SSB07 Fire Engineering

STEEL BUILDINGS IN EUROPE
Single-Storey Steel Buildings

Part 7: Fire Engineering

Single-Storey Steel Buildings

7 - ii

7 - iii
FOREWORD
This publication is the seventh part of the design guide, Single-Storey Steel Buildings.
The 11 parts in the Single-Storey Steel Buildings guide are:
Part 1: Architects guide
Part 2: Concept design
Part 3: Actions
Part 4: Detailed design of portal frames
Part 5: Detailed design of trusses
Part 6: Detailed design of built up columns
Part 7: Fire engineering
Part 8: Building envelope
Part 9: Introduction to computer software
Part 10: Model construction specification
Part 11: Moment connections
Single-Storey Steel Buildings is one of two design guides. The second design guide is
Multi-Storey Steel Buildings.
The two design guides have been produced in the framework of the European project
Facilitating the market development for sections in industrial halls and low rise
buildings (SECHALO) RFS2-CT-2008-0030.
The design guides have been prepared under the direction of Arcelor Mittal, Peiner
Trger and Corus. The technical content has been prepared by CTICM and SCI,
collaborating as the Steel Alliance.
7 - iv

7 - v
Contents
Page No
FOREWORD iii
SUMMARY vi
1 INTRODUCTION 1
2 FIRE RISKS IN SINGLE-STOREY BUILDINGS 2
2.1 Fire safety objectives 2
2.2 Fire risk analysis 2
2.3 Main requirements of current fire regulations 3
3 PRACTICAL FIRE ENGINEERING OPTIONS IN THE EUROCODES 6
3.1 Current design approaches 6
3.2 Fire analysis 7
3.3 Heat transfer analysis 8
3.4 Structural analysis 8
4 GUIDANCE ON APPROPRIATE FIRE ENGINEERING SOLUTIONS 10
4.1 Field of application of different design methods 10
4.2 Choice of optimum design approach 11
5 DIRECT USE OF SIMPLE ENGINEERING OPTIONS FOR USE BY NON
SPECIALISTS 12
5.1 Fire models 12
5.2 Thermal Models 16
5.3 Structural Models 21
5.4 Specific design rules for single-storey buildings 31
5.5 Simplified design methods 33
5.6 Design recommendations 37
6 GUIDANCE ON THE USE OF MORE ADVANCED SOLUTIONS 47
6.1 Fire models 47
6.2 Thermal Models 50
6.3 Structural models 51
REFERENCES 56
APPENDIX A German fire safety procedure for single-storey industrial and
commercial buildings 57

7 - vi
SUMMARY
This document provides guidance for the fire design of single-storey steel building
structures. It contains detailed information to allow engineers and designers to be more
familiar with the current design approaches and calculation models, which can be
applied not only to meet the prescriptive requirements but also to develop the
performance-based fire safety design. The design methods introduced in the guide,
ranging from simple design rules to more sophisticated calculation models, are derived
from EN 1993-1-2 and 1994-1-2. They cover both steel and composite structures
(unprotected or protected). In addition, some specific design rules are given, allowing
simple verification of whether the behaviour of the steel structure of single-storey
industrial buildings in fire situation fulfils the safety objectives on the basis of
performance-based requirement.

7 - 1
1 INTRODUCTION
Due to the particularities of single-storey buildings, the life safety objective in
case of fire can be met easily without onerous fire resistance requirement for
the structure. However, other safety objectives have to be taken into account if
the collapse of these buildings or a part of them may be accepted. In
consequence, many European fire safety building regulations are moving
toward acceptance of alternative fire safety engineering designs. Prescriptive
rules can then be replaced with performance based requirements, such as
adequate fire behaviour of the structure, aimed at satisfying fire safety
objectives that include life safety of people (occupants and fire-fighters),
protection of environment, property protection and business continuity.
Benefits and successful application of the performance-based approach to
building fire safety designs have already been well demonstrated for single-
storey buildings, especially where fire resistance was required, allowing in
some cases more innovative, cost effective and safer solutions to be adopted.
To help the structural fire design of buildings, a new set of European Standards
has been developed, the Eurocodes. The Parts of the Eurocodes that are
relevant to the fire design of single-storey building consist of EN 1991-1-2
[1]

(which includes principal concepts and rules necessary for describing thermal
and mechanical actions on structures exposed to fire) and Parts of material
specific Eurocodes dealing with the fire design of structures, such as
EN 1993-1-2
,[2]
, related to steel structures and EN 1994-1-2
[3]
related to
composite steel and concrete structures.
The fire parts of Eurocodes provide at present a wide range of calculation
methods. They allow engineers to follow either a prescriptive approach to meet
the fire safety requirements, as specified in national building regulations, or to
carry out on the basis of performance-based rules, a fire safety engineering
design that involves in general more complex computational analysis and
provides more accurate answers to fire safety objectives.
The present guide provides an overview of the current design methods
available for evaluating the fire performance of single-storey buildings
composed of either steel or composite structure as well as their application
fields. Simple calculations methods, easy to use, and more advanced
calculations models are dealt with separately. Moreover, to allow quick
assessment, simple design rules are given to assess quickly whether the
structural behaviour of steel structures of storage and industrial buildings fulfils
the fire safety objectives required by the fire safety regulations for industrial
buildings.
This guide aims also to help the engineer to understand more clearly the
different calculation methodologies and to carry out the structural fire design of
single-storey building according to the Eurocodes, from a relatively simple
analysis of single members under standard fire conditions to a more complex
analysis under real fire conditions.
7 - 2
2 FIRE RISKS IN SINGLE-STOREY BUILDINGS
2.1 Fire safety objectives
The primary objective of most fire safety regulations is to ensure the protection
of life (building occupants and fire fighters), environment and to some extent
property (building contents and building itself). Through a lot of measures
including a combination of active and passive fire protection systems, the
objectives are:
- To reduce and prevent the incidence of fire by controlling fire hazards in
the building.
- To provide safe escape routes for evacuation of building occupants.
- To prevent fire spread from the fire compartment to others parts of the
building and to neighbouring buildings.
- To ensure that the building remains structurally stable for a period of time
sufficient to evacuate the occupants and for the fire-fighters to rescue
occupants, if necessary.
2.2 Fire risk analysis
Single-storey buildings used as factories, warehouses or commercial centres
constitute a very common type of steel construction today. In the specific case
of warehouses, according to the storage arrangement (including free standing
storage, palletised rack storage, post-pallet storage or storage with solid or
slatted shelves) and the combustibility of materials being stored, fire may
develop very quickly and then might endanger occupants long enough before
the structural collapse of the building. Indeed, fire growth may be extremely
important, as the upward flame propagation is usually very rapid. Vertical and
horizontal shafts formed between adjacent pallets and racking behave as
chimneys, which increase the spread of flames up to the roof. The smoke
quickly forms a hot layer under the roof and then descends progressively with
fire development. Obviously, the rate at which this occurs varies according to
the combustible contents and the building arrangement. In unventilated
conditions, single-storey buildings can become smoke-logged in few minutes.
Although the smoke is largely made up of entrained air, it contains enough
toxic substances and asphyxiates to incapacitate or kill within minutes people
exposed to them. Moreover, the hot smoke layer will also radiate high heat flux
to people escaping from fire area. A hot gas layer at 500C leads to a heat flux
of about 20 kW/m (corresponding to the radiant energy emitted by a
blackbody at the temperature of 500C) and, under such thermal conditions,
skin burn will occur after only a few seconds
4
. Generally, it is agreed that the
tenability threshold is 2.5 kW/m
2
, which is much lower than heat flux needed
to lead to the failure of structural members. Consequently, buildings will
survive longer than occupants and the structural collapse of steel structures of
single-storey buildings generally does not provide additional threat to people
escaping from the fire area.
7 - 3
Regarding fire service operations, it is commonly accepted that fire-fighters
should not enter a single-storey building because of fast fire growth. Usually
they are forced to fight the fire from outside, covering neighbouring walls with
water. Hazard in this case for fire-fighters is then reduced to zero in the event
of structural collapse since it occurs at a level of temperature at which fire-
fighters can not withstand (provided that the progressive collapse, in the case
of compartmented buildings, and the collapse of the structure toward outside
do not occur
[5,6]
). In the event of, at the beginning of fire, they need to enter
within the building to rescue people, they cannot last within the building after
the heat flux is more than 7 kW/m, which is also very far for the risk of
collapse of the structure.
For these reasons, an increase of the intrinsic fire resistance of single-storey
buildings is unnecessary. However, the overall stability of the structure and the
stability of fire walls need to be accurately considered, to avoid any
progressive collapse. A single-storey building undergoes progressive collapse
when local failure of the heated part of the structure leads for the failure of
adjoining cold structures. In addition, to provide a safe situation to fire-fighters
located around the building, the structure of single-storey building (including
faade elements) must collapse towards the inside of the building.
Many National Regulations have taken into account previous remarks for
industrial single-storey buildings as well as for public buildings by not
requiring any fire resistance rating for such works but introducing specific
safety requirements in terms of overall structural behaviour and concentrating
requirements on egress facilities and early fire detection and/or suppression.
With regards to other single-storey buildings with relatively low fire loads, the
risk of life in the event of fire is reduced as egress of occupants and fire-ground
operations are straightforward.
2.3 Main requirements of current fire regulations
2.3.1 Fire resistance of structural members
Despite the comments above, fire resistance ratings are sometimes required for
the structure of single-storey buildings
[7]
.
The fire resistance is expressed as the time during which a building element
can withstand exposure to fire without losing its function (load-bearing
elements or separating element). Usually, building elements are classified
using three performance criterion:
- The load bearing capacity, R, which is the ability for a load-bearing element
to resist a fire without losing its structural stability
- The integrity, E, which is the ability of a separating element, when exposed
to fire on one side, to prevent the passage through it of flames and hot gases
- The insulation, I, is the ability of a separating element, when exposed to fire
on one side, to restrict the temperature rise on the unexposed face below
specified levels (in general a average value of 140C).
7 - 4
In prescriptive fire regulations, required fire resistance for a building element is
expressed in terms of the minimum period of time during which the building
element would function satisfactorily while subject to the standard fire.
When fire stability requirements are given for single-storey buildings, they
usually range from 15 minutes (R15) to 60 minutes (R60), depending on the
occupancy class of the building, the provision of sprinklers, the building height
and the compartment size.
2.3.2 Compartmentation and building separation
Single-storey building must be subdivided into compartments separated by fire
walls when the floor area of the building exceeds the allowed maximum
compartment size. Limits on the compartment size may be removed by fitting
the building with sprinklers.
The effects providing compartmentation on property loss is that direct damage
is confined to the content of the compartment in which the fire starts, reducing
the chances of the fire growing large. As regards the life safety, people in other
parts of the building can use escape routes to get out safely without being
exposed to the smoke or gases from the fire.
When considering fire walls between compartments, fire resistance is generally
in the range of REI 60 to REI 120.
Fire spread to neighbouring buildings also needs to be prevented. This is
achieved traditionally by sufficient separating distances or faade elements
with adequate fire resistance. In the French research project Flumilog, a design
method has been recently developed to assess the thermal radiant effects of
fires in single-storey storage buildings. The method allows calculation of the
safe separating distances, taking into account the main characteristics of the
building, such as the building content, the type of faade elements and roof,
etc.
2.3.3 Fire suppression
Sprinklers may be required by national fire regulations. In addition to their
obvious effect in the reduction of the fire growth, their use leads usually to a
reduction of the fire resistance rating required for the structure. They allow also
larger fire compartment sizes.
2.3.4 Smoke control systems
National fire regulations may require that smoke control systems are
implemented in public buildings, storage building and industrial buildings in
order to facilitate escape, by minimising risks of smoke inhalation and injury
and to some extend to enable fire-fighters to better see the fire and therefore to
extinguish it more speedily and effectively. Smoke control systems help in
removing smoke from the fire area, and in limiting the spread of hot gas
beneath the roof, which increases the time for the compartment to become
smoke-logged, giving people more time to escape safely from the building.
This can be achieved by a combination of smoke exhaust systems (mechanical
or natural) and screens (which contain the smoke in specific areas).
7 - 5
2.3.5 Fire detection and fire alarms
Adequate measures are necessary for detecting any outbreak of fire and for
alerting the building occupants and the fire department of the occurrence of
fire. In small single-storey buildings where all exits are visible, it is likely that
any fire will be quickly detected by the occupants and a shout of Fire! may be
sufficient. In larger single-storey buildings, a simple sounder such as a battery
powered alarm or rotary bell may be adequate. In an industrial building, the
ambient noise has to be considered, to ensure that the alarm will be heard by
the occupants.
2.3.6 Egress facilities
For safe evacuation, appropriate means of escape are needed, such as a proper
number and width of emergency exits and proper length, width and height of
passages and evacuation accesses. Escape routes in small single-storey
buildings generally lead directly to a safe location outside the building; they do
not normally require any special treatment. In larger buildings, where travel
distances are greater and where the fire is likely to make a single escape routes
unusable, an alternative means of escape may be necessary. Consideration of
disabled people must also be made
7 - 6
3 PRACTICAL FIRE ENGINEERING OPTIONS
IN THE EUROCODES
3.1 Current design approaches
Using the fire parts of Eurocodes
[8,9]
, single-storey buildings can be designed
using either the prescriptive approach or the performance-based approach
applying fire safety engineering principles
[10]
.
The prescriptive approach is mostly applied to fulfil standard fire resistance
requirements usually prescribed in national fire regulations. It gives a safety
level that is relatively easy to achieve and implement. However it may be
conservative, in requiring the use of important passive fire protection to fulfil
the required fire resistance rating. This approach is usually carried out for the
design of relatively simple buildings and structures.
As an alternative or when allowed by national regulation, the performance-
based approach can allow to assess adequate measures to satisfy a set-out of
defined fire safety objectives, such as stated in paragraph 2.1, and the
corresponding performance criteria. Using structural fire engineering,
engineers can assess the necessary fire resistance to structure in order to avoid
the spread of fire and/or to prevent a premature structural collapse. As regards
the single-storey buildings, the main structure could be designed to remain
stable under fire exposure conditions long enough for the occupants to escape.
Such an approach takes into account the severity of fire exposure by
appropriate estimations of actual fire loads and fire development parameters,
which may be calculated from the building activity.
The performance-based approach provides flexibility when selecting technical
solutions to meet the fire safety objectives, but usually requires the use of
sophisticated design tools. Engineers and designers using advanced
calculations models need to be properly educated in their use and in their
limitations. As fire safety engineering allows for highly efficient designs, with
little unassigned reserve capacity, an experienced user is required to ensure that
appropriate models are used.
Where national fire regulations authorise the performance-based approach,
regulatory bodies may require that the fire design is checked by a third party.
The fire performance of a whole structure, or a part of it, is carried out by
following, for a given design fire scenario, three successive steps of structural
fire engineering
[1]
.
- Fire Analysis. To calculate the thermal actions/exposure - Fire models.
- Thermal analysis. To determinate the heating rate and temperatures on
structural members - Thermal models.
- Structural analysis. To calculate the mechanical response of structural
members- Structural models.
7 - 7
Available design methods to evaluate the fire performance of structure are
briefly described below. These methods range from simple hand calculations to
the use of sophisticated computer models. The overall complexity of the fire
safety design will depend on the assumptions and methods adopted to predict
each of the three design steps.
3.2 Fire analysis
The main objective of the fire modelling is the simulation of the fire
development and the prediction of thermal actions (gas temperature, heat flux)
on the structural members (in order to determinate, in a following step, the
temperature in the structural members).
Although common practice is to represent a fire by a standard fire curve,
structural fire design may be based on a design fire that provides more realistic
conditions in fire compartment. In this way, parameters such as the magnitude
of the fire load, the rate of heat release and the ventilation factor, which play an
important role in fire severity, are taken into account. Moreover, the
identification of relevant and realistic design fire scenarios is a crucial aspect of
the fire safety design. The design fire scenarios used for the analysis of a
building fire have to be deduced from all the possible fire scenarios. In most
buildings, the number of possible fire scenarios is infinite and need to be
reduced. Only credible worst case fire scenarios will need to be studied.
When the design fire scenarios are chosen, a number of fire models are
available to assess the fire severity and calculate the corresponding thermal
actions
Different levels of fire models are relevant to the various stages of fire
development. When a fire is initiated, it is localised within a compartment and,
according to the characteristics of the compartment and of the fire load, it can
remain localised or becomes generalised to the whole compartment. In the case
of small compartments or compartments with small ventilation openings
relative to the size of the compartment, the fire develops into to a fully
engulfed fire.
Three levels of modelling are available to describe both localised and fully
generalised fires, as shown in Table 3.1.
Table 3.1 Levels of fire models
Levels of the model Localised fire Generalised fire
Simplified model Hasemi model
Heskestad model
Parametrical fires
Zone models 2 zone model 1-zone model
Field model CFD CFD

