Fire Performance

of Thin-Walled
Steel Structures
Fire Performance
of Thin-Walled
Steel Structures

Yong Wang
Mahen Mahendran
Ashkan Shahbazian
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2020 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper

International Standard Book Number-13: 978-1-138-54085-9 (Hardback)

This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher
cannot assume responsibility for the validity of all materials or the consequences of their use. The
authors and publishers have attempted to trace the copyright holders of all material reproduced in
this publication and apologize to copyright holders if permission to publish in this form has not
been obtained. If any copyright material has not been acknowledged please write and let us know
so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.
copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC),
222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that
provides licenses and registration for a variety of users. For organizations that have been granted a
photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks,

and are used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com

and the CRC Press Web site at

http://www.crcpress.com
Contents

Preface ix
Authors xi
List of notations xiii

1 Introduction: Fire Safety Requirements and Implications

for Thin-Walled Steel Construction 1
1.1 Thin-Walled Steel Structures 1
1.2 Fire Safety Requirements and Their Implications for
Thin-Walled Steel Structures 3
1.3 Determination of Fire Resistance of Thin-Walled Steel
Structures 6
1.4 Scope and Layout of This Book 8

2 Applications of Thin-Walled Steel Structures 11

2.1 Introduction 11
2.2 Recent Innovations and Advances in Cold-Formed Steel
Industry 14
2.2.1 Innovative Lightweight and Structurally Efficient
Sections 14
2.2.2 Prefabricated Structural and Modular Units 16
2.3 Fire Resistance of Thin-Walled Steel Structures and
Methods of Enhancement 18
2.3.1 Fire Protective Boards 19
2.3.2 Insulation 22
2.3.3 Steel Studs and Joists 22
2.3.4 Steel Sheathing 23
2.4 Summary 23

3 Fire Resistance Tests 25

3.1 Standard Fire Resistance Test 25
3.2 Fire Resistance Tests of Walls 28
3.2.1 General Behaviour of LSF Walls in Fire 28
3.2.2 Effects of Cavity Insulation on Fire Resistance 34
3.2.3 Effects of External Insulation on Fire Resistance 35

v
vi Contents

3.2.4 Effects of New Stud Sections on Fire Resistance 36

3.2.5 Effects of Plasterboard Joints on Fire Resistance 37
3.2.6 Effects of Other Types of Boards on Fire Resistance 39
3.2.7 Effects of Steel Sheathing on Fire Resistance 42
3.3 Fire Resistance Tests of Floors 44
3.3.1 General Behaviour of LSF Floors in Fire 44
3.3.2 Effects of New Joist Sections on Fire Resistance 46
3.4 Other Fire Resistance Tests on Thin-Walled Steel Structures 47
3.5 Final Comments 50

4 Numerical Modelling of Fire Resistance of Thin-Walled

Steel Structures 51
4.1 Fire Behaviour 51
4.1.1 Nominal Fires 53
4.1.2 Parametric Fires 54
4.1.3 Travelling Fires 55
4.1.4 Computer Programs 55
4.2 Heat Transfer Modelling 57
4.2.1 Basics of Heat Transfer 57
4.2.2 Thermal Boundary Conditions for Heat Transfer
Modelling 58
4.2.3 Computer Programs 59
4.2.4 A Simplified Heat Transfer Model for Thin-Walled
Steel Structural Panels 60
4.3 Modelling Behaviour of Thin-Walled Steel Structures at
Elevated Temperatures 63
4.4 Summary 64

5 Elevated Temperature Properties of Materials 65

5.1 Introduction 65
5.2 Mechanical Properties of Cold-Formed Steels at Elevated
Temperatures 66
5.3 Thermal Properties of Fire Protection Materials at Elevated
Temperatures 69
5.3.1 General 69
5.3.2 Gypsum 70
5.3.2.1 Density 71
5.3.2.2 Specific heat 72
5.3.2.3 Thermal conductivity 72
5.4 Additional Issues 73
5.5 Summary 74
 Contents vii

6 Performance-Based Design Methods of Thin-Walled

Steel Members at Elevated Temperatures 75
6.1 Thin-Walled Steel Members with Uniform Temperature
Distribution 76
6.2 Simplified Methods for Thin-Walled Members with
Non-uniform Temperature Distribution 76
6.2.1 Limiting Temperature Method 76
6.2.2 Extension of Fire Test Results 77
6.2.2.1 Members in walls 77
6.2.2.2 Members in floors 78
6.3 Thin-Walled Steel Columns with Non-uniform
Temperature Distribution in the Cross-Section 79
6.3.1 Effective Width Method 79
6.3.2 Direct Strength Method 80
6.3.3 Simplified Effective Width/Direct Strength Method 83
6.3.3.1 Effective width method 83
6.3.3.2 Direct strength method 84
6.4 Thin-Walled Steel Beams with Non-uniform Temperature
Distribution in the Cross-Section 85
6.5 Illustrative Design Example Using Direct Strength Method 85
6.6 Summary 87

