Fire Endurance Testing of Floor Systems –

Effects of Scale and Restraint

by

John L. Gross1

ABSTRACT used in the WTC towers. These tests duplicated,

as closely as practical, the steel truss-supported
Three standard fire resistance tests (ASTM E composite concrete floor system. In practice, a
119) were conducted on a composite floor system floor assembly such as that used in the WTC
to study the effects of 1) test restraint conditions, towers is neither restrained nor unrestrained but is
and 2) scale of the test. Two full-scale tests, 35 ft2 likely somewhere in between. Testing under both
(10.7 m) span, and one half-scale test, 17 ft (5.2 restraint conditions, then, bounds expected
m) span, were conducted. Two tests were performance under the standard fire exposure, and
conducted under restrained conditions and one provides a comparison of unrestrained ratings
under unrestrained conditions. Results showed developed from both restrained and unrestrained
that the full-scale restrained floor system test conditions. Also, the spans of the WTC floor
obtained a fire resistance rating of 1½ h, while the system were up to 60 ft (18.3m) while furnaces
full-scale unrestrained floor system achieved a 2 used in establishing fire resistance ratings in the
h rating. Past experience with the ASTM E119 US allow spans of approximately 18 ft (5.5 m).
test method would lead one to expect that the Thus the extrapolation of results of a
unrestrained floor assembly would not perform as reduced-scale test may be an issue.
well as the restrained assembly, and therefore,
would receive a lower fire rating. For the full- Full-scale tests were conducted in both the
and half-scale floor systems tested in the restrained and unrestrained support conditions.
restrained condition, the full-scale specimen Further, a roughly one-half scale test of the
obtained a fire resistance rating of 1½ h, while the restrained floor system was also conducted. A
half-scale specimen achieved a 2 h rating. Both description of the WTC floor system, scaling and
tests would be expected to produce the same fire conduct of the tests, and results (fire resistance
resistance rating. ratings) are presented here.

KEYWORDS: ASTM E119, fire; fire resistance; 2.0 FIRE RESISTANCE TESTING
fire testing; floor systems; standard fire resistance
tests; testing The fire rating of structural materials and
assemblies is generally determined through testing,
1.0 INTRODUCTION and in the

NIST’s World Trade Center (WTC) investigation 

2
(NIST, 2005) allowed the opportunity to conduct The policy of the National Institute of Standards and
full- and reduced-scale tests of the floor system Technology is to use the International System of Units
(metric units) in all its publications. However, in North
 America in the construction industry, certain non-SI
1
National Institute of Standards and Technology, units are so widely used instead of SI units that it is
Building and Fire Research Laboratory, 100 Bureau more practical and less confusing to include
Drive, Gaithersburg, MD 20899 USA
measurement values for customary units as the an unrestrained rating determined from a
primary units of measure. restrained test sample, the conditions of
United States, such testing is frequently acceptance are based on the same criteria and, in
conducted in accordance with ASTM E 119, addition, temperature limitations are placed on the
“Standard Test Methods for Fire Tests of main structural members.
Building Construction and Materials” (ASTM,
2000). This standard was first published in 1917 In addition, since 1971, the ASTM E 119 Standard
as a tentative standard ASTM C 19 and was first describes a means to establish unrestrained ratings
adopted as ASTM E 119 in 1933. Since its from unrestrained test samples. For unrestrained
introduction, the test method has been modified samples, the fire resistance rating is again based on
and updated, although its essential character has limiting flame spread, exceeding temperatures on
remained unchanged. The test methods described the unexposed surface of the slab, and failure to
in ASTM E 119 prescribe a standard fire sustain the applied load; there are no limiting
exposure for comparing the test results of temperatures on the steel structural members when
building construction assemblies. For the tests of the test sample is installed in an unrestrained
floors and roofs, a test assembly is structurally condition.
loaded and the standard fire exposure is applied
to the underside of the specimen. The assembly 3.0 CONDUCT OF TESTS
is evaluated for its ability to contain the fire by
limiting flame spread and heating of the The floor system design consisted of a lightweight
unexposed surface while maintaining the applied concrete floor slab supported by steel trusses. The
load. The assembly is given a rating, expressed main composite trusses were spaced at 6 ft - 8 in.
in hours, based on these conditions of acceptance. (2.0 m) on center (o.c.) and had a nominal clear
span of either 35 ft (10.7 m) or 60 ft (18.3 m). The
Since 1971, versions of the ASTM E 119 steel trusses were fabricated using double-angles
Standard differentiate between testing and for the top and bottom chords, and round bars for
classifying thermally restrained and unrestrained the webs. The web members protruded above the
floor assemblies. According to Appendix A4 of top chord in the form of a “knuckle” which was
ASTM E 119-73 (ASTM, 1973), a restrained embedded in the concrete slab to develop
condition is “one in which expansion at the composite action. Additionally, the floor system
support of a load carrying element resulting from included bridging trusses (perpendicular to main
the effects of fire is resisted by forces external to trusses) spaced 13 ft - 4 in. (4.0 m) o.c. Figure 1 is
the element.” In an unrestrained condition, the a cut-away of the composite floor system showing
element is free to expand and rotate at its supports. the main and bridging trusses, metal deck and
It is customary in the United States to conduct concrete slab.
standard fire tests of floor assemblies in the
restrained condition. 3.1 Test variables

