TERM PAPER
ON
FIRE RESISTANCE
BY
AFNAN ALI
i
Introduction
In 2015 alone, there were 501,500 structure fires, causing 2,685 civilian deaths,
13,000 civilian injuries, and $10.3 billion in property damage in United States. It is
evident that fire incidents in building structures are very high as fire incident in
building is reported every 83 seconds in United States (NFPA 2016). There are
different ways to protect human life and property such as active & passive fire
protection and fire prevention i.e. by educating users and operators of the facility. Fire
prevention methods are mostly used in industries like oil & gas, it normally does not
apply on residential structures. Active measures include fire alarm, detection system
or sprinklers etc. National Fire Protection Associations annual report: The NFPA
has no record of a fire killing more than two people in completely sprinkled building
where the system was properly operating, except in an explosion or flash fire, or
where industrial fire brigade members or employees were during fire suppression
operations. But it is possible for the private and public investors to invent on
expansive sprinklers system? This question brings us back to passive fire protection
methods.
Passive fire protection measures are built into the structural system by choice of
design. While the current prescriptive methodology has proven adequate for
structural engineers and by the steel industry because steel is penalized by the cost of
fire protection (Ellingwood 2015). Furthermore, a wide spread perception exists that
1
prescriptive methods are not cost effective. The shortcoming in the current practices
for providing fire resistance have motivated the recent shift towards performance-
based fire engineering that requires a better response when exposed to elevated
The 1 New York Plaza (New York, NY) a 50-storey high rise office building, having
been constructed on a steel frame and a core of R.C.C, with a passive fire protection
system of steel using SFRM (Spray applied Fire Resistive Material) lacked any
sprinkler system for fire protection. The fire which broke out on August 5, 1970
continued for more than six hours affecting extreme damage to the bolted
connections. Some filler beams dropped on the bottom flange of the main girder due
to the sheared bolts, but the building survived the fire. After an inquiry, it was
determined that during the initial phase when the fire broke out, the light spray on fire
resistive material peeled off that produced damage to the floor system.
The First Interstate Bank (Los Angeles, CA) is a 62-level office building, has a central
concrete core which is surrounded by steel column trees. There was a fire incident on
May 4, 1988 in an open-office area on the 12th floor of this building. Office furniture,
paper documents, and computer terminals were the source of fuel. The fire devastated
12 to 16 floors within 3.5 hours causing an estimated loss of about $200 million. The
fire could not be controlled in its early stage because the sprinkler system that cost
around $3.5million was being installed in the building and was only 90% completed at
2
the time of fire. The building survived the fire although its five floors were completely
It is a 14-floors steel frame office building located in London, England. Its structure is
made of W shape steel beams and steel trusses composite with a concrete slab. The
fire incident took place in this building in 1990 when the building was constructed but
lacked fire protection. Timber products in this building were the main driving source
for fire. This fire lasted for about 4.5 hours and completely destroyed 5 floors of the
building. There was enormous damage to the structural elements of the building but it
One Meridian Plaza is a 38 story high rise building in Philadelphia, Pennsylvania. Its
framing was of steel while core was made up of concrete. It does not had a sprinkler
system, in fact it had fire proofing material sprayed on structural steel. There was a
fire incident on the 22nd floor of the building on February 23, 1991 and lasted for 18
hours. It caused about $100 million in damage by damaging the 8 floors of the
building. In spite of a great damage on the exterior of the building, the main structure
survived this catastrophe and the building did not collapsed. Temporary support
The World Trade Center Fire Building 1 and 2 (NIST NCSTAR 1-6)
The World Trade Center Buildings 1 and 2 which were attacked by terrorists on
September 11th, 2001 were also destroyed by fire. The fire erupted in both the
buildings as soon as the airplanes hit the building. The fuel carried by the airplanes
3
were the driving source of the fire. WTC 1 collapsed after 1.5 hours of the fire, while
WTC 2 did so 1 hour after the fire. The investigation report of WTC collapse by NIST
stated that the collapse was due to the expansion of the floors because of the
temperature increased by fire at an early stage. This expansion pushed the exterior
columns out, further weakening the framing system. Floors showed great sag. The
catenary forces produced in sagged floor systems dragged the exterior columns inside.
