer — 20–90% coverage, even graduated zones. The standard tool to cut solar gain 30–60% without losing the sky view.
Best move on a seastead: replace lower-ring or north-facing triangles with insulated sandwich panels (U ≈ 0.2–0.3). Privacy, real insulation, and a cheaper envelope — hybrid domes are common practice.
Silver frit or a metallised mid-layer reflects infrared and slashes heat build-up — worth it in the tropics, at some cost to sparkle.
Grey or bronze-tinted films exist and trim glare modestly. Less common than fritting, which is more controllable.
Translucent aerogel granules in the cavity have been prototyped and keep appearing in research, but aren't a reliable off-the-shelf option. Practically: add layers, or go hybrid-opaque.
Durability
ETFE is a fluoropolymer — UV barely touches it, and the surface is so non-stick that rain mostly cleans it. Plans usually assume inspection at ~20 years and re-skinning somewhere between 25 and 35. Crucially for a seastead, individual cushions unclip and swap in under an hour with two people — so keep 5–10 spares on board, sized for your most common triangle.
Solar behind the skin
Your instinct is right: protecting flexible panels inside a clear dome trades away some yield for a huge maintenance win — no salt spray, no wind uplift, no gull damage, trivially serviceable wiring. Here's where the losses come from.
Bottom line: with clear, unfritted cushions, panels inside a 2-layer dome make roughly 72–82% of outdoor yield; a 3-layer skin lands around 66–76%. Any shading frit over the array zone cuts that further, so keep the facets directly above the array crystal-clear. Mitigations that claw a few percent back: mount panels just below the skin with airflow behind them (heat is the sneaky loss), and pick facets facing your sun.
Assumptions: 94% transmission per clear ETFE foil; 92% averaged geometry factor for a curved envelope (oblique sun angles reflect more); 98% for interior cleanliness; −0.35% per °C of cell temperature above 25°C (typical crystalline silicon). Lightweight flexible modules (CIGS etc.) behave similarly. Add 3–5% back if the array is mounted with a ventilated gap just under the top facets.
Packing audit
An order-of-magnitude stowage plan for the Ø15 m, 4V dome. Interior of a 40′ high-cube: ≈76 m³ / ≈26.5 t payload.
| Item | Volume share | m³ | kg |
|---|---|---|---|
| ETFE cushions, folded (~170 pcs, 2–3 layers) | ~1.1 | ~520 | |
| Perimeter extrusions & clamps | ~3.5 | ~900 | |
| Struts, aluminium (~250, 4 lengths ≤4.6 m) | ~9 | ~1,600 | |
| Hubs, fasteners, air unit + spares | ~1.2 | ~530 | |
| Deck panels, flat-pack (182 m²) | ~22 | ~6,500 | |
| Interior/bulkhead flat-packs | ~12 | ~2,200 | |
| Flexible solar rolls, batteries, wiring | ~2.5 | ~900 | |
| TOTAL (vs 76 m³ / 26,500 kg available) | ~51 m³ | ~13,150 |
Volume and weight both clear with room to spare — the floor deck is the hog, so design it as nested cassettes. The one honest caveat: flotation doesn't fit in this budget. Most builders either ship sectional pontoon hulls as a second load, or build the float locally (concrete pontoon, steel barge, or HDPE cubes) and treat the container as the "house kit." That keeps the concept intact: one container carries everything above the waterline.