The simplified models are generally empirical models based on conventional
assumptions. The zone models take into account the main parameters
controlling the fire, but introduce simplified assumptions that limit the domain
of application. They would be used in simple easily defined compartment
geometries. The field models are more accurate but are rather complex for use
7 - 8
as a general design tool; they would be required in compartments with complex
geometries or with high and irregular ceilings.
Conditions of use will be briefly detailed in Chapter 6.
3.3 Heat transfer analysis
Once the thermal actions are calculated, the thermal transfer to the structural
elements has to be calculated. Thermal models, which will be used, should be
based on the acknowledged principles and assumptions of the theory of heat
transfer.
Different modelling can be used according to the assumptions and needs. In the
thermal models, there are the analytical rules allowing obtaining an estimation
of uniform temperature across-section, mainly for steel elements. There are
also advanced calculation methods based on either finite elements or the finite
difference method, allowing determination of the 2D or 3D temperature
distribution in structural members (through the cross-section and along the
length). Advanced models can be applied for any type of structural member
analysis in fire design.
Thermal models will be briefly detailed in following chapters.
3.4 Structural analysis
From the temperature fields previously obtained in the structural members and
from the combination of the mechanical actions loads in case of fire the
structural behaviour can be assessed following one of the three possible
approaches:
- Member analysis, in which each member of the structure will be assessed
by considering it fully separated from other members. The connection
condition with other members will be replaced by appropriate boundary
conditions.
- Analysis of parts of the structure, in which a part of the structure will be
directly taken into account in the assessment by using appropriate boundary
conditions to reflect its links with other parts of the structure
- Global structural analysis, in which the whole structure will be used in the
assessment
7 - 9

Member
analysis
Analysis of part
of the structure
Global structural
analysis

Figure 3.1 Different design approaches for mechanical response of
structures in fire
Member analysis is easy to use particularly with simplified calculation methods
and therefore largely used under standard fire condition. The analysis of the
whole structure or its subassemblies considers at least several structural
members together, so that the interaction effect between them will be directly
dealt with. In this way, load redistribution from heated parts (weakened parts
inside fire compartment) to cold parts (stronger parts outside fire compartment)
can be taken into account in accurate way and global analysis provides
therefore a much better understanding of overall behaviour of structure under
fire condition.
According to the Eurocodes, three types of design methods can be used to
assess the mechanical behaviour of structures under fire situation in the
different design approaches explained above. Fire design can be carried out by
means of:
- A simple calculation method, based on predefined tabulated data, as given
in EN 1994-1-2
[3]
. This method is only applicable to steel and concrete
composite structures. The tables were evaluated by numerical models and
experiments for basic types of structures, such as slabs, beams and
columns, for certain time of fire resistance, for heating according to the
nominal fire curve and for defined level of loading. The tables are easy to
use and safe but cover only a limited range of section types.
- Simple calculation models. This type of design method can be divided into
two different families. The first one is the critical temperature method
widely applied to steel structural member analysis. The second is the use of
simple mechanical models (verification in strength domain) developed for
both steel and composite structural member analysis. Models have been
developed for standard structural elements, e.g. slabs, beams, and columns.
- Advanced calculation models. This kind of design method can be applied to
all types of structures and the models are, in general, based on either finite
element method or finite difference method. They should provide a realistic
analysis of structures. The results of the analysis are generally obtained in
terms of deformation of structure during the whole fire period.
Structural models will be briefly detailed in following chapters.
7 - 10
4 GUIDANCE ON APPROPRIATE FIRE
ENGINEERING SOLUTIONS
4.1 Field of application of different design methods
The following table shows the field of application of the available fire design
methods, considering either design according to prescriptive requirements
based on the standard fire or a performance-based fire design
[11]
.

Table 4.1 Field of application of different design methods
Approach Tools
Thermal
actions
Thermal
modelling
Structural
modelling
Pre-engineered data
from standard fire
tests (Data from
manufacturers)

Tabulated data from
EN 1994-1-2
EN 1994-1-2, 4.2
Steel
EN 1993-1-2
4.2.5
Steel
EN 1993-1-2
4.2.3 4.2.4
Simplified calculation
models given in
Eurocodes
Composite EN 1994-1-2 4.3
Steel and composite
P
r
e
s
c
r
i
p
t
i
v
e

a
p
p
r
o
a
c
h

(
S
t
a
n
d
a
r
d

f
i
r
e

d
e
s
i
g
n
)

Advanced calculation
models
Standard ISO
curve

EN 1991-1-2
FEA* or FDA** FEA*
Simplified calculation
models
Fully engulfed
fire (Parametric
fire, standard
ISO curve***)

Localized fire
Steel

EN 1993-1-2
4.2.5
Steel

EN 1993-1-2
4.2.3 4.2.4

Specific rules
based on fully
engulfed fire
5.4
Steel and composite
P
e
r
f
o
r
m
a
n
c
e

b
a
s
e
d

a
p
p
r
o
a
c
h

(
n
a
t
u
r
a
l

f
i
r
e

d
e
s
i
g
n
)

Advanced calculation
models
Zone models

Field models
FEA* or FDA** FEA*
*FEA : Finite element Analysis **FDA : Finite Difference analysis
*** Collapse of single-storey buildings usually occurs when the building structure (a part of it or
the whole structure) is fully engulfed in fire. In such fire condition, because the gas temperature
rise has no significant effect on the failure mode of the building structure, a performance-based
approach referring to thermal actions based on standard fire curve is appropriate to investigate
the fire behaviour of single-storey buildings. This approach can be used to demonstrate the
non-progressive collapse and the failure inwards of the building structure.
7 - 11
4.2 Choice of optimum design approach
The choice of the design approach depends on the type of building (storage
building, industrial building, commercial building, etc.), the requirements
specified in the corresponding national fire regulation and the acceptance or
not by the regulatory authorities of applying a performance-based approach as
an alternative to prescriptive rules.
Some suggestions on the choice of fire design approach are given below.
With the diversity of requirement, the most important first step is to answer the
following:
- What is the required fire resistance rating, if any?
- Is it possible to carry-out a performance-based approach?
When a prescriptive approach is to be used (with reference to standard fire
design):
- It may be appropriate to use simplified calculation models where low fire
resistance ratings (R15 or R30) are required for structural members
- Advanced calculation models must be used where structural members are
not covered by the simplified calculation models. They can also be
employed with some economic benefits for steel structure where high fire
resistance ratings (higher than R60) are required, reducing the thickness of
fire protection on steel members.
Where the performance-based approach is accepted by the regulatory
authorities and structural stability is needed:
- A performance-based approach is most likely to be beneficial where the
structure is unusual and may not be well covered by traditional prescriptive
methods
- Localised fire protection may be needed, considering the overall behaviour
of the whole structure in a real fire, to ensure adequate life safety for the
building occupants and firemen.
National fire regulations may require the use of the performance based
approach for single-storey buildings with significant fire risks (high fire loads).
National fire regulations may allow a performance-based fire safety design to
refer to simple rules and design recommendations for single-storey buildings.
Such approaches are given in 5.4 and Appendix A. Other design guidance and
recommendations can be found in reference
[12]
.
Active fire protection measures (installation of sprinklers, fire detectors, fire
alarms, smoke exhaust systems) and passive fire protection measures
(compartmentation, egress facilities, etc.) are usually implemented in buildings
in accordance with the requirements in fire national regulations.
7 - 12
5 DIRECT USE OF SIMPLE ENGINEERING
OPTIONS FOR USE BY NON SPECIALISTS
This chapter gives an overview of current easy-to-use simple calculation
design rules, for assessing the fire resistance of steel and composite steel and
concrete structural members.
Specific simple design rules and design recommendations to satisfy specific
safety requirements in terms of structural behaviour introduced recently in fire
safety regulations of many European countries for single-storey storage and
industrial buildings are given. It is noted that these methods are also applicable
to other type of single-storey buildings.
5.1 Fire models
5.1.1 Nominal temperature-time curves
EN 1991-1-2
[1]
provides three standard fire curves, defining arbitrary hot gas
temperature-time relationships in which no physical parameters of the fire load
or fire compartment are taken into account. The most commonly used
relationships in building design and in regulation prescriptions is the standard
temperature-time curve (standard ISO fire) which represents a fully developed
compartment fire. The second curve, the external fire curve, is intended for
faade elements and the third curve is the hydrocarbon fire curve, representing
a fire with hydrocarbon or liquid type fuel.
The nominal temperature-time curves are defined as follows:
- For standard temperature-time curve (standard ISO fire ):
) 1 8 ( log 345 20
10
+ + = t
g
u (1)
- For the external fire curve:
20 ) 313 , 0 687 , 0 1 ( 660
8 , 3 32 , 0
g
+ =
t t
e e u (2)
- For the hydrocarbon fire curve:
20 ) 675 , 0 325 , 0 1 ( 1080
5 , 2 167 , 0
g
+ =
t t
e e u (3)
where:
g
is the gas temperature in the fire compartment [C]
t is the time [min]
It is important to note that the previous curves are reference curves. They do
not represent the real thermal effect of a fire. The temperatures given by these
curves always increase with time, without considering the limited fire load.
The standard fire resistance rating required for structural members (expressed
in terms of time) does not therefore indicate the actual time for which they will
survive in a building fire.
7 - 13
5.1.2 Parametric fires
Parametric fire models provide a rather simple design method to estimate gas
temperature in fire compartment, taking into account in a simplified way the
main parameters that influence the fire development, such as the fire
compartment size, the fire load (corresponding to the mass of combustible
materials in the fire compartment), ventilation conditions (openings) and
thermal properties (such as thermal conductivity and specific heat) of the
compartment walls and ceilings.
Like nominal temperature-time curves, parametric temperature-time curves
provide gas temperature-time relationships for design. They are based on the
hypothesis that the temperature is uniform in the compartment, which limits
their field of application to post-flashover fires (fires generalised to the whole
compartment) in compartments of reasonable dimensions. The predicted fire
curve comprises a heating phase represented by an exponential curve up to a
maximum temperature, followed by a linearly decreasing cooling phase to a
residual temperature that is usually the ambient temperature. The maximum
temperature and the corresponding fire duration are the two main parameters
affecting the fire behaviour of structural members. Consequently, they were
adopted as the governing parameters in the design formulae for the parametric
fires.
Such a model is given in Annex A of EN 1991-1-2. It is valid for
compartments up to 500 m of floor area, without openings in the roof, and a
maximum compartment height of 4 m, for compartment linings with thermal
inertia between 100 and 2200 J/m
2
s
1/2
K, for an opening factor in the
range 0,02 to 0.20 and for compartments with mainly cellulosic type fire loads.
Due to these limitations, the model is mainly used for the office part of single-
storey buildings.
7 - 14
Time
u
g
u
max
t*max
heating
phase
cooling
phase
ug=20+1325(1-0,324e
-0,2t*
-0,2e
-1,7t*
-0,427e
-19t*
)

with t*= t.C where t is the time (hours) and
) 1160 / 04 . 0 /( ] b / O [ R =

Main parameters:
- Wall characteristics : thermal inertia c b =
- Opening characteristics: opening factor
t v
A / h A O =
umax= ug (t*max) = ug (tmax . I ) (C)
with tmax = max{ (0.2.10
-3
qt,d / O). I/ O, tlim } (hours)
where tlim is a function of fire growth rate (according to building type):
- tlim =25 min if slow fire growth rate
- tlim =20 min if medium fire growth rate,
- tlim =15 min if fast fire growth rate,
- qt,d is the design value of the load density [MJ/m]
ug = ug (t*, t*max, x) (C)
= umax 625.(t* - t*max.x) if t*max s 0,5
= umax 250.(3- t*max).(t* - t*max.x) if 0,5 < t*max s 2
= umax 250.(t* - t*max) if t*max > 2
with t*= t.I t*max = (0.2.10
-3
qt,d / O). I
and x is a function of tmax as follows:
x = 1 if tmax > tlim
x = tlim. I / t*max if tmax = tlim