References 89
Index 95
Preface

The behaviour of thin-walled structures is complex and fire resistance of struc-

tures is a niche subject. Therefore, fire resistance of thin-walled structures has
attracted a relatively low level of interest and investment from the research
community. As a result, the current design guidance for the evaluation of fire
resistance of thin-walled structures is rudimentary, still dominated by fire
resistance testing.
However, thin-walled steel structures are increasingly used in building
construction and large numbers of thin-walled steel structure products are
developed owing to flexibility of the manufacturing methods. Research and
development of thin-walled structures by fire testing will no longer be tenable
due to prohibitive cost and limited scope of application, and methods of calcu-
lation will become essential.
Developing calculation-based methods for fire resistance of thin-walled
structures demands a thorough understanding of the fundamentals of a number
of subjects and availability of reliable input data for material properties. To
aid this process of development, this book provides a single source of infor-
mation which at present is scattered in academic articles. This book attempts
to provide an authoritative account of the latest developments in the different
aspects of the subject matter, including fire resistance requirements, behav-
iour of thin-walled structures in fire and measures of improving fire resistance
of thin-walled structures, fire behaviour, simplified heat transfer modelling to
obtain temperatures in thin-walled structures, temperature-dependent thermal
and mechanical properties of materials, and latest developments in evaluating
the load carrying capacity of thin-walled steel structures at elevated tempera-
tures by calculation.
The book is jointly written by two research groups who have had a long
and distinguished record of extensive research on fire resistance of thin-walled
steel structures and are active in developing calculation methods. The authors
hope that this book will be an indispensable reference to researchers of this
increasingly important field. We also believe that the book will be valuable
to fire protection engineers who want to optimise fire resistant design of thin-
walled steel structures through developing a firm understanding of the first
principles of the subject matter, and specialist manufacturers of thin-walled

ix
x Preface

steel structures in their efforts to develop more efficient products by helping

them understand the factors that control fire resistance of thin-walled steel
structural systems.
Transforming our motivation to action of writing this book was greatly
helped by the decision of our publisher, Taylor & Francis, to launch the CRC
Focus book format which is ideal for presenting a complete treatment of this
specialist subject in a concise manner. Writing a book in our spare time for a
deadline long in the future was never going to make it a priority until reaching
the last minute, which would surely be a disaster. We were spared of this crisis
by Gabriella Williams, who regularly checked that we were making progress
according to our plan.

Yong Wang,
Mahen Mahendran,
Ashkan Shahbazian
Authors

Yong Wang is professor of structural and fire engineering at the University

of Manchester, UK, where he leads the Structural Resilience research group.
He  is author of Performance-Based Fire Engineering of Structures and
Steel and Composite Structures: Behaviour and Design for Fire Safety, also
published by Taylor & Francis.

Mahen Mahendran is professor of structural engineering at Queensland

University of Technology, Australia, where he leads the Wind and Fire research
lab. He has served for more than 35 years at six universities.

Ashkan Shahbazian is an adjunct assistant professor at the Institute for

Sustainability and Innovation in Structural Engineering, University of
Coimbra, Portugal and Head of Research and Development at the Iranian
Society of Structural Engineering.

xi
List of notations

LOWER CASE
b width
d depth of steel cross-section
e eccentricity/moisture content
f a constant
f y yield strength of steel
h convective heat transfer coefficient
k thermal conductivity
t time
x,y,z coordinates

UPPER CASE
A area
Cp specific heat
E Young’s modulus
L length/span/height
N compression resistance
Pn design resistance under compression
∙
Q heat
R thermal resistance
T temperature
Tf fire/furnace temperature
Ts surface/steel temperature
W width of panel

xiii
xiv  List of notations

SUBSCRIPT
0 ambient condition
b buckling
cap related to capacitance
cond conduction
conv convection
cr critical
d distortional
e effective/Euler buckling
Ed design effect
f flange/fire
l local
rad radiation
Rd design resistance
ref reference
s surface
y yield

GREEK LETTER
α coefficient of thermal expansion
ε emissivity
ρ density
σ Stefan-Boltzmann constant
λ slenderness
Δ difference
Σ sum
Introduction
Fire Safety
Requirements and
1
Implications for
Thin-Walled Steel
Construction
This chapter presents the general requirements of fire safety and, in particular,
fire resistance, the implications for thin-walled steel construction, methods of
evaluating fire resistance of thin-walled steel structures and the contents of
this book.

1.1  THIN-WALLED STEEL STRUCTURES

Thin-walled steel structures (alternatively referred to as light gauge steel struc-
tures, light steel framing, lightweight framing) are increasingly used in building
construction worldwide, not only as secondary structural members such as sheet-
ing and purlins, but also as primary load-bearing members forming walls and
floors. For example, Figure 1.1 shows an example of thin-walled steel studs as
part of the wall construction of a residential building. Thin-walled steel structural
components are normally cold-formed using 0.55–6.35 mm thick low- and high-
strength steels (nominal yield strength in the range of 250–550 MPa at ambient
temperature). C-section profiles are commonly used, however, with significantly

1
2  Fire Performance of Thin-Walled Steel Structures

FIGURE 1.1  Thin-walled steel studs in a residential building. (Courtesy of Metek,

Elmshorn, Germany.)

advanced manufacturing technologies; other innovative and structurally efficient

steel sections are being produced and used to suit specific applications.
Owing to their lightweight, a variety of construction methods can be used
to build thin-walled steel walls, floors and roofs. They include:

1. Stick building: Whereby individual studs are framed on site, one at

a time. Interior insulations and panels are then installed to form the
complete walls and floors. This method gives flexibility on site, but
is time consuming.
2. Panel construction: Thin-walled steel members are pre-assembled
in a factory, together with interior insulations and panels, and then
transported and erected on site, for example, wall and floor pan-
els and trusses. This method reduces the on-site construction time,
improves material efficiency and construction quality.
3. Volumetric construction: Whole rooms are constructed in a factory,
together with fittings and electrical/mechanical services, and then
transported to and connected on site. This method of construction
further improves on-site construction speed and material efficiency.
Although volumetric construction has been promoted for some
time, often in the context of modern methods of construction to help
 1 • Introduction  3

FIGURE  1.2  Examples of thin-walled steel structure construction. (From Steel

Construction Institute (SCI), Fire resistance of light steel framing, SCI Publication
P424, Steel Construction Institute, Ascot, UK, 2019. With permission.)

solve the problem of lack of qualified skills, its market share is still
relatively small. User confidence in the quality and flexibility of this
type of construction appears to be an issue.

Whatever the method of construction, the final product is a panel with enclosed
thin-walled steel members, with or without interior insulation. Figure 1.2 shows
examples of thin-walled steel structural floors and wall panels. It  is the fire
performance of the complete panel assembly that will be the topic of this book.

1.2  FIRE SAFETY REQUIREMENTS

AND THEIR IMPLICATIONS FOR
THIN-WALLED STEEL STRUCTURES
Although different countries have different specific requirements for fire safety of
buildings, the requirements can generally be grouped into the following five sets,
as in the UK’s Approved Document B to the Building Regulations (BR2010).
4  Fire Performance of Thin-Walled Steel Structures

However, as far as thin-walled steel structures are concerned, only a small num-
ber of these requirements affect their design and construction, as explained next.

• B1 Means of warning and escape: The building shall be designed

and constructed so that there are appropriate provisions for the early
warning of fire, and appropriate means of escape in case of fire from
the building to a place of safety outside the building capable of being
safely and effectively used at all times. Means of escape consider-
ations include local means of escape from the initial fire enclosure
when the fire is at the pre-flashover stage and global means of escape
after the fire in the initial fire enclosure has developed into the post-
flashover stage. For  thin-walled steel structures, as for any other
types of structures, the post-flashover fire stage governs their fire
safety design and construction, which is the subject of fire resistance
(B3, see later). As for detailed assessments for means of warning
and escape, this requirement does not affect and is not influenced
by the use of thin-walled steel structures.
• B2 Internal fire spread (Linings): To inhibit the spread of fire within
the building, the internal linings shall:
• Adequately resist the spread of flame over their surfaces
• Have, if ignited, a rate of heat release or a rate of fire growth,
which is reasonable in the circumstances.
Here, ‘internal linings’ mean the materials or products used in lining
any partition, wall, ceiling or other internal structure. The lining mate-
rials on thin-walled steel structures are generally non-combustible,
such as gypsum plasterboards (see Figure 1.2).
• B3 Internal fire spread (Structure): The building shall be designed
and constructed so that, in the event of fire, its stability will
be  maintained for a reasonable period. This  is to ensure that
the fire does not spread out of the initial compartment where the
fire breaks out.
This  requirement is concerned with fire resistance, which is the
ability of the construction to prevent fire spread by achieving ade-
quate insulation, integrity and load-bearing performance. Insulation
performance is achieved if temperature on the unexposed surface of
the construction is low, defined as no more than 140°C on average,
and no more than 180°C at the maximum above the ambient tem-
perature. Adequate integrity performance ensures that the fire does
not burn through the construction. Load-bearing capacity is necessary
to ensure that the structure of the construction does not fail/collapse,
thereby leading to fire spread from one compartment to another.
 1 • Introduction  5

Thin-walled steel structures usually form the walls and floors

of fire-resistant construction. Therefore, they need to achieve all
of  the above three fire resistance requirements. At  present, integ-
rity of construction in fire still cannot be quantified by calculations;
therefore, detailing instructions according to the instructions of spe-
cialist suppliers of the construction, fire testing or expert opinions
should be adhered to. This  includes penetrations for services and
cavity spaces in the construction.
It  is possible to demonstrate fulfilment of the insulation and
load-bearing requirements by calculations, and these are the main
focuses of this book.
• B4 External fire spread: This  performance requirement mainly
deals with lining materials on the external surface of the construc-
tion, similar to requirement B2 for the internal lining materials.
Gypsum boards are non-combustible materials. If other types of lin-
ing materials are used on top of the plasterboards, they should meet
the requirements of the regulations. However, these requirements are
the same for thin-walled steel structures as for other construction
types. It does not impose any additional requirement on thin-walled
steel structures.
• B5 Access and facilities for the fire service: This  requirement is
about providing reasonable facilities to assist firefighters and mak-
ing reasonable provisions within the site of the building to enable
firefighting appliances to gain access to the building. However, as
thin-walled steel structures are usually enclosed by gypsum plas-
terboards to form walls and floors, one concern is that firefighting
water may soak up the gypsum plasterboards. Water itself is ben-
eficial to fire resistance as evaporation of water takes a lot of heat
away from the construction, thus slowing down temperature rise.
The concern is the amount of firefighting water increasing weight
of the plasterboards and reducing mechanical properties of the con-
nection, thereby causing premature collapse of the floor. However,
because the amount of firefighting water is low, in comparison with
weight of the plasterboards, and there is no historical evidence of
firefighting water causing plasterboard failure in fire, this issue is
minor.
Therefore, the main implications of fire safety of buildings on thin-walled
steel construction are fire resistance requirements. Of the three specific fire
resistance requirements of insulation, integrity and load-bearing, meeting the
integrity requirement is not by calculations. Since this book is to present cal-
culation methods, it will only cover insulation and load-bearing.
6  Fire Performance of Thin-Walled Steel Structures

