Designation: E 84 – 00a An American National Standard

Standard Test Method for

Surface Burning Characteristics of Building Materials1
This standard is issued under the fixed designation E 84; the number immediately following the designation indicates the year of original
adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A superscript
epsilon (e) indicates an editorial change since the last revision or reapproval.

This standard has been approved for use by agencies of the Department of Defense.

1. Scope assessment of the materials, products, or assemblies under

1.1 This fire-test-response standard for the comparative actual fire conditions..
surface burning behavior of building materials is applicable to 1.8 This standard does not purport to address all of the
exposed surfaces such as walls and ceilings. The test is safety concerns, if any, associated with its use. It is the
conducted with the specimen in the ceiling position with the responsibility of the user of this standard to establish appro-
surface to be evaluated exposed face down to the ignition priate safety and health practices and determine the applica-
source. The material, product, or assembly shall be capable of bility of regulatory limitations prior to use.
being mounted in the test position during the test. Thus, the 2. Referenced Documents
specimen shall either be self-supporting by its own structural
quality, held in place by added supports along the test surface, 2.1 ASTM Standards:
or secured from the back side. A 390 Specification for Zinc-Coated (Galvanized) Steel
1.2 The purpose of this test method is to determine the Poultry Fence Fabric (Hexagonal and Straight Line)2
relative burning behavior of the material by observing the C 36 Specification for Gypsum Wallboard3
flame spread along the specimen. Flame spread and smoke D 4442 Test Methods for Direct Moisture Content Measure-
developed index are reported. However, there is not necessarily ment of Wood and Wood-Base Materials4
a relationship between these two measurements. D 4444 Test Methods for Use and Calibration of Hand-Held
1.3 The use of supporting materials on the underside of the Moisture Meters4
test specimen has the ability to lower the flame spread index E 69 Test Method for Combustible Properties of Treated
from those which might be obtained if the specimen could be Wood by the Fire-Tube Apparatus4
tested without such support. These test results do not neces- E 160 Test Method for Combustible Properties of Treated
sarily relate to indices obtained by testing materials without Wood by the Crib Test5
such support. E 162 Test Method for Surface Flammability of Materials
1.4 Testing of materials that melt, drip, or delaminate to Using a Radiant Heat Energy Source6
such a degree that the continuity of the flame front is destroyed, E 176 Terminology of Fire Standards6
results in low flame spread indices that do not relate directly to E 286 Test Method for Surface Flammability of Building
indices obtained by testing materials that remain in place. Materials Using an 8-ft (2.44-m) Tunnel Furnace7
1.5 The values stated in inch-pound units are to be regarded 3. Terminology
as the standard.
1.6 The text of this standard references notes and footnotes 3.1 Definitions—For definitions of terms used in this test
that provide explanatory information. These notes and foot- method refer to Terminology E 176. The term flame spread
notes, excluding those in tables and figures, shall not be index from Terminology E 176 is of particular interest to this
considered as requirements of the standard. standard and is defined in 3.1.1.
1.7 This standard is used to measure and describe the 3.1.1 flame spread index, n—a number or classification
response of materials, products, or assemblies to heat and indicating a comparative measure derived from observations
flame under controlled conditions, but does not by itself made during the progress of the boundary of a zone of flame
incorporate all factors required for fire-hazard or fire-risk under defined test conditions.

2
Annual Book of ASTM Standards, Vol 01.06.
1 3
This test method is under the jurisdiction of ASTM Committee E-5 on Fire Annual Book of ASTM Standards, Vol 04.01.
4
Standards and is the direct responsibility of Subcommittee E05.22 on Surface Annual Book of ASTM Standards, Vol 04.10.
5
Burning. Discontinued; see 1993 Annual Book of ASTM Standards, Vol 04.09.
6
Current edition approved April 10, 2000. Published May 2000. Originally Annual Book of ASTM Standards, Vol 04.07.
7
published as E 84 – 50 T. Last previous edition E 84 – 00. Discontinued; see 1991 Annual Book of ASTM Standards, Vol 04.07.

Copyright © ASTM, 100 Barr Harbor Drive, West Conshohocken, PA 19428-2959, United States.

1
 E 84
3.2 Definitions of Terms Specific to This Standard: Width: 17 3⁄4 6 1⁄4 in. (451 6 6.3 mm) measured between the top
3.2.1 smoke developed index, n—a number or classification ledges along the side walls, and 17 5⁄8 6 3⁄8in. (448 6 10
mm) at all other points.
indicating a comparative measure derived from smoke obscu- Depth: 12 6 1⁄2 in. (305 6 13 mm) measured from the bottom of
ration data collected during the test for surface burning the test chamber to the top of the ledges on which the
characteristics. specimen is supported. This measurement includes the 1⁄8
in. (3.2 mm) thickness of the 1 1⁄2 in. (38 mm) wide woven
3.2.2 surface flame spread, n—the propagation of a flame fiberglass gasket tape.
away from the source of ignition across the surface of the Length: 25 ft 6 3 in. (7.62 m 6 76 mm).
specimen. 5.1.2 The sides and base of the chamber shall be lined with
4. Significance and Use an insulating firebrick with the dimensions of 4 1⁄2 in. by 9 in.
by 2 1⁄2 in. thick as illustrated in Fig. 2. The insulating firebrick
4.1 This test method is intended to provide only compara- shall have the following properties:
tive measurements of surface flame spread and smoke density Maximum Recommended Temperature 2600°F (1424°C)
measurements with that of select grade red oak and inorganic Bulk Density 50 6 3 lb/ft3 (0.77 6 0.046 g/cm3)
reinforced cement board surfaces under the specific fire expo- Thermal Conductivity at Mean
Btu•in./hr•ft2•°F W/m•°C
Temperature of
sure conditions described herein. 400°F (205°C) 1.7 0.24
4.2 This test method exposes a nominal 24-ft (7.32-m) long 800°F (425°C) 1.9 0.27
by 20-in. (508-mm) wide specimen to a controlled air flow and 1200°F (650°C) 2.2 0.32
1600°F (870°C) 2.6 0.37
flaming fire exposure adjusted to spread the flame along the 2000°F (1095°C) 3.2 0.46
entire length of the select grade red oak specimen in 51⁄2min. 2400°F (1315°C) 3.9 0.56
4.3 This test method does not provide for the following:
5.1.3 One side of the chamber shall be provided with double
4.3.1 Measurement of heat transmission through the tested
observation windows8 with the inside pane flush mounted (see
surface.
Fig. 2). Exposed inside glass shall be 2 3⁄4 6 3⁄8 by 11 + 1, −2
4.3.2 The effect of aggravated flame spread behavior of an
in. (70 6 10 by 279 + 25 − 50 mm). The centerline of the
assembly resulting from the proximity of combustible walls
exposed area of the inside glass shall be in the upper half of the
and ceilings.
furnace wall, with the upper edge not less than 2.5 in. (63 mm)
4.3.3 Classifying or defining a material as noncombustible,
below the furnace ledge. The window shall be located such that
by means of a flame spread index by itself.
not less than 12 in. (305 mm) of the specimen width can be
5. Apparatus
5.1 Fire Test Chamber—See Figs. 1-5. 8
Heat-resistant glass, high-silica, 100 % silica glass, nominal 1⁄4-in. thick has
5.1.1 The fire test chamber is a rectangular horizontal duct been found suitable for the interior pane. Borosilicate glass, nominal 1⁄4-in. thick has
with a removable lid. The inside dimensions are as follows: been found suitable for the exterior pane.

