Axis 01 · MATERIALS
Materials Intelligence
For any surface or element — indoor, outdoor, floor, wall, ceiling, built-in, custom-made or modular — which material is correct, and why?
Status
PARTIAL
named, but the answer is incomplete
Rules in force
5
RULE-MAT-WET-004 · RULE-TIER-DISPLAY-001 · RULE-MAT-CLAIM-001 · RULE-MAT-THINPANEL-001 · RULE-ENV-SANDWICH-001
Open gaps
8
each with a stated fix
Decision purposes
4
MATERIAL · FLOOR · WALL · WET_ZONE
Rules answering this axis
| ID | Rule | Status | Ev. | Publication |
|---|---|---|---|---|
| RULE-MAT-WET-004 | Sink-base / wet-zone carcass material selection | APPROVED | B | PUBLIC |
| RULE-TIER-DISPLAY-001 | Tier naming — display labels vs code keys | APPROVED | B | PUBLIC |
| RULE-MAT-CLAIM-001 | A performance claim needs a standard, a class and a document — or it is not said | CANDIDATE | B | PROFESSIONAL |
| RULE-MAT-THINPANEL-001 | Ultra-thin large-format panels are a substrate job before they are a tiling job | CANDIDATE | B | PROFESSIONAL |
| RULE-ENV-SANDWICH-001 | An insulated sandwich roof panel is specified by its core, not by its face | CANDIDATE | B | PROFESSIONAL |
Taxonomy
- materials.json ladder (5 rungs, zone + cost_tier)
- rooms.json (12 rooms)
- material-facets.json EXPOSURE_CONTEXT (PROPOSED)
- assembly-systems.json (2 candidate assemblies, WEAKEST-LAYER principle)
Canon cited
- ELS-BOOKS Book 3 CAR (carcass) · Book 4 CT · Book 5 TF
- /api/material-zones feed
- ELS-BOOKS Book 1 — REF-Z / REF-M / REF-H / REF-C / REF-S / REF-U / REF-SE
What is missing
8 openCEILING has no decision purpose and no rule
decision-purposes.json carries FLOOR and WALL but no CEILING. Book 7 is the estate's ceiling authority (B7-G gypsum, B7-H height, B7-C cove, B7-VC01 coordination sequence, B7-BM beam, B7-CB01 curtain box) and none of it is reachable from this repo.
Fix — Add a CEILING purpose; cite Book 7 codes rather than restating them.
No fabrication-route axis: built-in vs custom-made vs modular vs site-assembled
The customer-visible difference between a fitted built-in, a bespoke one-off and a modular carcass drives lead time, recovery, price and who installs it — and nothing in the repo records which route a material or element takes. Sharpened 2026-08-26: the two assemblies raised that day are none of the three original values — they are SITE-ASSEMBLED, built up in layers in position, where the sequence is part of the specification and cannot be recovered afterwards.
Fix — Propose a FABRICATION_ROUTE facet (built-in · modular · custom-made · site-assembled · supply-only) to ELS alongside the existing seven.
The material ladder is 5 rungs against a 315-SKU catalogue
materials.json carries 5 ladder rungs; the ALUMAX Sell Catalog carries 315 SKUs across 23 lines. The ladder is a teaching device, not an index — that is fine, but nothing states it, so the gap reads as under-population.
Fix — State the ladder's job in its _meta, and cite the Sell Catalog as the index.
No FIRE_PERFORMANCE facet anywhere in the estate
material-facets.json proposes seven facets and none of them asks how a material behaves in a fire. That was tolerable while the canon was kitchen carcasses; it is not tolerable now that envelope build-ups are arriving. It is also the only facet in the estate with a regulator (UBBL 1984) and an insurer behind it, and the only one whose wrong answer is not recoverable by a credit note.
Fix — PROPOSED to ELS in assembly-systems.json — values non-combustible-certified · limited-combustibility-certified · combustible · untested, where anything but 'untested' requires a named standard and class on file.
No ASSEMBLY_ROLE facet — the weakest-layer principle is prose, not data
Nothing records whether a material is a finish, substrate, core, barrier or fixing. Without that, there is no way for a surface to ask which layer governs a claimed property, so the principle stays wording that a spec sheet can ignore.
Fix — PROPOSED to ELS in assembly-systems.json. 'standalone' as a value keeps every existing ladder material valid with no migration.
MATERIAL_RISK cannot express the two failure modes envelope work introduces
interstitial-condensation and substrate-telegraphing are both slow, invisible and expensive — the profile of a failure most worth naming — and neither is expressible with the current ten values. Condensation in particular surfaces two to three years in, which is after the warranty conversation has got harder.
Fix — PROPOSED as additions to the existing facet in assembly-systems.json.
Two assemblies are CANDIDATE with blocking questions unanswered
ASSY-ROOF-IMP-SHINGLE-001 carries four blocking open questions — core type, what the shingles fix into, vapour control, and panel span against the inherited purlin layout — and its 防燃 claim is recorded CONTRADICTED. ASSY-WALL-THINPANEL-001 is two systems described as one. Neither may be quoted until those are closed.
Fix — Answer the core question first; it blocks the other three on the roof build-up.
REF-M has no envelope materials
The material permissions table gates what may be used in which zone and has been amended seven times. It carries no row for coated steel sandwich panel, mineral wool, organic insulation cores, bitumen shingle, terracotta rainscreen, or ultra-thin porcelain. Envelope work is unrepresented in the estate's central material gate.
Fix — Propose REF-M Amendment 8 through the established pattern — the owner's table, not a private rule here.