1.3  DETERMINATION OF FIRE RESISTANCE

OF THIN-WALLED STEEL STRUCTURES
The  fire resistance of thin-walled steel construction can be demonstrated in
two generic ways, by standard fire resistance testing or by calculation. Testing
is necessary because thin-walled construction involves many systems and
details. However, standard fire testing is expensive and the results can only be
applied to the particular construction system tested.
The scope of calculation methods can range greatly. Options include:

1. Extension based on fire test results.

2. Limiting temperature method.
3. Extension of ambient temperature design methods.
4. Simplified heat transfer and load-carrying capacity models.
5. Advanced method by numerical modelling.

In standard fire resistance tests, only a very limited number of configurations

can be tested due to high cost of testing, and the test specimens are limited
to specific dimensions and boundary conditions due to the constraints of fire
testing furnaces. Therefore, manufacturers of thin-walled steel construction
systems always attempt to extend the range of their fire test results to different
situations via expert opinion and engineering judgement. This  may be pos-
sible under very limited circumstances (e.g. from a thinner section to a thicker
section), and the extensions are within a small deviation from the tested con-
figuration, for example, to extend the height of application by no more than
10%. Because this method is not  based on a thorough investigation of heat
transfer, load-bearing capacity and fire performance of the construction, it is
by its nature very approximate and tends to be conservative in general, but can
also be unsafe.
Both the limiting temperature method and the extension to ambient tem-
perature design methods are covered in Eurocode EN 1993-1-2 (CEN 2005)
and AS/NZS 4600 (SA 2018). The limiting temperature method simply states
that the temperature in a class 4 (thin-walled) section should not exceed 350°C.
This is a very conservative value in the majority of cases. Extending the ambi-
ent temperature design method consists of using ambient temperature effec-
tive widths for thin-walled steel structures (EN 1993-1-3 (CEN 2006a)/EN
1993-1-5 (CEN 2006b)) and replacing the ambient temperature properties of
steel by those at elevated temperatures. The implicit assumption of this exten-
sion method is that the temperature distribution in the thin-walled (class 4)
 1 • Introduction  7

steel members is uniform. This assumption is not suitable in the majority of

cases in which thin-walled steel members are protected by fire-rated boards
and have steep non-uniform temperature distributions due to fire  exposure
from one side only. With non-uniform temperature distribution, it would
not be sensible to assume that the temperature distribution is uniform, even
at the maximum temperature. This is because this assumption could lead to
very conservative answers in some cases and very unsafe solutions in other
cases when the effects of thermal bowing play a significant role. Extending
the ambient temperature method also assumes similar mechanical property
characteristics at ambient and elevated temperatures, which is not  true for
cold-formed steels.
On the other hand, sophisticated finite element methods can deal with any
type of thin-walled structures. However, using them requires deep and special-
ist knowledge as well as skills. They are best used as research tools rather than
for everyday design.
This book will thus focus on simplified, but flexible, calculation methods.
The simplified methods involve the following general steps:

1. Quantifying the fire exposure condition: For heat transfer and sub-

sequent structural/ mechanical calculations, the fire exposure con-
dition is presented in the form of fire temperature–time relationship.
For  simplicity and to cover the majority of common applications,
only nominal fire temperature–time relationships will be used in
this book.
2. Calculating the temperature field in the structural member under
the above fire exposure: Thin-walled steel structures usually form
part of panel construction (walls/floors) and are exposed to fire from
one side. This  makes the simple analytical solutions in Eurocode
EN 1993-1-2 (CEN 2005) not  relevant, because they are for uni-
formly heated members. This  book will present an alternative
simplified method.
3. Calculation of the remaining load-carrying capacity of the struc-
tural member at elevated temperatures and comparison with the
applied load: The  direct strength method, recently developed by
Shahbazian and Wang (2014a), has been demonstrated to be suit-
able and relatively simple to implement. This  will form the basis
of the calculation method of this book. AS/NZS 4600 (SA  2018)
has included simple methods in its appendix for thin-walled steel-
framed walls and floors based on the research at Queensland
University of Technology, which is an extension of the earlier work
by Feng et al. (2003d).
8  Fire Performance of Thin-Walled Steel Structures