FIG. 1 Test Furnace, Showing Some Critical Dimensions (Not a Construction Drawing)

2
 E 84

FIG. 2 Test Furnace Showing Critical Dimensions (Not a Construction Drawing)

observed. Multiple windows shall be located along the tunnel 5.1.5.2 The lid shall be insulated with a minimal thickness
so that the entire length of the test sample is observable from of 2 in. (51 mm) castable insulation or mineral composite
outside the fire chamber. The windows shall be pressure tight material having physical characteristics comparable to the
in accordance with 7.2 and 7.2.1. following:
5.1.4 The ledges shall be fabricated of structural materials9 Maximum effective use temperature of
capable of withstanding the abuse of continuous testing. The at least: 1200°F (650°C)
Bulk density 21 lb/ft3 (336 kg/m3)
ledges shall be level with respect to the length and width of the Thermal conductivity at 300 to 700°F 0.50 to 0.71 Btu·in./h·ft2·°F (0.072 to
chamber and each other. The ledges shall be maintained in a (149 to 371°C) 0.102 W/m·K)
state of repair commensurate with the frequency, volume, and
5.1.5.3 The entire lid assembly shall be protected with flat
severity of testing occurring at any time.
sections of high-density (nominal 110 6 10 lb/ft3 or 1761 6
5.1.5 Lid:
160 kg/m3) 1⁄4-in. (6.3-mm) inorganic reinforced cement board.
5.1.5.1 The lid shall consist of a removable noncombustible
This protective board shall be maintained in an unwarped and
metal and mineral composite structure as shown in Fig. 2 and
uncracked condition through continued replacement. The pro-
of a size necessary to cover completely the fire test chamber
tective board is to be secured to the furnace lid or placed on the
and the test samples. The lid shall be maintained in an
back side of the test specimen.
unwarped and flat condition. When in place, the lid shall be
completely sealed to prevent air leakage into the fire test 5.1.6 Gas Burners:
chamber during the test. 5.1.6.1 One end of the test chamber shall be designated as
the “fire end”. This fire end shall be provided with two gas
burners delivering flames upward against the surface of the test
9
High-temperature furnace refractory. Zirconium silicate, or water-cooled steel sample (see Fig. 2). The burners shall be spaced 12 in. (305
tubing have been found suitable for this purpose. mm) from the fire end of the test chamber, and 7 1⁄26 1⁄2 in.

3
 E 84

FIG. 3 Typical Exhaust End Transition (Not a Construction Drawing)

(190 6 13 mm) below the under surface of the test sample. Gas shutter shall be positioned so as to provide an air inlet port 3 6
to the burners shall be provided through a single inlet pipe, ⁄ in. (76 6 2 mm) high measured from the floor level of the
1 16

distributed to each port burner through a tee-section. The outlet test chamber at the air intake point.
shall be a 3⁄4 in. NPT elbow. The plane of the port shall be 5.1.7.2 To provide air turbulance for proper combustion,
parallel to the furnace floor, such that the gas is directed turbulance baffling shall be provided by positioning six refrac-
upward toward the specimen. Each port shall be positioned tory firebricks (as defined in 5.1.2) along the side walls of the
with its centerline 4 6 1⁄2 in. (102 6 13 mm) on each side of chamber. With the long dimension vertical, 4 1⁄2in. (114-mm)
the centerline of the furnace so that the flame is distributed dimension along the wall, place the bricks as follows from the
evenly over the width of the exposed specimen surface (see centerline of the burner ports:
Fig. 2). On the window side at 7, 12, and 20 6 1⁄2 ft (2.1, 3.7, and 6.1 6 0.2 m)
5.1.6.2 The controls used to assure constant flow of gas to On the opposite side at 4 1⁄2, 9 1⁄2, and 16 6 1⁄2 ft (1.3, 2.9, and 4.9 6 0.2 m)
the burners during period of use shall consist of a pressure
regulator, a gas meter calibrated to read in increments of not 5.1.7.3 The movement of air shall be by an induced draft
more than 0.1 ft3 (2.8 L), a manometer to indicate gas pressure system having a total draft capacity of at least 0.15 in. (3.8
in inches of water, a quick-acting gas shut-off valve, and a gas mm) water column with the sample in place, the shutter at the
metering valve. fire end open the normal 3 6 1⁄16 in. (76 6 2 mm), and the
5.1.7 Air Intake: damper in the wide open position. A draft gage tap to indicate
5.1.7.1 An air intake shutter shall be located 54 6 5 in. static pressure shall be inserted through the top at the midwidth
(1372 6 127 mm) upstream of the burner, as measured from of the tunnel, 1 6 0.5 in. (25 6 12 mm) below the ceiling, 15
the burner centerline to the outside surface of the shutter (see 6 0.5 in. (381 6 12 mm) downstream from the inlet shutter
Fig. 1). The air intake is to be fitted with a vertically sliding (see Fig. 1).
shutter extending the entire width of the test chamber. The 5.1.8 Exhaust End:

4
 E 84
in.2 (1290 cm2) at any point (see Fig. 3).
5.1.8.2 The transition piece shall in turn be fitted to a 16 in.
(406 mm) diameter duct pipe. A typical duct system shown in
Fig. 4 contains two 90° elbows (see Fig. 5) with the exhaust
duct running beside the fire test chamber. In order to comply
with this typical design, the vertical centerline of the exhaust
duct system is identical to that of the fire test chamber.
5.1.8.3 The exhaust duct is to be insulated with at least 2 in.
(51 mm) of high temperature mineral composition material
from the exhaust end of the fire chamber to the photometer
location.
5.1.8.4 An exhaust fan shall be installed at the end of the
exhaust duct. The air flow shall be controlled as specified in
5.1.11.
5.1.8.5 An alternative exhaust duct layout design shall
demonstrate equivalency by meeting the requirements speci-
fied in Section 7.
5.1.9 Photometer System:
5.1.9.1 A photometer system consisting of a lamp10 and
photocell11 shall be mounted on a horizontal section of the
16-in. (406-mm) diameter vent pipe at a point where it will be
preceded by a straight run of pipe (at least 12 diameters or 16
ft (4.88 m) and not more than 30 diameters or 40 ft (12.19 m))
from the vent end of the chamber, and with the light beam
directed upward along the vertical axis of the vent pipe. The
vent pipe shall be insulated with at least 2 in. (51 mm) of
high-temperature mineral composition material, from the vent
end of the chamber to the photometer location. The photoelec-
tric cell of which the output is directly proportional to the
amount of light received shall be mounted over the light source
and connected to a recording device having a minimum
operating chart width of 5 in. (127 mm) with an accuracy
within 61 % of full scale, for indicating changes in the
attenuation of incident light by the passing smoke, particulate,
and other effluent. The distance between the light source lens
and the photocell lens shall be 36 6 4 in. (914 6 102 mm). The
cylindrical light beam shall pass through 3-in. (76-mm) diam-
eter openings at the top and bottom of the 16-in. diameter duct,
with the resultant light beam centered on the photocell.
5.1.9.2 Linearity of the photometer system shall be verified
periodically by interrupting the light beam with calibrated
neutral density filters. The filters shall cover the full range of
the recording instrument. Transmittance values measured by
the photometer, using neutral density filters, shall be within
63 % of the calibrated value for each filter.
5.1.10 Draft Regulating Device:
5.1.10.1 An automatically controlled damper to regulate the
draft pressure shall be installed in the vent pipe down-stream of

10
The sole source of supply of the apparatus known to the committee at this time
is 12-V sealed beam, clear lens, auto spot lamp, No. 4405, from General Electric,
FIG. 4 Plan View—Typical Duct System (Not a Construction
Nela Park, OH. If you are aware of alternative suppliers, please provide this
Drawing) information to ASTM Headquarters. Your comments will receive careful consider-
ation at a meeting of the responsible technical committee,1 which you may attend.
11
5.1.8.1 The other end of the test chamber is designated as The sole source of supply of the apparatus known to the committee at this time
is No. 856BB from Weston Instruments, Wauconda, IL. If you are aware of
the exhaust end. The exhaust end shall be fitted with a gradual alternative suppliers, please provide this information to ASTM Headquarters. Your
rectangular-to-round transition piece, not less than 20 in. (508 comments will receive careful consideration at a meeting of the responsible
mm) in length, with a cross-sectional area of not less than 200 technical committee,1 which you may attend.

