OVERVIEW

• This section will:

– provide information to calculate performance of steel
DESIGN OF STEEL buildings exposed to fires
– describe simple methods for designing individual steel
STRUCTURES IN FIRE members to resist fire exposure, including:
• calculations of elevated temperatures
• methods of fire protection
• information on the thermal and mechanical properties of steel
at elevated temperatures

Winter 2003 Design of Steel Structures 1 Winter 2003 Design of Steel Structures 6-2

BEHAVIOUR OF STEEL BEHAVIOUR OF STEEL

STRUCTURES IN FIRE STRUCTURES IN FIRE
• When steel structures are under fire exposure: • Protected steel members can have excellent fire
– steel temperatures increase resistance
– strength and stiffness of the steel are reduced • Unprotected steel members perform poorly in fires
• This leads to deformation and failure depending • Steel has high thermal conductivity values than
on applied loads and support conditions most other materials
• Thermal expansion of steel members can cause
• Increase in steel temperatures depends on:
damage in other parts of the building
– fire severity
– area of steel exposed to fire
– amount of applied fire protection

Winter 2003 Design of Steel Structures 6-3 Winter 2003 Design of Steel Structures 6-4

BEHAVIOUR OF STEEL
FIRE-RESISTANCE RATINGS
STRUCTURES IN FIRE
• The main factors affecting the behaviour of steel • Verification Methods
structures in fire are as follows: • Generic Ratings
– elevated temperatures in the steel members
• Proprietary Ratings
– applied loads on the steel members
• Calculated Ratings
– mechanical properties of steel members
– geometry of the steel members

Winter 2003 Design of Steel Structures 6-5 Winter 2003 Design of Steel Structures 6-6

1
 Verification Methods Verification Methods
• Design for fire resistance requires: provided fire • Required fire resistance is then compared with the
resistance > design fire severity fire-resistance rating of the selected assembly
• The verification may be in the time domain, • In temp. domain, comparison is made between
the limiting steel temp. and the maximum temp.
temperature domain or strength domain
reached during fire exposure
• Time domain method is traditionally used for fire- • The limiting temperature is usually provided by
resistance ratings codes and standards
• In the time domain, required fire resistance may • In strength domain, load-bearing capacity of the
be prescribed by codes or calculated by formulae steel member is compared with the expected load
on the member during fire

Winter 2003 Design of Steel Structures 6-7 Winter 2003 Design of Steel Structures 6-8

Generic Ratings Proprietary Ratings

• Generic ratings assign a time of fire resistance to
materials with no reference to manufacturers • Manufacturers of passive fire protection products
provide listings of ratings
• Many codes provide lists of generic ratings for
• The ratings are similar to generic ratings but they
structural steel members
relate to more closely defined products
• The table below is an example taken from NBC • Proprietary ratings usually include reference to
Minimum thickness of solid concrete protection to the size and shape of the member
steel columns to provide fire resistance (NBC) • Below is an example of proprietary ratings
Time (hours) 1/2 3/4 1 1.5 2 3 4
Thickness (mm) 25 25 25 25 39 64 89

Winter 2003 Design of Steel Structures 6-9 Winter 2003 Design of Steel Structures 6-10

Proprietary Ratings Proprietary Ratings

Steel beam and column protected with spray-on
• Thickness of proprietary spray-on protection material
required to provide fire resistance to a steel beam
or column (ASFPCM, 1988)
• F/V is section factor and V/F effective thickness
Section size Fire resistance
F/V (m-l) V/F (mm) 1 hour 2 hours 3 hours 4 hours
70 14.3 10 22 36 50
110 9.1 10 28 47 65
150 6.7 12 33 54 75
190 5.3 13 37 60 83
230 4.3 14 39 64 89
270 3.7 15 41 68 94

Winter 2003 Design of Steel Structures 6-11 Winter 2003 Design of Steel Structures 6-12

2
 Calculated Ratings STEEL TEMPERATURES

• Calculation methods for steel structures under • To design steel structures for fire, the temperature
fire exposure will compare loads with load of the steel must be known
capacity in the strength domain • The exposure may be standard time-temp. curve
• There are two main types of calculation: or real curves
– simple calculation model which is used for single
members
– general calculation model which requires use of
computer program for analysis of complex structures

Winter 2003 Design of Steel Structures 6-13 Winter 2003 Design of Steel Structures 6-14

Calculation Methods Calculation Methods

• In simple hand calculations, best-fit empirical • Available design charts can provide graphical
formulae are used to obtain the temperature results of simple calculations
• Simple hand calculations assume uniform • Advanced computer-based methods can calculate
temperature over the cross section of steel temperatures within a cross section
• A simple computer program can be used with a • In most calculations, 2-D models are suitable,
step-by-step calculation technique: assuming uniform temp. along the member
– with a constant temperature over the cross section • 3-D heat transfer calculations may be useful, e.g.,
– assuming a lumped mass of steel at uniform temp. at member junctions
• This method can be used with any design fire
curve as input
Winter 2003 Design of Steel Structures 6-15 Winter 2003 Design of Steel Structures 6-16

Section Factor Section Factor

• The rate of temp. rise in structural steel members • The section factor can be expressed in one of
exposed to fire depends on the section factor following four different methods:
• The section factor is a measure of ratio of heated – ratio of heated surface area to volume, both per unit
length, F/V (m-1)
perimeters to the area/mass of the cross sections
– ratio of heated perimeter to cross section area, Hp/A
• The section factor is important because the rate of (m-1)
heat input is directly proportional to the area – ratio of heated surface area to mass, both per unit
exposed to the fire environment length, F/ M (m2/tonne)
– effective thickness, V/F or A/Hp (m or mm)
• The first two ratios are identical (see Table below)

Winter 2003 Design of Steel Structures 6-17 Winter 2003 Design of Steel Structures 6-18

3
 Section Factor Section Factor
Table defining section factor in the Eurocode
• F = surface area of unit length of member (m2)
• V = volume of steel in unit length of member (m3)
• Hp = heated perimeter of cross section (m)
• A = cross-sectional area of section (m2)
• M = mass per unit length of member (tonne)

Winter 2003 Design of Steel Structures 6-19 Winter 2003 Design of Steel Structures 6-20

Section Factor WORKED EXAMPLE 1

• The ratio, V/F or A/Hp is an effective thickness of • Calculate the section factor for a steel H-section
the cross section column, of dimensions 300x300 mm. The column
is exposed to fire on all four sides.
• The section factor is:
• Make calculations for (a) box-type protection and
– V/F = t/2 for a steel plate of thickness t, exposed to a
fire on both sides
(b) spray-on protection
– V/F = t for a hollow tube of thickness t GIVEN:
– V/F is one half of the average thickness of the • Height of section h = 300 mm
different parts for an I-beam • Width of section b = 300 mm
• See Tables provided from appendix C in • Flange thickness T = 20 mm
Textbook • Web thickness t = 8 mm
Winter 2003 Design of Steel Structures 6-21 Winter 2003 Design of Steel Structures 6-22