1.4  SCOPE AND LAYOUT OF THIS BOOK

As has been mentioned previously, thin-walled steel construction involves
many details to ensure fire integrity and to satisfy various other performance
requirements such as sound and thermal insulation. An example is shown in
Figure 1.3. It is assumed in this book that the integrity of panels is not com-
promised due to any issues with detailing, and hence the boards and interior
insulation (if any) stay with the steel members throughout fire exposure.
In the majority of cases, thin-walled steel sections form part of walls and
floors, and fire exposure is from one side of the construction. In this book, it
is assumed that fire exposure is on the bottom surface of the floor and on the
interior surface of the wall. The situation of fire exposure on the top surface
of floors is unlikely to be as critical as on the bottom surface: the upward ther-
mal bowing of the floor under fire exposure on the top is counteracted by the
effects of gravity loading causing the floor to move downward. Similarly, if a

FIGURE 1.3  An example of detailing in thin-walled steel construction. (From Steel

Construction Institute (SCI), Fire resistance of light steel framing, SCI Publication
P424, Steel Construction Institute, Ascot, UK, 2019. With permission.)
 1 • Introduction  9

wall is exposed to an external fire, as in the case of external fire spread, the
external fire temperature will be much lower than the internal fire temperature.
Therefore, the external fire exposure will in general be less detrimental to the
structure than interior fire exposure.
Therefore, the general scope and assumptions of this book are:

• Simplified calculation methods to calculate temperatures and load-

bearing capacities of thin-walled steel members.
• Fire from underneath floors.
• Fire inside fire compartment.
• No integrity failure.

When performing calculations to evaluate the fire resistance of structures, it

is important to have accurate input data of temperature-dependent material
properties. This is even more so for thin-walled steel structures because the
elevated temperature properties of thin-walled steel (cold-formed) can vary
within a large range, and the thermal properties of the non-load-bearing mate-
rials in thin-walled construction (gypsum plasterboard, interior insulations)
are critically dependent on temperature.
Therefore, based on the aforementioned scope and important issues, this
book will cover the following topics:
Chapter 2 will present applications of thin-walled steel structures. In par-
ticular, it will introduce different types of thin-walled steel structures, focusing
on recent innovations.
Chapter 3 will present fire resistance tests of thin-walled steel construction.
It will provide guidance on the performance of different types of construction
and how detailing affects fire resistance/performance.
Chapter  4 will present various aspects of simplified modelling of fire
resistance for thin-walled steel construction, including an introduction to fire
behaviour and heat transfer modelling, guidance on the main features of heat
transfer in thin-walled steel construction and how to simplify thin-walled steel
construction for heat transfer modelling, and important features of modelling
thin-walled steel structures at elevated temperatures.
Chapter  5 will assess elevated temperature mechanical properties of
thin-walled (cold-formed) steels and present temperature-dependent ther-
mal properties of non-load-bearing materials relevant to thin-walled steel
construction.
Chapter  6 will present an assessment of different simplified calculation
methods and present robust and accurate design calculation methods for thin-
walled steel beams and columns at elevated temperatures. Suitable examples
will be provided to demonstrate how the various simplified calculation meth-
ods can be implemented.
Applications
of Thin-Walled
Steel Structures
2
Applications of thin-walled steel structures are expanding rapidly owing to
their many advantages as will be described in the next section. The increased
applications are assisted by advances made in the field including the flexibility
with which thin-walled steel structures can be made that lends to innovation
of thin-walled steel sections and systems. For thin-walled steel structures used
in the building sector, fire safety is a critical consideration. Before presenting
detailed methodologies for quantifying the fire resistance of thin-walled steel
structures, this chapter discusses applications of thin-walled steel structures in
the building sector, recent advances and innovations, and introduces the key
parameters that provide and enhance the fire resistance of thin-walled steel
construction.

2.1 INTRODUCTION
Thin-walled steel structural components are made using a cold-forming pro-
cess (roll-forming or press-braking) at ambient temperature. Unlike hot-rolled
and welded heavy steel sections, which are only available in limited shapes
and sizes, thin cold-formed steel sections can be custom designed and made
to enhance structural and cost efficiencies and to suit specific applications.
Unlipped and lipped channel sections (Figure  2.1) are still commonly used
in the construction of wall, floor and roof systems. Z-sections are also used
in applications such as purlins and joists. However, complex sections shown in
Figure 2.2 can be easily roll-formed using low- and high-strength cold-rolled
steel coils with thicknesses in the range of 0.55–6.35  mm since automated
advanced manufacturing technologies are available now.

11
12  Fire Performance of Thin-Walled Steel Structures

Unlipped and lipped sections

FIGURE 2.1  Simple channel sections.

Supacee Supazed

Sigma section Sections with multiple web stiffeners

FIGURE 2.2  Complex thin-walled steel sections.