5
 E 84

FIG. 5 Typical Duct Elbow (Not a Construction Drawing)

the smoke-indicating attachment. The damper shall be pro- 6.2 The upstream end of the fire test chamber shall be filled
vided with a manual override. with a 14 6 1⁄8-in. (356 6 3 mm) length of uncoated 16-gauge
5.1.10.2 Other manual or automatic draft regulation de- (0.053 to 0.060 in.) steel sheet placed on the specimen
vices, or both, are allowed to be incorporated to help maintain mounting ledge in front of and under the leading edge of the
fan characterization and air-flow control throughout the test. specimen.
5.1.11 Thermocouples: 6.3 The test specimen shall be conditioned to a constant
5.1.11.1 A No. 18 Awg (1.02-mm) thermocouple, with 3⁄8 6 weight at a temperature of 73.4 6 5°F (23 6 2.8°C) and at a
1⁄8 in. (9.5 6 3.2 mm) of the junction exposed in the air, shall relative humidity of 50 6 5 %.
be inserted through the floor of the test chamber so that the tip
7. Calibration
is 1 6 1⁄32in. (25.4 6 0.8 mm) below the top surface of the
gasketing tape and 23 ft 6 1⁄2 in. (7.0 m 6 13 mm) from the 7.1 Place a 1⁄4-in. (6.3-mm) inorganic reinforced cement
centerline of the burner ports at the center of its width. board on the ledge of the furnace chamber, then place the
5.1.11.2 Two No. 18 Awg (1.02 mm) thermocouples are removable lid of the test chamber in position.
embedded below the floor surface of the test chamber. These 7.2 With the 1⁄4-in. (6.3-mm) inorganic reinforced cement
thermocouples shall be mounted at distances of 13 ft 6 1⁄2 in. board in position on top of the ledge of the furnace chamber
(3.96 m 6 13 mm) and 23 1⁄4 ft 6 1⁄2in. (7.09 m 6 13 mm) and with the removable lid in place, establish a draft to produce
measured from the centerline of the burner ports. The thermo- a 0.15-in. (3.8-mm) water-column reading on the draft manom-
couples shall be inserted from below the fire test chamber eter, with the fire-end shutter open 3 6 1⁄16 in. (76 6 1.5 mm),
through the firebrick until the tip of the thermocouple is 1⁄8 6 by manually setting the damper as a characterization of fan
1⁄32in. (3.2 6 0.8 mm) below the floor surface. The tip of the performance. Then close and seal the fire-end shutter, without
thermocouples shall be covered with refractory or portland changing the damper position. The manometer reading shall
cement, carefully dried to avoid cracking. increase to at least 0.375 in. (9.53 mm), indicating that no
excessive air leakage exists.
6. Test Specimens 7.2.1 In addition, conduct a supplemental leakage test peri-
6.1 The test specimen shall be at least 2 in. (51 mm) wider odically with the tunnel sealed from the inlet end to beyond the
(nominally 201⁄4 6 3⁄4 in. (514 6 19 mm)) than the interior photometer system, by placing a smoke bomb in the chamber.
width of the tunnel. The length of the specimen shall be a total Ignite the bomb and pressurize the chamber to 0.375 6 0.125
of 24 ft6 1⁄2 in. (7.32 m 6 13 mm) and shall be butted against in. (9.53 6 3.18 mm) water column. Seal all points of leakage
the vent end of the fire test chamber. The specimen consists of observed in the form of escaping smoke particles.
a continuous, unbroken length, or of sections joined end-to- 7.3 Establish a draft reading within the range 0.055 to 0.100
end. Specimens shall be truly representative of the materials in. (1.40 to 2.54 mm) water column. The required draft gage
for which test results are desired. Properties adequate for reading will be maintained throughout the test by the automati-
identification of the materials or ingredients, or both, of which cally controlled damper. Record the air velocity at seven
the test specimen is made shall be recorded. points, 23 ft from the centerline of the burner ports, 6 6 1⁄4 in.

6
 E 84
(168 6 7 mm) below the plane of the specimen mounting of establishing the conditions that will exist following succes-
ledge. Determine these seven points by dividing the width of sive tests and for indicating the control of the heat input into
the tunnel into seven equal sections and recording the velocity the test chamber. If appreciable variation from the temperatures
at the geometrical center of each section. During the measure- shown in the representative preheat curve is observed, suitable
ment of velocity, remove the turbulence bricks (see 4.3) and the adjustments in the fuel supply may be necessary based on red
exposed 23-ft thermocouple and place 24-in. (670-mm) long oak calibration tests.
straightening vanes between 16 and 18 ft (4.88 and 5.49 m) 7.7 Allow the furnace to cool after each test. When the floor
from the burner. The straightening vanes shall divide the thermocouple at 13 ft (3.96 m) shows a temperature of 105 6
furnace cross section into nine uniform sections. Determine the 5°F (40.5 6 2.8°C), place the next specimen in position for
velocity with furnace air temperature at 73.4 6 5°F (23 6 test.
2.8°C), using a velocity transducer. The velocity, determined as 7.8 With the test equipment adjusted and conditioned as
the arithmetic average of the seven readings, shall be 240 6 5 described in 7.2, 7.3, 7.4, and 7.6, make a test or series of tests,
ft (73.2 6 1.4 m)/min. using nominal 23⁄32-in. (18.3-mm) select-grade red oak flooring
7.3.1 The following alternative to the velocity transducer as a sample. The red oak decks are to be constructed and
equipment and method of determining the tunnel air velocity conditioned as specified in Annex A1 and Annex A2. Make
has been found suitable: A 4–in.diameter low-speed rotary observations at distance intervals not in excess of 2 ft (0.6 m)
vane anemometer, having a resolution of 1 ft./min. with an and time intervals not in excess of 30 s, and record the time
accuracy of 6 2 %, is attached to the steel stand and placed in when the flame reaches the end of the specimen 191⁄2 ft (5.94
the tunnel 22.5 ft downstream of the burners. Three trials shall m) from the end of the ignition fire. The end of the ignition fire
be conducted and their values averaged. The average is shall be considered as being 41⁄2 ft (1.37 m) from the burners.
rounded to the nearest unit. The centerline of the vane The flame shall reach the end point in 51⁄2min6 15 s.
anemometer shall be aligned with the vertical centerline of the Automatically record the temperatures measured by the ther-
tunnel by placing it on the steel stand. Trial 1 is run with the mocouple near the vent end at least every 15 s. Automatically
vane edge 1 in. from the non-window wall; Trial 2 is with the record the photoelectric cell output immediately prior to the
center axis at the tunnel center point; and Trial 3 is run with the test and at least every 15 s during the test.
vane edge 1 in. from the window wall. 7.8.1 Another means of judging when the flame has reached
the end point is when the exposed thermocouple at 23 ft
7.4 The room in which the test chamber is located shall have
registers a temperature of 980°F (527°C).
provision for a free inflow of air during test to maintain the
7.9 Plot the flame spread distance, temperature, and change
room at atmospheric pressure during the entire test run.
in photoelectric cell readings separately on suitable coordinate
Maintain the air supply at a temperature of 73.4 6 5°F (23 6
paper. Figs. 4-6 are representative curves for red oak flame
2.8°C) and a relative humidity of 50 6 5 %.
spread distance, time-temperature development, and smoke
7.5 Supply the fire test chamber with natural (city) or density, respectively. Flame spread distance shall be deter-
methane (bottled) gas fuel of uniform quality with a heating mined as the observed distance minus 41⁄2 ft (1.37 m).
value of nominally 1000 Btu/ft3 (37.3 MJ/m3). Adjust the gas 7.10 Following the calibration tests for red oak, conduct a
supply initially at approximately 5000 Btu (5.3 MJ)/min. similar test or tests on samples of 1⁄4-in. (6.3-mm) inorganic
Record the gas pressure, the pressure differential across the reinforced cement board. These results shall be considered as
orifice plate, and the volume of gas used in each test. If a representing an index of 0. Plot the temperature readings
temperature- and pressure-compensating mass flowmeter is separately on suitable coordinate paper. Fig. 7 is a representa-
utilized, record only the volume of gas used. Unless otherwise tive curve for time-temperature development for inorganic
corrected for, when bottled methane is employed, insert a reinforced cement board.
length of coiled copper tubing into the gas line between the
supply and metering connection to compensate for possible 8. Procedure
errors in the flow indicated due to reductions in gas tempera- 8.1 With the furnace draft operating, place the test specimen
ture associated with the pressure drop and expansion across the on the test chamber ledges that have been completely covered
regulator. With the draft and gas supply adjusted as indicated in with nominal 1⁄8-in. (3.2-mm) thick by 11⁄2-in. (38-mm) wide
7.3 and 7.4, the test flame shall extend downstream to a woven gasketing tape. Place the specimen as quickly as is
distance of 41⁄2 ft (1.37 m) over the specimen surface, with practical. Place the removable top in position over the speci-
negligible upstream coverage. men.
7.6 Preheat the test chamber with the 1⁄4-in. (6.3-mm)
inorganic reinforced cement board and the removable lid in
place and with the fuel supply adjusted to the required flow.
Continue the preheating until the temperature indicated by the
floor thermocouple at 231⁄4 ft (7.09 m) reaches 150 6 5°F (66
6 2.8°C). During the preheat test, record the temperatures
indicated by the thermocouple at the vent end of the test
chamber at intervals not longer than 15 s and compare these
readings to the preheat temperature shown in the time- FIG. 6 Representative Time-Absorption Curve for Smoke Density
temperature curve in Fig. 3. This preheating is for the purpose of Red Oak