WORKED EXAMPLE 1 WORKED EXAMPLE 1

CALCULATION:
• Surface area of 1 m length F = Hp x 1.0m = 1.2 m2
(a) Box-type protection
• Section factor Hp/A = 1.2/0.01408 = 85.2 m-1
• Area of cross section:
• Section factor F/V = 1.2/0.01408 = 85.2 m-1
A = 2(b x T) + t(h-2T) = 14080 mm2 = 0.01408 m2
• Effective thickness V/F = 1000/(F/ V) = 11.7 mm
• Volume of 1 m length:
V = A x 1.0m = 0.01408 m3
• Perimeter of section:
Hp = 2 (b+h)= 1200 mm = 1.2 m

Winter 2003 Design of Steel Structures 6-23 Winter 2003 Design of Steel Structures 6-24

4
 WORKED EXAMPLE 1 Thermal Properties
(b) Spray-on protection • For calculating temperatures in fire-exposed
• Perimeter of section structures, knowledge of materials thermal
Hp = 2 (b + h + (b-t)) = 1784 mm = 1.78 m properties is necessary
• Surface area of 1 m length F=Hp x 1.0m = 1.78 m2 • The density of steel is 7850 kg/m3 and remains
essentially constant with temperature
• Section factor Hp/A = 1.78/0.01408 = 126m-1
• The specific heat of steel varies with temp.
• Section factor F/ V = 1.78/0.01408 = 126 m-1
• For example, Eurocode gives the relation shown
• Effective thickness V/F = 1000/(F/V)=7.9 mm
in the Figure below (peak at ~ 730°C)

Winter 2003 Design of Steel Structures 6-25 Winter 2003 Design of Steel Structures 6-26

Thermal Properties Thermal Properties

Specific heat of steel as a function of temperature • For simple calculations, the specific heat can be
taken as Cp = 600 J/kg-K
• It is more accurate to use the equations below:
Cp= 425+0.773T-1.69x10-3T2+2.22x10-6T3
for 20°C ≤ T < 600°C
Cp= 666+13002/(738-T) for 600°C ≤ T < 735°C
Cp= 545+17820/(T-731) for 735°C ≤ T < 900°C
Cp= 650 for 900°C ≤ T ≤ 1200°C
• T is the steel temperature (°C)
Winter 2003 Design of Steel Structures 6-27 Winter 2003 Design of Steel Structures 6-28

Thermal Properties Thermal Properties

Thermal conductivity of steel as a function of
• The thermal conductivity, k, of steel varies with temperature
temp. as shown in the Figure below (Eurocode)
• k reduces linearly from 54 W/m-K at 0°C to 27.3
W/m-K at 800°C
• For simple calculations, thermal conductivity can
be taken as k = 45 W/m-K

Winter 2003 Design of Steel Structures 6-29 Winter 2003 Design of Steel Structures 6-30

5
 Thermal Properties Temperature Calculation for
Unprotected Steel
• It is more accurate to use the equations below: • Under fire, unprotected steel members can reach
k = 54 - 0.0333 T for 20°C ≤ T < 800°C high temp. quickly
k = 27.3 for 800°C ≤ T ≤ 1200°C • Below, two 'lumped mass' methods of calculating
• T is the steel temperature (°C) temperature are explained
• With significant temp. gradients within the cross
section, the lumped-mass method is not valid, e.g.
I-beam with a concrete slab on top
• In this case a FEM should be used

Winter 2003 Design of Steel Structures 6-31 Winter 2003 Design of Steel Structures 6-32

Unprotected Steel - Unprotected Steel -

Best-fit method Best-fit method
• An empirical formula to predict the time t (min) for • The equation can be used to calculate the
unprotected steel members to reach a limiting expected temperature at a particular time
temp. Tlim (°C) when exposed to standard fires is:
• This is an approximate best-fit empirical equation
t = 0.54(Tlim - 50)/(F/V)0.6 • For more accurate design, it is recommended that
• F/V is the section factor (m-1) temperatures be calculated using the step-by-step
• This expression is valid for: method (below) or a FEM
– 10 ≤ F/V ≤ 300 m-1 (3.3 ≤ V/F ≤ 100 mm)
– 400°C ≤ Tlim ≤ 600°C
– 10 min ≤ t ≤ 80 min
Winter 2003 Design of Steel Structures 6-33 Winter 2003 Design of Steel Structures 6-34

Unprotected Steel - Unprotected Steel -

Step-by-step method Step-by-step method

• The step-by-step method for unprotected steel is • cs is the specific heat of steel (J/kg K)
based on the principle that heat entering the • ∆Ts is the change in steel temperature in the time
steel over the exposed surface area in a small step (°C or K)
time step ∆t (s) is equal to heat required to raise • q” is the heat transfer at the surface (W/m2) as:
the temp. of the steel by ∆Ts (°C) assuming that
q" = hc(Tf -Ts) + σε(Tf4 – Ts4)
the section is a lumped mass at uniform temp.:
• hc is the convective heat transfer coefficient
heat entering = heat to raise temperature (W/m2K)
q” F ∆t = ρs cs V ∆Ts
• σ is the Stefan-Boltzmann constant (56.7x10-12
• ρs is the density of steel (kg/m3) kW/m2K4)
Winter 2003 Design of Steel Structures 6-35 Winter 2003 Design of Steel Structures 6-36

6
 Unprotected Steel - Unprotected Steel -
Step-by-step method Step-by-step method
• ε is the resultant emissivity • Eurocode recommends a value of resultant
• Tf is the temperature in the fire environment (K) emissivity of 0.50
• Ts is the temperature of the steel (K) • Eurocode suggests a time step of no more than
30 s, and a minimum value of the section factor
• Re-arranging the equations give:
F/V of 10 m-1 (maximum value of V/F of 100 mm)
∆Ts = (F/V)(1/(ρs cs)) [hc(Tf -Ts) + σε(Tf4-Ts4)] ∆t • This type of calculation can give good predictions
• The convective heat transfer coefficient, hc, is of unprotected steel beam temp. in standard fire-
recommended to have a value of 25 W/m2K resistance tests
• Eurocode recommends 25 W/m2K for standard • See table below for the calculation method
fires and 50 W/m2K for hydrocarbon fires
Winter 2003 Design of Steel Structures 6-37 Winter 2003 Design of Steel Structures 6-38

Unprotected Steel -
WORKED EXAMPLE 2
Step-by-step method
Table showing a spreadsheet calculation for • Use the step-by-step method to calculate the steel
temperatures of unprotected steel sections temperature of an unprotected beam exposed to
the ISO 834 standard fire.
Time Steel temperature Fire temperature Difference in Change in steel
Ts Tf temperature temperature ∆Ts
t1 = ∆t Initial steel Fire temperature T f - Tso Calculate from
temperature Tso halfway through time
step (at ∆t/2)
equation of ∆Ts
with values of Tf
• The beam section factor F/V is 200 m-1. Use a
and Tso from this convective heat transfer coefficient hc = 25 W/m2K
row
t2 = t1 + ∆t Ts from previous time Fire temperature half T f - Ts Calculate from and emissivity =0.6. The density of steel is 7850
step + ∆Ts from way through time step Equation of ∆Ts
previous row (at t1 + ∆t/2) with values of Tf kg/m3 and the specific heat is 600 J/kg-K. Use a
and Ts from this
row time step of 0.5 min.