The current standard describes a means to NIST studied two factors, the effect of (1) scale of
establish restrained and unrestrained ratings for the test, and (2) test restraint conditions. To this
floor assemblies from restrained test samples. end, three tests were designed and conducted as
For restrained ratings from restrained test follows:
samples, the conditions of acceptance are based
on limiting flame spread, limiting temperatures Test #1: Full-scale, restrained test condition
on the unexposed surface of the slab, and failure Test #2: Full-scale, unrestrained test condition
of the assembly to sustain the applied load. For Test #3: Reduced-scale, restrained conditions
 off to prevent them from damaging the fire brick
and instrumentation in the event of a catastrophic
3.2 Preparation of Test Assemblies failure of the floor system.
3.3.2 Reduced-scale Test - 17 ft (5.2 m) span
Original shop drawings by Laclede Steel 3
(manufacturer of the steel trusses) were used for For the reduced-scale test specimens, the size of
the design of the 35 ft (10.7 m) span and 17 ft (5.3 the truss members and thickness of concrete slab
m) span test assemblies. The steel trusses were selected to allow the most information to be
faithfully duplicated the geometry of the original extracted as practicably possible considering the
design. Since equipment for making the Standard Fire Resistance Test is a test of the
resistance welds is not available in the United assembly’s ability to contain a fire and is based on
States, metal inert gas (MIG) welding was used both thermal response (flame spread and heating
and the welds were designed per American of the unexposed surface) and structural response
Institute of Steel Construction (AISC) (support the applied load) to the standard fire. The
Specification (AISC, 2001) to develop the web sizes of the steel members, thickness of concrete
diagonal capacities in tension or compression. slab, and truss spacing were selected to be the
This requirement was based on available test data same as in the full-scale tests. Otherwise, the
indicating that weld capacities exceeded proof geometry was scaled by roughly half. This scaling
loads by a factor in excess of 2.0. In addition, the required that the loading be increased by a factor
steel angles and round bars, reinforcing steel, of 2.
welded wire fabric, metal deck, lightweight
concrete, and primer paint were all matched as The reduced-scale tests were conducted at the UL
closely as practical. Sprayed Fire Resistant furnace facility in Northbrook, Illinois. The
Material (SFRM) was applied to the steel trusses superimposed uniform load was applied through a
at a specified thickness of 0.75 in (19 mm). combination of concrete blocks, water-filled
containers and hydraulic actuators located along
3.3 Description of Tests and Loading the trusses.