7
 E 84

FIG. 7 Representative Time-Temperature Curve for Fuel Contribution of Inorganic Reinforced Cement Board

8.2 Keep the completely mounted specimen in position in 8.7 Plot the flame spread distance, temperature, and change
the chamber with the furnace draft operating for 120 6 15 s in photoelectric cell readings separately on the same type of
prior to the application of the test flame. coordinate paper as used in 7.9 for use in determining the
8.3 Ignite the burner gas. Observe and record the distance flame-spread and smoke-developed indexes as outlined in
and time of maximum flame front travel with the room Section 9. Flame front advancement shall be recorded at the
darkened. Continue the test for a 10-min period. Termination of time of occurrence or at least every 30 s if no advancement is
the test prior to 10 min is permitted if the specimen is noted. Flame spread distance shall be determined as the
completely consumed in the fire area and no further progres- observed distance minus 41⁄2 ft (1.37 m).
sive burning is evident and the photoelectric cell reading has
returned to the baseline. 9. Interpretation of Results
8.4 Record the photoelectric cell output immediately prior 9.1 The flame spread index (FSI) shall be the valve,
to the test and at least every 15 s during the test. determined as follows, rounded to the nearest multiple of five.
8.5 Record the gas pressure, the pressure differential across 9.1.1 In plotting the flame spread distance-time relationship,
the orifice plate, and the volume of gas used in each test. If a all progressive flaming as previously recorded shall be in-
temperature- and pressure-compensating mass flowmeter de- cluded at the time of occurrence. A straight line shall be used
vice is used to monitor the gas flow, record only the volume of to connect successive points. The total area (AT) under the
gas. flame spread distance-time plot shall be determined by ignor-
8.6 When the test is ended, shut off the gas supply, observe ing any flame front recession. For example, in Fig. 8 the flame
smoldering and other conditions within the test duct, and spreads 10 ft (3.05 m) in 21⁄2 min and then recedes. The area is
remove the specimen for further examination. calculated as if the flame had spread to 10 ft in 21⁄2 min and

FIG. 8 Example of Time-Distance Relationship with Flame Front Recession

(Total Area, AT 5 A1 + A2)

8
 E 84
then remained at 10 ft for the remainder of the test or until the 11.1.5 Graphical plots of flame spread and smoke developed
flame front again passed 10 ft. This is shown by the dashed line data.
in Fig. 8. The area (AT) used for calculating the flame spread
index is the sum of areas A1 and A2 in Fig. 8. 12. Precision and Bias 12
9.1.2 If this total area (AT) is less than or equal to 97.5 12.1 Precision—A series of interlaboratory tests for this test
ft·min, the flame spread index shall be 0.515 times the total method was run using eleven laboratories and six materials.
area (FSI 5 0.515 AT). Four replicates of each material were tested. The complete
9.1.3 If the total area (AT) is greater than 97.5 ft·min, the results have been placed on file at ASTM Headquarters as a
flame spread index shall be 4900, divided by the difference of Research Project entitled “Interlaboratory Test Study on ASTM
195 minus the total area (AT). (FSI 5 4900/(195 − AT)). E84, Standard Test Method for Surface Burning Characteristics
9.2 The test results for smoke shall be plotted, using the of Building Materials.” Data was calculated in accordance with
same coordinates as in 7.9. The area under the curve shall be Practice E 691 and ISO 5725.
divided by the area under the curve for red oak, multiplied by 12.2 Even though Test Method E 84 provides measurement
100, and rounded to the nearest multiple of five to establish a of a Flame Spread Index and a Smoke Developed Index, only
numerical smoke-developed index. The performance of the the Flame Spread Index is considered in this precision state-
material is compared with that of inorganic reinforced cement ment because the test series utilized a smoke measurement
board and select grade red oak flooring, which have been system that was a variation from that described in the test
arbitrarily established as 0 and 100, respectively. For smoke- method. Data on precision of the smoke developed index is
developed indexes 200 or more, the calculated value shall be being worked on by Task Group No. 1 of Subcommittee E5.22
rounded to the nearest 50 points. and will be included in a future revision of this test method.
12.3 Within-laboratory (repeatability) data is given in Table
NOTE 1—Allowance should be made for accumulation of soot and dust
on the photoelectric cell during the test by establishing a revised base line. 1.
The revised base line shall be a straight line drawn from the zero point 12.4 Between-laboratory (reproducibility) data is given in
(point on base line where incipient light attenuation occurs) to the point Table 2.
established after the sample has been removed. 12.5 Bias—No information is presented on the bias of the
procedure in this test method because correct values for
10. Analysis of Products of Combustion fire-test-response characteristics of building materials can only
10.1 Samples for combustion product analysis, when analy- be defined in terms of a test method. Within this limitation, this
sis is requested, shall be taken downstream from the photom- test method has no known bias and can be accepted as a
eter, or shall consist of not more than 1 % of the total flow. reference method.
Analysis of the products of combustion is not required in this
test method. 13. Keywords
13.1 flame spread; flame spread index; smoke developed;
11. Report
smoke developed index; Steiner tunnel; surface burning char-
11.1 Report the following information: acteristics; 25 ft tunnel; tunnel test
11.1.1 Description of the material being tested, including its
composition or generic identification, thickness, and any rel-
evant additional details,
11.1.2 Test results as calculated in Section 9, 12
Supporting data have been filed at ASTM Headquarters and may be obtained
11.1.3 Details of the method used in placing the specimen in by requesting PCN 33000014-31.
the chamber, to include the following:
11.1.3.1 A statement whether a continuous or sectioned TABLE 1 Within-Laboratory (Repeatability) Precision Data
specimen is used, Parameter—Flame Spread Index
11.1.3.2 A description of the number of sections and their Material Repeatability Relative
sizes, when the specimen consists of sections joined end-to- Standard Standard
Mean Value Deviation, Sr Deviation,%
end,
11.1.3.3 The mounting method employed, Douglas Fir Plywood 91 15 16
Fire Retardant Treated 17 3 17
11.1.3.4 The method of placement of the cement board Douglas Fir Plywood
protecting the furnace lid assembly. Type X Gypsum Board 9 2 19
11.1.4 Observations of the burning characteristics of the Rigid Polystyrene Foam 7 1 18
Rigid Polyurethane Foam 24 3 13
specimen during test exposure, such as delamination, sagging, Composite Panel 17 2 12
shrinkage, fallout, etc., and

9
 E 84
TABLE 2 Between-Laboratory (Reproducibility) Precision Data
Parameter—Flame Spread Index

Material Reproducibility Relative

Standard Standard
Mean Value Deviation, SR Deviation,%
Douglas Fir Plywood 91 23 25
Fire Retardant Treated 17 6 33
Douglas Fir Plywood
Type X Gypsum Board 9 3 36
Rigid Polystyrene Foam 7 4 60
Rigid Polyurethane Foam 24 5 23
Composite Panel 17 4 21

ANNEXES

(Mandatory Information)

A1. CONSTRUCTION GUIDELINES OF RED OAK DECKS

A1.1 Introduction
A1.1.1 General construction outline of the red oak decks is
shown in Fig. A1.1.