Figure 5.1 Parametric Fire (Annex A of EN 1991-1-2)
The inputs for the parametric fire curves are the design fire load density, the
fire growth rate, the ventilation conditions (described by the size and the
position of the openings) and the thermal properties (heat capacity, the density
and the conductivity) of walls to evaluate the heat losses which occur by
convection and radiation at the compartment boundaries. For the fire load
density, it is common practice in design to refer to the characteristic values
given in EN 1991-1-2.
Even though these parametric fire curves offer a significant improvement
compared to the standard ISO-fire, the parametric fires are not yet able to
provide a very accurate evaluation of the fire severity. Consequently, some
European countries recommend their use only for pre-design calculation.
5.1.3 Localised fire
EN 1991-1-2 provides simple approaches for determining thermal actions of
localised fires in Annex C. Two situations are distinguished according to the
height of the fire flame relative to the ceiling of the compartment: where the
flame is not impacting the ceiling (based on Heskestads method); and where
the flame is impacting the ceiling (based on Hasemis method).
7 - 15

Flame axis
L
z
D
f
H
Z
0
= 1,02 D + 0,00524 Q
2/5
z
0
Flame axis
L
z
D
f
H
Z
0
= 1,02 D + 0,00524 Q
2/5
z
0
Flame axis L
h
D
H
r
Flame axis L
h
D
H
r
The flame is not impacting the ceiling The flame is impacting the ceiling
Required data:
- Rate of heat realase: Q (W)
- Distance fire Source-ceiling: H (m)
- Diameter of the fire: D (m)
Results:
- Flame length Lf (m) :
L
f
= -1,02 D + 0,0148 Q
2/5

-Temperature u(z) in the plume along
the symmetrical vertical axis:
u(z) = 20 + 0,25 (0.8Q)
2/3
(z-z0)
-5/3

u(z) s 900C
Results:
- Horizontal flame length Lh
- heat flux received by the fire exposed unit surface
area at the level of the ceiling at the distance r from
the flame axis:
h
= 100000 if y s 0,30
h
= 136300-121000 y if 0,30 < y < 1,0

h
= 15000 y
-3,7
if y > 1,0
with

'
'
h
z H L
z H r
y
+ +
+ +
=

where
r: is the distance from the flame axis to the
point where the thermal flux is calculated (m)
z: is the vertical position of the virtual heat
source (m)
D: is the diameter of the fire (m)

Figure 5.2 Localised Fires (Annex C of EN 1991-1-2)
For situations where the fire is not impacting the ceiling, a design formula is
given to calculate the temperature in the plume at heights along the vertical
flame axis. For situations where the fire is impacting the ceiling, some simple
steps are given to calculate the heat flux received by the fire-exposed surfaces
at the level of the ceiling.
These models are most often used to calculate thermal actions (expressed in
terms of heat flux resulting from a radiation part and a convection part) on
horizontal structural members, such as beams. At the present time, no method
is available for vertical steel members affected by a localised fire.
The input data are the rate of heat release (RHR), the distance between the fire
source and the ceiling, and the diameter of fire. The RHR is usually determined
by using EN 1991-1-2 section E.4.
These approaches are limited to cases where the diameter of fire D is less than
10 m and the rate of heat release of fire Q is less than 50 MW.
7 - 16
5.2 Thermal Models
Considering the high thermal conductivity of steel and the small thickness of
steel profiles commonly used in the construction, it is sufficiently accurate to
ignore thermal gradients within the cross-section of structural members and
assume a uniform temperature when uniformly heated.
Consequently, simple design equations can be used to predict the temperatures
of steel members that are fully exposed to fire or steel members that support a
concrete slab and are exposed on three sides. Similar rules exist for fire-
protected steel sections, although the thermal properties of the proposed
protection material are needed, which can be difficult to obtain.
For the composite steel-concrete members, strictly speaking, there are no
simplified models to estimate the evolution, as a function of time, of
temperature distribution through members. To simplify the design, information
on temperature distribution at current time of standard fire exposure (i.e. 30,
60, 90 and 120 minutes) is given in EN 1994-1-2.
5.2.1 Unprotected steel member
Heating of the unprotected steel members can be determined by means of the
simple analytical approach given in EN 1993-1-2. In this method, the
temperature rise depends on the thermal actions (expressed in terms of net heat
fluxes), the thermal properties of the steel and the section factor of the element
A
m
/V defined as the ratio between the surface area exposed to the heat flux A
m

[m/m] and the volume of the element by unit length V [m
3
/m]. The section
factors for some unprotected steel members are shown in Figure 5.3.

b
h
t
t
t
A
m
/V=Perimeter exposed to fire
/Cross-section area
A
m
/V=1 / t A
m
/V=2 / t

Figure 5.3 Example of section factor for unprotected steel members
Assuming an equivalent uniform temperature distribution in a cross-section,
the increase of temperature Au
a,t
in an unprotected steel member during a time
interval At may be determined from:
t
h
c
/V
A
k A = A

d net,
a
a
m
sh t a,
u with At s 5 s (4)
where:
sh
k is the correction factor for the shadow effect caused by local
shielding of radiant heat transfer due to shape of steel profile
a
C is the specific heat of steel [J/kgK]
7 - 17
a

is the unit mass of steel [kg/m
3
]
h
d net,
is the net heat flux per unit area [W/m]
Solving the incremental equation step by step gives the temperature
development of the steel element during the fire. In order to assure the
numerical convergence of the solution, some upper limit must be taken for the
time increment At. In EN 1993-1-2, it is suggested that the value of At should
not be taken as more than 5 seconds.
The thermal actions are determined by the net heat flux
r net,
h
absorbed by the
steel member during the fire exposure. It is expressed in terms of the hot gas
temperature as the sum of two distinct fluxes: a convective component
c net,
h

and a radiant component
r
h
net,
.
Convective heat flux is expressed as:
) (
m g c c net,
u u o = h
(5)
where:
c
o is the coefficient of heat transfer by convection [W/mK]
g
u is the gas temperature [C]
m
u is the surface temperature of the member [C]
Radiant heat flux is given by:
) 273) ( ) 273 ((
4
m
4
r m 0 r net,
+ + = u u c |o h
(6)
where:
| is the configuration factor, including position and shape effect (<1)
m
c is the surface emissivity of the member
r
u is the radiation temperature of the fire environment [C] (u
r
ug)
m
u is the surface temperature of the member [C]
0
o is the Stephan Boltzmann constant [= 5,67 10
-8
W/m
2
K
4
]
According to EN 1991-1-2, for many practical cases the configuration factor
may be taken equal to unity. The coefficient of convection (
c
o ) varies from
25 W/mK (standard fire conditions) to 50 W/mK (hydrocarbon fire
conditions). The emissivity of carbon steel and composite steel and concrete
members may be taken as 7 , 0
m
= c .
7 - 18
For cross-section with a convex shape, such as hollow steel sections, fully
embedded in fire, the shadow effect does not play a role and it can be taken
that k
sh
= 1. Otherwise, the correction factor for the shadow effects k
sh
is given
by:
=
cases others for
actions fire nominal under sections - I for
/
] / [
/
] / [ 9 , 0
m
b m
m
b m
sh
V A
V A
V A
V A
k (7)
where:
b m
] / [ V A

is the box value of the section factor [m
-1
].
Application of the EN 1993-1-2 calculation method with standard ISO fire
exposures of 15 and 30 minutes leads to the temperature curves illustrated in
Figure 5.4 and given in Table 5.1 as function of the section factor including
shadow effect (A
m
/V)
sh
= k
sh
A
m
/V.

Figure 5.4 Temperature of unprotected steel members after 15 and
30 minutes of standard ISO fire exposure
0
100
200
300
400
500
600
700
800
900
0 50 100 150 200 250 300 350 400 450 500
(A
m
/V)
sh
= k
sh
(A
m
/V) (m
-1
)
Temperature (C)
15 minutes
30 minutes
10
7 - 19
Table 5.1 Temperature of unprotected steel members after 15 and
30 minutes of standard ISO fire exposure
Steel temperature (C) Steel temperature (C) Section
factor
(A
m
/V)
sh

15 min 30 min
Section
factor
(A
m
/V)
sh

15 min 30 min
10 113 257 130 621 802
20 194 431 140 634 809
30 265 554 150 646 815
40 328 636 160 655 819
50 383 690 170 664 822
60 432 721 180 671 825
70 473 734 190 677 827
80 509 741 200 682 828
90 539 753 250 699 833
100 565 767 300 708 835
110 586 781 400 716 837
120 605 792 500 720 838

5.2.2 Protected steel member
EN 1993-1-2 also provides a simple design approach for insulated members
with passive fire protection materials. In such cases, the temperature rise
depends on the section factor A
p
/V for the steel member insulated by fire
protection material (A
p
is the appropriate area of fire protection material per
unit length and V is volume of the steel member per unit length) and the
insulation characteristics. The insulating materials can be in form of profiled or
boxed systems, but this simple approach does not cover intumescent coatings.
Assuming uniform temperature distribution, the temperature increase Au
a,t
in
an insulated steel member during a time interval At may be determined from:
( ) ( )
t g,
10 /
t a, t g,
p
a a
p p
t a,
1 e
3 / 1
1
/
u u u
|
u
|
A A
|
|
.
|
\
|
+
= A t
V
A
c
d
(8)
with
V
A
d
c
c
p
p
a a
p p
| = (9)
where:
p
d is the thickness of fire protection material [m]
p
C is the specific heat of fire protection material [J/kgK]
p
is the thermal conductivity of the fire protection material [W/mK]
p
is the unit mass of the fire protection material [kg/m
3
]
7 - 20
g
u is the gas temperature [C]
Figure 5.5 gives expressions to calculate the section factor of protected steel
members.
A
m
/V= (P-b) / A
s
b
A
m
/V= (2h+b) / A
s
A
m
/V= 2(2+b) / A
s
A
m
/V= P / A
s

h
b
b
h
P : perimeter ; A
s
: cross-section area

Figure 5.5 Example of section factor for insulated steel members
It is important to note that thermal characteristics of fire protection materials
are usually determined from fire tests performed under standard fire conditions.
Consequently, referring to thermal actions based on natural fires, the use of
Equation (8) for the fire design situation of protected steel members should be
handled with some caution. The calculation should be performed only if
appropriate data are available or if it can be shown that fire conditions have no
significant effects on thermal characteristics and integrity of fire protection
materials. Nevertheless, it is commonly assumed that thermal properties of an
insulation material can be used under natural fire conditions when the
temperatures of hot gases remain lower than the maximum temperature reached
during the standard fire test for the insulation material (For example, about
1100C for 4 hours of the standard temperature-time curve).
The material properties given in Table 5.2 may be used as a first approximation
to calculate heating of protected steel members. These average values are
derived from fire tests by material manufacturers.
7 - 21
Table 5.2 Average materials properties of main fire protection materials
Material
Density
p
[(kg/m
3
]
Conductivity
p
[W/mK]
Specific heat
p
C [J/kgK])
Mineral fibre 300 0,12 1200
Vermiculite and
cement
350 0,12 1200 Sprays
perlite 350 0,12 1200
vermiculite (or
perlite) and cement
550 0,12 1100
High density
sprays
vermiculite (or
perlite) and gypsum
650 0,12 1100
vermiculite (or
perlite) and cement
800 0,2 1200
fibre-silicate or fibre
calcium-silicate
600 0,15 1200
fibre-cement 800 0,15 1200
Boards
gypsum board 800 0,2 1700
Compressed
fibre boards
fibre-silicate,
mineral, stone-wool
150 0,2 1200
5.3 Structural Models
According to the Eurocodes, several simple design methods can be used to
assess the fire resistance of structures under fire conditions. The first one is the
critical temperature method widely applied to steel structural member analysis
and the second one is the simple mechanical models developed for both steel
and composite steel and concrete structural members.
It is important to remember that the design methods available for composite
members are only valid for the standard fire exposure. Moreover, design
methods given for columns should be only applied to members of braced
frames (where the column ends have no horizontal displacement).
5.3.1 Critical temperature method
The critical temperature is calculated by using applied mechanical actions,
design resistance in the normal temperature condition and the strength loss of
steel at elevated temperature. This critical temperature generally varies
between 500C and 800C. It can be obtained by calculation according to the
simple rules given in the EN 1993-1-2 or by referring to default values.
According to the critical temperature method, the fire resistance of a steel
member without instability effect is satisfied after a time t if the steel
temperature
t , a
u does not exceed the critical temperature
cr
u of the element:
cr t , a
u u s (10)
7 - 22
The critical temperature of the member can be calculated from the degree of
utilization
0
as follows:
482 1
9674 . 0
1
ln 19 , 39
833 . 3
0
+
(
u
cr
(11)
The degree of utilization
0
is obtained from:
d,0 fi,
d fi,
0
R
E
= (12)
where:
d fi,
E is the design effect of actions for the fire design situation, according
to EN 1991-1-2
d,0 fi,
R is the corresponding design resistance of the steel member, for the
fire design situation, at time t = 0 (at normal temperature) but with
safety factor
fi , M
in fire situation
The expression for
cr
can be used for all classes of section except the very
slender Class 4 sections, for which a single conservative critical temperature of
350C should be used.
In principle, Expression (11) applies for members in pure bending, short
columns without buckling and members in tension, heated uniformly or with
slight temperature gradient. However, in situations of instability (slender
columns, unrestrained beams), the method becomes applicable by calculating
the design resistance for the fire design situation at time t = 0 with a value of
the slenderness that takes into account temperature effects on the slenderness
of structural members. As a simplification, the slenderness in fire situations can
be taken as
u
3 . 1 = (where
is the non dimensional slenderness at

normal temperature).
As an alternative, to relation (11) nationally determined critical temperatures
can be given in the National Annex to EN 1993-1-2.
A simple conservative expression for
0
can also be used for tension members
and restrained beams (where lateral-torsional buckling is not a potential failure
mode):
2 1
M
fi , M
t , fi 0
k k
q = (13)
where:
t , fi
q is the load level at time t
fi , M
is the relevant partial safety factor for fire situation ( 1
fi , M
= )
0 M
is the partial safety factor at normal temperature ( 1
0 M
= )
7 - 23
1
,
2
are the adaptation factors to take account a non-uniform
temperature distribution on steel member.
The load level at time t is defined as:
d
d fi,
t fi,
R
E
= q (14)
where:
d fi,
E is the design effect of actions for the fire design situation, according
to EN 1991-1-2
d
R is the ultimate resistance in room temperature
For a given fire duration t, assuming that
cr t , a
u u = , the maximum value of
utilization level
0
of unprotected steel members to satisfy the required fire
resistance may be easily calculated from (11), as function of section factor
including the shadow effect (A
m
/V)
sh
. In this way, it may be assumed that fire
resistance of unprotected steel members is satisfied after a time t if:
max 0
s (15)
Maximum degrees of utilisation
max
calculated for standard fire resistance
R15 and R30 are given in Figure 5.6. It should be noted that for a fire
resistance R30, unprotected members with a section factor (A
m
/V)
sh
higher than
50 m
-1
can only achieve very low values of the degree of utilisation.

