Research
A tempting one-board wall fails the storm check on every house we generate.
Fewer parts is a design goal: every deleted board is a board a family never has to hang. A storm-rated exterior gypsum board promises to do wind protection, weather exposure and bracing in one layer. Before changing the wall, we swept the idea through the same structural gate every Kult house passes. The one-board version passes 0 of 14 test houses. This paper shows the numbers, then the version of the simplification that survives them.
Finding
As a racking (wind-bracing) board, a storm-rated 9.5 mm gypsum board carries roughly one third of the shear an 11 mm OSB board carries per metre of wall: 0.80–0.96 kN/m against 3.02 kN/m at standard screw spacing. Used as the only bracing layer it fails the roof-diaphragm shear check on all 14 houses in our regression corpus (utilisation 1.1–4.2); today's OSB wall passes 9 of 14, the rest being width-limited under any board.
The board manufacturer's own design method never uses the board alone either. Its worked wall always credits the interior gypsum lining. The honest simplification therefore reads "storm board replaces the dedicated OSB layer, and bracing is shared with the interior lining the wall already has" rather than "storm board replaces OSB one-for-one". That version clears every normal-width house (up to ~7.75 m) and removes one full sheathing layer from the build.
Method
Kult's structural gate checks wall racking to Eurocode 5 §9.2.4 (Method A) and the roof diaphragm to §9.2.3. Both draw their capacity from one fastener design value and one screw spacing. We took the storm board's published fastener design value from the manufacturer's structural handbook (Gyproc Handbok 10, "Skivverkan — elastisk metod": Fd = 0.16 kN on C24 timber, minimum screw spacing 80 mm), substituted it into both checks, and re-ran the full 14-house regression corpus plus a width sweep on the reference house, everything else held constant.
Where the handbook and our model disagree, we chose the reading that favours the storm board, so the failure cannot be an artefact of pessimistic assumptions.
Numbers
| Wall solution | vRd and result |
|---|---|
| OSB 11 mm, screws s150 (today's wall) | 3.02 — passes 9/14 corpus houses; width limit ~9.5 m |
| Storm board alone, s200 (book value) | 0.80–0.96 — passes 0/14; utilisation 1.1–4.2 |
| Storm board alone, minimum spacing s80 | 2.40 — passes 8/14; width limit ~7.5 m; ~1.9× more screws per edge |
| Storm board + one interior gypsum layer, s200 | 2.49 — passes 8/14; width limit ~7.75 m; no extra screws |
| Storm board + two interior layers (handbook wall) | 4.02 — passes 10/14; width limit ~12.75 m |
Material cost is close to a wash: today's OSB plus a separate wind-barrier board costs ~180–214 kr/m² in two layers; the storm board is ~150–189 kr/m² in one (Swedish retail, July 2026; trade prices lower). The real deletions are the dedicated racking board and the separate membrane pass, and a board a novice can cut with a knife instead of a saw — carried by two and hung flat at bench height, since board joints must land on the stud grid, which fixes the format at 1200 mm.
Typical corpus houses use only 29–46 % of OSB's capacity today, which is why the shared-bracing version clears them with margin. The one house it loses against OSB is the widest one, and that house needs an interior bracing line under any board choice.
What this does not show
- The decision has since been taken and shipped (July 2026): the OSB layer and the separate wind membrane are deleted, and the model credits both board faces — storm board outside, a structural gypsum board screwed directly to the studs inside — with the handbook's asymmetry cap (stronger side + 75 % of the weaker). At the dense screw schedule the system lands at 3.03 kN/m, parity with the OSB it replaced; the generated-house corpus is unchanged. One honest nuance the sweep below did not capture: the interior board that earns the credit must sit on the studs — the service-cavity finish board rides on battens, which is not the tabulated connection, so a second, structural gypsum layer was added at the stud line. The net deletion is OSB + wind membrane out, one gypsum layer in.
- All values are calculated rather than tested: they are design values from the manufacturer's Eurocode-based handbook (derived from EN 594 panel tests), run through our checks. We have not racked a physical wall.
- In the sweep below the conservatism runs one way only: it omitted the handbook's asymmetry cap, which would only make the lone storm board look worse. The failure stands either way. (The shipped check applies the cap.)
- Anchorage details (hold-downs, sill fixings) are a separate calculation and unchanged by board choice.
Sources
- Gyproc Handbok 10 — "Statik: Skivverkan, elastisk metod" (fastener design values, minimum spacings, worked wall example): gyproc.se
- Glasroc X Storm wind-protection system — product data, 12-month weather exposure: gyproc.se/glasroc-x-storm
- EN 1995-1-1 (Eurocode 5) §9.2.3–9.2.4 — diaphragm and racking design; EN 594:2011 — the racking test method behind the handbook values.
- TräGuiden — racking capacity of timber wall panels: traguiden.se
- Kult internal research, "Storm board as sole racking sheathing — verification" (2026-07-10) — full tables, prices and corpus results.
Read next
The energy code does not set the wall thickness
The same wall judged by the U-value engine: what actually drives the 220 mm build-up.
Fit and assembly order, verified before cutting
The geometric checks that run on every generated house before files are exported.
All papers
The research hub collects every published paper, and the note on why the work is open.