Winter 2003 Design of Steel Structures 6-39 Winter 2003 Design of Steel Structures 6-40

Temperature Calculation for

WORKED EXAMPLE 2
Protected Steel
• The first two minutes of the solution are shown in
the table below. The results are plotted in Figure • With applied thermal insulation, protected steel
6-54. members heat up more slowly than unprotected
members
Time Time at half Steel ISO fire Difference in Change in
(minutes) step temperature temperature at temperature steel
Ts half step Tf temperature • Two approximate calculation methods are given
0.0
0.5
0.25
0.75
20.0
26.8
184.6
311.6
164.6
284.7
6.8
13.8
below (as in unprotected steel)
1.0 1.25 40.6 379.3 338.7 18.2
1.5 1.75 58.8 425.8 366.9 21.5 • For steel members protected with heavy
2.0 2.25 80.3 461.2 380.9 24.0 insulating materials or with temp.-dependent
2.5
3.0 thermal properties, a FEM should be used

Winter 2003 Design of Steel Structures 6-41 Winter 2003 Design of Steel Structures 6-42

7
 Protected Steel - Protected Steel -
Best-fit method Best-fit method

• The following approximate formula predicts the • This equation is valid in the following ranges:
time t (min) for a steel member protected with – 30 ≤ t ≤ 240 min
light, dry insulation to reach a limiting temp. Tlim – 400°C ≤ Tlim ≤ 600°C
(°C) when exposed to the standard fire: – 10 ≤ F/V ≤ 300 m-1 (3.3 ≤ V/F ≤ 100 mm)

t = 40 (Tlim - 140) [(di / ki)/(F/V)] 0.77 – 0.1 ≤ di/ki ≤ 0.3 m2K/W

• ki is the thermal conductivity of insulation (W/m-K)
• di is the thickness of the insulation (m)

Winter 2003 Design of Steel Structures 6-43 Winter 2003 Design of Steel Structures 6-44

Protected Steel - Protected Steel -

Best-fit method Step-by-step method
• For insulation containing moisture, a time delay tv • The iterative calculation method for protected
(min) can be added to the time t (previously steel is similar to that for unprotected steel
calculated), using: • It is assumed that the internal surface of the
tv = m ρi di2 / (5ki) insulation is at the same temp. as the steel
• ρi is the insulation density (kg/m3) • The equation is:
• m is the insulation moisture content (%) ∆Ts = (F/V)(ki/di ρs cs)
• For more accurate design, it is recommended that [ρscs / (ρscs+(F/V) di ρi ci / 2)] (Tf -Ts) ∆t
temperatures be calculated using the step-by-step
method (below) or a FEM • ci is the specific heat of the insulation (J/kg K)

Winter 2003 Design of Steel Structures 6-45 Winter 2003 Design of Steel Structures 6-46

Protected Steel - Protected Steel -

Step-by-step method Step-by-step method
• The table used for unprotected steel can be used • Ai is the insulating material cross-sectional area
to calculate protected steel with ∆Ts changing and A is the steel cross sectional area
• Eurocode suggests a time step of 30 s • The effect of time delay for moist materials can
• For insulation with low mass and specific heat, i.e. be incorporated into the calculation by modifying
insulation heat capacity will not highly slow temp. the specific heat of the insulating material to
increase of steel, then the term in [ ] is ignored include a local increase of specific heat at 100°C
• The insulation heat capacity can be ignored if it is • The table below shows typical values of thermal
less than 1/2 of that of the steel section, i.e.,
properties of insulating materials
ρs cs A/2 > ρi ci Ai

Winter 2003 Design of Steel Structures 6-47 Winter 2003 Design of Steel Structures 6-48

8
 Protected Steel -
WORKED EXAMPLE 3
Step-by-step method
Table of thermal properties of insulation materials • Use the step-by-step method to calculate the steel
Thermal Equilibrium temperature of a protected beam exposed to the
Density conductivity Specific heat moisture
ρi ki ci content ISO 834 standard fire. The beam is the same as
Material (kg/m3) (W/m-K) (J/kg K) %
Sprays: 300 0.12 1200 1
in Worked Example 2. The beam is protected with
Sprayed mineral fibre
Perlite or vermiculite plaster 350 0.12 1200 15
50 mm of lightweight insulating material which has
High-density perlite or 550 0.12 1200 15 thermal conductivity of 0.2 W/m-K, specific heat
vermiculite plaster
Boards: 600 0.15 1200 3 1100 J/kg K and density 300 kg/m3.
Fibre-silicate or fibre-calcium
silicate
Gypsum plaster 800 0.20 1700 20
Compressed fibre boards: 150 0.20 1200 2
Mineral wool, fibre silicate

Winter 2003 Design of Steel Structures 6-49 Winter 2003 Design of Steel Structures 6-50

WORKED EXAMPLE 3 Typical Steel Temperatures

• The first two minutes of the solution are shown in
the table below. The results are plotted in Figure • The Figure (a) below shows steel temperatures
6-54, with another curve for 15 mm of insulation. for a beam with F/V = 200 m-1 (V/F = 5 mm),
exposed to the ISO standard fire, calculated using
Time
(minutes)
Time at half
step
Steel
temperature
ISO fire
temperature at
Difference in
temperature
Change in
steel
the step-by-step method:
Ts half step Tf temperature
0.0 0.25 20.0 184.6 164.6 0.62
– the top curve is the ISO fire temperature
0.5 0.75 20.6 311.6 290.9 1.10 – the second curve is the temperature of an unprotected
1.0 1.25 21.7 379.3 357.6 1.35
1.5 1.75 23.1 425.8 402.7 1.52 steel beam
2.0 2.25 24.6 461.2 436.6 1.65
2.5 – The lower two curves are protected with insulating
3.0 material, using thickness values of 15 and 50 mm