Figure 5.6 Maximum utilization level as a function of section factor (A
m
/V)
sh

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 50 100 150 200 250 300 350 400 450 500
(A
m
/V)
sh
= k
sh
(A
m
/V)(m
-1
)

max
10
15 minutes
30 minutes
practical field of
0
7 - 24
5.3.2 Simple design method for steel members
According to EN 1993-1-2, the load-bearing function of a steel member should
be assumed to be maintained at a time t if:
t fi, d, d fi,
R E < (16)
where:
d fi,
E

is the design effect of actions for the fire design situation, according
to EN 1991-1-2
t fi, d,
R

is the corresponding design resistance of the steel member, for the
fire design situation, at time t
The following simplified calculation methods allow the designer to assess the
design fire resistance (buckling resistance, resistance moment) of steel
members. They are mainly based on the assumption of constant temperature
within the section.
Steel columns under compression only
The design resistance for the fire design situation at time t of a compression
member with a Class 1, 2 or 3 cross-sections at a uniform temperature
a

should be determined from:
( )
Rd y,
f M,
M0
f Rd t, fi,
N k N
i
i
_
u
= (17)
where:
y,
k is the reduction factor for the yield strength of steel at the steel
temperature reached at time t
fi , M
is the partial safety factor for fire situation ( 1
fi , M
= )
0 M
is the partial safety factor at normal temperature ( 1
0 M
= )
Rd
N is the design resistance of the cross-section N
pl,Rd
for the normal
temperature design according to EN 1993-1-1
fi
_ is the reduction factor for flexural buckling in the fire design
situation
The reduction factor
fi
_

for flexural buckling is obtained from the non-
dimensional slenderness
u
at temperature using:
2
2

f
1

_
+
=
i
but _
fi
s 1.0 (18)
with
| |
2

1
2
1
o + + =
7 - 25
where:
o is the imperfection factor for the appropriate buckling curve given
by
y
/ 235 65 . 0 f = o with f
y
is characteristic yield strength of steel.
The non dimensional slenderness at temperature is given by:
E, y,
/ k k =
(19)
where:
y,
k
is the reduction factor for the yield strength of steel at the
temperature u
E,
k
is the reduction factor for the slope of the linear elastic range at the
temperature u
The non dimensional slenderness at normal temperature, according

to EN 1993-1-1
The non dimensional slenderness at normal temperature is given by:
E
f
i
y
cr
1
= (20)
where:
cr
is the buckling length in the buckling plane considered
i is the radius of gyration about the relevant axis, determined using
the properties of the gross cross-section
For a practical use, the reduction factor
i f
_ for flexural buckling can be directly
calculated from values given in Table 5.3, according to the steel grade and the
non dimensional slenderness of steel member at normal temperature
. Values
of reduction factor
fi
_ in Table 5.3 were calculated assuming a slenderness in
the fire situation equal to 3 . 1
= . For intermediate value of non-

dimensional relative slenderness, linear interpolation may be used.
7 - 26
Table 5.3 Values of reduction factor _
fi
as function of non dimensional
slenderness at normal temperature and the steel grade
Steel grade Steel grade

S235 S275 S355

S235 S275 S355
0,2 0,8480 0,8577 0,8725 1,7 0,1520 0,1549 0,1594
0,3 0,7767 0,7897 0,8096 1,8 0,1381 0,1406 0,1445
0,4 0,7054 0,7204 0,7439 1,9 0,1260 0,1282 0,1315
0,5 0,6341 0,6500 0,6752 2 0,1153 0,1172 0,1202
0,6 0,5643 0,5800 0,6050 2,1 0,1060 0,1076 0,1102
0,7 0,4983 0,5127 0,5361 2,2 0,0977 0,0991 0,1014
0,8 0,4378 0,4506 0,4713 2,3 0,0903 0,0916 0,0936
0,9 0,3841 0,3951 0,4128 2,4 0,0837 0,0849 0,0866
1 0,3373 0,3466 0,3614 2,5 0,0778 0,0788 0,0804
1,1 0,2970 0,3048 0,3172 2,6 0,0725 0,0734 0,0749
1,2 0,2626 0,2691 0,2794 2,7 0,0677 0,0686 0,0699
1,3 0,2332 0,2387 0,2473 2,8 0,0634 0,0642 0,0653
1,4 0,2081 0,2127 0,2200 2,9 0,0595 0,0602 0,0612
1,5 0,1865 0,1905 0,1966 3 0,0559 0,0565 0,0575
1,6 0,1680 0,1714 0,1766

Steel beams
The design moment resistance for the fire design situation of a laterally
unrestrained beam with a Class 1, 2 or 3 cross-section, at a uniform
temperature u
a
is given by:
( )
Rd y,
fi M,
M0
f LT, Rd t, fi,
M k M
i
_
u
= (21)
where:
y,
k is the reduction factor for the yield strength of steel at the steel
temperature reached at time t
Rd
M is the moment resistant of the gross cross-section (plastic moment
resistant
Rd pl,
M or elastic plastic moment resistant
Rd el,
M for the
normal temperature design calculated using EN 1993-1-1
fi LT,
_ is the reduction factor for lateral-torsional buckling in the fire
design situation. It may be calculated in the same way as the
reduction factor for flexural buckling but using the appropriate non-
dimensional slenderness
For laterally restrained beams, the same design method can be used, adopting
1
fi LT,
= _ .
Often structural members will not have a uniform temperature. An adaptation
factor
1
can be introduced to take account a non-uniform temperature
distribution over the height of the steel section. A further adaptation factor
2

can be also introduced to account for variations in member temperature along
the length of the structural member when the beam is statically indeterminate.
7 - 27
The value of the adaptation factors
1
and
2
should be taken according to
EN 1993-1-2.
Members subject to combined bending and axial compression
A simplified design method is also available to verify the fire resistance of
steel members subjected to combined bending and axial compression, such as
slender columns under eccentric load and long beams with lateral buckling. For
this situation, the simple calculation model takes into account the combination
effect of bending and compression by combining above two models for the
simple loading condition. Detailed information is given in EN 1993-1-2.
5.3.3 Determination of fire protection material thickness
In situations where requirements with respect to fire resistance are high
(generally more than R30), the application of prescriptive rules usually leads to
the fire protection of steel structures. When passive fire protection is necessary,
the knowledge of the critical temperature, the section factor and the fire
resistance time required, allow for a given fire protection system (spray, board,
intumescent coating), determination of the thickness to apply. Only products
which were tested and assessed in standard fire tests according to the European
standard EN 13881 may be used in practice.
The required thickness can usually be determined from manufacturers
published data. Such manufacturers data can be given in form of table or
diagram as illustrated in Figure 5.7. The data generally relates the thickness of
fire protection material to the section factor of the steel member (A
p
/V), the
critical temperature and the fire resistance time required. For typical building
construction using standard I and H steel profiles, the value of A
m
/V is usually
in the range 30 450 m
-1
.
Fire resistance rating R60
Section factor Ap/V (m
-1
)
S
t
e
e
l

t
e
m
p
e
r
a
t
u
r
e

(
C
)

Figure 5.7 Example of French diagram for boarded fire protection
In practical design, for a given fire protection material, the thickness may be
determined according to following steps:
- Choose the data related to the fire resistance time required
- Calculate the section factor according to the shape of the steel profile, the
presence of any shading of the structural member against heat transfer from
7 - 28
the fire during the fire duration (for example a concrete slab put on the
upper flange of the profile), the type of fire protection (according to the
outline of the steel profile or in box)
- Determine the thickness from the manufacturers data using the critical
temperature and the section factor. Linear interpolation is permissible to
determine thickness.
The European Convention for Constructional Steelwork (ECCS) has developed
so-called Euro-nomograms
[13]
, which relate for a given time of standard fire
exposure, the temperature reached by insulated steel members to the factor
(
p
/d
p
) (A
p
/V) depending on the fire protection characteristics (
p
and d
p
) and
the section factor A
p
/V. Note that these Euro-nomograms are determined on the
basis of the ENV version of the fire part of Eurocode 3. Also for this reason
they should be used with some caution. Other nomograms based on
EN 1993-1-2 have been recently developed
[14]
.
5.3.4 Design tables for composite members
Design tables for composite members are given in EN 1994-1-2. They are
applicable only to steel and concrete composite members (composite beams
with partially or fully concrete encasement of steel beam, composite columns
with partially or fully concrete encased profiles, composite columns with
concrete filled rectangular or circular steel hollow sections). They use
predefined values, based mainly on standard fire test results, improved with
analytical investigation. The tables allow the designer to quickly obtain the
member size (minimum dimensions of cross-section, the necessary reinforcing
steel area and its minimum concrete cover) as a function of the load level for
common standard fire resistances. The most important advantage of this
method is the ease of application. However it is limited by a very strict set of
geometrical rules and it gives more conservative results compared to other
simple calculation models or advanced calculation models. As a consequence,
it should only be applied for the pre-design of a building.
Detailed information is given in EN 1994-1-2.
5.3.5 Simplified calculations models for composite members
The following design methods have been developed to predict the resistance of
individual members when exposed to a standard fire curve. Therefore they are
not applicable to natural fires.
Only the design methods for the most commonly used composite members in
single-storey building (composite columns and partially encased concrete
beams) are described here.
Composite columns
The simple design methods for columns allow the designer to assess the fire
resistance of a composite column by calculating its buckling resistance using
the temperature distribution through the cross-section and the corresponding
reduced material strength defined at the required fire resistance time. This
method is based on the buckling curve concept: the plastic resistance to axial
compression N
fi,pl,Rd
and the effective flexural stiffness (EI)
fi,eff
, are used to
derive a reduction factor for buckling. The method is applicable to all types of
7 - 29
composite column provide that an appropriate buckling curve is used.
Checking the column consists of proving that the axial compression (for the
combination of actions considered in fire situation according to EN 1991-1-2)
is less than the buckling resistance of the column.
For a given temperature distribution across the cross-section, the design
resistance of a composite column N
fi,Rd
can be determined from the appropriate
buckling column curve relating the load capacity N
fi,Rd
to the plastic load
N
fi,pl,Rd
and the elastic critical load N
fi,cr
as follows:
( )
Rd pl, fi, Rd fi,
.N N _ = (22)
_ is the reduction factor for flexural buckling depending on the slenderness in
fire situation
.For composite columns,
may be defined as:

cr fi, R pl, fi,
/ N N = (23)
where:
cr fi,
N is the Euler buckling load
R pl, fi,
N

is the value of N
fi,pl,Rd
according to (24) when the partial security
factors
M,fi,a
,
M,fi,s
, and

M,fi,c
,of the materials are taken as 1.0
The reduction factor _ is determined as for normal temperature design but
using an appropriate buckling curve defined as function of column type
(partially encased steel section, filled hollow steel section).
The ultimate plastic load, N
fi,pl,Rd
of the cross-section is determined by
summing the strengths of every part of the cross-section (yield stress for steel
parts, compressive strength for concrete parts) multiplied by the corresponding
areas, taking into account the effect of temperature on these elements, without
considering their interaction (due to differential thermal stresses), i.e.:

+ + =
m
c
k
s
j
f
A
f
A
f
A N ) ( ) ( ) . (
c fi, M,
c,
s fi, M,
s,
a fi, M,
ay,
a Rd pl, fi,

(24)
N
fi,cr
is the Euler buckling load calculated as a function of the effective flexural
stiffness of the cross-section
eff fi,
) (EI and the buckling length
u
of the
column in fire situation, i.e.:
2
eff fi, 2
cr fi,
) (
EI
N = (25)
The effective rigidity (EI)
fi,eff
is determined from:

+ + =
m k j
I E I E I E EI ) ( ) ( ) ( ) (
c, sec, c, c, s, s, s, a, a, a, eff fi,
(26)
where:
, i
E is the characteristic modulus of material i at the temperature u. For
steel, it is the modulus of elasticity. For concrete: 2 / 3
sec c, c,
E E =
u

7 - 30
where
sec, c,
E is the characteristic value for the secant modulus of
concrete in the fire situation, given by the ration between f
c,
and
c
cu,u

I
i
is the second moment of area of material i related to the central axis
(y or z) of the composite cross-section
a,u
(for steel profile),
s,u
(for reinforcements) and
c,u
(for concrete) are
reduction coefficients due to the differential effects of thermal stresses.
Detailed information is given in EN 1994-1-2 4.3.5.
Partially encased steel beams
The simple design method for partially encased steel beams allows the designer
to assess the fire resistance by calculating its bending resistance at the required
fire resistance time. It is based on the simple plastic moment theory. The
method requires the calculation of the neutral axis and corresponding bending
resistance, taking into account temperature distribution through the cross-
section and corresponding reduced material strength. Distinction is made
between sagging moment capacity (usually at mid-span) and the hogging
moment capacity (at the support, if appropriate). If the applied moment is less
than the bending resistance of the beam, the member is deemed to have
adequate fire resistance.
The plastic neutral axis of the beam is determined such that the tensile and
compressive forces acting in the section are in equilibrium:
0
1
c fi, M,
c,
, c,
1
a fi, M,
, y
, y,
=
|
|
.
|
\
|
+
|
|
.
|
\
|

= =
m
j
j
j j
n
i
i
i i
f
k A
f
k A

u
(27)
where:
f
y,i
is the nominal yield strength for the elemental steel area A
i
taken as
positive on the compression side of the plastic neutral axis and
negative on the tension side
f
c,j
is the nominal compressive strength for the elemental concrete area
A
j
taken as positive on the compression side of the plastic neutral
axis and negative on the tension side
The design moment resistance
Rd t, fi,
M may be determined from:

= =
|
|
.
|
\
|
+
|
|
.
|
\
|
=
m
1 j c , fi , M
j c,
j , c, j j
n
1 i a , fi , M
i , y
i , y, i i Rd , t , fi
f
k z A
f
k z A M

u u
(28)
where:
z
i
, z
j
are the distances from the plastic neutral axis to the centroid of the
elemental area A
i
and A
j