The floor no longer acted as a brace to the fire-weakened columns and the interior
gravity columns were no longer capable of carrying the gravity loads. The loss of the
gravity load columns triggered the progressive collapse of whole building. (FEMA
403)
made up of steel while all of its floors are composite. This building consisted of
sprayed on fire proofing and in addition the sprinkler system was also installed. In
spite of all this safety measurements, fire broke out on the 34th floor of the building
and damaged 26 of its floors and lasted 17 hours. This event occurred on October 17,
2004. After investigation, it was concluded that the sprinkler system failed during the
fire. Both sprayed on fire proofing and sprinklers were installed in the building but the
sprinkler system malfunctioned during fire. There was a terrific amount of fire
4
The Windsor Building Fire (INTEMAC, 2005)
constructed with a steel exterior 10 frame and steel-reinforced concrete core. On the
night of February 12, 2005, a fire erupted on its 21st storey and burned the building
for almost 48 hours. The building survived the fire but, the exterior frame from 21 st to
32nd floors collapsed. Due to a shortage of water, it took great amount of time to
5
Behavior of Steel under fire
Steel is a non-combustible material, but its yield strength and Youngs modulus de-
grade quickly at high temperature, which make the steel have a low fire resistance. At
the temperature of 1100F, the steel will lose most of its strength and stiffness (Li
2013). The fire in the building may cause global collapse to the steels structure or to
the structural components. Steel has high thermal conductivity and low specific heat
combine with the loss of stiffness and strength at elevated temperatures (Buchanan
2002). The basis for these rules in the US is the standard fire resistance test used for
building frame elements specified in ASTM E119. This approach has changed little in
the last 100 years, and has been widely recognized as an inadequate basis for
structural fire safety design (Hu 2011). It is more important to evaluate the thermal
and structural interaction between structural components during a realistic fire and
their influence on the overall fire resistance of the building structure. Most columns
test in the United States are being conducted without axial loading in the furnaces due
to expensive facilities required to perform test. These unloaded column tests were
conducted by uniformly heating the column length in a gas furnace following ASTM
E119-2010 T-t (Temperature Time) curve. The design fire resistance rating (FRR)
value is obtained as the time when the average temperature of the member cross
section exceeds 1000F. Unfortunately, these standard fire tests provide insufficient
information regarding the structural behavior and failure of members subjected to fire
loading. In addition, the ASTM E119 (2010) T- t curve does not compare favorably
Over last fifteen years notable analytical and experimental work has been carried out.
Most of which have been carried out in Europe and China. Cardington program was
6
the most significant programs for investigating steel structure under fire. It was an
eight story steel framed structure with composite floors and beams were not fire
protected. Despite the absence of fire protection, the floor system and the entire
demonstrated the potential for significant cost saving in the fire protection while still
maintaining the safety of the steel structure under fire exposure (Hu 2011). Outcome
of the research was that the ability of the floor system to survive a fire by
development of tensile catenary and membrane action resulting from the large vertical
This is a highly beneficial system of a full composite floor system is not recognized
in current US standards and typical practice, where fire protection requirements for
structural steel beams are based on a standard fire test of isolated floor assemblies
(Usmani 2001).
7
Stress/Strain relation of steel in fire
There are two ways to test steel in tension, Steady-state test in which steel sample is
heated without load or any restrained. Then sample is loaded is loaded at which stress
and strains are required. Other method is transient test in which sample is loaded to
required stress level. Constant required stress is applied with increase in temperature
until fracture. Stress measured in the steady-state tests were somewhat longer than
those measured in the transient tests when only small strains are considered. For larger
strains, however, the two testing methods produced similar results (Kirby 1988).