FIG. A1.1 Red Oak Calibration Deck Construction

10
 E 84

A2. PROCEDURE FOR DETERMINING MOISTURE CONTENT IN RED OAK

A2.1 Introduction tioning atmosphere that will result in an average moisture

A2.1.1 This procedure shall be used for the determination of content of 7 6 0.5 %. Using either a conductance or dielectric-
moisture content of the select-grade red oak calibration mate- type meter (calibrated per Test Methods D 4444), monitor
rial. From trimmed sections of the calibration decks, prepare a moisture content until the desired level is reached. Subject the
minimum of six specimens 4 + 1⁄16 -0 inches (100 + 2 -0 mm) trimmed sections only to the secondary oven-drying method
long. The specimens shall be free from visible irregularities of (Method B) in Test Methods D 4442 for the final determination
knots, decay, reaction wood, and resin concentration. Place the of moisture content.
trimmed sections adjacent to the calibration decks in a condi-

APPENDIXES

(Nonmandatory Information)

X1. GUIDE TO MOUNTING METHODS

X1.1 Introduction (a) Mounting methods should be considered for building

X1.1.1 Discussion: products that normally incorporate joint details either in design
X1.1.1.1 This guide has been compiled as an aid in selecting or installation. A nonhomogenous product containing underly-
a method for mounting various building materials in the fire ing core (Note X1.1) material that may adversely affect the test
test chamber. These mountings are suggested for test method results should be tested with a joint. This joint should be
uniformity and convenience; they are not meant to imply located longitudinally between the burners.
restriction in the specific details of field installation. NOTE X1.1—Core is defined as: a central and often foundational part
X1.1.1.2 For some building materials none of the methods usually distinct from the enveloping part by a difference in nature
described may be applicable. In such cases, other means of (Webster’s New Collegiate Dictionary).
support may have to be devised. (b) The surface burning behavior should be determined using
X1.1.1.3 These suggested mounting methods are grouped the manufacturer’s recommended joint detail.
according to building materials to be tested which are broadly
(c)If a joint detail is not recommended by the manufacturer,
described either by usage or by form of the material.
the product should be tested both with a separation of 3⁄16 6
X1.1.2 Support Pieces: 1⁄16in. (4.26 1.5 mm) and with the edges in direct contact.
X1.1.2.1 Whenever inorganic reinforced cement board is
specified as a backing in this appendix, the material shall be X1.2 Acoustical and Other Similar Panel Products Less
nominal 1⁄49 (6.3 mm) thick, high density (110 6 5 lb/ft3) and Than 20 in. (508 mm)
uncoated.
X1.1.2.2 Whenever metal rods or bars are specified in this X1.2.1 For acoustical materials and other similar panel
appendix as supports they should be: products whose maximum dimension is less than 20 in. (508
Steel rods, 1⁄4 in. (6.3 mm) diameter
mm), metal splines or wood furring strips and metal fasteners
Steel bars, 3⁄16 by 2 in. (5 by 51 mm) shall be used.
(a) The rods or bars should span the width of the tunnel. X1.2.2 Steel tee splines for mounting kerfed-acoustical tile
Rods should be placed approximately 2 in. (51 mm) from each shall be nominal 1⁄2-in. (13-mm) web by 3⁄4-in. (19-mm) flange,
end of each panel and at approximately 2-ft (0.6-m) intervals formed No. 24 MS gage sheet metal.
starting with the fire end of each panel. X1.2.3 Wood furring frames for mounting acoustical mate-
(b) Bars are used instead of rods only when they are required rials and other similar panel products less than 20 in. (508 mm)
to support the sample. The bars should be placed approxi- shall be nominal 1 by 2-in. (20 by 41-mm) wood furring joined
mately 2 in. (51 mm) from each end of each panel and at with corrugated-metal fasteners. Use two frames as shown in
approximately 2-ft (0.6-m) intervals starting with the fire end Fig. X1.1.
of each panel.
X1.1.2.3 Whenever netting is specified as a support in this X1.3 Adhesives
appendix, the material shall be 20-gage, 2-in. (51-mm) hex- X1.3.1 To determine the surface burning characteristics of
agonal galvanized steel netting conforming to Specification adhesives, they shall be mixed as specified in the manufactur-
A 390. er’s instructions and shall be applied to inorganic reinforced
X1.1.3 Joints: cement board in the thickness or at the coverage rate recom-
X1.1.3.1 Products that are normally installed adjoining mended by the manufacturer. The adhesive application shall be
themselves longitudinally are evaluated under this paragraph. cured prior to testing.

11
 E 84

FIG. X1.1 Wood Frame for Acoustical Materials and Other Similar Panel Products Less Than 20 in. (508 mm)

X1.4 Batt or Blanket-Type Insulating Materials steel screening13 with approximate 3⁄64-in. (1.2-mm) openings
X1.4.1 Batt or blanket materials that do not have sufficient supported on a test frame 20 in. (508 mm) wide by 2 in. (51
rigidity or strength to support themselves shall be supported by mm) deep, made from 2 by 3 by 3⁄16-in. (51 by 76 by 5-mm)
metal rods inserted through the material and positioned such steel angles. Three frames are required. See Fig. X1.2. The
that the bottom of the rod is approximately 1⁄4 in. (6.3 mm) insulation shall be packed to the density specified by the
from the surface to be exposed to the flame. It is recommended manufacturer.
that batt or blanket materials less than 1 in. (25.4 mm) thick not X1.7 Plastics
be mounted for testing in this manner.
X1.7.1 The term plastics includes foams, reinforced panels,
X1.5 Coating Materials, Cementitious Mixtures, and laminates, grids, and transparent or translucent sheets.
Sprayed Fibers X1.7.2 When any plastic will remain in position in the
X1.5.1 Coating materials, cementitious mixtures, and tunnel during a fire test, no additional support will be required.
sprayed fibers shall be mixed and applied to the substrate as Thermoplastic and other plastics that will not remain in place
specified in the manufacturer’s instructions at the thickness, are to be supported in accordance with X1.1.2.2 and X1.1.2.3.
coverage rate, or density recommended by the manufacturer. X1.8 Thin Membranes
X1.5.2 Materials intended for application to wood surfaces X1.8.1 Single-layer membranes or thin laminates consisting
shall be applied to a substrate made of nominal 23⁄32 in. of a limited number of similar or dissimilar layers not intended
(18-mm) select grade, red oak flooring which is also used as for adherence to another surface may be supported on poultry
the calibration material. Test decks placed end to end shall be netting placed on steel rods in accordance with X1.1.2.2 and
used. Construct and condition in accordance with Annex A1 X1.1.2.3. Netting shall be 20-gage, 2-in. (51-mm) hexagonal
and Annex A2. galvanized steel poultry netting conforming to Specification
X1.5.3 Materials intended for application to particular com- A 390. If so tested, the specimen shall be additionally tested,
bustible surfaces shall be applied to the specific surfaces for bonded to a substrate representative of a field installation.
which they are intended.
X1.5.4 Materials intended for only field application to X1.9 Wall Coverings
noncombustible surfaces shall be applied to 1⁄4-in. (6.3-mm) X1.9.1 Whenever an adhesive is used to attach a wall
inorganic reinforced cement board. covering, adhesive specified by the manufacturer shall be used
X1.5.5 Since the nature of the substrate may significantly in the test in a manner consistent with field practice.
affect the performance of the fire retardant coating, an indica- X1.9.2 Mount wall coverings intended for application di-
tion of the performance of a fire retardant coating can be rectly to a noncombustible wall surface to 1⁄4 in. (6.3 mm)
determined by comparing the surface flammability of the
coated substrate with that of the uncoated, specific substrate. 13
The use of galvanized steel screening normally lowers the flame spread index
X1.6 Loose-Fill Insulation values obtained for some materials that are tested in this manner and, therefore, the
results do not necessarily relate directly to values obtained for other materials
X1.6.1 Loose-fill insulation shall be placed on galvanized- mounted without galvanized steel screening.

FIG. X1.2 Steel Frame for Loose Fill Materials

12
 E 84
inorganic-reinforced cement board. Temperature Bonding Mortar14 or the equivalent. In the event
X1.9.3 Mount wall coverings intended to be applied over the specimen cannot be adhered using Sairmix #7, Kentile #9
epoxy15 has been found to be a suitable alternative. The
gypsum wallboard to 5⁄8 in. (15.9 mm) Type X gypsum
application shall be determined by a 3⁄32-in. (2.4-mm) notched
wallboard complying with specification Specification C 36. trowel held at an 80 to 90° angle using a random pattern. The
There is no need to mount the gypsum wallboard on studs. adhesive shall be applied only to the specimen back. The
X1.9.4 Mount wall coverings intended for application over specimen shall then be placed on the smooth side of the
a combustible substrate to one representative of that substrate. inorganic reinforced-cement board and rolled using a 100-lb
(45.4-kg) roller (nominal 5-in. (127-mm) diameter, three 5-in.
X1.9.5 Mount wall coverings not intended to be adhered long sections placed end to end for a total length of 15 in. (381
directly to a wall surface, but hung or otherwise supported by mm)). The prepared samples can be dead stacked overnight but
framing or a track system, in a manner that is representative of should be transferred to separate storage racks until tested.
their installation. Where this is not practical, support the Each sample shall be vacuumed prior to test.
sample on netting placed on metal rods as provided in X1.1.2.2 X1.10.2 Textile materials intended for application to walls
and X1.1.2.3. or ceilings should be mounted in accordance with X1.9.

