Construction Concerns: Materials Testing

By Greg Havel

One of the most common tests cited in the technical bulletins on building materials is ASTM E84 Standard Method of Test of Surface Burning Characteristics of Building Materials, which classifies a building material as Class A, B, or C depending on its rate of flame spread.

ASTM E84, UL 723, NFPA 255, and ULC S102 Standard Method of Test of Surface Burning Characteristics of Building Materials describe the same test procedure, using an apparatus called a Steiner Tunnel (Photo 1). This apparatus was first developed by A. J. Steiner at Underwriters Laboratories in 1922.

Steiner Tunnel

Photo 1. Steiner Tunnel: By Herbert Blenstein (Own work) [GFDL ( ) or CC BY-SA 3.0 ( ], via Wikimedia Commons

The Steiner Tunnel is 25 feet (7.62 m) long, not including the air intake and gas burner assembly at one end, and the vent pipe connection at the other. It has an internal cross section 17-5/8 inches wide by 12 inches high (448 by 305 mm), inside its lining of refractory fire brick. The entire top of the tunnel is removable in one piece, so that the test sample can be attached to it. The gas flames are 4.5 feet (1.37 meters) long, and calibrated to 5,000 British Thermal Units (BTU) (88 KW), and produce a temperature of 1,600ºF (900ºC).

The test is conducted with the sample attached to the underside of the horizontal removable top of the tunnel, after calibration with test samples of red oak (flame spread and smoke developed = 100) and inorganic cement board (flame spread and smoke developed = 0). When the gas burner at one end of the tunnel is ignited, the flame spread is observed and timed as it progresses along the surface toward the vented end of the tunnel. The speed of flame spread is compared to that of the cement board and red oak samples, to determine the flame spread rating for the material. The amount of smoke developed is determined from readings of the photoelectric cell at the vent end of the tunnel, and compared to that of the cement board and red oak.

The test duration is 10 minutes, unless the test panel is completely consumed, and there is no further progression of the flame spread, and the photoelectric cell readings have returned to baseline. Under these circumstances, the test may be concluded in less than 10 minutes and the gas burners turned off.

The classification of materials tested, compared to the standard of red oak, is usually stated as:

  • Class A: Flame spread 0-25; smoke developed 0-450
  • Class B: Flame spread 26-75; smoke developed 0-450
  • Class C: Flame spread 76-200; smoke developed 0-450

The ASTM E84 test is the one most frequently cited by building product manufacturers for flame spread. However, this test is for building materials in a horizontal configuration, like acoustic ceiling tile. It was never intended to be the authority on building materials in a vertical configuration. We have all seen or heard about how rapidly flame spreads on carpeting when it is used as a wall covering. (Photo 2)


Carpeting as wall cover

Photo 2. Photo by author.

NFPA 285 is the Standard Test Method for Evaluation of Fire Propagation Characteristics of Exterior Non-Loadbearing Wall Assemblies Containing Combustible Components, 2012 Edition. (According to manufacturer literature on the aluminum composite panels used at Grenfell Tower in London, England, the polyethylene (PE) panels used were untested; but the more expensive fire-resistant (FR) products that were not used did pass this test.)

The NFPA 285 test is not a test of flame spread over the surface of a material, but a test of the performance of an entire wall assembly, based on a fire in the first-floor room that is extending out of the window.

The NFPA 285 test facility must be at least 30 feet by 30 feet by 23 feet high (9.1 by 9.1 by 7 m), containing a test apparatus 15 feet 8 inches (4.8 m) high, with a test room on each of two stories. Each room must be 10 feet by 10 feet by 7 feet high (3.05 by 3.05 by 2.13 m), with concrete floors and three permanent walls of concrete or concrete masonry units, protected by layers of Type X gypsum drywall board. The fourth wall opening is designed to be closed with a removable test frame holding the wall sample that will be tested.

One gas burner is placed in the first-floor test room, while the second burner is placed near the top of the window opening in the first floor of the test specimen. Each gas burner is calibrated to flow a specified quantity of gas, producing the specified amount of heat expressed in kilowatts and British thermal units (BTU), which are increased at the intervals according to Table 4.4.13 in NFPA 285.