Yu (2006) tested A572 Gr. 50 steel of dia. 3/8 and 1/2 inch. Test samples were heated
to target temperature under zero loads. Then load was applied at constant temperature
until facture was occurred. Figure 4a and figure 4b represents the stress strain curves
of dia. Gr. 50ksi reinforcement bar. From temperature 84F to 392F a clear
plateau can be found on the curve. Beyond 572F yield plateau cannot be found. At
temperatures 572F, 752F and 932F steel entered hardening status right after elastic
part. At 1112F, 1292F and 1472F, hardening part became shorter, followed by
crosshead movement. As temperature went up, the V-Shapes became larger with
wider opening which shows more relaxation and creep behavior in steel at higher
temperature. The creep and relaxation behavior of steel from 752F to 1472F
indicated that loading rate would have significant effects on yield and ultimate
8
Figure 4a: Stress-Strain curve of steel at elevated temperature (Yu 2006)
9
Increase in temperature leading to catenary action in beam
Thermal strains induced in the members through heating the most important factor
that determines a real structure response to the unavoidable thermal strains due to
gradient in the members whose end are rotationally restrained can lead to large
hogging (negative) bending moments throughout the length of the member without
deflection. The effect of induced curvature in members whose ends are rotationally
situation in real structures under fire have a complex mix of mechanical strains due to
applied loading and mechanical strains due to restrained thermal expansion. These
lead to combined mechanical strains which often far exceed the yield values, resulting
in extensive plastification.
Thermal expansion of the simple supported beam which is free to expand from one
end will extend. Only thermal strain will be induced which will result in expansion in
Clearly, usually beams in real life do not have the freedom to expand as shown in
Figure 2a. When the beam shown in Figure 2b is subjected to increase in temperature
10
there is no thermal expansion in longitudinal direction because it is cancelled out by
When the beam has enough capacity, at some point axial stresses will reach the yield
stress and then beam will continue to yield without any further increase in stress
(elastic-plastic behavior). If the temperature is allowed to rise further, then the total
restraining force will stay constant (assuming an elastic material and no thermal
behavior which is the combination of both may also occur. After the significant
increase in compression forces at the restrained points, these forces will change into
11
Similar phenomenon is also explained by Rotter J.M. in his paper, he compares two
beams. In the first case is heated simply supported beam which is laterally
unrestrained with uniformly distributed load. The second case is the beam which is
laterally restrained but rotationally unrestrained beam with the same load. In the first
beam which is laterally unrestrained deflections are lower because the supports are
able to translate outward due to expansion. Runaway occurs at around 842F (even
supports when the flexural stiffness of the beam reduces to a point where it cannot
sustain the imposed load and there is nothing to restrain the growing deflection
(Rotter 1999). In the second beam which is laterally restrained initially deflection
occurs because the beam buckles due to restraining forces very early on 158F and
deflection. But runaway does not occur until much later 1652F when the steel
properties are completely lost (Rotter 1999). This runaway in laterally restrained beam
12
would replace the lost in flexural strength of the beam due to temperature gradient.
Figure 3: Behavior of restrained and unstrained beam under fire load (Rotter 1999).
Thermal Gradient
Variation in behavior of steel member will occur if the member is not under same
thermal gradient on all four sides. As shown in figure 5. In real life thermal gradient is
usually different on all four sides for example if there is fire in the room, the bottom
flange of the roof beam will be under larger temperature than the top flange of the
beam.
13
Figure 5: Fire scenarios and P-M curves at different gradients (Mahmoud 2010).
Stiffness and strength properties of steel will be different during the depth of the
section due to unequal heat distribution in the section. Due to this variation in stiffness
and strength of steel this leads to eccentricity between center of geometry and center
of stiffness, the axial force will act eccentrically on the section and greater bending
moment will be occurred. This bending moment can cause a shift in the plastic P-M
diagram. W24x64 section of the beam, load-moment interaction curve for different
maintained at 932F and the temperature gradient of the member is increase from 0F
to 150F. Through direct integration of the ultimate stresses across the cross section of
Fire Scenario
type of fire scenario. Current information is based on the fire resistance test on
restrained columns or beams typical exposed to the standard fire. But in reality the
behavior of the fire is dependent on the fuel load and ventilation characteristics of the
compartment (Buchanan 2002). Other main different is that in standard fire exposure
14
the cooling effect is ignored which has significant effect of the behavior of members
especially restrained beam columns. As the result of the exhaustion of fuel and/or
ventilation the compartment enters the cooling phase as the fire the in the building
dies down. This change of temperature from high to low has significant effect on the
behavior restrained members particularly. During decay phase of design fire steel
starts to recover its strength and stiffness as the temperature goes down to normal
results in contraction of beam and recovery of its deformation. The recovery depends
on the stress state of steel before the start of cooling period. Under realistic fire there
During the heating process of the restrained beam local flange buckling near the
supports is occurred due to large compressive forces due to restrains, but at the time of
cooling beam-column connection were damaged where the bolt on the header plate on
the header plate of the beam sheared because of contraction of beam (Wang 2000).