Thin-walled steel sections are not limited to non-load-bearing units such

as partitions in buildings and are increasingly used in many primary load-
bearing applications. They  possess many beneficial characteristics such as
high strength-to-weight ratio (lightweight), easier and faster fabrication and
construction, enhanced durability through galvanizing (zinc or zincalume coat-
ing), termite-resistance, non-combustibility and sustainable features such as
recyclability, reduced energy consumption and emissions, which have led to
increased development and use of thin-walled steel construction systems. Most
importantly, American, European and Australia/New Zealand cold-formed
steel design standards, AISI S100 (AISI 2016), EC3 Part 1.3 (CEN 2006a) and
AS/NZS 4600 (SA  2018), have been significantly enhanced in recent years
through the adoption of many useful outcomes from research and development
projects, ­allowing engineers to undertake economical and safe building designs.
All of these have led to the expansion of thin-walled cold-formed steel construc-
tion from single ­storey residential buildings (Figure 2.3) to mid-rise residential
and non-­residential buildings of up to 12 storeys (Figure 2.4) in many countries
around the world.
 2  •  Applications of Thin-Walled Steel Structures  13

FIGURE 2.3 A residential building.

FIGURE  2.4  A mid-rise building. (Courtesy of Super Stud Building Products,

Edison, NJ.)
14  Fire Performance of Thin-Walled Steel Structures

2.2  RECENT INNOVATIONS AND

ADVANCES IN COLD-FORMED
STEEL INDUSTRY
Thin-walled steel sections and structural systems can be easily manufactured.
Because of which, there have been many research and development studies to
develop optimised thin-walled steel structural products and systems, which
helped expand the applications of thin-walled steel structures.

2.2.1 Innovative Lightweight and

Structurally Efficient Sections
Lipped channel sections (Figure 2.1) are commonly used as studs, joists and chord
and web members in wall and floor panels and roof trusses and their dimensions,
especially web height and thickness, are varied accordingly to suit the application.
For example, 90 × 35 × 10 × 0.75 mm lipped channels are commonly used as
wall studs while floor joists are commonly 180 × 40 × 15 × 1.15 mm lipped
channels. Such thin-walled steel sections under compression or bending actions
are subject to buckling modes such as local buckling, distortional buckling
and flexural buckling (compression only) and flexural torsional buckling and
their interactions. However, innovative sections can be developed to eliminate
or delay one or more of the above-mentioned buckling modes to enhance their
load-bearing capacities. For example, to eliminate local buckling effects, longi-
tudinal web and flange stiffeners are used while complex return lips are used to
enhance distortional buckling capacity.
Figure 2.2 shows two such sections (Supacee and Supazed) developed by
the University of Sydney researchers in collaboration with an industry partner
(Pham and Hancock 2013). Their research demonstrated the enhancements in
bending capacity of these sections (up to 22% compared to conventional chan-
nel sections) and led to thinner and lightweight sections enabled also by the use
of high-strength steels (G450-nominal yield strength of 450 MPa). These sec-
tions are currently being manufactured and distributed by Bluescope Lysaght
in the Asia-Pacific region. Similar complex sections shown in Figure 2.5 have
been researched extensively and used in different countries. In recent times,
optimisation techniques such as simulated annealing algorithm with fabrica-
tion and geometric end-use constraints are being used to develop innovative
and structurally efficient thin-walled steel sections for use in various applica-
tions (Leng et al. 2014).
 2  •  Applications of Thin-Walled Steel Structures  15

FIGURE 2.5  Other complex sections.

In general, thin-walled cold-formed steel sections are open, monosym-

metric or asymmetric sections, and are therefore subjected to more com-
plicated buckling modes. To overcome the shortcomings associated with
such sections, a series of hollow flange sections (HFS) were produced from
a single strip of steel by an Australian steel product manufacturer using a
patented manufacturing process based on simultaneous cold-forming and
dual electric resistance welding. The  first such section known as hollow
flange beam (HFB) with torsionally rigid, triangular hollow flanges was a
doubly symmetric ­section with no free edges and was used as flexural mem-
bers (Figure  2.6). It  eliminated local and distortional buckling effects and
was shown to be 40% lighter than conventional cold-formed steel sections
(Dempsey 1990). This  was followed by a mono-symmetric hollow flange
channel (HFC) section with rectangular flanges (Figure  2.6), which was
produced to eliminate the connection problems associated with triangular
flanges. Both sections were successfully used in Australia and New Zealand
for many years in a range of applications. However, due to commercial rea-
sons, their manufacturing operations in Australia were discontinued by 2014.
Structural behaviour and moment and shear capacities of both these sec-
tions were investigated extensively using experimental and numerical stud-
ies by Queensland University of Technology researchers (Avery et al. 2000,
Keerthan and Mahendran 2011, Anapayan and Mahendran 2012), which led
to useful design capacity equations and tables that were used by their manu-
facturers and engineers.
Despite discontinuation of such structurally efficient HFS, there were
industry expectations to replace them with alternative sections. Hence
a new rivet-fastened HFC section (Figure  2.6) was developed, and their
bending, shear and web crippling capacities were extensively researched
by Queensland University of Technology researchers (Siahaan et al. 2016).
This thin-walled steel section can be manufactured at a reduced cost in com-
parison with welded HFC and thus offers a cost-effective and safe solution to
the building sector. Research on the use of HFS as wall studs and floor joists
has demonstrated their efficiencies and enhancements over conventional sec-
tions (Jatheeshan and Mahendran 2016a, Kesawan and Mahendran 2017a).
16  Fire Performance of Thin-Walled Steel Structures

FIGURE 2.6  Hollow flange sections.