3.3.1 Full-scale Tests - 35 ft (10.7 m) span 4.0 TEST RESULTS

The full-scale tests were conducted at the As noted above, prior to 1971, the ASTM E 119
Underwriters Laboratories (UL) furnace facility Standard did not differentiate between testing and
in Toronto, Canada. Loading of the floor slab classifying thermally restrained and unrestrained
with an applied load to “simulate a maximum floor assemblies. The 1961 revision of ASTM E
load condition” as required by ASTM E 119, was 119, the revision referenced in the 1968 New York
accomplished through a combination concrete City Building Code, is used here for reporting the
block and water-filled containers which were tied Standard Fire Test ratings. Using this revision, a
single rating is developed. The year 2000 revision
of the Standard is used here for reporting
3
Certain trade names and company products are restrained and unrestrained ratings.
mentioned in the text or identified in an illustration in
order to adequately specify the experimental A photograph of the underside of the full-scale,
procedure and equipment used. In no case does such restrained test specimen after almost 2 h of
an identification imply recommendation or exposure is shown in Figure 2. Buckling of the
endorsement by the National Institute of Standards and compression diagonals can be seen as well as
Technology, nor does it imply that the products are bowing of the metal deck between supports. Upon
necessarily the best available for the purpose.
cooling, the test specimen recovered at least half
of the deflection achieved during the test. The tested floor assemblies are similar, though not
Sectioning of the slab revealed that the bowing identical, to steel joist and concrete floor systems
resulted from spalling of the concrete. that are widely used in low rise construction. The
Table 1 shows results for all three tests giving the test results provide valuable insight into the
times (in minutes) to reach the conditions of behavior of these widely used assemblies and also
acceptance, and the Standard Fire Test rating (in identify issues regarding scaling and restraint that
hours) for both the 1961 and 2000 revisions of require further study for floor systems and
ASTM E1194 . Note that in none of the tests did possibly other types of structural component.
the floor assembly fail to support the applied
load. The tests show temperature damage to the
Several observations can be made from the bridging trusses and buckling (in the full scale
results (ratings) shown in Table 1 as follows: tests) of compression diagonals and the vertical
strut near the supports. No evidence of knuckle
• The restrained full-scale floor system failures was seen in the tests.
obtained a fire resistance rating of 1½ h while
the unrestrained floor system achieved a 2 h The ASTM E 119 standard test method has been
rating. Past experience with the ASTM E119 used for several decades and has, for the most part,
test method would lead one to expect that the served its intended purpose well when taken
unrestrained floor assembly would receive a together with the fire rating requirements. This is
lower fire rating. supported by historical fire loss data for more than
half a century for different high-rise building
• A fire rating of 2 h was determined from the occupancies. In addition, there are extensive data
reduced-scale test while a fire rating of 1½ h and experience that have been developed using the
was determined from the full-scale test. test method.
• The above result raises the question of The NIST tests have identified areas where further
whether or not a fire rating based on the study related to the standard test method may be
ASTM E119 performance of a 17 ft (5.2 m) warranted. The issues related to the test method
span floor assembly is scalable to a larger that NIST will consider in formulating its
floor system of, say, up to 60 ft (18.3 m). recommendations include:
5. OBSERVATIONS AND CONCLUSIONS • the scale of the test for prototype assemblies
that are much larger than the tested
4
ASTM E 119 contains the following statement assemblies,
regarding precision and bias: No comprehensive
test program has been conducted to develop data on • the effect of restraint conditions on test results,
which to derive statistical measures of repeatability
(within-laboratory variability) and reproducibility • the repeatability of test results (e.g., do
(among-laboratory variability). The limited data
multiple fire resistance tests conducted under
indicate that there is a degree of repeatability and
reproducibility for some types of assemblies.
the same conditions yield the same results?),
Results depend on factors such as the type of
assembly and materials being tested, the • the acceptance criteria to evaluate the load
characteristics of the furnace, the type and level of carrying capacity of the tested assemblies
applied load, the nature of the boundary conditions (currently tests are stopped before the load
(restraint and end fixity), and details of carrying capacity of the assembly is reached
workmanship during assembly.
 because other acceptance criteria are met or
because the deflection becomes excessive
and assembly failure could damage the
furnace).

REFERENCES

American Institute of Steel Construction (AISC)

(2001). Manual of Steel Construction: Load
and Resistance Factor Design, 3rd Edition,
Chicago, Il.

ASTM (2000). “Standard Test Methods for Fire

Tests of Building Construction and
Materials,” ASTM E 119-00, ASTM
International, Conshohocken, Pa.

ASTM (1973). “Standard Test Methods for Fire

Tests of Building Construction and
Materials,” ASTM E 119-73, ASTM
International, Conshohocken, Pa.

Beitel, J, and Iwankiw, N. (2002). “Analysis of

Needs and Existing Capabilities for
Full-Scale Fire Resistance Testing,” NIST
GCR 02-843, National Institute of Standards
and Technology, Gaithersburg, Md.

NIST (2005). Federal Building and Fire Safety

Investigation of the World Trade Center
Disaster: Final Report on the Collapse of the
World Trade Center Towers. NIST NCSTAR
1. Gaithersburg, MD, September

.
 Table 1. Results of ASTM E119 Standard Fire Tests
Times to Reach Conditions of Acceptance (min) Standard Fire Test Rating (h)
Temperature on Unexposed Test ASTM
Steel Temperatures ASTM E119-00
Surface Failure to Termin-a E 119-61
Test Description Average Maximum Average Maximum Sup-port ted
Load (min) Unrestr-ai
Temp Temp Temp Temp Restr-aine
Rating ned
Ambient+ Ambient+ 1100 ºF 1300 ºF d Rating
Rating
250 ºF (121ºC) 325 ºF (163 ºC) (593 ºC) (704 ºC)
35 ft (10.7 m),
1 --- 111 66 62 (3) 116(1) 1½ 1½ 1
restrained
35 ft (10.7 m), (2)
2 --- --- 76 62 (3) 146 2 --- 2
unrestrained
17 ft (5.3 m), (2)
3 180 157 86 76 (3) 210 2 2 1
restrained
(1) Imminent collapse
(2) Vertical displacement exceeded capability to measure accurately
(3) Did not occur

Figure 1. Floor system tested

Figure 2. Buckled truss webs and bowed metal deck