7 - 31
For the calculation of the design value of the moment resistance, the cross-
section of the beam is divided into various components, namely:
- the flanges of the steel profile
- the web (lower and upper parts) of the steel profile
- the reinforcing bars
- the encased concrete.
To each of these parts of the cross-section, simple rules are given which define
the effect of temperatures and allow calculation of the reduced characteristic
strength in function of the standard fire resistance R30, R60, R90 or R120.
Detailed information is given in EN 1994-1-2 4.3.4.
5.4 Specific design rules for single-storey buildings
National fire regulations of many European countries have been changed
recently to introduce, for single-storey storage and industrial buildings with
significant fire risks (high fire loads), specific safety requirements in terms of
structural behaviour as an alternative to standard prescriptive requirements.
The following criteria relating to the structural behaviour of storage and
industrial buildings (load-bearing structure, faade elements, roofing and fire
walls) must be satisfied to ensure adequate life safety for building occupants
and firemen:
- In case of fire occurring in one of the cells of the building, its structure
(including faade elements) must not collapse towards the outside.
- In case of fire occurring in one of the cells of the building, the localized
failure of the cell in fire must not lead to the collapse of the neighbouring
cells.
To help the design of storage and industrial buildings with a steel structure,
several simple design methods can be used
5,6
. These design methods allow the
designer to easily prove that the behaviour of the steel structure of these
buildings in fire situations fulfils the above criteria. The methods are
implemented in the LUCA software
[15]
.
The design methods enable the designer to:
- Evaluate forces induced by the collapse of the heated part of the structure.
These forces should be used as additional horizontal load for the stability
check of the part of the frame that remains cold during the fire. That part
can be assessed using normal conditions design tools for structure analysis.
- Provide maximum horizontal displacements developed at the ends of the
compartment affected by the fire. These displacements are used to ensure
that movements of the structure in the event of fire do not adversely affect
the stability of fire walls or building faades. Design methods used for this
verification depend on the type of the wall (such as in lightweight concrete,
reinforced concrete, hollow block, steel sheeting with insulator,
plasterboard, bricks, etc.) and connection to the steel frame.
7 - 32
The following buildings can de designed by these methods:
- Storage and industrial buildings with steel structure. Either steel portal
frames with standard H or I hot rolled profiles or equivalent welded plate
girders, or steel frames based on lattice beams with columns in standard H
or I hot rolled profiles or equivalent welded plate girders
- Storage and industrial buildings of portal frame construction divided in
several cells, separated one from each other by fire walls. These walls can
be either perpendicular to the steel portal frames or parallel to the steel
portal frames (see Figure 5.8).
These methods were specifically developed for storage and industrial buildings
but they can also be applied to other type of single-storey buildings.

fire wall perpendicular
to the steel frame
fire wall parallel to the
steel frame

Figure 5.8 Location of fire wall compared to steel frames
Calculation methods (see Section 5.5) are only required when fire walls are
perpendicular to steel frames of the building and the building height exceeds
20 m
5
. When fire walls are parallel to steel frames, the risks of collapse
towards the outside and progressive collapse (between different fire
compartments) can be simply avoided by following the recommendations in
Section 5.5.3.
7 - 33
5.5 Simplified design methods
A flowchart showing simplified calculation methods is given in Figure 5.9.

yes
No
No
yes
(*) For all the possible fire scenarios according to the building arrangement
Industrial hall
Checking of failure modes
Choice of fire scenario
(*)
(see Figure 5-14)
Calculation of displacements of the steel structure i
(see expression (30))
Calculation of tensile force Fi
(see expression (29))
Checking of the compatibility of displacements
Steel structure and partition elements
Steel structure and facade elements
Checking of the stability at the
ultimate limit states of the cold
parts of the steel structure
End of
checking
yes
Change in the
steel structure
Change in the design of partition
or facade elements to ensure the
compatibility of displacements
No
Is it a simple isolated
portal frame?
Design recommendations
at the bottom of columns
(see end of 5.6.2)
yes

Figure 5.9 Application flowchart of calculation methods
The calculations of tensile force and lateral displacements at compartment ends
must be performed for all possible fire scenarios. Examples of scenarios are
given in Section 5.5.3. Calculation methods are given in Sections 5.5.1 and
5.5.2.
5.5.1 Tensile force at compartment ends
m
1
= 1 m
2
= 2 n = 1
K
2

F
F
K
1

Figure 5.10 Horizontal tensile force at the fire compartment ends
When a fire occurs in a compartment of the building, the horizontal tensile
force F at the compartment ends resulting from the collapse of the roof
structure (see Figure 5.10), which is needed to verify the stability of the cold
part of the structure can be obtained from:
q n c F
eff p
= (29)
7 - 34
where:
p
c is an empirical coefficient (depending on the slope of the roof and
the type of steel structure)
=
Frames Lattice for
Frames Portal for
45 , 1
slope 10% for 10 , 1
slope 5% for 16 , 1
slope 0% for 19 , 1
p
c
n
eff
is a coefficient related to the total number of heated bays n in the
fire compartment (see Table 5.4)
q is the linear load on roof [N/m] (equal to the load density multiplied
by the spacing between frames) applied on the beam and calculated
in fire situation (q = G +
1
S
n
), where G is the permanent load
including self-weight of the steel frame and service overloads, S
n
is
the snow load and
1
is the load factor according to load
combination coefficients defined in EN 1990 and corresponding
national annexes.
is the span of on heated bay connected to the column [m]
Table 5.4 Values of coefficient n
eff

Portal frame Lattice Frame
Setting of compartment in fire Setting of compartment in fire
Number of bay
in fire
end middle end middle
n = 1 n
eff
=0,5 n
eff
=1,0 n
eff
=0,6 n
eff
=1,0
n > 2 n
eff
=1,0 n
eff
=2,0 n
eff
=1,0 n
eff
=1,0
Where columns of the steel frame support a boundary fire wall, columns
should be designed (providing adequate robust base to columns) to resist a
horizontal force calculated according to equation (29) but using n
eff
= 1,0.
5.5.2 Lateral displacements at the fire compartment ends
In the event of fire, movements of steel single-storey buildings can be of the
order of several tens of centimetres and therefore could lead to the failure of
faade or the partition element if it is not sufficiently ductile or not accurately
fixed. So it is important to check that faade elements and fire walls in contact
with the steel structure are compatible with the lateral displacements developed
at the ends of fire compartments and that they keep their integrity to avoid the
collapse towards outside and the progressive collapse between different fire
compartments
7 - 35
Maximum lateral displacements
i
(i = 1, 2) induced at the top of columns
located at the compartment ends can be obtained using the following
expression (see Figure 5.11):
)
`
=
building the of middle in the is fire when the ; Max
building the of end at the is fire when the
th
t
th
t
i i
i
i
K
F
nl c
K
K
nl c
K
K
o (30)
where:
n is the number of heated bays
K
i
is the equivalent lateral stiffness of the considered part i of the
structure [N/m]
K
t
is the equivalent stiffness (depending on equivalent stiffnesses
1
K and
2
K ) given by:

2 1
2 1
t
K K
K K
K
+
=
is the span of one heated bay connected to the column [m]
F is the tensile force [N]
c
th
is an empirical coefficient (dependent on the slope of the roof and
the type of steel structure)
=
Frames Lattice for
Frames Portal for
009 , 0
slope 10% for 015 , 0
slope 5% for 011 , 0
slope 0% for 01 , 0
th
c
Lateral stiffness K for fire in the middle of a frame
If the fire compartment is in the middle of the frame as illustrated in
Figure 5.11 , K
1
and K
2
should be calculated by an elastic method.
1
o
2
o
m
1
= 1 m
2
= 2 n = 1
K
2

K
1

Figure 5.11 Fire located in a cell at the middle of the building
However, for usual steel frames (constant range, even standard steel profiles
from one span to another), the equivalent lateral stiffness
i
K on either side of
the fire can be calculated approximately according to the number of cold spans
on that side (m
i
) using the following relationships:
7 - 36
>
=
=
2 for
1 for
i
i
i
m ck
m k
K (31)
)
6 , 0
1 (
2 1
1 2
2
2
1
) (
12
2 1
with
c
b
3
c
h
f
l
f h
I
I
j
i
m
j
j
c
f h
EI
k
+
=
+
+
=
+ =
+
+
=
o
o
o
o
o
(32)
where, for each side in turn (i = 1, 2):
h is the height of the columns
f is the ridgepole
l is the length of the span
I
b
is the second moment of area of the beam
I
c
is the second moment of area of the column
E is the modulus of elasticity of steel for normal temperature

f
h
mi=2
Ib
Ic

Figure 5.12 Definition of parameters of cold parts on side i of the frame
Lateral stiffness K for fire at the end of a frame
If fire compartment is at the end of the frame, K
2
should be calculated as for
fire in the middle compartment. K
1
, which is defined as the lateral stiffness of
the steel frame of the heated fire compartment, should be calculated as follows:
)
`
>
=
>
=
=
=
frames lattice for
2 for 3 , 0
1 for 2 , 0
frames portal for
2 for 13 , 0
2 for 13 , 0
1 for 065 , 0
2
2
1
n K
n K
n k c
n k
n k
K (33)
where k and c are calculated from equation (32) with m
1
= n 1, where n is the
number of heated bays, as shown in Figure 5.13.
7 - 37
K
1

1
o
2
o
n = 1 m
2
= 3
K
2

Figure 5.13 Fire in a compartment at the end of the building
5.5.3 Example of fire scenarios
The above calculations must be performed for all possible fire scenarios. These
scenarios are defined in accordance with the arrangement of the storage
building (structure and partitioning) as illustrated in the example in
Figure 5.14.

Configuration of storage building: 5 spans and 3 cells
Cell 1 Cell 2 Cell 3
Fire wall Fire wall
Scenario 1: fire in cell 1
3 fire scenarios need to be considered

Figure 5.14 Fire scenarios according to the arrangement of the building
5.6 Design recommendations
Additional design recommendations for fire walls, faade elements and bracing
systems must be put into practice to avoid the collapse toward the outside of
the building and the progressive collapse of the steel structure. Obviously,
recommendations allow also the collapse of the steel structure under fire
condition on either side of fire wall without causing any damage to this wall.
7 - 38
5.6.1 Fire walls
To limit the fire spread to a neighbouring compartment from the fire
compartment, a solution that requires the building to be subdivided into
independent compartments can be achieved by implementing one of the
following construction details:
- Two independent fire walls (such as sandwich panels, prefabricated panels,
etc.) each fixed to an independent structural frame (see Figure 5.15 (a)). In
this case, when one structure and its fire wall collapse during a fire, the fire
cannot spread to the neighbouring structure, which remains stable and fire
protected by the second fire wall
- A single fire wall inserted between both structures. This fire wall can be a
self-stabilized wall and fully independent. The fire wall can be also fixed at
its top to both structures by means of fusible ties (see Figure 5.15 (b))
which, in case of fire near the wall, releases the connection to the hot
structure (usually when a temperature from 100 to 200C is reached in
bolts) without causing any damage to the wall (it one remains attached to
the steel structure located on the cold side) and the stability of the
neighbouring cold structure.
Self-stabilized walls are commonly used in practice. However during a fire,
this solution can be dangerous for people (occupants and firemen) because they
collapse away from the fire as a consequence of thermal bowing effect. So,
they should be used only if their behaviour has been evaluated by advanced
calculation model taking into account second order effects. Moreover, where
spacing from the self-stable wall to the neighbouring steel structure is not
sufficient, it is important to make sure that the fire wall can bear the force
which may be induced by the movements of the building due to the thermal
elongation of the roof structure (beams and purlins) due to the increase of
temperature in the cell with the fire.
As an alternative to the previous solutions, it is possible to insert the fire wall
into the steel structure of the single-storey building as illustrated in
Figure 5.15(c). Such wall can be either perpendicular to the steel frame or
parallel to the steel frame. Several solutions can be then considered: fire wall
inserted into a line of columns, fire wall attached to columns or fire wall
moved from a line of columns. For these solutions, adequate measures must be
implemented to avoid the collapse of the wall as a result of significant lateral
displacements of the steel structure. These measures concern:
- The attachment of fire walls to the steel structure
- The fire protection of the steel structure near fire walls,
- The roof system above fire walls
- The bracing system.
7 - 39
a) Doubling of the structure as
well as fire walls
b) Doubling of structure with
fire wall fixed by fusible
ties
c) Example of fire wall
inserted into the steel
structure

Figure 5.15 Some solutions of fire walls
Attachment of faade elements and fire walls to steel structure
Fire walls and faade elements fixed to steel structure of single-storey-
buildings have to remain solidly attached in order to prevent any failure of
these elements due to significant lateral displacements of structure in the event
of fire, and so to avoid risks of progressive collapse and collapse towards the
outside of the building.

3m
3m
3m
3m
Fire wall
Facade element

Figure 5.16 Design detail for faade elements and fire walls
One solution consists of fixing these elements to the columns of the load-
bearing structure by means of suitable attachment systems uniformly
distributed over the building height. The maximum spacing of these
attachments will be fixed by the manufacturer of the walls; it is recommended
that the spacing should not exceed 3 m for walls constructed on-site walls
(concrete, masonry, etc.).
In addition, fastenings used to connect fire walls and faade elements on the
columns must be designed to resist the forces produced due to wind and self-
weight of partition elements under the effect of the lateral displacement
induced by the steel frame of the building. If these fastenings are in steel and
unprotected against fire, each of them must be designed at ambient temperature
to resist the following force:
n d p W F
i
/ 5 o + = (34)
where:
W is the characteristic wind load used for the design at ambient
temperature and applied to each fastening [N]
p is the self-weight of the wall [N/m]
d is the spacing between frames [m]
7 - 40
n is the total number of fastenings (uniformly distributed along the
height)
i
o is the maximum lateral displacement obtained from relation
(26) [m]
Fire protection of steel elements near to fire walls
The requirement that there should be no fire propagation between different
compartments and no progressive collapse (i.e. the integrity condition of fire
walls must be preserved and the cold parts of the structure must remain stable),
leads to the requirement that that columns used as supports of fire walls must
achieve the same fire resistance as required for fire walls. In common cases,
these fire requirements lead to the application of fire protection to the columns.
On the other hand, columns which do not support fire walls will not require fire
protection.
Additionally, structural members that could damage fire walls (such as beams
and purlins near or crossing the walls) will also have to be fire protected.
5.6.2 Recommendations for steel portal frames
Fire wall perpendicular to steel frame
Figure 5.17 illustrates the situation where the fire wall is perpendicular to the
steel frame. For this situation:
- Columns that are built into or near a wall must be fire protected.
- Where fire wall is inserted between the flanges of the columns, no
additional fire protection is needed for the roof beams (Figure 5.17 (a)).
- Where portal frames do not have haunches and fire wall is fixed to one
flange of columns, fire protection must be applied to any beam crossing the
fire wall (on the side of the wall) over a minimum length of 200 mm
beyond the wall limit. This protection allows a shift of the plastic hinges
away from the walls and thus prevents damage to the wall as a result of the
collapse of the beam (see Figure 5.17 (b)). Where portal frames have
haunches, no fire protection is needed for the beams.
- Purlins do not cross the fire wall in this situation and no special
considerations are required.
The thickness of fire protection material applied to columns may be calculated
assuming a critical temperature of 500C and the same required fire resistance
as the fire walls. Fire protection should be provided over the full height of
columns.
If beams are partially protected, the thickness of fire protection material may
be calculated assuming a steel section exposed on four faces for the section
factor, a standard fire exposure of one hour and a critical temperature of 500C.
7 - 41
beam
purlin
fire
wall
protected
column

purlin
fire
wall
protected
column
fire
protection
d > 200 mm
beam

a) Wall inserted between the flanges of
columns
b) Wall fixed to one flange of columns

