Smoke kills more people in fires than flame does, and it does so by taking away sight: escape routes disappear behind optical density long before temperatures become lethal. The NBS smoke density chamber is the instrument regulators chose to measure that property, and it has sat at the centre of transport fire approval since the aviation industry adopted it. This post explains the chamber, the number it produces, and how a test actually runs.
Inside the chamber
The chamber is a sealed box - in the Motis SDB, internal dimensions of 914 x 914 x 610 mm, with Teflon-coated walls (smoke is corrosive and the coating is what keeps the box measurable over years) and a full-opening door for specimen access and cleaning. Two features define it. First, it is sealed during the test: unlike flow-through instruments, smoke accumulates in a known volume, so the measurement is of the total smoke the specimen produces, concentrated. Second, a vertical light beam crosses the chamber to a photometric detector - in the SDB, a photomultiplier-based system - recording transmittance continuously as the beam fades.
Supporting details matter more than they sound: a pressure-relief arrangement with an explosion-relief foil device at the top (sealed boxes of burning polymer can pressurise), and automatic cylinder-driven intake and exhaust so the chamber can be purged between tests without opening it to the laboratory.
The exposure: two modes
The specimen is exposed horizontally to radiant heat - in the ASTM E662 arrangement documented for the SDB, a furnace delivering 25 kW per square metre. In the non-flaming mode that is all: the material pyrolyses and smokes without flame, modelling smouldering and radiant-only scenarios. In the flaming mode a row of pilot burners (the standard six-burner arrangement) plays on the specimen, modelling direct flame involvement. Most specifications demand both, because some materials smoke worst in one mode and some in the other. Surface temperature for the exposure is set with a foil-backed copper-disc calorimeter.
The number: specific optical density
The chamber reports specific optical density, Ds - a normalised measure that accounts for the chamber volume, the exposed specimen area and the light path length, so results are comparable between laboratories. Reports quote the curve over time and its landmarks: the maximum (Dm) and the values at the minutes the specification names. Aviation's classic criterion reads the four-minute value; railway methods accumulate the early minutes into the VOF4 figure. Smoke numbers are only meaningful with their method attached - a Ds from ASTM E662 and one from ISO 5659-2 come from the same chamber family but different exposure and reporting rules, as our comparison post explains.
From density to toxicity
The chamber is also the sampling point for smoke toxicity. Options documented for the SDB include Drager gas detection tubes (the aviation smoke toxicity tradition) and an FTIR interface for full quantitative gas analysis - the configuration railway CIT index work is built on. A chamber bought with its sampling interface planned is a toxicity station in waiting.
Conditioning, cleaning and the discipline of a dirty instrument
Smoke chambers live a hard life: every test coats the interior with the specimen's condensed products, and the method's repeatability depends on returning the box to the same state each time. The Teflon coating on the chamber walls exists because smoke residues are corrosive and because a surface that can be wiped back to a consistent condition is a metrological feature, not housekeeping. The photometric system likewise depends on clean optics and verified zero - the light path check before a test series is the smoke equivalent of the cone calorimeter's methane calibration. Laboratories new to smoke work are often surprised that chamber time is dominated by conditioning specimens (smoke output is moisture-sensitive, like every fire property) and by cleaning and verification between runs; plan staffing and throughput on that basis, not on the 20 or 30 minutes of the test itself.
Dm, Ds(4) and VOF4: reading the landmarks
Because specifications quote different landmarks, confusion between them is the commonest reporting error in smoke work. Dm is the maximum specific optical density reached in the test - the worst the smoke gets. A time-point value such as Ds at 4 minutes asks how bad things are early, which is what escape-time thinking cares about; aviation's classic limit is written this way. VOF4, used in railway work, accumulates the density over the first four minutes, penalising materials that smoke early even if they later clear. A material can rank differently on each landmark, and none of the three is "the smoke number" without its method and mode (flaming or non-flaming) attached. Quote all three pieces - value, landmark, method - whenever smoke data change hands.
A short history, because it explains the design
The chamber's name is its origin: it was developed at the US National Bureau of Standards (NBS, now NIST) in the 1970s, when aviation and building regulators needed a smoke measurement that different laboratories could reproduce. Its design choices - a sealed, accumulative volume; a photometric rather than gravimetric measurement; a small specimen with a standardised exposure - were all reproducibility decisions, trading the realism of a flowing system for agreement between laboratories. Half a century of specification history has accumulated on those decisions, which is why the chamber's geometry survives essentially intact in today's standards and in instruments like the SDB: in regulatory testing, comparability with the archive of past results is itself a technical requirement.
Flaming mode details that catch new operators
Two practical points from the chamber's working life. The flaming mode's burner row must be lit and stable before exposure timing begins, and its behaviour with the specimen's own flames is part of the method's choreography - a burner that is blown out by the specimen's outgassing invalidates assumptions the report is built on, which is why the arrangement is specified so completely. And the first minutes matter disproportionately: most specification landmarks sit between ninety seconds and four minutes, so operator attention and recording integrity early in the test decide the quality of the evidence, even though the test continues well beyond the landmarks.
The chamber in one sentence
The NBS chamber endures because it answers the question regulators actually ask - how much sight will this material take away, in the first minutes, in a closed space - with a number that laboratories on different continents can agree on; everything else about it is engineering in service of that agreement.
Who runs one
Aviation material suppliers proving FAR 25.853 smoke compliance; railway suppliers proving EN 45545-2 smoke limits; marine suppliers under the IMO code; and building material laboratories working to the chamber methods. If your specification says "smoke density", this chamber is very probably the instrument behind the words. The SDB product page documents the chamber's construction and full specification.