Li and Guo (2007) tested the two different specimens of beams. Both members were
double channel beams (symmetric) and top flange was covered with the ceramic
blanket to depict the behavior similar to a composite beam supporting a slab over it.
Ceramic blanket is provided to protect top flange from heating. Only difference
between samples was the distance between the test sample and beam over it. It was
done to change unsupported length of the column which is change the behavior of the
restrains of the beams. The results showed that the behavior of the restrained beam is
wholly different from typical simply supported beams. When the heating is initiated
the compressive forces are generated in the steel beam due to restrains. Then if the
temperature keeps on increasing the deflection in the beams will increase and cause
15
the catenary action and tension forces at the supports this is when thermal expansion
overcomes over the thermal action. Local buckling near the supports were noticed
which resulted in reduction in the stiffness of the beam hugely. During the cooling
period, the tension forces on the restrained are increased due to contraction. Larger is
the axial stiffness of the restrains, larger the tension the force will be (Li 2007).
Behavior of structural bolts has significant impact on the performance of end framing
steel under fire. Normally, structural bolts are made of high strength steel and they act
the steel used in beam or column because of the difference in proportions of alloys in
Testing methods are similar to ambient temperature, the strength of structural bolts are
determined using bolt shear factor test and bolt fraction tension test. Hardness is used
Eurocode 3 (2006) gives the strength reduction factors for bolts under high
temperature. These factors can be seen in table 1. It is evident from the table below
that strength of bolts reduces to almost fifty percent at temperatures above 500C.
16
Table 1 Strength Reduction factor for bolts (Euro 3 2006)
Behavior of ASTM bolts under double has been studied by (Yu 2006) at elevated
temperatures. ASTM A325 and A490 were tested under shear at temperature up to
800C. He also investigated the residual shear capacity of A325 and A490 bolts at
Outcome of this study was that the bolts of both type A325 and A490 lost their
cooling rate and heating time are not significant which were where found during the
test.
17
Figure 6: Shear capacities of A325 and A490 bolts (Yu 2006)
Bolted Connections
Single bolt connections having plate thickness of 3/8 and grade 50 tested by (Yu
2006). The dynamic yield strength was found to be 57.1ksi and dynamic ultimate
strength of plate is 78.1 ksi at ambient temperature. Bolt has dia. of 7/8 and the
end distance is 1xDia and 1.5xDia was tested under axial tensile load. Bolt type is
A325 with 3 length. The bolts were snugged tight only which reduced friction
18
force between plates connected. Both type of end distances 1xD and 1.5x D were
Figure 8 summarizes the results of the tests. At different temperature intervals load
capacity of single bolt connection and end hole distances are shown. It can be seen
that from ambient temperature to 300C the capacity increase from approximately
19
Conclusion
Frequency of the building structure fire in United State in substantially high and
with popularity of steel structure, there are many areas to explore to in the of fire
It has been realized that the behavior of steel under fire is quite dissimilar at
steel beams and composite floors have higher capacity than the prediction from
conventional fire in the past. More experimental data is required to improve the
predictive methods steel under fire. Catenary action plays an important role in fact
it increases the load carrying capacity of the floor beams but the forces induced
from catenary action may become the reason of failure for connections. Shear
connections suspected to fire must be designed for shear force as well as axial
forces induced due to restrain and catenary actions. Post fire effects on the
structure should be also taken into account because steel tries to regain its stiffness
while cooling. Better understand of the different fire scenarios is also required
because action such as bowing effect may occur in roof beam with bottom flange
20
References
Li, G., Wang, P., 2013. Advance Analysis and Design for Fire Safety of Steel Structures,
Berlin, Germany. Springer-Verlag Berlin Heidelberg.
