Why Coatings Need a Furnace at All<\/h2>\n
A coating’s job is to slow heat reaching the substrate. Bunsen-burner and cone-heating tests can measure ignition, flame spread and heat release of the coating film itself \u2014 and they matter \u2014 but none of them reproduces the question a structural fire design actually asks: how long does this protected member stay below its critical temperature in a compartment fire?<\/em> That question has three ingredients only a furnace can supply:<\/p>\n This is why coating approvals read like “1,110 g\/m\u00b2 dry film thickness gives R 60 on a section factor up to 250 m\u207b\u00b9” \u2014 each line of such a table traces back to furnace data (see the companion article on fire retardant coating testing methods and equipment<\/a> for the full method landscape, and on the difference between fire retardant and fire resistant coatings<\/a> for the vocabulary).<\/p>\n There is also a scaling argument. A coating film on a small plate can be heated uniformly by a cone or a radiant panel, but a structural section in a compartment fire receives heat from a furnace atmosphere whose temperature, emissivity and pressure all evolve together for an hour or more. Intumescent char grows, cracks and re-densifies over that timeline; its insulating value at minute 40 is not predictable from its behaviour in the first two. Furnace exposure is the only laboratory environment that forces the coating through that whole timeline under controlled, repeatable conditions \u2014 which is why every framework that assigns structural ratings, EN or ASTM, routes the final decision through a furnace.<\/p>\n The default exposure for building products is the cellulosic (standard) fire curve, defined in ISO 834-1<\/strong> and adopted for European fire resistance testing by EN 1363-1<\/strong>:<\/p>\n Working from that equation: about 842 \u00b0C at 30 minutes<\/strong>, 945 \u00b0C at 60 minutes<\/strong>, 1,006 \u00b0C at 90 minutes<\/strong> and 1,049 \u00b0C at 120 minutes<\/strong> [2]. The current edition is ISO 834-1:2025<\/strong>, in force since 30 May 2025, which replaced the 1999 edition the same day [2].<\/p>\n EN 1363-1:2020 adds the laboratory requirements around the curve \u2014 furnace pressure control, thermocouple types and placement, and the rules for recording failure [1]. A furnace does not simply follow the curve on paper: EN 1363-1 defines how closely the furnace control must track the target temperature and how deviations are recorded, which is why the furnace log is part of every coating test report. Laboratories comparing results across years should check which edition each historical campaign was run under, because the 2025 re-issue is recent enough that older reports predate it.<\/p>\n Where the fire risk is a fuel spill or a tunnel fire, coatings are evaluated against faster, hotter exposures. EN 1363-2<\/strong> defines the alternative and special curves \u2014 the hydrocarbon curve among them \u2014 that run far above the cellulosic values in the first minutes [3]. Coatings qualified for petrochemical or tunnel applications are therefore usually furnace-tested twice: once cellulosic, once hydrocarbon, because intumescent char behaviour differs sharply between the two regimes.<\/p>\n For structural coatings, the specimen is a steel section \u2014 commonly a column or beam representative of the product’s intended use \u2014 coated at the dry film thickness (DFT) under assessment. The section’s heated perimeter to cross-sectional area ratio (section factor, A\/V or Hp\/A)<\/strong> is recorded, because it is the variable the assessment will extrapolate: compact sections heat slowly and need less coating, slender sections heat quickly and need more 5<\/a>.<\/p>\n Specimen build is a quality gate in its own right. The coating must be applied with the manufacturer’s own system \u2014 primer, mesh if the datasheet requires it, and the declared number of coats \u2014 because the assessment covers the system<\/em>, not the intumescent layer alone. Application records, DFT survey maps and curing logs travel with the specimen into the furnace file; a test whose coating application cannot be reconstructed is not repeatable, and repeatability is the whole point of a standardised furnace.<\/p>\n Embedded and surface thermocouples track steel temperature at defined positions along the member; plate thermocouples in the furnace track the exposure; and for loaded tests, hydraulic actuators hold the design load throughout. Under EN 13381-8<\/strong>, the European assessment standard for reactive fire protection on steel members, the test data are processed into the relationship between coating thickness, section factor and achieved fire resistance \u2014 the format in which coating manufacturers publish their design tables [5]. The assessment framework sits within the structural fire design environment of the Eurocodes (EN 1992-1-2<\/strong> for concrete structures and its steel-side companions), which consume those ratings in design 6<\/a>.<\/p>\n For a loadbearing coated member the failure criterion is the R criterion<\/strong> \u2014 loss of loadbearing capacity as the steel reaches its critical temperature under load. (How R ratings are structured across steel protection routes is covered in our guide to fire resistance testing for steel structures<\/a>.) For separating assemblies coated with a protective product, E and I criteria apply instead, exactly as in any fire resistance test [1]. A coating test furnace therefore needs the same control and instrumentation chain as any fire resistance furnace; what distinguishes it is the specimen programme built around coating variables (DFT levels, section factors, application conditions).<\/p>\n Before a manufacturer commits to full-scale furnace work, screening tests rank formulations cheaply. The long-established tool is ASTM D3806<\/strong>, the Small-Scale Evaluation of Fire-Retardant Paints (2-Foot Tunnel Method)<\/em>, current as the 2019 edition [4]. A coated panel faces a small gas-fired tunnel flame, and the method reports intumescence and flame spread performance on a laboratory scale. It is deliberately not<\/strong> a fire resistance test: a paint that performs well in the 2-foot tunnel has earned a screening result, not an R, E or I rating \u2014 the furnace campaign is still ahead of it.<\/p>\n That division of labour is a feature, not a limitation. Formulation screening lives on iteration speed: dozens of variants, small specimens, same-day results. Furnace work is the opposite \u2014 few variants, expensive specimens, and days of furnace time per cell. Laboratories that push formulation questions into the furnace waste budget; laboratories that skip screening and go straight to the furnace discover their worst formulations at the highest possible price per data point. The two scales are complementary, and the specification for each should say which job it is doing.<\/p>\n\n
The Heating Curves the Furnace Runs<\/h2>\n
The cellulosic curve<\/h3>\n
T = 20 + 345 \u00d7 log10(8t + 1) (T in \u00b0C, t in minutes)\n<\/code><\/pre>\n![\"Fire](\"https:\/\/brt-testing.com\/wp-content\/uploads\/2026\/10\/fire-resistance-test-furnace.avif\")
Hydrocarbon and tunnel curves<\/h3>\n
What Happens Inside During a Coated-Section Test<\/h2>\n
The specimen<\/h3>\n
The instrumentation<\/h3>\n
The criteria<\/h3>\n
The Small-Scale Alternative: ASTM D3806<\/h2>\n
Choosing and Specifying a Coating Test Furnace<\/h2>\n
![\"Steel](\"https:\/\/brt-testing.com\/wp-content\/uploads\/2026\/09\/Steel-Fire-Retardant-Coating-Test-Furnace.jpg\")
<\/a>