Winter 2003 Design of Steel Structures 6-51 Winter 2003 Design of Steel Structures 6-52

Typical Steel Temperatures Typical Steel Temperatures

Typical steel temp. for unprotected/protected steel
beams exposed to standard (a) and real (b) fires
• The Figure (b) shows steel temperatures for the
same beam exposed to a parametric fire,
calculated using the step-by-step method:
– the top curve is the fire temperature
– the second curve is the steel beam with no protection
– the lower two curves are protected with insulating
material, using thickness values of 15 and 50 mm

Winter 2003 Design of Steel Structures 6-53 Winter 2003 Design of Steel Structures 6-54

9
 WORKED EXAMPLE 4 WORKED EXAMPLE 4

• Use the step-by-step method to calculate the steel • Length of room l1 = 5.0 m
temperature of an unprotected beam exposed to a • Width of room l2 = 5.0 m
parametric fire. The beam is the same as in • Height of room Hr = 3.0 m
Worked Example 2. The fire compartment is
• Area of internal surfaces:
made from lightweight concrete with density 2000
kg/m2, specific heat 840 J/kg-K and thermal At = 2 (l1l2 + l1 Hr + l2 Hr)
conductivity 0.8 W/m-K. The room is 5 m square At = 2 (5 x 5 + 5 x 3 + 5 x 3) = 110 m2
and 3 m high with one window 2.4 m wide and 1.5 • Height of window Hv = 1.5 m
m high. The fuel load is 800 MJ/m2 floor area. • Width of window B = 2.4 m
• Area of window Av = B Hv = 2.4 x 1.5 = 3.6 m2
Winter 2003 Design of Steel Structures 6-55 Winter 2003 Design of Steel Structures 6-56

WORKED EXAMPLE 4 WORKED EXAMPLE 4

• Ventilation factor
Fv = Av √ Hv /At = 3.6 x √1.5/110 = 0.04 m-1/2 • Thermal inertia b = √kρcp = 1160 W s0.5/m2K
(medium)
• Fuel load (floor area) ef = 800 MJ/m2
• Gamma factor Γ =(Fv/0.04)2/(b/1900)2 = 2.69
• Fuel load (total area)
• The parametric fire can be calculated using this
et = ef Af / At = 800 x 25.0/110 = 182 MJ/m2
value of gamma.
• Thermal conductivity k = 0.8 W/m-K
• The first two minutes of the solution are shown in
• Density ρ = 2000 kg/m3 the table below. The results are plotted in Figure
• Specific heat cp = 840 J/kg-K 6-54.

Winter 2003 Design of Steel Structures 6-57 Winter 2003 Design of Steel Structures 6-58

WORKED EXAMPLE 4 WORKED EXAMPLE 5

• Repeat Worked Example 4 with the beam

Time
(minutes)
Parametric fire
temperature
Steel
temperature Ts
Fire temperature
(average of this and
Difference in
temperature
Change in steel
temperature
protected with 50 mm of the insulation as in
0.0 0.0 20.0
next step) Tf
114.6 94.6 3.7
Worked Example 3.
0.5 229.2 23.7 305.9 282.2 13.5
1.0 382.6 37.2 434.5 397.3 23.1 • The first two minutes of the solution are shown in
1.5
2.0
486.4
557.8
60.3
91.8
522.1
582.8
461.8
491.0
31.5
38.1
the table below. The results are plotted in Figure
2.5
3.0
607.8
6-54.

Winter 2003 Design of Steel Structures 6-59 Winter 2003 Design of Steel Structures 6-60

10
 WORKED EXAMPLE 5 Typical Steel Temperatures

• Temp. of any steel members exposed to the

Time Parametric fire Steel Fire temperature Difference in Change in steel
(minutes) temperature temperature Ts (average of this and temperature temperature standard fire can easily be obtained from charts
next step) Tf
0.0 0.0 20.0 114.6 94.6 0.36 • The Figure below gives the temperature of
0.5 229.2 20.4 305.9 285.6 1.08
1.0 382.6 21.4 434.5 413.1 1.56 unprotected steel members after any time of
1.5 486.4 23.0 522.1 499.1 1.88
2.0 557.8 24.9 582.8 557.9 2.11 exposure, as a function of F/V (m-1)
2.5 607.8
3.0 • This can be used for any member for which the
section factor is known

Winter 2003 Design of Steel Structures 6-61 Winter 2003 Design of Steel Structures 6-62

Typical Steel Temperatures Typical Steel Temperatures

Chart for calculating temperatures of unprotected
steel members exposed to the standard fire • The Figure below gives the temperature of
protected steel members after any time of
exposure, as a function of the modified section
factor (F/V)(ki/di) (W/m3K)
• This can be used for any member for which the
section factor and insulation are known

Winter 2003 Design of Steel Structures 6-63 Winter 2003 Design of Steel Structures 6-64

Typical Steel Temperatures Temperature Calculation for

Composite Construction
Chart for calculating temperatures of protected steel
members exposed to the standard fire • Composite construction refers to combined
structural systems of steel and concrete both
contributing to the load-bearing capacity
• In many composite structures, steel members
are protected from fire exposure by concrete
• Figure below shows an example of a composite
system (a concrete slab on a steel beam)
• Above equations may be used as rough
calculations

Winter 2003 Design of Steel Structures 6-65 Winter 2003 Design of Steel Structures 6-66

11
 Temperature Calculation for Temperature Calculation for
Composite Construction Composite Construction
Composite construction with concrete slab on steel • Heat-transfer programs are necessary for
deck and steel beam accurate predictions of temperatures
• For light steel framing members used in wall and
floor assemblies, accurate prediction of
temperatures requires a heat transfer computer
program (will be studied in future lectures)

Winter 2003 Design of Steel Structures 6-67 Winter 2003 Design of Steel Structures 6-68

PROTECTION SYSTEMS Concrete Encasement

• A number of alternative passive fire protection • A traditional method for fire protection of steel is
systems are available to reduce temperature encasement in poured concrete
increase in steel structures exposed to fire • The required concrete thickness to achieve the
required standard ratings is given in codes
• Concrete encasement is not widely used because
it is expensive and time-consuming

Winter 2003 Design of Steel Structures 6-69 Winter 2003 Design of Steel Structures 6-70

Board Systems Spray-on Systems

• Spray-on protection is often the cheapest form of
• Board systems can be used to protect structural passive fire protection for steel members
steel • Spray-on materials are usually cement-based
• Gypsum board has good insulating properties with some form of glass or cellulosic fibrous
• Board systems have the advantages that they are reinforcing to hold the material together
easy to install and finish • Spray-on protection has the disadvantages of
being wet and messy
• Board systems are more often used for columns
• Spray-on materials are more often used for
than for beams beams than for columns
• Boards are usually glued or screwed to metal or • Spray-on protection is easy to apply to
wood framing complicated details such as connections

