Test

The fire test

A timber fire door in a BS EN 1634-1 furnace, followed for 60 minutes. The 3D explorer runs it live, with the furnace, the graph and the instruments.

Simulated for teaching. The curve and the pressure regime are the real ones; the temperatures read off the specimen and the failure times are indicative of a timber leaf, not measurements from any particular test, and nothing here predicts how any particular door would perform.

The run

Ignition Minutes 0 to 1: The burners light and the furnace begins climbing the curve. The specimen is closed and latched, mounted in the supporting construction it was tested in. Nothing about the door has been helped: it is on its own from here.
Smoke moves first Minutes 1 to 4: Long before anything glows, pressure does the work. Above the neutral axis the furnace pushes hot gases out through every gap; below it, air is drawn in. Cold smoke seals are the only thing in the way. This is the S in FD30S.
The seals answer Minutes 4 to 10: The curve passed 250 °C inside the first two minutes, above the activation point of any intumescent on this door. The seals have fired, expanding many times their packed size and closing the perimeter gap the smoke was using.
Holding the line Minutes 10 to 25: The core chars from the exposed face and the char layer insulates what is left. Steel hinges, melting well above the furnace temperature so far, the expanded seals and the latch this specimen was tested with keep the bowing leaf seated in its frame.
Thirty minutes Minutes 25 to 35: Integrity has held to the half hour. A fire door that reaches here has earned E 30, and with the right evidence FD30. Note what the protected face is reading: hot enough to matter, which is the point most people miss.
Deeper into the hour Minutes 35 to 50: Past 800 °C. Distortion grows, the leaf bows toward the fire, and every component is now working at once. The second half hour is not a repeat of the first: it starts from a furnace already this hot rather than from a cold room, which is part of why sixty-minute performance is so much harder to achieve than thirty. This is why one missing screw, one wrong hinge, one gap left unsealed changes the outcome.
The limit Minutes 50 to 60: An FD60 fire door holds integrity to the hour. A laboratory test does not stop here: the burners stay on until the specimen gives way, and that overrun is part of what the evidence records. This simulation stops at sixty minutes because that is the period the classification describes, not because the door has necessarily reached its limit.

What can be got wrong before the burners light

Failure times are indicative. The mechanisms are the real ones.

Acrylic vision panel

The glazing was broken, and clear plastic cut to size is a tenth of the price of fire-rated glass.

Acrylic softens near 100 °C and burns readily. The furnace passes both inside the first minutes; the pane slumps, ignites, and hands the fire an open hole through the leaf.

The pane softened, slumped out of its beads and ignited. With flame through the aperture the leaf is breached; nothing else about the door gets the chance to matter.

Failed against: sustained flaming on the unexposed side for more than 10 seconds. Indicative failure at minute 4; timing indicative, mechanism real.

Excessive perimeter gaps

The frame was out of square, the leaf was never adjusted, and 6 to 8 mm gaps got signed off as “it still closes”.

Intumescent seals are proven closing the gap they were tested at, typically a few millimetres. Ask them to bridge double that and the expanded material cannot span it: hot gas jets through the head.

The seals fired but could not close a gap twice the size they were tested at. Hot gases jetted through the head until the cotton pad ignited. The seal was never the problem; the gap was.

Failed against: cotton pad ignition at a gap. Indicative failure at minute 13; timing indicative, mechanism real.

Lock fitted without intumescent protection

The lock was swapped and the intumescent kit that came with it stayed in the box. The door looks identical from either face.

A mortice lock is a steel box in a large hole in the core. Unprotected, it conducts furnace heat straight into the timber around it, and the charring it causes runs far ahead of the rest of the leaf.

The steel case cooked the core from the inside. Charring around the mortice ran ahead of the rest of the leaf until flame showed at the closing edge. The kit left in the box was the fire stopping.

Failed against: sustained flaming at the closing edge for more than 10 seconds. Indicative failure at minute 27; timing indicative, mechanism real.

Behind the test

The test

BS EN 1634-1: The method for putting a complete door assembly in a furnace and measuring how long it holds: run under the general rules of BS EN 1363-1, classified through BS EN 13501-2, and the test behind UKCA and CE marked fire doors because BS EN 16034 calls for it. Everything simulated on this screen follows it.
The heating: The furnace is driven along the standard cellulosic time/temperature curve plotted on this screen, defined in BS EN 1363-1: past 550 °C inside five minutes, past 800 °C before the half hour is out.
Overrun, and why this stops: A laboratory test does not stop at the rating. The burners stay on past the target period and the specimen is run on until it gives way; that overrun is part of what the evidence records. This simulation stops at sixty minutes because that is the period the classifications describe, not because the door has necessarily reached its limit.
Other test routes exist: This is not the only furnace test a fire door can carry evidence from, and the methods differ in detail, in pressure regime and in how results are expressed. Those differences matter when reading certification, but they are a certification subject, not this one. Everything here is BS EN 1634-1.

The neutral pressure axis

What it is: A furnace is not at one pressure. Hot gas is buoyant, so pressure rises with height, and BS EN 1363-1 sets that gradient at about 8.5 Pa per metre. Somewhere up the specimen the inside and outside pressures are equal. That height is the neutral pressure axis. Above it the furnace pushes out through any gap; below it, air is drawn in.
Where BS EN 1634-1 puts it: 500 mm above the notional floor. On a 2 m leaf that leaves roughly three quarters of the door under positive pressure, with smoke and hot gas driven hardest at the head and the top of the closing edge. That is why the top of a fire door is where the early action is, and why the pressure probe reads negative only near the threshold.

What is being judged

Integrity, the E: Can the specimen keep flame and hot gas on the fire side? Three kinds of observation decide it. Gap gauges: a 6 mm gauge must not pass through and run 150 mm along an opening, and a 25 mm gauge must not pass through at all. Sustained flaming: no flame on the unexposed face may last more than 10 seconds. The cotton pad: a pad held close to any gap or crack on the unexposed face must not ignite or glow, which is how hot gases escaping through an opening too small for the gauges are caught.
Insulation, the I: Can it keep the protected face cool enough to stand next to? The limits are an average rise of 140 K across the measuring points and 180 K at any single one. Watch the second line on the graph.
E, or EI: Many fire doors are classified for integrity alone, E; others carry an insulation classification too, EI. Which a door achieves is a property of its evidence, not of its thickness, and it cannot be read off a nominal 44 mm leaf. The face temperature on this screen is indicative of a timber leaf and is there to show one thing: the protected face gets hot, well before integrity is lost.

Fire Door Anatomy is an introduction to fire doors, not the whole subject. Always defer to the fire door’s certification, its test evidence and a competent inspection.