Fire-rated door installation is governed less by the door leaf itself and more by what surrounds it: a listed frame, gasketing, closing hardware, and a certification label that survives the installer's drill bit.
Specifying a 90-minute door on a non-listed frame, or skipping the intumescent strip, is the single most common way a passive fire assembly fails a final inspection on a commercial build [S2][S4].
What a "Fire-Rated Door" Actually Buys You
A fire-rated door is a listed assembly, not a slab: it has demonstrated fire-resistance capacity of 20 minutes to 3 hours when tested as a complete unit (leaf + frame + hardware) under positive-pressure furnace protocols such as UL 10C, NFPA 252, or BS 476 Part 22 [S2][S6].
Three pass/fail criteria are recorded per the furnace test: structural stability (no collapse), integrity (no flame passage through gaps or glass), and insulation (temperature rise on the unexposed face held below 180 °C average / 250 °C at any single point) [S6][S9]. The leaf material — wood, steel, or glass-clad — is secondary to the tested assembly; swapping a listed wood core for an unlisted MDF core invalidates the label the moment the screw holes are drilled [S7][S9].
For background on how ratings and hardware interact, the Fire-Rated Door Types, Ratings, and Hardware Map breaks the difference between FD30, FD60, FD90, and FD120 designations across regions.
Site Survey Before a Single Anchor Goes In
Field installation starts with verifying four items on the rough opening: substrate material, plumb and square within 1.5 mm/m, clearances, and the wall's fire-resistance rating matching the door listing [S4][S6].
Openings in masonry, stud wall, or drywall partitions must each be framed per the door manufacturer's listed installation instructions — UL10(c) technical manuals and FRX fire-rated jamb sheets show clip spacing at typically 600 mm centres and a maximum anchor embedment depth per the tested design [S6]. Wood or steel stud frames require structural reinforcement behind hinge and strike locations; without it, the assembly racks under door-closer cycling and exceeds the 3 mm diagonal clearances permitted in the listing [S4][S6].
Substrate must be dry, plumb, and capable of carrying the door mass — a 90-minute steel-clad leaf can exceed 60 kg per leaf, and closer arm forces of 15-25 N·m at the hinge jamb translate directly into fastener shear if the framing is light [S4].
Frame Set, Anchor Pattern, and Shim Strategy
Steel frames are typically anchored with masonry T-anchors or wire anchors at hinge, strike, and head locations; minimum anchor count is 8 per single door frame per the SDI / HMMA 867 referenced listing instructions, with additional anchors at the top and bottom of each jamb for openings above 2.1 m [S6].
Shims behind jambs must be load-bearing and non-combustible (steel or mineral composite), not organic timber wedges that char out within minutes of fire exposure and leave a clear smoke path. Maximum shim gap is 6 mm; any wider gap is filled with fire-rated mineral wool or intumescent backer, not polyurethane foam [S4][S6].
Plumb tolerance after setting: 1.5 mm in 2.0 m height on the strike jamb, 3 mm maximum face twist across the opening. Out-of-plumb frames bind the door, force the closer to over-extend, and accelerate the failure of the listed self-closing device [S4].
Hardware: Where the Assembly Either Stays Listed or Stops Being One
Every piece of ironmongery on a fire door — hinges, latch, closer, coordinator, kick-plate, viewer, letterplate — must be listed for the door's fire-resistance rating and installed per its own certificate [S1][S8].
A common failure: a CE / UL-listed closer is fitted, but the through-bolts are replaced with self-tappers because the frame reinforcement wasn't pre-drilled. The result is a closer that pulls out of the jamb during a fire, leaving the door open and defeating the rating [S1][S8]. Likewise, installing a non-rated deadbolt above the listed latch on a 60-minute door requires a fire-rated mortice lock case with intumescent gasketing under the forend [S8].
Hardware-set minimums as commonly called out on commercial schedules: 3 hinges per leaf for 90+ minute ratings, self-closing device sized to door mass (EN size 3-5 for ≤80 kg / ≤1100 mm, EN 6-7 for heavier leaves), and a coordinator on pairs of doors with overlapping astragals [S1][S6][S8]. See the Fire-Rated Door Types, Ratings, and Hardware Map reference for the typical hardware package by rating band.
Seals, Gaskets, and the Smoke-Leak Problem
Smoke leakage, not flame, is the dominant cause of fire-door failure in occupied buildings: most fatalities in compartmented fires are smoke-related, and a door that passes the furnace test can still pass 5 m³/min of cold smoke at 25 Pa pressure differential unless fitted with a tested smoke seal [S4][S6].
Intumescent strips are typically 10×2 mm or 15×4 mm graphite-based seals mounted in the frame reveal or door edge; they expand to 20-30× their free thickness at 180-200 °C to seal gaps that would otherwise pass hot gas. Combined intumescent + cold-smoke brush or silicone gaskets carry dual certification to BS 476 Part 31.1 / EN 1634-3 for smoke leakage under both ambient and medium-temperature exposure [S6][S9].
Continuous seals are required at head, both jambs, and meeting stiles — any 50 mm gap in the seal run (commonly left under the closer arm cut-out) becomes a measured smoke path during acceptance testing [S4][S6].
Adjust, Test, and Label: The Acceptance Path
After installation, the closing speed, latch engagement, and seal compression are checked against values on the hardware schedule; closer sweep time is typically adjusted to 5-7 seconds from 90° to closed, latching within the final 200 mm of travel [S1][S4].
Acceptance values from Australian BCA / AS 1905.1 and equivalent EN / UL practice: maximum 3 mm gap at head and jambs (2 mm on smoke-control doors), maximum 8 mm at meeting stiles, door must self-close from any position including fully open against the closer back-check [S4][S6]. Field testing usually pairs a 25 Pa smoke-leakage check with a 100-cycle operational test on closers and coordinators before the certification tag is affixed [S4].
The label — UL, Certifire, Activar, or the regional equivalent — is the inspector's only proof that the installed configuration is the one that was tested. Covering it with paint, relocating it to a hinge, or worse, installing a non-listed leaf into a listed frame, removes the assembly from the regulatory record and the building fails handover [S4][S6].
Common Installation Defects That Void the Listing
Defects most often caught at handover: gaps above 3 mm at head/jambs (60% of failed inspections), missing or discontinuous intumescent seals, non-listed kick-plates or viewers drilled after certification, and closers set too weak to overcome the latch under fire-induced door warping [S4][S6].
Field repairs are limited: a frame out of square by more than 4 mm must be reset, not shimmed; a closer that fails the 100-cycle test must be replaced with the same listed model, not a "similar" one; a damaged intumescent strip must be replaced with the identical product from the same manufacturer (mixing seals from different listings invalidates the assembly test) [S4][S6][S8].
When in doubt, escalate to the certification body before signing off — a re-test is cheaper than a failed handover, and a door that won't stay closed in a fire is no door at all.
For the relevant spec sheets and selection criteria, see fire rated door, fire door, and linear guide.