Figure 5.17 Design detail near fire walls perpendicular to portal steel frame
Fire wall parallel to steel frame
Figure 5.18 illustrates the situation where the fire wall is parallel to the steel
frame.
For this situation:
- The fire wall either be located between two frames or in the plane of the
frame, between faces of the columns and beams.
- Columns and beams that within the fire wall or near a fire wall must be fire
protected.
- Purlins will cross the fire walls. It is therefore necessary to fire protect
continuous purlins (over a distance of 200 mm from the wall) or to design a
non-continuous purlin system. For example, where fire wall is in the plane
of a frame, steel elements fixed to the beams should be inserted through the
wall to support the purlins.
The thickness of fire protection material applied to columns and beams may be
calculated assuming a critical temperature of 500C and the same required fire
resistance as the fire walls. Fire protection should be provided over the full
height of columns.
7 - 42
fire
wall
protected
beam
purlin
protected
column
purlin
flexible fire
protection
material
continuous purlin
protected
beam
d > 200 mm
protected
column fire wall
fire
protection
purlin
Rigid
support
fire wall
purlin
protected
beam
protected
column
a) Fire wall inserted between
the flanges of columns
b) Fire wall fixed to one flange of columns

Figure 5.18 Design detail near fire walls parallel to portal steel frame
If purlins are partially protected, the thickness of fire protection material may
be calculated assuming a steel section exposed on four faces for the section
factor, a standard fire exposure of one hour and a critical temperature of 500C.
Additional design recommendations for simple portal steel frames
In the case of single-storey buildings with simple portal steel frame where the
column height/beam span ratio of the frame (h/l) is greater than 0,4, the failure
mode towards the outside can be avoided by designing the connections
between columns and foundation, and the foundation itself, to have sufficient
resistance to sustain the vertical loads in the fire situation together with an
additional bending moment equal to 20% of the ultimate plastic moment of the
column at normal temperature.
Fire wall
simple portal steel frame
simple portal steel frame
h
L

Figure 5.19 Single-storey buildings with simple portal steel frame
Examples of fire walls
Illustrations of fire walls adopting some of the above recommendations are
shown in Figure 5.20. They show clearly that the fire walls were not damaged,
despite the collapse of the steel structure.
7 - 43

a) Self-stable fire wall inserted between two
independent steel framework
b) Partially fire protected steel beam crossing
a fire wall fixed to steel columns

Figure 5.20: Views of fire walls after fire disaster in steel single-storey building
5.6.3 Recommendations for steel frames based on lattice beams
Fire wall perpendicular to steel frame
Figure 5.21 illustrates the situation where the fire wall is perpendicular to the
steel frame. For this situation:
- Columns that are built into or near a wall must be always fire protected.
- Where fire wall is inserted between the flanges, the lattice beams should be
fire protected on both side of the wall (see Figure 5.21 (a)).
- Were the fire wall is fixed to one flange, only the lattice beams on the wall
side have to be protected. Fire protection must be applied to the beams over
a minimum length equal to the distance separating the wall with the first
vertical member of lattice frame (see Figure 5.21 (b)).
- Purlins do not cross the fire wall in this situation and no special
considerations are required.
The thickness of fire protection material applied to columns may be simply
calculated assuming a critical temperature of 500C and the same fire
resistance as required for fire walls. Fire protection should be provided over the
full height of the columns.
If lattice beams are partially protected, the thickness of fire protection material
may be calculated assuming for the section factor: a steel section exposed on
four faces for bottom chords, vertical members and diagonals; and on three
faces for top chords. A standard fire exposure of one hour and a critical
temperature of 500C may be used.
7 - 44
first vertical
member
fire wall
protected
column
fire protection
lattice beam

fire
Wall
protected
column
fire protection
first vertical
member
lattice beam

a) Fire wall inserted between the flanges
of columns


Figure 5.21 Design detail near fire walls perpendicular to steel frame with
lattice beam
Where fire wall is parallel to steel frame
Figure 5.22 illustrates the situation where the fire wall is parallel to the steel
frame. For this situation:
- It is not practical to provide a wall in the plane of a frame, because it is
difficult to make it continuous through the depth of the lattice beam. roof,
Fire walls parallel to a frame are therefore usually either beside and in
contact with the steel frame or between two independent steel structures.
- Where the fire wall is attached to a steel frame, the columns and beams
must be fire protected (see Figure 5.22 (b)). Moreover purlins and beam
stays near the wall must be fire protected over a minimum length
corresponding to the distance from the wall to the joint purlin/beam stay
when the roof structure is made of purlins.
- Where the fire wall is inserted between two independent steel structures, no
fire protection is needed (see Figure 5.22 (a)).
If columns are protected, the thickness of fire protection material may be
calculated assuming a critical temperature of 500C and the same fire
resistance as required for fire walls. Fire protection should be provided over the
full height of the columns.
If lattice beams are protected, the thickness of fire protection material may be
calculated assuming for the section factor: a steel section exposed on four faces
for bottom chords, vertical members and diagonals; and three faces for top
chords. A standard fire exposure of one hour and a critical temperature of
500C may be assumed. Fire protection should be provided over the full length
of the lattice beams.
The thickness of fire protection material applied to purlins and beam stays may
be simply calculated assuming a steel section exposed on four faces for the
section factor, a standard fire exposure of one hour and a critical temperature of
500C.
7 - 45
purlin
fire wall

beam stay
column

protected
column
purlin
fire
wall
protected
lattice beam

beam stay

a) Fire wall inserted between two independent
steel framework


Figure 5.22 Design detail near fire walls parallel to steel frame with lattice
beam
5.6.4 Recommendations for bracing system
The requirement for no collapse towards the outside of the building in the the
longitudinal direction (perpendicular to steel frames) can be satisfied using
appropriate bracing systems. Specifically, each compartment must have its own
bracing system.
Fire wall perpendicular to the steel frame
Figure 5.23 (a) illustrates the situation where the fire wall is perpendicular to
the steel frame. For this situation:
- Use additional vertical bracing systems at both ends of fire wall, to ensure
integrity of wall. These bracing systems should be designed to support a
lateral load taken as 20% of that due to normal wind actions (according to
the combination of actions for the fire situation), calculated for a gable area
that is limited to the width between gable posts.
- Provide double bracing systems (i.e. have bracing systems on both sides of
fire walls) or protect the bracing system.
- The bracing systems must be arranged in a way that they will not cause
problems for normal temperature design, for example by compromising
movement of an expansion joint.
Fire wall parallel to the steel frame
Figure 5.23 (b) illustrates the situation where the fire wall is parallel to the steel
frame. For this situation:
- Install bracing systems (vertical bracing and horizontal bracing on roof) in
each compartment. This solution may lead to additional bracing systems for
normal conditions.
- Design each bracing system to provide adequate stability in normal
condition and to support in fire condition a horizontal uniform load [N/m]
taken as F = 1,19 (G +
1
S
n
)l
f
, where l
f
is the spacing between steel
frames, G is the permanent action, including service overloads, S
n
is the
7 - 46
snow load and
1
is the frequent combination factor according given in the
relevant National Annex to EN 1990.
- Where the fire wall is fixed to one flange of the columns, the elements of
bracing systems must be fixed to rigid steel elements supporting the purlins
on the side of the wall.

Fire wall
Building end
Doubling of additional
bracing system put at the
end of fire wall
Bracing system for
normal temperature
Building
end

Wall perpendicular to steel frame

Fire wall
Bracing system

Wall parallel to steel frame

Figure 5.23 Recommendations for bracing system
5.6.5 Recommendations for roof systems above the separation
elements
The roof should be independent from one compartment to the next, adopting
the following recommendations (see Figure 5.24 (a)):
- Purlins should be provided either side of the fire wall.
- The roof should be stopped on both sides of the fire wall
- The roof should be provided with fire protection over a width of 2,50 m
either side of the wall.
Alternatively, the wall may be extended above the roof, up to a specific
distance d (see Figure 5.24 (b)).
National regulations may specify other special requirements for roof covering
adjacent to fire walls.
protected
column
Fire wall

beam
purlin
roof with fireproof material >2x2,50m
roof

part of roof between purlins
fire
protection

Protected
column
fire wall
beam

purlin
roof

d
fire
protection

a) Roof with fireproof material b) wall above the roof

Figure 5.24 Roof system above the separation elements
7 - 47
6 GUIDANCE ON THE USE OF MORE
ADVANCED SOLUTIONS
This chapter gives an overview of advanced calculation models available for
fire modelling, thermal modelling, and structural modelling that can be used in
fire engineering design
[9,16]
.
6.1 Fire models
Two kind of numerical models are available to model the development of real
fires: zone models and field models. These models and allow temperatures,
smoke descent, flame spread, time to flashover and many other effects to be
calculated.
6.1.1 Zone models
The simplest model is a one-zone model for fully developed fires (post-
flashover fires), in which the conditions within the compartment are assumed
to be uniform and represented by a single temperature.
Two-zone models may be used for pre-flashover situations, mainly in the
growth phase of a fire. The model is based on the hypothesis of smoke
stratification, separating the fire compartment into two distinct layers: a hot
upper layer (containing most of the fires heat and smoke), and a cool lower
layer (which remains relatively uncontaminated by smoke). A fire plume feeds
the hot zone just above the fire. The temperature of each layer is calculated
from conservation of energy; the amount of toxic combustion products in each
layer is calculated from conservation of chemical species; and the size of each
zone is calculated from conservation of mass. Simple rules govern plume
entrainment, heat exchange between zones and mass flow through openings to
adjoining compartments. As a result of the simulation the evolution of gas
temperature in each of the two layers, the evolution of wall temperatures,
evolution of flux through the openings and the evolution of the thickness of
each layer are given as a function of time. The thickness of the lower layer,
which remains at rather cold temperature and contains no combustion products,
is very important to assess the tenability of the compartment for the occupants.
Often, the local effect near the fire may be studied using a simple model such
as Hasemi methodology with the two-zone models. The combination of both
models then allows the determination of the gas temperature field near and far
from the fire (see Figure 6.1).
When the thickness of the lower layer is too small compared to the height of
the compartment, the two-zone assumption becomes inapplicable and a one
zone model becomes more appropriate. Moreover if the fire area is big
compared to the floor area, the one-zone model assumption is usually better
than the two-zone one.
Some zone models include the possibility of a switch from a two-zone model
to a one-zone model when some conditions for temperatures, fire area and
smoke layer thickness corresponding to flashover) are encountered.
7 - 48
It is also still possible to choose to follow a two-zone or a one-zone strategy for
the entire duration of a fire. With these strategies, the whole simulation is made
considering two or one zones, from the initial time to the end of the calculation.
No modification of the rate of heat release is made, except via the combustion
models.
Localised fire
2 zone
model
roof
u
g
Beam
Hasemi methods
u
20C
z
u
g
2-zone model
u at beam l evel
x

Figure 6.1 Combination of two-zone model with Hasemi methods
Some of the more complex zone models allow radiation calculations between
the upper layer and room objects. They may also allow multiple fire plumes
and multiple compartment analysis with mass exchange between each
compartment (see Figure 6.2).
The input data are usually the room geometry, room construction (including all
walls, floors and ceilings), number of vents (or holes) and their sizes, room
furnishing characteristics, and fire data (such as RHR curve, pyrolisis rate,
combustion heat of fuel). The output data are usually the prediction of sprinkler
and fire detector activation time, time to flashover, upper and lower layer
temperature, smoke layer height, and species yield.
The fire load can be considered to be uniformly distributed if the combustible
material is present more or less over the whole floor surface of the fire
compartment and when the fire load density (quantity of fuel per floor area) is
more or less uniform. By contrast, the fire load should be localised if the
combustible material is concentrated on quite a small surface compared to the
floor area with the rest of the floor area being free of fuel.
An essential parameter in advanced fire models is the rate of heat release. For
design it is common practice to refer to the values given in EN 1991-1-2.
For irregular or complex building geometry, complex ventilation systems, or
where more detail is required on convective or radiant heat exposure levels at
specific targets, the use of a field model should be considered.
7 - 49
A7
12.5m*9m
A13
25m*54m
A3
12.5m*9m
A4
12.5m*9m
A8
12.5m*9m
A2
25m*18m
A1
25m*18m
A12
25m*18m
A9
12.5m*9m
A14
25m*54m
A10
12.5m*9m
A5
12.5m*9m
A6
12.5m*9m
A11
25m*18m