14
The sole source of supply of the apparatus known to the committee at this time
X1.10 Mounting Method for Heavy Textile Materials is A. P. Green Refractories, Green Blvd., Mexico, MO 65265. If you are aware of
alternative suppliers, please provide this information to ASTM Headquarters. Your
X1.10.1 When the surface burning characteristics of the comments will receive careful consideration at a meeting of the responsible
material itself are required, specimens shall be mounted on technical committee,1 which you may attend.
15
The sole source of supply of the apparatus known to the committee at this time
inorganic reinforced cement board with Sairmix #7 High is Kentile Floors, Inc., Brooklyn, NY 11215. If you are aware of alternative
suppliers, please provide this information to ASTM Headquarters. Your comments
will receive careful consideration at a meeting of the responsible technical
committee,1 which you may attend.

X2. DERIVATION OF FLAME SPREAD AREA FORMULAS APPEARING IN 8.1

X2.1 Introduction X2.2 Formula 1—Constant

X2.1.1 This appendix contains an abbreviated discussion of X2.2.1 In Fig. X2.1, an idealized straight-line flame spread
the derivations of the flame spread area formulas used to distance-time plot is drawn. Lines OA, OA8, and OA9 produce
calculate the flame spread index in this test method. This a family of areas ORA having a maximum possible area of 97.5
appendix will show not only the derivations of the formulas, ft·min (1⁄2 by 10 min by 19.5 ft). These represent a steady
but will illustrate the relationship between this method of flame progression of the flame front to a maximum distance at the
spread calculation and a previous method. end of the 10-min test.
X2.1.2 In these calculations, it is assumed that the flame X2.2.2 When the flame front spreads its maximum distance
front never recedes. Hence, in Fig. 8 there is an imaginary line (19.5 ft) in 10 min, a formula used in Test Method E 84 would
bounding the upper edge of area A2. yield the following:

FIG. X2.1 Idealized Straight-Line Flame Spread Distance-Time Curve for Total Areas Less than or Equal to 97.5 min·ft

13
 E 84
550 550 to 195 − ORBI.
FSI 5 t 5 10 5 55 (X2.1)
Thus, a new flame spread index formula may be derived as
X2.2.3 Also, when the flame front is maximized at 19.5 ft in follows:
10 min, the area in Fig. X2.1 ORA is maximized to 97.5 ft·min. K K K
X2.2.4 To relate the current formula, which is of the straight FSI 5 OIA 5 195 2 ORBI 5 195 2 A (X2.7)
T
line, origin intercept form, to the previous (Test Method E 84) X2.3.3 To establish K, a relationship between the current
formula, it is necessary to equate the two as follows: and the previous Test Method E 84 formulas will be established
550 at the red oak calibration point of 19.5 ft progression at 5.5 min
FSI 5 t 5 KA (X2.2)
as follows:
where: 550 K
FSI 5 t 5 195 2 A (X2.8)
K 5 proportionality constant for equations of the T

current formula’s type, and where:

AT 5 total area under area ORA. AT 5 195 − (9.75 (5.5)) 5 141.38 ft·min, and
If AT 5 97.5 ft·min at t 5 10 min, then t 5 5.5 min.
550 Thus:
FSI 5 100 5 K 3 97.5, and (X2.3)
550 K
550 FSI 5 5.5 5 195 2 141.38 , or (X2.9)
K 5 10 3 97.5 5 0.564 (X2.4)
550 3 ~53.63!
K 5 5.5 5 5363
X2.3 Formula 2—Constant
X2.3.1 In the idealized straight-line flame spread distance- X2.4 Formulas 1 and 2
time curve of Fig. X2.2, lines OI, OI8, and OI9 produce a
family of trapezoidal areas ORBI ranging from 97.5 to 195 X2.4.1 To account for the disproportionate increase which
ft·min (1⁄2 by 10 min by 19.5 ft to 10 min by 19.5 ft). This can occur in FSI values at the lower end of the index scale, for
represents a flame front progression to the end of the specimen K 5 0.564 in Formula 1 and 5363 in Formula 2, a further
within the 10 min of the test. The area (AT) of ORBI may be mathematical modification is made.
expressed as follows: X2.4.2 In order to establish a relationship between the

S D S D
constants (K) in X2.2 and X2.3, it is necessary to consider the
1 1
form of the basic formulae, which are as follows:
2 by 19.5 by OR 1 2 by 19.5 by ~102AI! (X2.5)
K1
which is equal to: FSI 5 195 2 A ~A . K2! (X2.10)
T
195 2 9.75 AI (X2.6) FSI 5 K3AT ~A, K2!
since OR is always 10 min.
X2.3.2 The triangular area OIA divided into a proportional- where:
ity constant K will determine a relationship between flame K1 5 100 (195 − R),
spread indexes and the rate and distance of flame propagation. R 5 the area associated under the curve that is to be
The total area available is 195 ft·min, hence area OIA is equal associated with an index of 100,

FIG. X2.2 Idealized Straight-Line Flame Spread Distance-Time Curve for Total Areas Greater than 97.5 min·ft

14
 E 84

K2 5 an arbitrary choice within the limits of 0 and 195, and K3 5 4900/~97.5 3 97.5! 5 0.515 (X2.12)
K3 5 K1/(K2[195 − K2]). X2.4.5 Thus, the formula for flame spread index in 8.1.2 is
X2.4.3 Choosing K2 5 195/2 produces a minimum value of as follows:
K3, that is, any other K2 value will result in a higher K3 value, FSI 5 0.515 AT (X2.13)
and choosing R, the area under a red oak calibration plot, as a
median value of 146, implies the following: X2.4.6 Thus, the formula for flame spread index in 8.1.3 is
as follows:
K1 5 100 ~195 2 146! 5 4900 (X2.11)
4900
FSI 5 195 2 A (X2.14)
X2.4.4 Then using 97.5 as the value for K2, K3 would be: T