Winter 2003 Design of Steel Structures 6-71 Winter 2003 Design of Steel Structures 6-72

12
 Intumescent Paint Concrete Filling

• Intumescent paint is a special paint that swells up • Hollow steel sections can be filled with concrete to
into a thick charry mass when heated improve the fire performance
• Several coats of paint may have to be applied to • The filling concrete can be plain or reinforced
obtain the necessary thickness • The steel tube can provide excellent structural
• A disadvantage of intumescent paint is the high confinement to the concrete at ambient
cost compared to board and spray-on materials • It is essential to provide vent holes to prevent
• All intumescent paints are proprietary products excessive steam pressure from exploding the
and many are under continual development hollow member during heating

Winter 2003 Design of Steel Structures 6-73 Winter 2003 Design of Steel Structures 6-74

Water Filling MECHANICAL PROPERTIES OF

STEEL AT ELEVATED
• A less common but effective way of preventing TEMPERATURE
rapid heating of hollow steel sections is to fill them
with water • In the following, the effects of mechanical
• A plumbing system is necessary to ensure that properties of the steel on the behaviour of steel
the system works structures in fire are reviewed
• Additives may be necessary to prevent corrosion,
and to prevent freezing in cold climates
• This method of protection is expensive and is
used for special structures only

Winter 2003 Design of Steel Structures 6-75 Winter 2003 Design of Steel Structures 6-76

Components of Strain Components of Strain

• The deformation of steel at elevated temp. can • For simple structural members such as S.S.
be formulated by the change in strain ∆ε as: beams, only the stress-related strain needs to be
∆ε = ε -εi = εth(T) + εσ (σ,T) + εcr(σ,T,t) considered (reduced strength at elevated temp.)
• ε is the total strain at time t • For more complex structural systems, the thermal
and creep strains must also be considered (use of
• εi is the initial strain at time t = 0
computer models)
• εth(T) is the thermal strain
• εσ (σ,T) is the stress-related strain
• εcr(σ,T,t) is the creep strain

Winter 2003 Design of Steel Structures 6-77 Winter 2003 Design of Steel Structures 6-78

13
 Thermal Strain Thermal Strain
• The thermal strain is the thermal expansion that • This gives the thermal elongation of steel ∆L/L
occurs when materials are heated as function of temperature T (°C) by:
• At ambient, the coefficient of thermal expansion ∆L / L = 14 x 10-6 (T - 20)
is usually taken to be 11.7x10-6/°C
• For the design of simple single beams and
• At high temperatures, the coefficient increases columns, it is not usually necessary to calculate
• For normal design purposes, Eurocode and include the effects of thermal strains
recommends a linear coefficient of 14x10-6/°C • Thermal restraint forces developing in beams
are usually beneficial to fire performance

Winter 2003 Design of Steel Structures 6-79 Winter 2003 Design of Steel Structures 6-80

Creep Strain Creep Strain

• At ambient, creep is relatively insignificant in Creep of steel tested in tension
structural steel
• At high temp., creep becomes very significant
especially over temp. of 400 or 500°C
• The Figure below shows that the creep is highly
dependent on temp. and stress level
• The creep deformations accelerate rapidly when
creep strain curves becomes almost vertical
• Creep is usually not included explicitly in fire
design models (lack of data and knowledge)
• Usually implicitly included in stress-strain curves
Winter 2003 Design of Steel Structures 6-81 Winter 2003 Design of Steel Structures 6-82

Stress-related Strain Stress-related Strain

Stress-strain curves for typical-hot rolled steel at
• Stress-strain relationships at elevated temp. can elevated temperature
be obtained from tests
• Figures below show typical stress-strain curves
for structural steel at elevated temperatures

Winter 2003 Design of Steel Structures 6-83 Winter 2003 Design of Steel Structures 6-84

14
 Stress-related Strain Stress-related Strain
Stress-strain curves for prestressing steel at
elevated temperature • In the figure for hot-rolled, it is seen that:
– yield strength and modulus of elasticity both
decrease with increasing temp.
– ultimate tensile strength increases slightly at
moderate temp. before decreasing at higher temp.
• In the figure for cold-drawn prestressing, the
yield point is not well-defined and behaviour at
elevated temp. is slightly different

Winter 2003 Design of Steel Structures 6-85 Winter 2003 Design of Steel Structures 6-86

Stress-related Strain - Stress-related Strain -

Proof strength and yield strength Proof strength and yield strength
• At ambient, designing structural steel members Stress-strain curves for steel illustrating yield
requires knowledge of the steel yield strength strength and proof strength
• At ambient, most structural steels have a well-
defined yield strength
• At high temp., this good definition disappears
• The Figure below shows a sketch of stress-strain
relationships for a typical steel, showing:
– a well-defined yield strength at normal temp.
– a much softer curve at elevated temp.

Winter 2003 Design of Steel Structures 6-87 Winter 2003 Design of Steel Structures 6-88

Stress-related Strain - Stress-related Strain -

Proof strength and yield strength Design values
• A value of yield strength is required for design at • There is a large scatter in the reporting of steel
elevated temperatures properties at elevated temperatures
• 1% proof strength, as the effective yield strength • A review of data by Harmathy (1993) is shown in
in fire calculations, is recommended the Figures below
• In the Figure, line AB is constructed to pass • The first figure is for hot-rolled steel and the
through 1% strain and is parallel to the linear second for cold-worked steel
elastic portion of the 400°C curve • The scatter is clear, maybe due to lack of clear
• The vertical value of point B defines the 1% proof definition of yield strength
strain, and can be used for any steel temp. • The dotted lines are suggested design values

Winter 2003 Design of Steel Structures 6-89 Winter 2003 Design of Steel Structures 6-90

15
 Stress-related Strain - Stress-related Strain -
Design values Design values
Scatter in published results of hot-rolled steel Scatter in published results of cold-worked steel

Winter 2003 Design of Steel Structures 6-91 Winter 2003 Design of Steel Structures 6-92

Stress-related Strain - Stress-related Strain -

Design values Design values
Design curves for reduction in yield strength and
• For design purposes, different national building modulus of elasticity of steel with temperature
codes proposed slightly different design values
• The Figure below shows examples of typical
relationships for structural, reinforcing and
prestressing steel

Winter 2003 Design of Steel Structures 6-93 Winter 2003 Design of Steel Structures 6-94