Fire source

Figure 6.2 Example of fire modelling using zone models for an industrial
building
6.1.2 Field models
Field models (computational fluid dynamics models) are the most sophisticated
deterministic models for simulating enclosure fires. They incorporate sub-
models for turbulence, heat transfer and combustion.
The CFD modelling technique is based on a complete, time-dependent, three-
dimensional solution of the fundamental conservation laws (conservation of
mass, momentum, and energy). The volume under consideration, usually a fire
compartment, is divided into a very large number (sometimes hundreds of
thousands or even millions) of cells. The approximate number of cells
appropriate for the studied compartment will depend on the compartment
geometry, the accuracy required, and from a practical standpoint, the computer
speed and memory.
Three cases of field models, according to the turbulence method implemented
in model, exist:
- Direct numerical simulations (DNS): The basic equations are directly
solved but need very short time and spatial steps in order to simulate all
time and spatial scales coming from the turbulent and the chemical
processes. DNS require particularly powerful computers and are used for
academic studies or are confined to simple applications.
- Large Eddy Simulation (LES): Large scale motions of the flow are
calculated while the effect of smaller scales is modelled using sub-grid
scale model. The most commonly used sub-grid model is the Smazorinsky
model.
7 - 50
- Reynolds-averaged Navier Stokes (RANS): The basic equations are
averaged and all turbulent scales are modelled. The most frequently model
used is c k model.
The input data are the same as those required for a zone model but they have to
be supplied with a higher degree of detail. They are the detailed room
geometry, room construction (including all walls, floors and ceilings), number
of vents (or holes) and their sizes, room furnishing characteristics,
fuel/combustion characteristics, turbulence parameters, and radiation
parameters.
The output data are the smoke and heat movements, prediction of sprinkler and
fire detector activation time, time to flashover, temperatures in the domain,
velocities, smoke layer height, and species yield.
Due to their complexity and the CPU time needed, field models are very little
used for evaluating fire resistance of structures, particularly for fully developed
fire. In the fire domain, the use of a field model is often reduced to specific
cases with sophisticated geometry.
6.2 Thermal Models
Advanced heat transfer models can be used to calculate temperature
distribution in a structure in a fire. They are mostly based on either finite
difference methods or finite element methods. They are often used to estimate
temperature gradients through structural members primarily made of materials
with a low thermal conductivity and/or high moisture content, such as concrete.
Moreover, they can be applied to structural members under nominal fire
conditions or natural fire conditions.
Such methods have to take into account non-linearity due to temperature
dependence of material properties and boundary conditions. As commonly
assumed in fire design, heat transfer from fire to exposed surfaces is essentially
by convection and radiation. Inside homogeneous materials such as steel, heat
is only transferred by conduction. On the other hand, for porous materials such
as concrete or where internal cavities exist, heat transfers are more complex.
The three processes: conduction, convection and radiation can occur together,
to which may be added mass exchange. However, by way of simplification,
only the dominating process is explicitly introduced in thermal analysis, taking
into account secondary processes through adequate adjustment. In fire design,
it is usually assumed that concrete is a homogeneous material and that heat
transfer occur mainly by conduction. Heat transfer by convection and radiation
occurring in pores are considered as secondary processes and are implicitly
taken into account in thermal properties available for concrete (conductivity,
specific heat). Moreover, mass-exchange is generally neglected and only
moisture evaporation in concrete is taken into account. The effects of moisture
(assumed uniformly distributed in the concrete) is treated in a simplified way,
assuming that when the temperature in a concrete part reaches 120C, all of the
heat transferred to that part is used to evaporate water. Moisture movements
are rarely modelled. For composite members, contact between steel parts and
concrete parts can be assumed to be perfect (no gap). Radiation in internal
7 - 51
voids (such as hollow steel section) should be considered in the thermal
analysis.
In principle, where the effects of a fire remain localised to a part of the
structure, temperature distributions along structural members can be strongly
non-uniform. So a precise calculation of temperatures should be determined by
a full 3D thermal analysis. However, due to the prohibitive computing time of
such analysis, it is often considered an acceptable simplification to perform a
succession of 2D thermal analyses through the cross-sections of the structural
members. Calculations are then performed at relevant location along the length
of each structural member and the temperature gradients are obtained, assuming
linear variation between adjacent temperature profiles. This approach gives
usually a reasonable approximation to the actual temperature profile through
members and allows significant reduction of the modelling and numerical
effort. In 2D thermal analysis, cross-sections of members are commonly
discretised by means of triangular or quadrilateral plane elements with thermal
conduction capability. All sections encountered in civil engineering can thus be
modelled. Each plane element describing the cross-section can have its own
temperature-dependent material such as steel, concrete or insulation materials.
Boundary conditions can be either prescribed temperatures or prescribed
impinging heat flux to simulate heat transfer by convection and radiation from
fire to the exposed faces of structural members. Effects of non-uniform thermal
exposure may be introduced in modelling with appropriate boundary
conditions.
Effects of mechanical deformations (such as buckling of steel element,
cracking and crushing of concrete, etc.) on the temperature rise of structural
members is neglected, which is the standard practice. Consequently geometry
of structural members does not vary during the analysis
As for simple models, the use of advanced models require knowledge of the
geometry of structural members, thermal properties of the materials (thermal
conductivity, specific heat, density, moisture...) and heat transfer coefficients at
the members boundaries (emissivity, coefficient of heat transfer by
convection).
Usually for fire design, temperature-dependent thermal material properties of
concrete and steel are taken from EN 1992-1-2 and EN 1993-1-2 and heat
transfer coefficients are those given in EN 1991-1-2 respectively.
6.3 Structural models
Advanced numerical models for the mechanical response should be based on
the acknowledged principles and assumptions of the theory of structural
mechanics. They are usually finite element models. They can simulate a partial
or a whole structure in static or dynamic modes, providing information on
displacements, stress and strain states in structural members and the collapse
time of whole building if collapse occurs within the period of the fire. The
changes of mechanical properties with temperature, as well as non-linear
geometrical and non-linear material properties, can be taken into account in the
structural fire behaviour. The transient heating regime of structures during fire
7 - 52
is modelled by use of step-by-step iterative solution procedures, rather than a
steady state analysis.
This Section outlines some of the primary considerations in modelling the
behaviour of single-storey buildings with steel or composite frames in the fire
situation, notably features related to material models, computation procedure,
structural modelling, etc.
Advanced calculation models can be used in association with any heating
curve, provided that the material properties are known for the relevant
temperature range and that material models are representative of real
behaviour. At elevated temperature, the stress-strain curve of steel is based on a
linear-elliptic-plastic model, in contrast to the elasto-plastic model adopted for
normal temperature design. The steel and concrete stress-strain relationships
given in EN 1993-1-2 and EN 1994-1-2 are commonly used.
In the fire situation, the temperature field of structural members varies with
time. As stress-strain relationships of materials are non-linear and temperature
dependant, an appropriate material model has to be adopted in advanced
numerical modelling to allow the shift from one behaviour curve to another, at
each step of time (and thus of temperature). The so-called kinematical material
model is usually used for steel structures, assuming that the shift from one
stress-strain curve to another one due to the change of temperature is made by
staying at a constant plastic strain value (see Figure 6.3). This model can be
used at any stress state of steel (tension or compression). For concrete, it is
much more complicated, since the material has a different behaviour in tension
and in compression. Therefore, different shift rules are needed for when the
material is in tension or in compression. Generally, this kinematic model is
used in most advanced calculation models for fire safety engineering
applications.
Behaviour of steel is often modelled with a Von Mises yield contour including
hardening. Behaviour of concrete in compression is modelled with a
Drucker-Prager yield contour, including hardening.

c
) , (
d
d
0
1
= |
.
|
\
|
(t) =
1
t) (t + =
2
a) Behaviour law of structural steel
Parallel to
) 0
2
= |
.
|
\
|
, (
d
d
Parallel to
o
Compression
b) Behaviour law of concrete
c
o
(t) =
1
t) (t + =
2
tensile

Figure 6.3 Kinematic material models for steel and concrete
Another aspect to be noted in the application of advanced calculation models
for steel and composite structures under natural fire conditions is the material
behaviour during cooling phase. It is well known that for commonly used steel
grades, the variation of mechanical properties with temperature are considered
7 - 53
as reversible, which means that once they cool down they will recover their
initial mechanical properties. However, this phenomenon is not true with
concrete, whose composition will be totally modified when heated to an
elevated temperature. After cooling down, it cannot recover its initial strength.
Indeed, its strength might even be less after cooling than at maximum
temperature.
The effects of thermal expansion should be taken into account. This is done by
assuming that the total deformation of structural members is described by the
sum of independent terms:
r tr c th t
) ( c c c c c c + + + + = (30)
where
th
c ,
c ,
r
c and
c
c are the strains due to thermal expansion, stress, residual
stress and creep, respectively.
tr
c is the strain due to transient and non uniform
heating regime for concrete (usually neglected).
In Eurocodes, the creep strain is considered to be included implicitly in stress-strain
relationships of steel and concrete. The residual stress is usually neglected except for
some special structural analysis. The thermal strain is the thermal expansion (AL/L)
that occurs when most materials are heated. Thermal strains are not important for fire
design of simply supported steel members, but they must be considered for
composite members, frames and complex structural systems, especially where
members are restrained by other parts of the structure (as for single-storey
building divided into cells separated from one another by fire walls) since
thermally induced strains, both due to temperature rise and temperature
differential, can generate significant additional internal forces.

Distribution of temperature
for z = cte
Unit strain
Cross-section
(x = cte)
z
y
G
u
c
c
th
c
o
c
t
c
r
c

Figure 6.4 Strain composition of material in advanced numerical modelling
In general, the structural analysis in the fire situation is based on ultimate limit
state analysis, at which there is equilibrium of the structure between its
resistance and its applied loading. However, significant displacement of the
structure will inevitably occur, due to both material softening and thermal
expansion, leading to large material plastification. Therefore, advanced fire
analysis is a non-linear elasto-plastic calculation in which both strength and
stiffness vary non-linearly. From a mathematical point of view, the solution of
such analysis cannot be obtained directly and has to be achieved using an
iterative procedure:
- A step-by-step analysis is carried out in order to find the equilibrium state
of the structure at various instants (at different temperature fields).
- Within each time step, an iterative solution procedure is carried out to find
the equilibrium state of the structure behaving in elasto-plastic way.
7 - 54
Different types of convergence procedure are usually employed, such as the
pure Newton-Raphson procedure and the modified Newton-Raphson
procedure. The pure Newton-Raphson procedure is recommended for
structures made of beam elements, and the modified Newton-Raphson
procedure is recommended for structures made of shell elements.
Static analysis is normally sufficient for modelling the behaviour of a structure
in fire. However, local failure or instability of a structural member (such as
lateral buckling of purlin) does not lead to overall structural failure.
Consequently, analysis should be performed by a succession of subsequent
static and dynamic analyses to pass instabilities and to obtain the complete
failure mechanism to predict the influence of a local failure on the global
behaviour of the structure and to follow eventually progressive collapse. It has
to be kept in mind that here the aim is not the precise modelling of dynamic
effects. So, default values of the main parameters fixed in models to
determinate acceleration and damping effects can be used.
Existing boundary conditions should be rightly represented. It is common to
design structure by assuming pinned support conditions at the column bases.
However, as fully pinned bases of columns are never achieved in reality, it is
also possible, when data are available, to introduce semi-rigid connections.
Where only a part of the structure is modelled, some restrained conditions from
unmodelled part of the structure should be taken into consideration in
appropriate way. The choices of restrained conditions that have to be applied at
the boundaries between the modelled substructure and the rest of the structure
have to be chosen by the designer. For example, in case of symmetry boundary,
restraints to translation across the symmetry boundary and rotational restraint
about the two major axes on the plane of symmetry are introduced in
modelling.
Usually, beam-to-column joints are assumed to be fully rigid in the fire design
of steel and steel-concrete composite frames. However, in the case of steel
frames based on lattice beams, joints between members of lattice beams and
connections between top and bottom chords of lattice beams and columns can
be assumed pinned or fully rigid according to the type of truss.
Two types of action need to be applied to heated structures. The first type is
static loading. It must correspond to that for fire situation. The second type
consists of the temperature increase (above ambient) of the structural members
obtained, from previous thermal analysis. Boundary conditions at supports as
well as applied gravity loads are assumed to remain unchanged throughout the
fire exposure
It is important to choose an appropriate structural modelling strategy.
Simulation of the mechanical behaviour of single-storey building in fire
conditions can be performed either by a 2D or a 3D analysis.
In a 2D analysis, simulation are performed in the plane of each portal frame,
assuming a three dimensional behaviour of the frame to take into account the
lateral instability of the members (columns, beams). In such modelling,
adequate restraint conditions should be introduced to stabilize the frame
laterally. In reality, these out-of-plane restraints are provided by roof structure
7 - 55
(as purlins) as well as faades elements fixed on columns (concrete walls,
sandwich panels, steel sheeting), so that out-of-plane collapse does not occur.
In a 3D analysis, several parallel portal frames, the roof structure (purlins) and
eventually bracing system are explicitly modelled (see Figure 6.5). The main
difference in this 3D analysis is that the interaction effects between members
will be directly dealt with; load redistribution from heated parts (weakened
parts inside fire compartment) to cold parts (stronger parts outside fire
compartment) can be taken into account in an accurate way and the global
behaviour of structures will be analysed, providing a more realistic situation of
mechanical response of structures in fire. Computation cost with a three-
dimensional analysis is high because of significant number of elements used in
the modelling.
The choice between 2D and 3D analysis will depend on several parameters,
such as the type of structure (steel or composite frame), the dimensions of the
single-storey building, the fire scenario and objectives of structural fire design
(to fulfil a prescriptive requirement, or to verify a failure mode).
Fire
wall

Figure 6.5 Example of 3D mechanical modelling
The basic finite element set-ups used to represent the structural members of
frame are given below. Solid elements are omitted. as they are numerically too
expensive.
7 - 56
REFERENCES