X3. COMMENTARY

X3.1 Introduction and specifically “white wash” were actively promoted. The
X3.1.1 This commentary has been prepared to provide the equipment consisted of a long bench with a noncombustible
user of Test Method E 84 with background information, top. The sample consisted of a wood trough about 16 ft long,
including literature references, on the development and use of 18 in. wide, and 18 in. deep (5.568 m long, 0.522 m wide, and
this test method. It also provides the reader and user with the 0.522 m deep), placed upside down on the bench. The inside of
basis for the methods that have been used for deriving the trough was coated with the paint. A known quantity of
numerical flame spread indexes; an appreciation of the vari- wood at one end furnished the ignition source.
ability of the test; and comments on its application and X3.2.2 In 1927 and 1928, chemically impregnated wood
limitations for testing selected types of materials. was being developed, and Underwriters Laboratories, Inc.,
X3.1.2 On Nov. 28, 1942, 490 people died in a fire in the used a tunnel 36 in. wide, 23 ft long, and 13 in. deep (1.044 m
Boston Coconut Grove Nightclub. On June 5, 1946, 61 persons wide, 8 m long, and 0.377 m deep) to evaluate its performance.
died in the La Salle Street Hotel fire. On Dec. 7, 1946, a fire in It was during this time that red oak flooring was selected as a
the Winecoff Hotel in Atlanta, Ga., claimed the lives of 119 control to calibrate the furnace. The sample formed the top of
persons. These fires had one thing in common. In all three fires, the tunnel. The fuel and draft were also controlled.
rapid flame spread along the surfaces of interior finish was X3.2.3 In the early 1940’s, a desire to reduce flammability
judged to be a major factor in the spread of fire. Two had burlap of wood-based products, and the introduction of new building
wall coverings, and the other an early type of plywood which materials and combinations of materials brought about the need
seriously delaminated. The fire protection authorities investi- to further improve the tunnel. The development of the third
gated several test methods with the objective of providing one tunnel furnace is explained fully in Underwriters Laboratories
that could be used to regulate interior finish materials and Bulletin of Research No. 32 (2). Subsequent refinements were
minimize repetition of such fires. These tests included: The incorporated, and the first formal test method was published as
Forest Products Laboratory Fire Tube Test (now Test Method Standard U.L. 723 by Underwriters Laboratories in August
E 69); Federal Specification SS A118b (acoustical tile/bunsen 1950. Revised editions were published in 1958, 1960, 1971,
burner test) (replaced by SS-A-118a-7/63-referencing Test 1977, and 1979. The National Fire Protection Association
Method E 84); New York City Timber Test and Shavings Test adopted the method as NFPA No. 255 in 1955 with revisions in
(now obsolete); Crib Test-Specification C 160 – 41 T (now 1958, 1961, 1966, 1970, 1972, and 1979. The test was adopted
Test Method E 160); and The Swedish Schlyter Test. (1)16 All by the American Society for Testing and Materials as a
of these were relatively small laboratory tests. Test Method tentative standard in 1950 and formally adopted in 1961 with
E 84 was developed on the premise that a large test would revisions made in 1967, 1968, 1970, and from 1975 through
provide a more realistic and comprehensive test, and it has 1980.
since been widely adopted for use by the building code X3.2.4 The tunnel has been designated the “Steiner Tunnel”
authorities to regulate the use of interior finish materials. by Underwriters Laboratories in honor of Albert J. Steiner (3)
Subsequently during this same period, two other test methods who had spent much time developing this and many other fire
were developed for use in research and development of new test methods.
materials, the NBS Radiant Panel (Test Method E 162) and the X3.2.5 Since 1950 the flame spread properties of materials,
FPL 8-ft tunnel (Test Method E 286). These test methods have as measured by this method have been reported as ratings,
been widely used for research and development purposes. classifications, or indices. The last is considered more indica-
tive of the nature of the results and is the present terminology
X3.2 History of Test Method E 84 used in the standard. The original method of determining
X3.2.1 The first “tunnel-type” furnace was built at Under- “flame spread index” was based on either the ratio of the time
writers Laboratories around 1922 when “fire-proofing” paints at which flames traveled the full tunnel length or the partial
flame travel distance relative to that of red oak. In 1968, a
change was made in the FSI calculation to account for an
16
The boldface numbers in parentheses refer to the list of references at the end anomaly between results for flame spread greater than or less
of this test method. than 131⁄2ft. In 1976, the flame spread index was changed to an

15
 E 84
area basis (4). Here the total area under the distance-time developed after repeated experiment tests, although not opti-
curve, ignoring any flame-front recession, was compared to a mum fire conditions, were selected to produce a flame spread
prescribed area typical of red oak flooring. The current over the entire length of the calibration material in about
calculation method (see Appendix X2) uses a formula that 51⁄2min (1). It was found that conditions could be changed so
takes the rate of flame travel into account. that flames would spread faster, but these conditions caused the
X3.2.6 The sensitivity study by Endicott and Bowhay (5) in flame to spread too fast to make the necessary observations of
1970 has led to a concerted effort by the “ASTM tunnel the flame spread, smoke density, and temperature rise of the
operators group” to address concerns identified by the report. thermocouple.
Since 1975 a series of changes have been specified in the
standard. These include defining duration of furnace preheat- X3.4 Furnace Calibration
ing, the incorporation of a floor thermocouple, as well as more X3.4.1 Select red oak was chosen as a control material
closely specifying details of furnace construction and standard- because this term denoted a fairly uniform grade of lumber
ization. nationally, whereas many other designations have a purely
X3.2.7 Particular attention is being paid to the refinement of local significance. It is readily available, usually uniform in
the apparatus and procedure involved in the measurement of thickness and moisture content, and generally gives repetitive
the smoke generated during testing. Round-robin tests that results. In recent years, experiments have been run using
have been conducted to date have indicated large differences in man-made materials such as particleboard in the hope of
smoke developed values for interlaboratory tests on replicate further refining the repeatability, however, red oak is still used
specimens. as a calibrating material.
X3.2.8 Some of these revisions include standardization of X3.4.2 The operating conditions of the tunnel are adjusted if
the smoke-density measuring equipment, its location in the necessary to ensure that the flame spreads to the end of the
exhaust duct, and its orientation. The measurement of smoke tunnel in 5.5 6 0.25 min, for a specimen of red oak flooring.
density is reported in terms of the area under the light Tests are run with an inorganic reinforced cement board (ACB)
absorption time curve related to a similar curve for red oak. specimen to establish the distance of the exposing flame at 4.5
Since the quality of vision obscuring particles in the smoke ft. It should be noted that the calibration specifies only the time
column is not linearly related to light absorption, this proce- at which the flame passes over the end of the specimen. The
dure has been criticized by some parties. The method does FSI depends on the area under the flame spread versus the time
however provide a basis for comparisons. curve. Therefore, the FSI of red oak is no longer exactly 100 as
X3.2.9 In 1970, a revision to the scope was adopted to originally specified.
emphasize that there was no direct relationship between the X3.4.3 Recognition of the importance of turbulence, includ-
flame spread index (FSI) and the fuel contributed or smoke ing the role of fire bricks and of window recesses, resulted in
density index (SDI). This revision was deemed necessary a revision in the method in 1976 (see 3.1 and section 3.3 ).
because some enforcement officials were assigning equal
significance to the values. X3.5 Repeatability and Reproducibility
X3.2.10 Prior to 1978, the report of tests included an X3.5.1 Four round-robin tests have been conducted: the first
evaluation of the fuel contribution as well as the FSI and SDI. in 1958 between Underwriters Laboratories and Southwest
However. it is now recognized that the rise in temperature of Research Institute; the second in 1959 sponsored by the former
the thermocouple located near the end of the tunnel, on which Acoustic Tile Assn. among four laboratories using four differ-
it is based, does not provide a valid measure of fuel distribu- ent tiles (6); the third in 1973 on floor coverings by the
tion. Therefore, although the data are recorded during the test, National Bureau of Standards with eleven cooperating labora-
this information is no longer normally reported. tories (7); the fourth in 1978 on loose-fill cellulosic insulation
X3.2.11 Appendix X1 adopted in 1968 is intended as a by the Consumer Product Safety Commission with six labora-
guide for the mounting of specimens. It is not a mandatory part tories (8); others are now in process under the auspices of
of the method, since the intent of the method is that the ASTM Committee E-5. A precision and bias statement is being
specimen be tested as closely as possible to the manner in prepared. In the interim, the reader is directed to the round-
which it will be applied in general use. In 1978, revisions were robin reports if information on precision and bias is needed.
made that dealt with the testing of adhesives, the description of X3.5.2 An ASTM task group of Subcommittee E05.22
a wood substrate for testing coatings, and the definition of the composed of tunnel operators is now working on comprehen-
properties of the inorganic reinforced cement board used as a sive design and on operational and procedural revisions to
standard backing and the metal rods used as supports. improve uniformity among facilities (9).
X3.3 Fire Exposure Conditions X3.6 Advantages and Problems
X3.3.1 The tunnel test fire exposure is provided by a 41⁄2-ft X3.6.1 Test Method E 84 results have generally showed
(1.35-m) long test flame, covering approximately 7 ft2 (0.63 performance similar to that observed during accidental build-
m2) of the 36 ft2 (3.25 m2) of the exposed specimen surface ing fires for some materials and exposure. It should be
during the 10-min test period. It releases heat at a rate of emphasized however that it is the intent of this test method to
approximately 5000 Btu/min (88 kW) and creates gas tempera- provide only comparative classifications.
tures near the specimen surface of up to 1600°F (900°C). X3.6.1.1 It provides a large flaming fire exposure, with
X3.3.2 The size and heat release rate of the exposing flame, specimen thermal exposure and area coverage sufficient to

