construction · manufacturing · family: it worked in the lab
eighteen stories ofwood, no fire test
Mass timber buildings above six stories have no full-scale fire test data to validate code assumptions
Problem statement
Cross-laminated timber (CLT) and other mass timber systems are now permitted for buildings up to 18 stories under the 2021 International Building Code (IBC Types IV-A, IV-B, IV-C). The fire safety provisions for these tall mass timber buildings were developed primarily from furnace tests on individual CLT panels and small compartment tests — no full-scale multi-story fire test has ever been conducted on a mass timber building above 6 stories. Code provisions rely on char-rate calculations and gypsum board encapsulation assumptions extrapolated from small-scale data, but full-scale fire behavior involves phenomena (floor-to-floor flame spread via connections, char delamination cascading across large exposed surfaces, structural load redistribution under fire) that cannot be captured by component-level testing.
Why this matters
Mass timber is the most significant structural building innovation in decades, with a global market projected at $1.4 billion by 2027. It offers 25–45% lower embodied carbon than concrete/steel equivalents. However, insurance underwriters, fire departments, and building officials in many jurisdictions are reluctant to approve tall mass timber projects because the fire safety evidence base does not match the code permissions. Several insurers have imposed exclusionary clauses or prohibitive premiums on mass timber buildings above 6 stories, effectively blocking projects that building codes technically permit.
What’s been tried and why it hasn’t worked
Small-scale furnace tests (ASTM E119) measure fire resistance of individual elements but cannot capture structural system behavior under fire. The few large-scale compartment fire tests conducted (notably the Carleton University/NRC tests in Canada and the APT Building fire test in Austria) used single-story or two-story configurations that do not replicate the vertical fire spread, connection behavior, and load path redistribution of tall buildings. Computer models (FEM fire simulations) fill the gap computationally but remain unvalidated against full-scale data. The fundamental barrier is cost: a full-scale multi-story fire test of a tall mass timber building would cost $10M–$30M and requires purpose-built test structures — neither industry nor government has funded one.
What would unlock progress
An intermediate-scale testing protocol that captures the critical multi-story fire phenomena (connection behavior, char delamination propagation, vertical fire spread through concealed spaces, structural load redistribution) without requiring a full 18-story test structure. This could involve a 3–4 story mass timber assembly tested with realistic fire scenarios and instrumented to validate the computational models that codes already depend on. The adjacent success of the Cardington steel frame fire tests (1990s, UK) — which validated steel fire engineering for a generation — provides a model for how a single well-instrumented test program can unlock an entire building technology.
Entry points for student teams
The validation gap can be attacked computationally with no facility at all: NIST's Fire Dynamics Simulator is free and open source and ships a library of validation reports with their underlying experimental datasets, so a semester can go into a structured sensitivity study of the assumptions the code already rests on — which char-rate, encapsulation-failure-time, and connection-conductance parameters, varied across their published ranges, actually flip whether a compartment flashes over or a load path is lost — delivering the ranked list of what a full-scale test would have to instrument. A materials team could take the delamination question to bench scale, comparing char depth and adhesive bond failure across CLT samples with different adhesive systems under sustained heating past the point where the rated fire-resistance test stops; that needs a cone calorimeter or a small furnace, ordinarily held by a university fire-protection or materials-testing lab, so a team without that access takes the modelling door instead. Third, the intermediate-scale 3–4 story test program this brief names as the unlock is itself a semester deliverable — fire scenario, instrumentation plan, load configuration, acceptance criteria, and cost envelope written as a Cardington-style program specification and handed to a fire research laboratory, insurer consortium, or code committee that could sponsor it — since the field's bottleneck is a funded test, not another argument for one. Relevant disciplines: fire protection engineering, structural engineering, materials science, computational mechanics.
Genome — every gene is a door
Structural cousins — same reason stuck, other fields
Sources
CIDB/BCA Construction Technology Roadmap 2025; IBC 2021 Type IV-A/B/C provisions; Zelinka et al., "Fire Safety Challenges of Tall Wood Buildings," USDA Forest Products Laboratory, FPL-GTR-270, 2018; Barber, "Fire Safety of Mass Timber Buildings," Arup, 2022. Accessed 2026-02-25.
verification notes (working record)
The collection team’s own sourcing notes for this brief, kept verbatim:
The barrier is genuinely static — the lack of full-scale fire test data has persisted since mass timber's introduction to codes and has not gotten worse; what has changed is the ambition of projects (taller buildings seeking approval), but the measurement gap itself is the same. Cluster targets: C4 (manufacturing scale-up — mass timber manufacturing at building scale), C10 (codes void — fire codes cannot validate what they permit). Related briefs: construction-3d-printed-concrete-code-void (same pattern of codes permitting technology without testing evidence).
Reconciliation 2026-08-21: Entry-point repair under the ≥2-doors rule (triage row: score 2, reduced-scale CLT furnace test plus instrumenting an existing building). Flag CONFIRMED, and worse than the score implies: BOTH doors failed, so the brief was at zero reachable doors. The fire-resistance furnace arm is a rare, expensive facility booked in commercial slots, and the second arm — instrumenting an occupied mass timber building with thermocouples and strain gauges — produces no fire-condition data at all absent a fire, which is the one thing that will not happen during the semester; it was deleted rather than repaired, since no access line rescues a door whose measurement never occurs. Replaced with three doors: (1) facility-free — a sensitivity study of the code's own char-rate, encapsulation-failure and connection-conductance assumptions in NIST's free open-source Fire Dynamics Simulator, delivering the ranked instrument list a future full-scale test needs; (2) bench-scale char and adhesive-delamination comparison, with the cone-calorimeter/small-furnace requirement stated as an explicit access line (university fire-protection or materials-testing lab) and the modelling door named as the fallback; (3) design-the-trial applied to the brief's own unlock — the intermediate-scale 3–4 story test program written as a Cardington-style specification and handed to a fire lab, insurer consortium, or code committee. Resources verified by fetch on 2026-08-21: FDS/Smokeview (https://pages.nist.gov/fds-smv/ — confirmed on the page as free and open-source NIST software, with validation test reports and their experimental datasets; downloads at https://pages.nist.gov/fds-smv/downloads.html). Declined to cite: the Fire Protection Research Foundation report library and the FPL Treesearch archive — both hosts resolve, but neither could be fetch-confirmed to hold the specific tall-wood fire reports at a stable URL, so the doors were built on FDS and on the compartment tests this brief already cites (Carleton/NRC, APT Austria) instead.