1 EN 1991-1-2:2002 Eurocode 1: Actions on structures - Part 1-2: General rules -
Actions on structures exposed to fire
2 EN 1993-1-2:2003 Eurocode 3: Design of steel structures - Part 1-2: General rules
Structural fire design
3 EN 1994-1-2:2003 Eurocode 4: Design of composite steel and concrete structures
Part 1-2: General rules - Actions on structures exposed to fire
4 HOCKEY, S.M., and REW, P.J.
Human response to thermal radiation
HSE Books, UK, 1996.
5 VASSART, O., CAJOT, L-G., ZHAO, B., DE LA QUINTA, J.MARTINEZ DE
ARAGON, J. and GRIFFIN, A.
Fire Safety of industrial halls and low-rise buildings: Realistic fire design, active
safety measures, post-local failure simulation and performance based requirements
ECSC research project 7210-PR-378.
6 RFCS Research: Fire safety of industrial hall, Design Guide, Arcelor Mittal,
CTICM, Labein tecnalia, ULG, Directorate-General for research, Research Fund
for Coal and Steel Unit, RFS2-CR-2007-00032, Luxembourg, 2007.
7 Report to ECCS: Fire building regulations for single-storey buildings in 9
European countries. Document RT915. Version 02 June 2002.
8 LENNON, T., MOORE,D., WANG, B. Y. C. and BAILEY, G.
Designers Guide to EN 1991-1-2, EN 1992- 1-2, EN 1993-1-2 and EN 1994-1-2
Actions on Structures Exposed to Fire and Structural Fire Design
Thomas Telford, 2007.
9 DIFISEK - Dissemination of Structural Fire Safety Engineering Knowledge
ECSC research project RFS-C2-03048.
10 PURKISS, J.A.
Fire safety design of structures
Butterworth-Heinemann, Oxford, UK
11 Risk Based Fire Resistance Requirements Competitive (RISK -REI), ECSC
research project 7210-PR-378.
12 SIMMS, W.I., and NEWMAN, G.M.
Single-storey steel framed building in fire boundary conditions (P313)
The Steel Construction Institute, 2002.
13 ECCS TC3: Euro-monograms for fire exposed steelwork.
14 SD005a-EN-EU, Data: Nomogram for protected members, www.steel-access.com
15 RFCS Research: Fire safety of industrial hall, Design Guide, Arcelor Mittal,
CTICM, Labein tecnalia, ULG, Directorate-General for research, Research Fund
for Coal and Steel Unit, RFS2-CR-2007-00032, Luxembourg, 2007.
16 FRANSSEN J. M., KODUR V. and ZAHARIA R.
Designing steel structures for fire safety
Balkema Book, 2009.
7 - 57
APPENDIX A German fire safety procedure for
single-storey industrial and commercial
buildings
In Germany, buildings for commercial and industrial use must conform to the
Musterbauordnung (MBO) and to all federal state building regulations
Bauliche Anlagen und Rume besonderer Art und Nutzung (Structural
facilities and spaces with special requirements and uses). In such cases, and in
order to meet essential requirements (concerning human safety, public security,
and protection of the natural environment), it is possible to adopt alternative
solutions to the prescriptive federal state building regulations.
This general statement has to be considered in the context of physical and
technical fire protection requirements for a building with reference to of
Wohngebude und vergleichbare Nutzungen (residential and similar uses)
according to the federal state building regulations. For commercial and
industrial uses, it is neither necessary nor appropriate to apply the requirements
of the federal state building regulations. When it comes to meeting general
structural fire protection objectives, it is more important to consider each
building on an individual basis.
A standard procedure for assessing requirements, using scientifically based
methods, is recommended.
Since industrial buildings are considered Sonderbauten (special buildings)
within the definition of 51 Abs.1 MBO and cannot usually be exempt from
the applicable regulations, the goal of MIndBauRl (the technical construction
regulation) is to determine the minimum requirements for structural fire
prevention. The MIndBauRl also uses design procedures according DIN
18230-1: Structural fire protection in industrial buildings fire resistance
design.
Regarding 3 Abs. 3, Satz 3 MBO, which permits variations from technical
construction standards, the procedure limits this to accepted methods for fire
protection engineering and requires that these are listed in accordance with
Annex 1.
The aim of the procedure is to regulate the minimum requirements for fire
protection of industrial buildings, in particular regarding:
- the fire resistance of components and the flammability of building materials
- the size of fire compartments and fire-fighting areas
- the availability, location and length of emergency escape routes.
The procedure will facilitate design for building owners, designers, draftsmen
and specialists; for the authorities it will provide justification for relaxation or
deviation from the alternatively applicable rules of the MBO. It offers building
control and approval bodies a benchmark for equivalent risks. A design method
that requires no detailed engineering analyses and no particular calculation has
been established. This responds to legal responsibilities and offers a
straightforward form of approval.
7 - 58
MIndBauRl applies to all industrial buildings regardless of their size. It does
not apply to:
- industrial buildings which are only used for storing technical equipment or
facilities and where only access is temporarily needed for maintenance and
inspection purposes
- industrial buildings that are mostly open, such as covered outdoor areas or
open warehouses
- buildings which can be assimilated due to their behaviour in fire.
In addition, the procedure does not apply to storage shelves more than 9.0 m
high (to the top of stored material).
This procedure may also be used for allowing and justifying relaxation of the
regulations according to 51 MBO for buildings and structural facilities, which
are not directly covered by the scope of MIndBauRl, although they are
comparable to industrial structures in respect to fire risk.
Justification for relaxation of conditions under 51 Abs. 1 MBO may be
provided with one of the following procedures.
- Simplified procedure
In the procedure according to Abs. 6, the maximum fire compartment
surface for a fire section area will depend on the fire-resistance
classification of the supporting and stiffening components as well as the
structures fire technical protection infrastructure.
- Complete verification procedure
In the procedure according to Abs. 7, the maximum surface area and the
requirements for the components in accordance with the fire safety classes
for a fire compartment will be based on the calculation procedure according
to DIN 18230-1.
- Engineering methods
Instead of proceeding according to Abs. 6 and 7, standard fire protection
engineering design methods may also be used.
The initiator of a fire protection concept has the choice which method (Abs. 6
or 7) will be implemented when using the MIndBauRl. However it is not
permissible to combine procedures.
Concerning the fire engineering methods, the MIndBauRl identifies the
principles and conditions for the hypotheses of such designs. It regulates the
verification and checking as well as documentation.
The MIndBauRl, which has been introduced as a standard in the Building
Regulations in all German states, is legally applicable. As part of the
application of IndBauRl, there are several procedural methods. The same
general requirements apply for all verifications; these are identical for all
procedures and must be respected. These include fire-fighting water
requirements, smoke evacuation, location and accessibility, emergency exits
and fire spread.
7 - 59
Fire-fighting water requirements must be agreed with the responsible fire
department taking into account the surface areas and fire loads. These
requirements should be assumed to last for a period of two hours.
- minimum 96 m/h for a surface area up to 2500 m
- minimum 192 m/h for a surface area greater than 4000 m.
Intermediate values can be linearly interpolated.
For industrial buildings with automatic fire extinguishing systems, a water
quantity of at least 96 m/h over a period of one hour is sufficient to extinguish
the fire.
Any factory or warehouse with an area of more than 200 m must have wall or
ceiling openings to allow smoke evacuation.
Individual spaces which are bigger than 1600 m must have a smoke evacuator,
so that fire fighting operations are possible. This is because a smoke layer of
2,5 m height has been mathematically proven.
In addition to the location and accessibility of each fire compartment, at least
one side has to be located at one outside wall and be accessible from there for
the fire department. This is not applicable for fire compartments which have an
automatic fire extinguishing system.
Stand-alone and linked industrial structures with foundations of greater than
5,000 m have to be accessible from all sides by fire fighting vehicles. These
access routes must meet the requirements for fire brigade usage.
The fire service access roads, operating areas and other routes should be kept
continuously free. They have to be permanently and easily recognizable.
Included in the emergency exits in industrial buildings are the main production
corridors and storage areas, the exits from these areas, staircases and exits to
the outside. Each room with an area of more than 200 m must have at least
two exits.
Regarding the maximum allowable length for emergency escape routes,
equipment and structural fire protection both influence each other.
The maximum length of emergency escape routes is limited as a rule to 35 m
for a clear height up to 5 m. However, if a fire alarm system is installed, then
this increases to 50 m.
The maximum increase in length in relation to free height up to 50 is 70 m.
The distances are measured as distances in space, but not through construction
elements or components. The real length should not be more than 1.5 times the
distance that was measured in space. Attention should be paid to the fact that
from any point in a room, a main gangway must be reachable within a
maximum of 15 minutes.
7 - 60
In case of fire, roofs often contribute significantly to fire spread; damage will
depend on which structural fire prevention measures were implemented for the
roof.
Regarding fire propagation in case of a fire from below, then the following
failure mechanisms are typical:
- The Durchbrand burn- through. This is the worst case, with fire spreading
on top of the roof, followed by the spread of fire down into other areas
through existing roof openings.
- Failure of the load-bearing roof shell by slipping from the supports, for
example with large spans.
- Fire propagation below the roof.
- Fire propagation within the roof shell. This is very dangerous because it
will not be seen from below. It becomes very critical when the fire services
are fighting at the fire source and suddenly it begins to burn behind them.
Table A.1 Fire compartment sizes
Maximum fire compartment size (m)
Safety category
Without fire resistance
requirement
R0
With fire resistance
requirement
R30
K1
Without requirements
1800* 3000
K2
Fire detection
2700* 4500
K3
Rescue service
3200 - 4500* 5400-7500
K4
Fire suppression
(Sprinkler system)
10000 10000
* heat extraction area > 5% and building width s 40m
The simplified method is based on the relationship between the permitted
surface area of the fire compartment and the safety category, the number of
storey and the fire rating classification of the components.
The surface area is given in Table A.1 and is well within extreme safety
measures.
For industrial buildings with an existing sprinkler system (safety category K4),
a maximum fire compartment surface area of 10000 m can be realized without
requirements for the fire resistance of structural components.
Without any fire protection requirements, surface areas up to 1800 m can be
left unprotected.
For industrial buildings which cannot be evaluated using the simplified
procedure, the entire verification procedure will be based in accordance with
DIN 18230-1.
7 - 61
First, the equivalent fire duration is determined using this method. With the
equivalent fire duration, a relationship between the incendiary effect of a
natural fire and the Einheitstemperaturzeitkurve (ETK standard temperature
time curve) is generated. The equivalence refers to the maximum temperature
of structural components under a natural fire.
Once the equivalent fire duration has been determined, two different methods
are available.
The first method is to determine the maximum floor surfaces using Table A.2.
No requirements for fire resistance of structural components are needed when
using this table.
The second method requires somewhat more effort. First, the maximum floor
surface is calculated using a formula. In this procedure, the fire resistance
rating of the structural components has to be proven. This is done with the
necessary fire resistance.
Table A.2 Maximum floor area (m
2
) according to safety category and
equivalent fire duration
Equivalent fire duration
Safety category
15 30 60 90
K1
Without requirements
9000* 5500* 2700* 1800*
K2
Fire detection
13500* 800* 4000* 2700*
K3
Rescue service
1600-22500* 10000-13500* 5000-6800* 3200-4500*
K4
Fire suppression
(Sprinkler system)
30000 20000 10000 10000
Minimum heat extraction
area
1 1 3 4
Maximum building width 80 60 50 40

In Table A.2, the maximum admissible floor surface can be defined with
reference to its safety category and the equivalent fire duration. In addition, the
corresponding heat extraction surface can be identified, indicated as a % of the
floor surface and the corresponding maximum width of the building.
Using the second method for the entire verification procedure, the maximum
floor area (m) is calculated using the base value for the surface area of
3000 m and factors F1 to F5.
A = 3000 F1 F2 F3 F4 F5
where:
F1 the equivalent fire duration
F2 the safety category
F3 : the height of the lowest floors
7 - 62
F4 : the number of storey
F5 : the type of floor openings
The sum of the total surface area shall not exceed 60000 m.
According to the Table A.2, when the simplified procedure is used for
structural components without requirements, the result is a maximum possible
surface area of 10 000 m.
When using the full verification procedure according to this table, a maximum
surface of 30000 m is possible. When using the full verification procedure in
addition to the fire resistance calculation, then a 60000 m surface area is
possible.
Under very special conditions, even larger surfaces, up to 120000 m can be
achieved.
Example:
The procedure and possibilities associated with MIndBauRl can best be shown
and explained by an example:
Building parameters
Length: 100 m
Width: 50 m
Average height: 6 m
Size: 5000 m
Number of storey: 1
Openings in the roof: 135 m
Doors, windows: 132 m
Fire load: q
R
= 126 kWh/m
Automatic fire alarm systems: Safety category K2
No internal fire walls
The first possibility is the simplified method according to Table A.1. The
industrial building must be equipped with an automatic sprinkler in order to
meet the above conditions.
In order to apply fully the full verification method, the equivalent fire duration
must first be determined. In this case, the heat extraction factor w is needed.
The heat extraction factor is determined by taking into account the related
opening surfaces. The related opening surfaces are auxiliary values. This is
simply a question of dividing the roof openings by the ground surface and then
the wall openings by the ground surface.
- Determination of the related horizontal opening surface a
h
:
a
h
= A
h
/ A = 135 m / 5000 m = 0,027
7 - 63
- Determination of the related vertical opening surface a
v
:
a
v
= A
v
/ A = 132 m / 5000 m = 0,026
The values of the related opening surfaces are introduced in Figure A.1 and the
value w
0
can be defined. In Figure A.2, the height of the hall is considered.
vertical opening area a
v
horizontal opening area ah

Figure A.1 Factor w
0
according to opening areas

height of the hall (m)

Figure A.2 Factor o
w
according to height of the hall
The heat extraction value of the buildings is:
w = w
0
= 1,70 1,0 = 1,70
The equivalent fire duration (t
) is based on the following factors: the fire load

density, the heat extraction factor and a factor c which takes into account the
7 - 64
heat extraction surface of the peripheral construction elements. In this example
c is given, for simplicity, the worst value.
t
= q
R
c w = 126 0,25 1,70 = 54,0 min
Through interpolation in Table A.2, in the safety category K2 for an equivalent
fire duration of 54 minutes, a maximum surface area of 4800 m can be
defined. At this point, some additional work by the designer could be useful in
reviewing the input data. Is the fire load case too high? What will happen when
the opening surfaces are modified and the ground floor is also modified at the
same time? Alternatively, what about the surfaces? Can the surface be reduced
by 200m? The onus is on the designer to present and explain the different
opportunities to the client and to list the comparison costs.
The second possibility using the full verification method is more precise. The
maximum floor surface is calculated using the basic value for the surface of
3000 m times factors F1 to F5. The factor values are taken from tables of DIN
18230-1 and do not need to be determined.
According to table 3 of DIN 18230-1 the factor F1 is: 1,9
According to table 7 of DIN 18230-1the factor F5 is: 0,7.
Inserted into the formula:
A = 3000F1F2F3F4F5 = 3000 1,9 1.5 1,0 1,0 0,7
A = 5989 m.
In this method, the fire resistance classification of the structural components
has to be calculated with the following equation:
Required fire resistance duration t
f
= t
o
L

The design of the fire resistance duration includes the following factors:
- the equivalent fire duration of 54 minutes
- the safety factor of 0,6 according to Table 2 of DIN 18230-1, and
- the factor alpha L takes into account the fire related infrastructure of 0,9
according to Table 4 of DIN.
Hence: t
f
= 54 0,6 0,9 = 29,16 min => R30

7 - 65
Table A.3 Summary of maximum compartment sizes
Area given by simplified method (m
2
)
Safety
category Without fire resistance
requirement
With fire resistance requirement
K1
K2 2700 4500
K3 5400-7500
K4 10000
R0 R30

A comparison of these methods, the options available and responsibilities of
the designer, can be seen in table A.3. In order to contain the industrial
building in one single fire compartment without requirements for the load-
bearing structure, it is necessary to install an automatic sprinkler system when
using the simplified method. When using the full verification method and
respecting the given conditions, a fire compartment of 4800 m is possible. To
achieve one fire compartment of 5000 m, at least one plant fire service must
be present.
With a fire resistance requirement of R30 for the load bearing structure, at least
one plant fire service is required for the simplified method (according to the
table). With a fire detector system, however, only one fire compartment area of
4500 m is possible. With the full verification method, a fire compartment
surface of 5989 m is possible.
Based on the results of the different methods, the designers task is clearly
defined. He should not only develop one fire protection concept, but has to
demonstrate alternative and more economical procedures to the client in
relation to the various production processes.

SSB07 Fire Engineering

Uploaded by

Copyright:

Available Formats

SSB07 Fire Engineering

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SSB07 Fire Engineering

Uploaded by

Copyright:

Available Formats

STEEL BUILDINGS IN EUROPE

Single-Storey Steel Buildings

= 136300-121000 y if 0,30 < y < 1,0

is the non dimensional slenderness at

The non dimensional slenderness at normal temperature, according

= . For intermediate value of non-

.For composite columns,

may be defined as:

) is based on the following factors: the fire load

You might also like