As this procedure tests the performance of an entire wall assembly, it requires thermocouples to measure temperature on both sides of the test specimen, within the cavities of the test specimen, and at other locations inside the test facility rooms. Heat flux gauges are also required. Before each test, the gas burners, thermocouples, and heat flux gauges must be calibrated according to the procedure outlined in NFPA 285.

The first-floor room gas burner is turned on first. The window gas burner is turned on five minutes later. Thirty minutes after the first-floor room burner is turned on, both burners are turned off, and residual burning is allowed to continue for at least 10 minutes, unless it is determined that immediate extinguishment is needed for safe conditions in the test facility. During the entire procedure, both still and video photos are taken, with elapsed timers on the screens.

After the test specimen has cooled, it is dismantled and examined with the photos and videos. The wall assembly is considered to have passed the test if these conditions are met during the 30-minute fire exposure:

  • No flame propagation on the surface beyond the area of flame impingement from the window burner; and
  • No vertical flame propagation inside the test specimen, its insulation, or within its cavities; and
  • No horizontal flame propagation inside the test specimen, its insulation, or within its cavities; and
  • No temperature readings in the second-floor test room more than 500ºF (278ºC) above the ambient temperature at the beginning of the test; and
  • No flames occurred in the second-floor test room; and
  • No flames occurred beyond the intersection of the test specimen and the side walls of the test rooms.

Each of the above bullet points has several additional parameters attached to it in the standard. The test report is in the format specified in NFPA 285, with supporting documents.

Please note that in the real world, a fire may not behave exactly as it did during the laboratory test because of the differing construction materials and methods used, wind speed and direction, and other factors.

Fire service personnel and code enforcement officials can no longer afford to let themselves be fooled by the citation of an ASTM or NFPA standard for flame spread. We must learn to recognize the parameters of each of these tests, and to recognize when the information supplied by a building materials manufacturer does not correspond to its intended use.

A citation of ASTM E84 or UL 723 for a material that will be used as a vertical cladding or interior finish is not valid for that material, since this test is for flame spread in a horizontal configuration.

It may be argued that obtaining a positive test result using the NFPA 285 test will be costly, and delay the construction of many buildings in the future, since there are so many variations in framing, insulation, cavities, and cladding. If that is true, we should be willing to live with that consequence; or use proven traditional non-combustible materials. One Grenfell Tower incident should be enough for all of us.

In the meantime, perhaps the people at ASTM, UL, and FM Global can develop a less expensive test for vertical flame spread, perhaps adapting a Steiner tunnel in a vertical orientation, with what is now the top as a removable wall; the same calibration standards of inorganic cement board and red oak; and shutting off the gas burners after a specified time interval, or after vertical flame spread is established on a specified percentage of the test sample. Although it would not be as definitive as the NFPA 285 test, it would at least provide information on which materials had a rapid self-supporting flame spread in a vertical configuration, and which could be depended upon to self-extinguish. And the test apparatus should be capable of accepting a test specimen both in direct contact with the removable side of the test apparatus, and with a cavity behind it like the cladding materials at Grenfell Tower. Once this test method is developed, it should be required by building and fire codes for all materials in addition to the ASTM E84 test that is already in common use.

The simple fact that we already have so many buildings with the potential to become another Grenfell Tower should not stop us from preventing the addition of more buildings to that number. We owe this to the future of the fire service, and to our society. Life is fragile at best; and when we allow building construction that is a recipe for disaster, we make it cheap besides.

For more detailed information on the building material tests described above, visit, and log-in for free read-only access; or use the log-in from your fire department’s or technical college’s National Fire Code subscription. Look for NFPA  255 and 285.

Download this article as a PDF HERE (213 KB)


Gregory HavelGregory Havel is a member of the Town of Burlington (WI) Fire Department; retired deputy chief and training officer; and a 35-year veteran of the fire service. He is a Wisconsin-certified fire instructor II, fire officer II, and fire inspector; an adjunct instructor in fire service programs at Gateway Technical College; and safety director for Scherrer Construction Co., Inc. Havel has a bachelor’s degree from St. Norbert College; has more than 35 years of experience in facilities management and building construction; and has presented classes at FDIC.



Combustible Cladding

Structural Steel Lintels


Wood and Concrete Composites

Composite Materials and Assemblies

No posts to display