There is great potential to use similar HFS sections and optimised C- and
Z-sections in a range of applications including wall, floor and truss systems.
Recently, an HFS similar  to HFC (Figure  2.6) was developed and used in
modular units in Korea due to their excellent performance and cost-effi-
ciency (Ha et al. 2016). Using simultaneous cold-formed and electric weld-
ing process, the so-called MCO s­ ections with 100–120 mm flange widths are
being produced with depths in the range of 200–390 mm and thicknesses in
the range of 4.5–10 mm. A smaller clinched HFS is also used as truss mem-
bers in Australia.

2.2.2 Prefabricated Structural
and Modular Units
The availability of new automated roll-forming facilities means that complex
and optimised thin-walled steel sections can be easily cold-formed to produce
them with tight tolerances and custom cut lengths. The new facilities can inte-
grate CAD design to produce pre-cut, punched and sized steel members, which
can then be readily used to make the basic structural wall and floor panels and
roof trusses inside a factory in a controlled environment. Figure 2.7 shows typi-
cal prefabricated wall and floor panels with suitable studs and joists at 600 mm
centres (also 300 and 400 mm). They can all be then transported to the site
(Figure 2.7) and assembled using, in most cases, the efficient self-drilling screw
fasteners in constructing low- and mid-rise buildings (Figures  2.3 and  2.4).
This off-site manufacturing of light steel framing (wall and floor panels and
roof trusses) will considerably reduce labour cost and waste in the factory and
 2  •  Applications of Thin-Walled Steel Structures  17

FIGURE 2.7  Wall and floor panels and roof trusses.

on site, and accelerate construction process (30%–35% faster). The basic wall

and floor panels can also be lined with the required linings and insulations
in the factory and then transported to site. They  can also be used to build
­modular units in the factory, and then transported and used to construct modu-
lar buildings (Figure 2.8), an increasing trend observed recently. Prefabricated
walls are also used as non-load-bearing infill walls in multi-storey concrete- or
steel-framed buildings. In  North America, Europe and Australia, there are
well-established manufacturers of thin-walled steel sections who are suppli-
ers of both prefabricated (planar) light steel framing systems and modular
(volumetric) units made of such sections for specific applications in the con-
struction sector.
Light steel buildings built using prefabricated steel framing or modular
units offer many advantages. The use of lightweight systems leads to signifi-
cantly reduced foundation loads. When appropriately designed, they possess
good performance characteristics in relation to fire resistance, thermal insula-
tion and acoustic insulation. The  fabrication and construction methods used
offer significant reduction in production and site waste, construction safety
benefits and improved quality of final products (Lawson and Way 2016). Load-
bearing and non-load-bearing light steel framing can be used in low- and
mid-rise buildings. Lateral stability is provided by walls that have in-built
K-strap or flat sheet cross-bracing or board lining. For mid-rise buildings, larger
stud sections or back-to-back or nested stud sections can be used in the lower
18  Fire Performance of Thin-Walled Steel Structures

FIGURE  2.8  A modular building. (From Steel Construction Institute (SCI), Fire
resistance of light steel framing, SCI Publication P424, Steel Construction Institute,
Ascot, UK, 2019. With permission.)

storeys, and then both stud size and thickness (0.75–3 mm) can be reduced for
the higher storeys. Light steel buildings can be easily modified and extended if
required, disassembled and reused at another site.

2.3  FIRE RESISTANCE OF THIN-WALLED

STEEL STRUCTURES AND
METHODS OF ENHANCEMENT
Except for some components such as wooden floor boards and EPS insulation
panels (if used as external cladding), all other components (cold-formed steel,
boards and insulations) of light steel wall and floor panels are non-combustible.
Hence, they do not contribute to the fire load of the building. However, the sec-
tion factor (exposed surface area to volume ratio) of thin-walled steel wall
studs and floor joists is high, and thus they heat up quickly in a fire unless
 2  •  Applications of Thin-Walled Steel Structures  19

FIGURE 2.9  Protected thin-walled steel wall systems.

protected adequately. Depending on the application, thin-walled steel wall and

floor systems may require fire resistance levels or fire resistance ratings up to
four hours under the three criteria of structural adequacy/stability, integrity and
insulation. For consistency, this book uses the term fire resistance level (FRL)
to refer to the period of standard fire exposure under which a fire-resistant con-
struction should maintain its fire-resistant function. Different names are used
in different standards, and other names include fire resistance period and fire
resistance rating. Non-load-bearing walls require only FRLs under the integ-
rity and insulation criteria. In order to provide the required FRLs, thin-walled
steel members are always protected by fire-rated boards such as gypsum-
based boards and calcium silicate boards. Some examples of board protection
are shown in Figure 2.9. Fire resistance depends on several key parameters,
and this section explains their effects briefly and discusses methods to pro-
vide enhanced FRLs. Chapter  3 will provide more details of fire resistance/
performance of thin-walled steel structures.