Mahmoud, H., Ellingwood, B., Turbert, C., Memari, M., 2015 Response of Steel Reduced
Beam Section Connections Exposed to Fire., J. Struct. Eng., 2016, 142(1): 04015076.
Phan, T., L., McAllister, P., T., Gross, L., J., Hurley, J., M., 2010 Best Practice Guidelines
for Structural Fire Resistance Design of Concrete and Steel Buildings Report NIST
Technical Note 1681.
Buchanan, A. 2002. Structural Design for Fire Safety, John Wiley and Sons, Ltd., England.
Hu, G., 2011. Behavior of Beam Shear Connections in Steel Buildings Subject to Fire
Ph.D. Dissertation, Dept. of Civil and Environmental Engineering, Univ. of Texas at Austin,
Austin, TX.
Varma, H., A., Choe, L., Agarwal, A., Surovek, A., 2011 Fundamental Behavior of Steel
Beam-Columns and Columns under Fire Loading: Experimental Evaluation J. Struct. Eng.,
2011, DOI: 10.1061/(ASCE)ST.1943-541X.0000446
Hu., G., Engelhardt M., (2011),"Investigations on the Behavior of Steel Single Plate Beam
End Framing Connections in Fire", Journal of Structural Fire Engineering, Vol. 2 Iss 3 pp.
195 204
Usmani, A.S., Rotter, J.M., Lamont, S., Sanad, A.M. and Gillie, M. 2001. Fundamental
Principles of Structural Behavior under Thermal Effects, Fire Safety Journal, Vol. 36,
Elsevier Science Ltd., p 721-744.
Kirby, B.R., Preston, R.R. 1988. High Temperatures Properties of Hot-rolled Structural
Steels for use in Fire Engineering Design Studies, Fire Safety Journal, Vol. 13, #1,
pp27-37.
Yu, L. 2006. Behavior of Bolted Connections during and After a Fire, Ph.D. Dissertation,
Department of Civil, Architectural and Environmental Engineering, University of Texas
at Austin.
Mahmoud, M., S., D., 2010. Response of Restrained Steel Beams Subjected to Fire Induced
Thermal Gradients. Ph. D. Dissertation, Department of Civil, Architectural and
Environmental Engineering, University of Texas at Austin.
Wang, Y., C., An analysis of the global structural behavior of the Cardington steel-framed
building during the two BRE fire tests. Engineering Structures 2000;22:40112.
Li., G., Q., Guo., S., X., 2007, Experiment on restrained steel beams subjected to heating
and cooling. c 2007 Elsevier Ltd. All rights reserved.
Eurocode 3. 2006. Design of steel structures. Part 1-2: General rules. Structural fire design,
EN 19931-2 European Committee for Standardization. CEN.
FEMA, World Trade Center Building Performance Study: Data Collection, Preliminary
Observations, and Recommendations, FEMA 403, May 2002
Beitel, J.J. and Iwankiw, N.R., Historical Survey of Multi-story Building Collapse Due to
Fire, Fire Protection Engineering, 2005
FEMA Technical Report Series, Interstate Bank Building Fire Los Angeles, California, May
4, 1988
FEMA Technical Report Series, High-rise Building Fire, One Meridian Plaza, Philadelphia,
Pennsylvania, February 23, 1991
NIST NCSTAR 1-6, Federal Building and Fire Safety Investigation of the World Trade
Center Disaster, Structural Fire Response and Probable Collapse Sequence of World Trade
Center Towers, NIST, September 2005
Fire in the Windsor Building, Madrid Survey of the Fire Resistance and Residual Bearing
Capacity of the Structure after the Fire, INTEMAC, NIT- 2(05)
Caracas Tower Floor Beams Deflected but Did Not Collapse, Engineering News Record,
November 15, 2004