Combustion testing of thermal insulation material specimens is a fundamental fire safety evaluation that measures the ignitability, flame spread, heat release, smoke production, and burning droplets behavior of insulation products used in buildings, industrial installations, and vehicles. Performed by accredited third-party laboratories in accordance with ISO, EN, ASTM, UL, and GB standards, these tests assign materials such as mineral wool, EPS, XPS, PUR/PIR, and phenolic foam to standardized fire performance grades. This article explains the specimen requirements, key testing items, standard test methods, and typical applications of insulation combustion performance testing.
Technical Overview
Thermal insulation materials are installed over large surfaces, in concealed cavities, and throughout building envelopes, which means their burning behavior directly influences fire growth, smoke development, and occupant safety. Lightweight polymer foams such as expanded polystyrene (EPS), extruded polystyrene (XPS), and polyurethane (PUR) are inherently combustible, while fibrous inorganic products such as rock wool and glass wool are essentially non-combustible. Because of these differences, most building codes and product certification schemes require laboratory combustion testing of representative specimens before insulation products can be placed on the market.
Combustion testing of insulation specimens subjects conditioned samples to controlled ignition sources, radiant heat fluxes, or full-scale burning conditions under standard atmospheres. Measured parameters include furnace temperature rise, mass loss, after-flame time, flame spread distance, heat release rate, smoke density, and the production of flaming droplets. The results are compared against classification thresholds defined in standards such as GB 8624 (the Chinese classification standard for burning behavior of building materials, harmonized with the EN 13501-1 hierarchy) and ASTM E84 or UL 723 for North American markets.
Under GB 8624-2012, insulation products are graded into four regulatory levels with sub-grades, broadly aligning with the European A1/A2/B/C/D/E/F system:
- Class A (A1, A2): non-combustible or limited-combustibility materials, typically verified through GB/T 5464 (ISO 1182) and GB/T 20284 (EN 13823).
- Class B1 (B, C, D): flame-retardant materials with restricted contribution to fire, verified through GB/T 20284 and GB/T 8626 (ISO 11925-2).
- Class B2 (E): combustible materials with limited ignitability under a small flame source, verified through GB/T 8626.
- Class B3 (F): readily ignitable materials that fail Class B2 criteria and are generally prohibited for load-bearing building applications.
Testing Samples
Inorganic Fibrous Insulation
Rock (mineral) wool, glass wool, and ceramic fiber boards, blankets, and mats are tested primarily to confirm their non-combustible classification. Specimens are cut from finished boards at the delivered density and thickness, since furnace temperature rise and mass loss criteria in the non-combustibility test must reflect the commercial product.
Organic Foam Insulation
EPS, XPS, PUR, PIR (polyisocyanurate), and phenolic foam specimens are tested for ignitability, flame spread, heat release, and dripping behavior. Because flame-retardant additives and facings significantly alter performance, samples must be taken from production batches in their end-use configuration, including any bonded facer layers.
Composite, Faced, and Coated Products
Aluminum-foil-faced boards, gypsum-sheathed panels, and metal-faced sandwich panel cores are tested as complete products. Standard classification rules require that the specimen represent the product as installed, since delamination of facings and adhesive failure frequently govern the observed burning behavior rather than the core insulation itself.
Specimen Preparation and Conditioning
Prior to testing, specimens are cut to the dimensions prescribed by each method, sealed at edges where required, and conditioned to constant mass in a standard atmosphere, typically 23 ± 2 °C and 50 ± 5% relative humidity in accordance with ISO 291 or GB/T 2918. Most methods require multiple replicate specimens (commonly three to six per condition) so that the worst result governs the classification.
Testing Items
Depending on the target classification and end use, a combustion test program for insulation specimens may include the following measurement items:
- Furnace temperature rise and sustained flaming duration in the non-combustibility test.
- Mass loss on ignition, expressed as a percentage of the initial specimen mass.
- Time to ignition and after-flame time under a small flame source.
- Flame spread distance and lateral flame spread (LFS) along the specimen surface.
- Fire growth rate index (FIGRA) and total heat release within 600 seconds (THR600s).
- Smoke growth rate index (SMOGRA), total smoke production (TSP600s), and smoke density rating by light attenuation.
- Formation of flaming droplets and particles that ignite filter paper.
- Gross heat of combustion (PCS) determined by bomb calorimetry.
- Limiting oxygen index (LOI) as an indicator of intrinsic flammability.
- Peak and total heat release rates, effective heat of combustion, and mass loss rate by cone calorimetry.
- Yields of fire gases such as carbon monoxide, carbon dioxide, hydrogen cyanide, and hydrogen chloride.
Testing Methods
Non-Combustibility Test (GB/T 5464 / ISO 1182 / ASTM E136)
A cylindrical specimen, typically 45 mm in diameter and 50 mm high, is inserted into a furnace stabilized at 750 °C for 30 minutes while thermocouples record the furnace, specimen surface, and specimen center temperatures. Classification as non-combustible generally requires that the mean temperature rise does not exceed 30 °C, mass loss does not exceed 50%, and no sustained flaming of 20 seconds or longer occurs. This method underpins the Class A1 and A2 grades for mineral wool and similar inorganic insulation products.
Single-Flame-Source Ignitability Test (GB/T 8626 / ISO 11925-2)
Rectangular specimens, typically 250 mm by 90 mm, are clamped vertically and exposed to a small propane diffusion flame applied to the bottom edge for 15 or 30 seconds. The specimen passes if the flame tip does not reach the 150 mm gauge mark within the specified period, flaming droplets do not ignite the filter paper placed beneath the specimen, and after-flame times remain within the classification limits. This screening test establishes the B2 (E) grade and forms part of the evidence for higher grades.