Stress-related Strain - Stress-related Strain -

Design values Design values
• The equations representing the lines are: • The limiting temp. (temp. at which an individual
ky,T = (905 - T)/690 for structural steel steel member is expected to fail) are given by:
ky,T = (720 - T)/470 for reinforcing steel • Tlim = 905 - 690 rload structural steel
ky,T = (700 - T)/550 for prestressing steel • Tlim = 720 - 470 rload reinforcing steel
• Tlim = 700 - 550 rload prestressing steel
• ky,T is the ratio of fy,T (yield strength at elevated
temp.) to fy (ambient yield strength)
• The equations above can be used to give the
steel limiting temp. for a given load ratio rload

Winter 2003 Design of Steel Structures 6-95 Winter 2003 Design of Steel Structures 6-96

16
 Stress-related Strain - Stress-related Strain -
Design values Design values
• Another example of curves given by Eurocode, Reduction in yield strength and modulus of elasticity
are shown in Figure below, defining reduction in with temperature
yield strength, elastic limit and elasticity modulus
• Based on a Eurocode critical temp. expression,
an approximation of yield strength reduction is:
ky,T = {0.9674(1 + exp[(T - 482)/39.19])}-1/3.833
• Other equations are published around the world
• Curves used for reduction in yield strength vary
from a country to another (variation in definition)
Winter 2003 Design of Steel Structures 6-97 Winter 2003 Design of Steel Structures 6-98

Stress-related Strain - Stress-related Strain -

Modulus of elasticity Modulus of elasticity
• Buckling calculations requires elasticity modulus • An example of relationship for modulus of
• Under fire exposure, calculations are often in the elasticity is given by the 1st figure as:
elastic range and therefore the modulus of kE,T =1.0+T/[2000 ln(T/1100)] 0 < T ≤ 600°C
elasticity would also be required for elastic kE,T =690(1-T/1000)/(T-53.5) 600 < T ≤ 1000°C
calculations in most cases
• The reduction in modulus of elasticity shows • kE,T is the ratio of ET (high temp. modulus of
similar trend as the reduction in yield strength elasticity) to E (ambient modulus of elasticity)

• Previous Figures showed the trend of E vs. T • The Eurocode reduction in modulus of elasticity
with temp. is shown in the 2nd Figure

Winter 2003 Design of Steel Structures 6-99 Winter 2003 Design of Steel Structures 6-100

DESIGN OF STEEL MEMBERS

Design Methods
EXPOSED TO FIRE
• Design Methods • There are two main methods for structural design
• Design of Individual Members of steel structures exposed to fire:
• Bolted and Welded Connections – simplified design method for single elements
– general method for restrained members, more
complex assemblies, or large frames
• The simplified method is described first

Winter 2003 Design of Steel Structures 6-101 Winter 2003 Design of Steel Structures 6-102

17
 Design Methods - Verification Design Methods - Verification
• Verification in the strength domain requires:
• The design force U*fire may be axial force N*fire,
U*fire ≤ Rfire
bending moment M*fire, or shear force V*fire
• U*fire is design force resulting from applied loads occurring singly or in combination
at the time of the fire
• The load capacity is calculated as axial force Nf,
• Rfire is load-bearing capacity in fire situation bending moment Mf or shear force Vf in the same
• Applied loads have been described earlier combination as design force
• Design forces are obtained from the applied • The calculations of the load capacity are based on
loads using structural analysis the mechanical properties of steel at elevated
• Strength reduction factor Φ is 1 at high temp. temperatures

Winter 2003 Design of Steel Structures 6-103 Winter 2003 Design of Steel Structures 6-104

Design Methods - Verification Design Methods - Verification

• The recommendations for member design, Stress-strain curve with elasto-plastic approximation
presented below, are based on the Eurocode
• The design process can be adapted anywhere
provided calculation equations are available
• The simplified method follows the ultimate
strength design method with values of modulus
of elasticity and yield strength of steel reduced at
elevated temperatures
• Design of steel structures assumes that steel is
ductile with the stress-strain curve below (dotted)

Winter 2003 Design of Steel Structures 6-105 Winter 2003 Design of Steel Structures 6-106

Design Methods - Verification Design of Individual Members

• At ambient, structural design requires prevention
• Tension members
of:
– collapse (strength limit state) and • Simply supported beams
– excessive deformations (serviceability limit state) • Continuous beams
• At ambient, most design process efforts are to • Columns
ensure that excessive deformations do not occur
• Design for fire resistance is mainly concerned with
preventing collapse

Winter 2003 Design of Steel Structures 6-107 Winter 2003 Design of Steel Structures 6-108

18
 Design of Individual Members - Design of Individual Members -
Tension members Tension members
• Single tension members are relatively simple • If the temperature is uniform, the tensile load-
elements to design and the stresses are often bearing capacity is obtained from:
uniform over the cross section Nf = A ky,T fy
• The design equation for tension members is: • A is the area of the cross section (mm2)
N*fire ≤ Nf • ky,T is the reduction factor for yield strength of the
• Design for fire depends on whether the temp. is steel at temperature T
uniform over the cross section • fy is the yield strength of the steel at ambient
(MPa)

Winter 2003 Design of Steel Structures 6-109 Winter 2003 Design of Steel Structures 6-110

Design of Individual Members - Design of Individual Members -

Tension members Tension members
• Often, there is a temperature gradient over the • The design equation is:
cross section, so the strength of the member Nf = ∑i=1,n Ai ky,Ti fy
can be obtained by summing contributions of • Ai is an elemental area of the cross section with a
different parts, considering the temp.-reduced temp. Ti and ky,Ti is the reduction factor for yield
yield strength of each part strength of the steel at temp. Ti
• This equation is based on the assumption that • It is also conservative to assume that the whole
steel is a ductile material, so that sufficient cross section is at the max. temp.
elongation can occur for each elemental area to
• The Figure below shows distribution of internal
develop its yield strength
forces for uniform and non-uniform temp.
Winter 2003 Design of Steel Structures 6-111 Winter 2003 Design of Steel Structures 6-112

Design of Individual Members - Design of Individual Members -

Tension members Simply supported beams
Internal forces in a steel tension member • The design equation for flexure is given by:
M*fire ≤ Mf
• The strength of bending members in fire depends
on temperature uniformity over the cross section
• In addition, susceptibility of the beams cross
section to local buckling should be considered

Winter 2003 Design of Steel Structures 6-113 Winter 2003 Design of Steel Structures 6-114

19
 Design of Individual Members - Design of Individual Members -
Simply supported beams Simply supported beams
• For uniform temperature, the design load-bearing • The decision to use elastic or plastic design
capacity can be obtained as: depends on the compactness of the cross section
Mf = S ky,T fy (plastic design) • The Figure below shows mid-span moment vs.
Mf = Z ky,T fy (elastic design) deflection for a S.S. steel beam showing plastic
• S is plastic section modulus (mm3) behaviour can be achieved for compact sections
• Z is elastic section modulus (mm3) • The equation for plastic design applies if the
• ky,T is reduction factor for yield strength of the shape of the steel section is such that full plastic
steel at temp. T moment can be achieved without local buckling
• fy is yield strength of the steel at ambient (MPa) occurring (Class 1 or 2)

