Every seat fabric, headliner, carpet, door panel trim and sun visor cover in a car is tested for horizontal burning rate. In the United States the requirement is Federal Motor Vehicle Safety Standard No. 302 (49 CFR 571.302). Most other markets use ISO 3795, or a carmaker specification built on it.
The two methods share the same geometry: a specimen held horizontally in a U-shaped frame, a small Bunsen flame applied to the free end, and a timer that measures how fast the flame front travels. The test looks simple. Its results are still disputed more often than they should be.
What FMVSS 302 specifies
The figures below come from the text of 49 CFR 571.302:
| Parameter | Requirement |
|---|---|
| Scope | any part of a material within 13 mm of the occupant compartment air space |
| Specimen | 102 mm wide × 356 mm long, where possible |
| Maximum thickness | 13 mm, measured from the face closest to the occupant compartment |
| Conditioning | 24 h at 21 °C and 50 % RH; tested under the same conditions |
| Flame | 38 mm, Bunsen burner with 10 mm tube, air inlet closed |
| Flame application | 15 s |
| Timing | from the point 38 mm from the open end, to 38 mm from the clamped end |
| Limit | burn rate not more than 102 mm per minute |
A material also meets the requirement if it stops burning within 60 seconds of the start of timing and has not burned more than 51 mm from the point where timing started.
ISO 3795:1989 — reviewed and confirmed by ISO in 2025 — describes the same principle and the same kind of apparatus. The important difference is that ISO 3795 is a test method without a pass/fail limit. The limit comes from the regulation or from the carmaker's material specification that calls up the method.
Error 1: cutting the specimen to 13 mm from the wrong side
For a composite thicker than 13 mm — a seat with foam, a carpet with acoustic backing — the specimen is cut down to 13 mm. The 13 mm is measured from the surface closest to the occupant compartment, so the layer you remove is the one furthest from the passenger.
Cut from the wrong side and you are testing the foam or felt instead of the face fabric, which can turn a passing material into a failing one, or the other way round.
Error 2: testing the wrong face down
The specimen is mounted so that the side facing the occupant compartment is exposed in the direction the method defines. Laminated headliners and backed carpets burn differently from each face. Mark the occupant side on every specimen when it is cut, before it goes into the conditioning room and loses its context.
Error 3: conditioning in the textile atmosphere
FMVSS 302 conditions at 21 °C and 50 % relative humidity. A textile laboratory's standard atmosphere under ISO 139 is different. If automotive fabrics are conditioned in the textile room and burned straight away, the moisture content of cellulosic and wool components will not be what the regulation assumes.
If you run both kinds of work, either keep a separate conditioning space for automotive specimens or record clearly that the conditioning followed the automotive method, not ISO 139. For ISO 3795 and OEM specifications, read the conditioning clause of the document you are reporting against — do not assume it matches either.
Error 4: starting the clock at ignition
Timing does not start when the flame touches the specimen. It starts when the flame front passes the first timing mark, 38 mm from the open end. The first part of the burn is deliberately excluded, because the ignition flame is still influencing it.
Operators who start the stopwatch at flame application report lower burn rates for fast-burning materials and longer burn distances for self-extinguishing ones. Put visible marks or wires at both measuring points and train operators to watch the flame front, not the burner.
Error 5: ignoring how the specimen behaves
Thin films and some synthetic knits melt, shrink back from the flame or drip. Some materials char and stop before the first mark. These are not "invalid" tests; they are results, and the report needs to say what happened:
- did the flame reach the first mark at all;
- did it stop before the second mark, and where;
- did the specimen drip, shrink or curl out of the frame.
Soft or sagging specimens sometimes need support wires under them. Use supports only in the way the method or specification you are reporting against describes them, and state it on the report — a support changes the heat loss and therefore the result.
Error 6: averaging away a failure
A specimen set is tested, and the regulation applies to each specimen. A mean burn rate below 102 mm/min does not rescue an individual specimen that burns faster. Report every specimen, and if a carmaker specification uses its own statistics, apply those explicitly.
Cabinet and burner details that matter
- Draughts. The test is run in a closed cabinet in still air. Turn off fume extraction during the burn and run it afterwards; an extraction fan running during the test pulls the flame and shortens times.
- Gas. FMVSS 302 calls for gas with a flame temperature equivalent to natural gas. Changing gas supply changes the flame. Record the gas type.
- Flame height. Check the 38 mm flame height with a gauge before each series, not by eye.
- Clean frame. Residue from earlier specimens on the U-frame acts as a heat sink or fuel. Clean it between tests.
Before you report
- Was the 13 mm measured from the occupant side?
- Was the occupant face marked and exposed correctly?
- Which conditioning atmosphere was used, and does it match the method?
- Was timing started at the first mark?
- Are individual results reported, not only the mean?
Our horizontal flammability tester for FMVSS 302, ISO 3795 and related automotive methods is manufactured in our own workshop in Istanbul; see the FMVSS 302 and ISO 3795 pages.