Single Burning Item Test (GB/T 20284 / EN 13823)
The SBI test exposes a corner assembly formed by a long wing and a short wing, approximately 1.5 m in height, to a 30 kW propane burner for about 20 minutes inside an exhaust hood instrumented with oxygen consumption gas analysis and laser smoke photometry. The method yields FIGRA, THR600s, LFS, SMOGRA, and TSP600s. Representative GB 8624-2012 thresholds include FIGRA0.2MJ of no more than 120 W/s and THR600s of no more than 7.5 MJ for grade B, and FIGRA of no more than 750 W/s for grade D, with additional smoke and droplet sub-criteria. This test is the cornerstone of B1 classification for foam insulation such as flame-retarded EPS and XPS.
Smoke Density Test (GB/T 8627)
Small specimens are exposed to radiant heating with or without pilot ignition inside a closed chamber, and the attenuation of a light beam passing through the accumulated smoke is converted into a smoke density rating. This method is applied where smoke obscuration criteria supplement the SBI results or where product specifications impose explicit smoke limits.
Cone Calorimeter Test (ISO 5660-1 / GB/T 16172)
A 100 mm by 100 mm horizontal specimen is exposed to a radiant flux, commonly 25 to 50 kW/m², with spark ignition. Oxygen consumption calorimetry quantifies the heat release rate, total heat release, effective heat of combustion, mass loss rate, smoke production rate, and yields of CO and CO2. Cone data are widely used for research, flame-retardant formulation development, and fire safety engineering input, including predictions of full-scale SBI performance.
LIMITING Oxygen Index Test (GB/T 2406.2 / ISO 4589-2 / ASTM D2863)
A candle-like specimen burns vertically in a glass chimney with a flowing nitrogen-oxygen mixture. The LOI is the minimum oxygen concentration, in volume percent, that just supports flaming combustion for the specified time or length. Typical values range from about 18 to 22% for untreated EPS, above 30% for PIR, and above 35% for phenolic foams, providing a sensitive comparative measure of flame retardancy.
Steiner Tunnel Test (ASTM E84 / UL 723)
For North American acceptance, insulation specimens are mounted on the ceiling of a 7.6 m tunnel and exposed to a controlled air-gas flame for 10 minutes. Flame spread and smoke developed indices are normalized against red oak flooring and inorganic board references, yielding Class A (index 25 or less), Class B (26 to 75), or Class C (76 to 200) surface burning characteristics, with smoke developed typically required not to exceed 450.
Small-Scale Component Flame Tests (UL 94)
Where insulation forms part of electrical equipment, appliances, or transport interiors, UL 94 vertical and horizontal burning tests classify specimens from HB through V-2, V-1, V-0, and 5VA/5VB based on after-flame times, afterglow, and flaming drips that ignite cotton. Heat of combustion may additionally be measured by bomb calorimetry per GB/T 14402 (ISO 1716) to support A1/A2 classification of homogeneous products.
Applications
Combustion test reports for insulation specimens are used throughout the product lifecycle by manufacturers, designers, regulators, and accredited third-party laboratories:
- Mandatory building code compliance for external wall insulation systems, curtain wall cavities, and roof assemblies where B1 or Class A materials are prescribed.
- Product certification and type approval for sandwich panels, faced boards, and insulated ducting.
- Factory production control, incoming material verification, and periodic surveillance testing of flame-retardant grades.
- Research and development of flame retardants, facings, and core chemistry, supported by LOI and cone calorimeter screening.
- Fire safety engineering, where heat release and smoke data feed zone and computational fluid dynamics fire models.
- Transport applications, including rail vehicle components tested to EN 45545-2 and marine installations evaluated under international fire test procedure codes.
- Dispute resolution and post-incident fire investigation requiring comparative burning behavior evidence.
FAQ
What is the difference between classes A, B1, B2, and B3 under GB 8624?
Class A denotes non-combustible or limited-combustibility materials verified mainly by furnace and heat of combustion tests; B1 denotes flame-retardant materials with restricted heat release and flame spread verified by the SBI test; B2 denotes combustible materials that resist a small ignition flame; and B3 denotes readily ignitable materials that fail the B2 criteria. The sub-grades A1, A2, B, C, D, and E align the Chinese system with the EN 13501-1 hierarchy.
Can EPS or XPS foam achieve a B1 classification?
Yes. Flame-retarded EPS and XPS formulations incorporating brominated or phosphorus flame retardants can pass the 30-second ignitability test and meet the FIGRA and THR600s thresholds of the SBI test, subject to regular surveillance testing. Untreated foams typically fall into B2 or B3 and require protective facings or barriers in end-use applications.
How many specimens are required for one determination?
Each method prescribes its own replication: commonly three to six conditioned specimens per test condition, with the least favorable result governing the classification. Additional spare specimens are recommended because specimen preparation quality and edge conditioning strongly influence flame spread results.
Is cone calorimeter data equivalent to SBI results for classification purposes?
No. Cone calorimeter results are bench-scale indicators used for research, screening, and modeling, whereas GB/T 20284 (EN 13823) SBI results are the prescribed basis for B1 grading under GB 8624 and EN 13501-1. Correlations exist, but classification must rely on the mandated methods.
What sample information should be submitted with specimens?
Laboratories typically require product name, nominal density and thickness, facing or coating configuration, production batch, intended application, and the target classification. Testing must be performed on the product in its end-use configuration, so any facings, adhesives, or laminations should remain intact on the submitted specimens.
How should specimens be conditioned before testing?
Specimens are conditioned to constant mass at 23 ± 2 °C and 50 ± 5% relative humidity in accordance with ISO 291 or GB/T 2918, unless the fire test standard specifies otherwise. Moisture content measurably affects ignitability and smoke production, so conditioning is a mandatory part of a valid determination.