Winter 2003 Design of Steel Structures 6-115 Winter 2003 Design of Steel Structures 6-116

Design of Individual Members - Design of Individual Members -

Simply supported beams Simply supported beams
Moment-deflection relationship for a steel beam
• The equation for elastic design should be used
for sections achieving the elastic moment only
without local buckling occurring (Class 3)
• For light cold-rolled sections susceptible to local
buckling (Class 4), a simple design approach is
to ensure that the steel temperature does not
exceed 350°C or 400°C

Winter 2003 Design of Steel Structures 6-117 Winter 2003 Design of Steel Structures 6-118

Design of Individual Members - Design of Individual Members -

Simply supported beams Simply supported beams
• With a temp. gradient over a cross section, it is • The plastic neutral axis of a flexural section with a
accurate to calculate the temp. of each part, so non-uniform temp. distribution is the axis
that the member strength is obtained by adding perpendicular to the plane of bending such that
up contributions of different parts, considering elemental areas yielding in compression and
temp.-reduced yield strength of each part as: tension on either side of the axis are in equal
Mf = ∑ i=1,n Ai zi ky,Ti fy • The axis is then located such that (at time t):
• zi is the distance from the plastic neutral axis to ∑ i=1,n Ai zi ky,Ti fy = 0
the centroid of the elemental area

Winter 2003 Design of Steel Structures 6-119 Winter 2003 Design of Steel Structures 6-120

20
 Design of Individual Members - Design of Individual Members -
Simply supported beams Simply supported beams
• The figure below shows distribution of internal Internal forces in a steel flexural member
forces in a simple rectangular flexural member
with a uniform and non-uniform temp. gradient
• As in the case of tension members, with a
temperature gradient over the cross section, it is
conservative to assume that the whole cross
section is at the maximum temperature

Winter 2003 Design of Steel Structures 6-121 Winter 2003 Design of Steel Structures 6-122

WORKED EXAMPLE 6 WORKED EXAMPLE 6

Fire calculation in strength domain
• Beam span L = 8.0 m
• For a simply supported steel beam of known span,
• Beam size 410 mm deep and 54 kg/m
load, yield strength, and section properties,
calculate the flexural strength after 15 minutes • This is a 'compact' section (type 1)
exposure to the standard fire. The beam has no • Plastic section modulus S = 1060 x 103 mm3
applied fire protection, and is exposed on 3 sides. • Section factors:
GIVEN • Area to volume ratio F/V = 190 m-1 (Effective
• Dead load Gk = 8.0 kN/m (including self weight) thickness V/F = 5.3 mm)
• Live load Qk = 15.0 kN/m

Winter 2003 Design of Steel Structures 6-123 Winter 2003 Design of Steel Structures 6-124

WORKED EXAMPLE 6 WORKED EXAMPLE 6

COLD CALCULATIONS FIRE CALCULATIONS

• Strength reduction factor Φ = 0.9 • Strength reduction factor Φ = 1.0 (hence not used
• Yield strength fy = 300 MPa in the calculations)
• Design load (cold) wc = 1.2Gk+ 1.6Qk = 33.6 kN/m • Design load (fire) wf = Gk+0.4Qk = 14.0 kN/m
• Bending moment M*cold = wcL2/8 = 269 kN-m • Bending moment M*fire = wfL2/8 = 112 kN-m
• Bending strength Mn = Sfy = 318 kN-m (assume • Temperature after time t:
adequate lateral restraint) T = 1.85t (F/V)0.6+50 (best fit Equation)
• Design flexural strength ΦMn = 286 kN-m • Temperature after 15 minutes:
• Design is OK (M*cold < ΦMn) T = 1.85 x 15 x 1900.6 + 50 = 696°C
Winter 2003 Design of Steel Structures 6-125 Winter 2003 Design of Steel Structures 6-126

21
 WORKED EXAMPLE 6 WORKED EXAMPLE 7
Fire calculation in time domain
• Yield strength reduction ky,T = (905-T)/690 = 0.30
• For a simply supported steel beam of known
• Flexural capacity:
span, load, yield strength, and section properties,
Mf = S ky,T fy (assume adequate lateral restraint) calculate the time to failure when exposed on
Mf = 1060 x 103 x 0.30 x 300/106 = 95 kN-m three sides to the standard fire: (a) unprotected,
• Design fails (M*fire > Mf) and (b) protected with insulation of known
(Note: For more accurate calculations, the thickness and properties.
maximum temperature should be calculated by GIVEN
the step-by-step method. The flexural calculation • Dead load Gk = 6.0 kN/m (including self weight)
method would be identical.) • Live load Qk = 12.5 kN/m
Winter 2003 Design of Steel Structures 6-127 Winter 2003 Design of Steel Structures 6-128

WORKED EXAMPLE 7 WORKED EXAMPLE 7

COLD CALCULATIONS
• Beam span L = 15.0 m
• Strength reduction factor Φ = 0.9
• Beam size 760mm deep and 147 kg/m
• Yield strength fy = 300 MPa
• This is a 'compact' section (type 1)
• Design load (cold) wc = 1.2Gk+1.6Qk = 27.2 kN/m
• Plastic section modulus S =4480 x 103 mm3
• Bending moment M*cold = wc L2/8 = 765 kN-m
• Section factors:
• Bending strength Mn = S fy = 1344 kN-m ( = Rcold)
Area to volume ratio F/V = 119 m-1 (Effective
(assume adequate lateral restraint)
thickness V/F = 8.4 mm)
• Design flexural strength Φ Mn = 1210 kN-m
• Design is OK (M*cold < Φ Mn)

Winter 2003 Design of Steel Structures 6-129 Winter 2003 Design of Steel Structures 6-130

WORKED EXAMPLE 7 WORKED EXAMPLE 7

FIRE CALCULATIONS
(a) Unprotected steel (three-sided exposure)
• Design load (fire) wf = Gk + 0.4Qk = 11 kN/m
• Time to reach limiting temperature:
• Bending moment M*fire = wf L2/8 = 309 kN-m
t = 0.54 (Tlim-50)/(F/V)0.6
• Load ratio rload = M*fire/ Rcold = 309/1344 = 0.23
t = 0.54 (746-50)/1190.6 = 21.4 minutes
• Limiting steel temperature:
• Design is OK if the equivalent fire severity is no
Tlim = 905-690 rload = 746°C
more than 21 minutes.