16
 E 84
bring about progressive surface burning and combustible recommended that the results of these tests be used alone to
volatile generation characteristic of the materials under evalu- describe material combustibility.
ation resulting in a moving, wind-aided flame front. X3.7.4 Composite assemblies or panels using metal or
X3.6.1.2 It involves a large specimen, nominally 36 ft2 (3.25 mineral facings and combustible interior cores, and which
2
m ) of exposed area, allowing for realistic fire involvement of remain essentially impermeable to flame throughout the test
material surfaces and the development of physical and struc- period, may not be completely evaluated for surface burning
tural failures (collapse, buckling, large ruptures or cracks, etc.) behavior by this method, since the interior cores are not fully
that may influence flammability performance during the test challenged.
period.
X3.6.1.3 It may be applied to a wide range of materials, X3.7.5 Some materials require support to remain in place
including composite constructions of faced or laminated during the test period, such as loose-fill insulation supported by
boards, panels, units, or sections in actual field-installed a metal screen. The supporting screen tends to provide low
thicknesses. flame spread indices (FSI) relative to those obtained for
X3.6.1.4 It may be used to measure the effects of density, materials that are not so supported. Conversely, materials that
thickness, surface contour, surface finish, delamination, are supported on rods, such as batt insulation may produce
strength, and joint design on the surface flammability of the higher FSI if retained on the ceiling rather than allowed to burn
specimen. on the floor.
X3.6.1.5 It does characterize most high-flame spread mate- X3.7.6 Some materials, such as composites may delaminate
rials identified as having been involved in rapidly developing during the test. This may cause two possible responses, the
field fires (for example, highly combustile coatings on wood material may expose two or more surfaces to the flame
products, certain cellulosic acoustical materials, insulation increasing the FSI; the material may sag or drop one end into
facings applied with combustible adhesives, etc.), as well as the fire chamber impeding further flame spread.
providing an accurate characterization of the performance of X3.7.7 Some materials, such as cellular plastics, and ther-
some low flame-spread materials in actual fires (for example,
moplastic and thermosetting materials, may be difficult to
gypsum and mineral products).
evaluate. Thermoplastic and thermosetting materials not me-
X3.6.1.6 Although this is a test to measure “surface burning
chanically fastened will often fall to the floor of the tunnel and
characteristics,” the visual observation of flame travel is based
also usually receive relatively low FSI (10). If supported on
on maximum flame extension anywhere within the tunnel
volume, not necessarily directly on the specimen surface that wire screen, rods or other supports, these materials may be
may not be clearly visible. Surface flammability measurements completely engulfed in flame, and a questionable comparison
of building materials do not yield a unique material property. is being made between the surface flame-spread of nominal
Rather, the measurement is influenced by the method of test to 1-in. (0.039-mm) thick red oak with the burning rate of these
a very considerable degree. materials. Where the entire specimen is consumed, as com-
pared to the surface burning of red oak, much more oxygen is
X3.7 Uses and Limitations used and higher smoke developed indexes are usually obtained
X3.7.1 The orientation of the specimen in this method is in for these materials.
a horizontal ceiling position. This orientation places some X3.7.8 The materials described above, that is, those that
limitations on the type of material that can be realistically drip, melt, delaminate, draw away from the fire, or require
mounted during testing. Prior to 1960, the tunnel was used artificial support present unique problems and require careful
primarily for the investigation of the surface burning charac- interpretation of the test results. Some of these materials that
teristics of homogeneous compositions of ceiling and wall are assigned a low FSI based on this method may exhibit an
finishes, such as acoustical tiles, wall coverings, coatings, and increasing propensity for generating flame-over conditions
various types of decorative panel, all being able to support during room fire test with increasing area of exposure of the
themselves in the ceiling position throughout the test. material and increasing intensity of the fire exposure. The
X3.7.2 Through adaptation (Appendix X1 on Guide to
result, therefore, may not be indicative of their performance if
Mounting Methods added in 1968), the procedure was ex-
evaluated under large-scale test procedures. Alternative means
panded to include the evaluation of composites and assemblies.
of testing may be necessary to fully evaluate some of these
The guide contains mounting suggestions for a number of
materials.
individual categories of product classifications, including:
acoustical and similar panel products; composite building X3.7.9 In order to provide needed technical data, flamma-
units; adhesives, batt- and blanket-insulating materials; fire- bility evaluations of cellular plastics for building construction
retardant and general-purpose coatings; loose-fill thermal in- using the Steiner tunnel began with the testing and classifica-
sulations; treated and untreated plywoods; lumber and wood tion of a flame-retardant formulation of polystyrene foam
composition boards; foamed, molded, reinforced, and lami- board in 1960, with subsequent evaluation of polyurethane-
nated plastics; and sheet-type wall coverings. type boards incorporating flame-retardant resin systems (first
X3.7.3 The difficulty of defining materials that contribute generation) beginning in 1964, polymerically and chemically
little or no fuel to a fire has in the past led to the use of Test modified flame retardant polyurethane type formulations (sec-
Method E 84 to provide information about the combustibility ond generation) in 1965, polyisocyanurate-type foams initiated
of materials. ASTM Committee E-5 does not and has not ever in 1968 and, most recently, urea-formaldehyde type, cavity-fill

17
 E 84
foams. Spray-applied and poured-in-place cellular foam sys- conditioning and tunnel air supply (both 73.4 6 5°F, 50 6 5 %
tems were first subjected to the test in 1968 and 1972, relative humidity) were selected to minimize these effects.
respectively.
X3.7.10 From 1960 through 1973, in excess of 2000 tunnel X3.8 Correlation with Other Fire Conditions
tests have been conducted on flame retardant and general-
purpose polystyrene, polyurethane, polyisocyanurate and urea- X3.8.1 Several studies have been made to examine the
formaldehyde cellular plastics, in board-stock, spray-applied or relationship of the FSI test results on materials with their
pour-in-place forms, yielding flame spread values ranging from performance in large scale fire growth experiments and with
less than 5 to over 2000. See Ref (11) for full report on one test their performance in other laboratory test methods. Some
series. comparisons with large scale experiments are given in the
X3.7.11 The flame spread index of some materials may vary references. Comparisons have also been made between Test
depending upon environmental conditions. The prescribed Methods E 84, E 162, E 286, the “2-ft tunnel test,” the “corner
limits on the temperature and relative humidity for specimen test” (12) and other tests.

REFERENCES

(1) Steiner, A. J., “Burning Characteristics of Building Materials,” Fire (8) Williamson, R. B. and Baron, F. M., “A Corner Test to Simulate
Engineering, May 2, 1951. Residential Fires,” Journal of Fire and Flammability, Vol 4, April
(2) Underwriters Laboratories, Inc.,“ Fire Hazard Classification of Build- 1973, pp. 99–105.
ing Materials,” Bulletin of Research, No. 32, Chicago, IL, September (9) Waterman, T. E., “Relation of the Building Occupancy and Design to
1947. Use of Combustible Materials on Construction,” IITRI Report, Society
(3) Steiner, A. J., Building Offıcials Conference of America Yearbook,
of Plastics Industry, Inc., IIT Research Institute, Chicago, IL, October
1949–1950, pp. 115–116.
1968.
(4) Fang, J. B., “Fire Buildup in a Room and the Role of Interior Finish
Materials,” NBS Technical Note 879, June 1975. (10) D’Souza, M. V. and McGuire, G. H., “ASTM E84 and the Flamma-
(5) Endicott, L. E. and Bowhay, R. B., “A Statistical Evaluation of the Fire bility of Thermosetting Plastics,” Fire Technology, Vol 13, No. 2,
Hazard Classification Furnace (ASTM E84 – 68),” ASTM Materials May 1977, pp. 83–94.
Research and Standards, May 1970, pp. 19–21, 50–52. (11) Underwriters Laboratories, Inc., Flammability Studies of Cellular
(6) “Round Robin Tests on Tunnel Type Flame Spread Furnaces” for Plastics and Other Building Materials Used for Interior Finish,
ASTM Project No. 1–811–2, Final Report, Southwest Research Subject 723, U.L., Inc., Northbrook, IL, June 13, 1975.
Institute, San Antonio, TX, April 16, 1959.
(12) McGuire, J. H., “The Spread of Fire in Corridors,” Fire Technology,
(7) Lee, T. G. and Huggett, C., “Interlaboratory Evaluation of theASTM
Vol 4, No. 2, May 1968, pp. 103–108.
E84 – 70 Tunnel Test Applied to Floor Coverings,” Journal of Testing
and Evaluation, Vol 3, No. 1, ASTM, 1975.

The American Society for Testing and Materials takes no position respecting the validity of any patent rights asserted in connection
with any item mentioned in this standard. Users of this standard are expressly advised that determination of the validity of any such
patent rights, and the risk of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM Headquarters. Your comments will receive careful consideration at a meeting of the responsible
technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should make your
views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.
Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above address or at
610-832-9585 (phone), 610-832-9555 (fax), or [email protected] (e-mail); or through the ASTM website (www.astm.org).

18