Winter 2003 Design of Steel Structures 6-131 Winter 2003 Design of Steel Structures 6-132

22
 WORKED EXAMPLE 7 WORKED EXAMPLE 7
(b) Protected steel • Time delay for insulation:
• Thickness of insulation di = 0.020 m (20 mm) tv = m ρi di2/(5 ki)
tv = 15x800x0.022/(5x0.10) = 9.6 minutes
• Thermal conductivity ki =0.10 W/m-K
• Total time ttolal = t + tv = 186 minutes
• Time to reach limiting temperature:
• Design is OK if the equivalent fire severity is no
t = 40 (Tlim-140) [(di/ki)/(F/V)]0.77 more than 186 minutes.
t = 40 (746-140) [(0.02/0.10)/119]0.77 = 177 min • (Note: For more accurate calculations, the
• Moisture content of insulation m = 15% maximum temperature should be calculated by the
step-by-step method. The flexural calculation
• Density of insulation ρi = 800 kg/m3 method would be identical.)

Winter 2003 Design of Steel Structures 6-133 Winter 2003 Design of Steel Structures 6-134

Simply supported beams -

Design of Individual Members -
Lateral torsional buckling
Simply supported beams
• Slender beams with no lateral restraint to the
• Lateral torsional buckling compression edge can fail by buckling before the
cross section flexural capacity is reached
• Shear
• Lateral torsional buckling does not occur if:
– compression edge is restrained against lateral
movement or
– cross section is reasonably compact and the
slenderness is not too large
• Buckling is allowed for using a strength reduction
factor, which reduces the design strength

Winter 2003 Design of Steel Structures 6-135 Winter 2003 Design of Steel Structures 6-136

Simply supported beams - Simply supported beams -

Lateral torsional buckling Shear
• The reduction amount depends on the: • The design equation to resist a shear force V*fire
– unrestrained length of the beam during fire is:
– compactness of the cross section V*fire ≤ Vf
• The Eurocode permits ignoring buckling for well • The design shear resistance Vf under fire is:
restrained fire-exposed beams of Class 1 or 2 Vf = ky,T Vc
• For beams with larger distances between lateral • Vc is the design shear resistance of the cross
restraint locations, the flexural capacity that section for normal temperature design
allows for buckling, is reduced by a factor • With a temp. gradient over the cross section, Vf
• See textbook for reduction factors should be based on the max. temp.

Winter 2003 Design of Steel Structures 6-137 Winter 2003 Design of Steel Structures 6-138

23
 Design of Individual Members - Design of Individual Members -
Continuous beams Continuous beams
• Continuous beams over many supports are • The Figure below shows 2 situations:
different from simply supported beams – a S.S. beam will fail once a plastic hinge forms at the
• The main advantage of continuity in fire design is centre
the possible moment redistribution during fire, – A continuous beam will fail once 3 plastic hinges form
often leading to fire resistance increase • Design of continuous beams is similar to that of
• A negative aspect of flexural continuity in steel simply supported beams, but including the
beams is the lack of lateral restraint at the redistribution of moments
compression regions near the supports

Winter 2003 Design of Steel Structures 6-139 Winter 2003 Design of Steel Structures 6-140

Design of Individual Members - Design of Individual Members -

Continuous beams Columns
Failure mechanisms for simply supported and
• Design of columns is often difficult because:
continuous beams
– usually, lateral buckling must be considered
– prediction of behaviour is not very reliable
• With a temp. gradient over cross sections, often
computer programs need to be used
• This is because thermal bowing and instability
considerations dominate the behaviour
• The Eurocode gives an approximate design
method based on the assumption that the whole
cross section is at the maximum temperature Tm
Winter 2003 Design of Steel Structures 6-141 Winter 2003 Design of Steel Structures 6-142

Design of Individual Members - Design of Individual Members -

Columns Columns
• For the calculation, the steel temperature must • In the approximate method, the compressive
be the maximum temperature load-bearing capacity is obtained from:
• This method may become unconservative when Nf = (χfi/1.2) A ky,Tm fy
thermal gradient causes significant bowing • χfi is the ambient buckling factor, calculated
• The design equations for a column subjected to using the effective length for fire design cases
an axial load N*fire is: • A is the area of the cross section, ky,Tm is the
N*fire ≤ Nf reduction factor for yield strength of steel which
is at the max. temp. Tm, and fy is the yield
strength of the steel at ambient
• 1.2 is an empirical correction factor
Winter 2003 Design of Steel Structures 6-143 Winter 2003 Design of Steel Structures 6-144

24
 Design of Individual Members - Design of Individual Members -
Columns Columns
Effective lengths of fire exposed columns in a multi-
• The design of columns in fire conditions should
storey frame
follow similar principles as for normal temp.
design with a factor of (1/1.2)
• The factor χfi may have a different name in other
codes
• The buckling length of a column is usually the
same as calculated in normal temp. design
• In braced frames, the buckling length changes as
shown in the following Figure

Winter 2003 Design of Steel Structures 6-145 Winter 2003 Design of Steel Structures 6-146

Bolted and Welded Connections DESIGN OF STEEL BUILDINGS

EXPOSED TO FIRE
• Welded connections perform well under fire
• Steel buildings design cannot be cost-effective
• Generally, bolted connections behave well
by the simple methods described previously
despite the bolt strength drop with high temp.
• It is necessary to use computer programs for
• There has been very limited research into the fire
analysis of the fire-exposed structure
performance of bolted steel connections
• Design of bolted beam-to-column connections • Such programs will impose deformations on the
should allow for very high tensile forces which structure and calculate the total strain in a
can occur during the fire decay stage member resulting from the deformations

Winter 2003 Design of Steel Structures 6-147 Winter 2003 Design of Steel Structures 6-148

Multi-storey Steel Framed

DESIGN OF STEEL BUILDINGS
Buildings
EXPOSED TO FIRE
• Recent large fires in steel framed buildings have
• The process is to calculate the stress-related showed good fire performance of steel frames
strain, calculate the internal forces in each compared to individual element performance
member, and compare with the applied loads • This excellent behaviour is due to steel ductility
• The general method is essential for structures • These observations have been supported by
with large displacements extensive computer analyses
• Calculated fire resistance of a structural steel • A large series of full-scale fire tests was
member is different when single vs. part of a conducted at Cardington Lab, BRE, England
frame (enhanced in a frame)

Winter 2003 Design of Steel Structures 6-149 Winter 2003 Design of Steel Structures 6-150

25
 Multi-storey Steel Framed
Buildings
• The actions can take place in 2-D or 3-D,
depending on the geometry of the building and
the layout of the structure
• The large deformations are often accompanied
by local buckling of the steel members
• The high axial tensile forces can result in
fractures of buckled beams after the fire
• Computer models were also used to help
interpret the behaviour of the Cardington tests

Winter 2003 Design of Steel Structures 6-151

26