2.3.1  Fire Protective Boards

Fire resistance of thin-walled steel wall and floor systems depends significantly
on fire protective boards used as linings and their arrangement. These boards
protect thin-walled members from direct fire exposure and delay the tempera-
ture rise in steel. The time-temperature profile developed in steel studs/joists
and across the depth of the wall/floor panel in fire depends on the thermal prop-
erties, fall-off temperature and joints of the fire protective boards. The time-
temperature profiles in studs/joists directly influence the structural resistance
in fire while the temperature profile on the unexposed side determines the insu-
lation performance. Therefore, the elevated temperature thermal properties
of fire protective boards, especially specific heat, relative density (mass loss)
and thermal conductivity, are critical. However, they vary among them due to
the differences in their chemical compositions and manufacturing processes.
20  Fire Performance of Thin-Walled Steel Structures

Although most of the gypsum plasterboards provide similar FRLs despite

some differences in their chemical composition, it may not be the same when
imported non-compliant boards are used. At present, board manufacturers do
not provide elevated temperature thermal properties, and therefore, they need
to be determined first, particularly if thermal modelling is to be undertaken for
fire design purposes. Some guidance is provided in Chapter 5 on temperature-
dependent thermal properties of materials.
Unlike gypsum plasterboard that provides consistent fire protection to
steel wall and floor systems, the use of other types of boards such as MgO
board and PCM (phase change material) board often reduces the FRLs of
light steel walls even though they provide better impact resistance, thermal
and acoustic insulation. Recent fire tests of MgO board-lined walls (Rusthi
et al. 2017) showed premature integrity failure within 30 min. This is due to
board joint opening and board cracking caused by significant mass loss with
increasing temperature (about 45% compared to 20% in gypsum plasterboard).
Hence it is clear that elevated temperature thermal property tests of boards
and/or fire tests of walls/floors should be undertaken before they are used.
The FRLs of wall and floor systems can be increased either by enhanc-
ing the specific heat and relative density values of gypsum plasterboards or by
reducing their thermal conductivity values. Baux et  al. (2008), Baspinar and
Kahraman (2011) and Keerthan et al. (2013) have shown that elevated tempera-
ture properties of plasterboard can be enhanced by adding chemical fillers and
additives. Further research is needed to develop new boards with desirable ther-
mal properties to provide considerably higher FRLs to wall and floor systems.
Gypsum plasterboards crack and fall-off due to dehydration and calci-
nation reactions, which leads to sudden localised temperature rise in steel
studs/joists and early failures in fire (Gunalan et  al. 2013). Enhanced plas-
terboards with reduced cracking and fall-off at elevated temperatures would
increase the fire resistance of light steel walls.
Board joints are unavoidable in thin-walled steel wall and floor systems
(Figure 2.10). They are staggered so that the same stud/joist does not have a
joint along the length on both flanges and are located in different directions
when multiple boards are used. Although joints are fully sealed, they open up
when exposed directly to fire and allow localised temperature rise, leading to
reduced FRLs as observed in many full-scale fire tests (Gunalan et al. 2013).
This  effect is significant in single board lined walls with joints along the
steel members. Good detailing and workmanship can eliminate the weak-
ness caused by board joints to some extent. Kesawan and Mahendran (2017b)
recommended improved joint methods (Figure 2.11) to eliminate the detri-
mental effects of joints; however, they may be considered labour intensive.
The new SCI guide (SCI 2019) provides many examples of good detailing
practice adopted by UK manufacturers of thin-walled steel structures.
 2  •  Applications of Thin-Walled Steel Structures  21

FIGURE  2.10  Plasterboard joints. (With kind permission from Springer

Science+Business Media: Fire Technol., A  review of parameters influencing the
fire performance of light gauge steel framed walls, 2017, Kesawan, S. and
Mahendran, M.)

FIGURE  2.11  Improved board joints. (With kind permission from Springer
Science+Business Media: Fire Technol., A  review of parameters influencing
the fire performance of light gauge steel-framed walls, 2017, Kesawan, S. and
Mahendran, M.)
22  Fire Performance of Thin-Walled Steel Structures

2.3.2 Insulation
Insulation (glass fibre, cellulosic fibre or rock fibre) is commonly used in
light steel wall and floor systems to enhance their thermal and acoustic
performance at ambient temperature. The use of cavity insulation has been
shown to increase the insulation-based FRL for non-load-bearing walls,
but it was found to be detrimental to the FRL of load-bearing walls (Kodur
and Sultan 2006, Gunalan et al. 2013). With cavity insulation, the consider-
able rise in temperatures on the fireside plasterboards and stud hot flanges,
together with a larger temperature gradient, led to premature stud failures.
To overcome this problem, Kolarkar and Mahendran (2008) proposed exter-
nally insulated light gauge steel-framed (LSF) walls, where an insulation
layer was sandwiched between two plasterboard layers (Figure  2.9). Fire
tests demonstrated 30% improvement in FRL when external insulation was
used. The  use of similar external insulation systems will enhance the fire
resistance of both non-load-bearing and load-bearing walls. If composite
panels consisting of boards and insulation can be prefabricated and installed,
the installation cost will remain the same. Enhanced insulation materials
with improved thermal properties can also be developed by varying their
compositions.

2.3.3  Steel Studs and Joists

Effects of stud profile on the time-temperature development across the wall
thickness were found to be negligible although the stud thickness influence
was visible (Kesawan and Mahendran 2016). The  structural performance
of light steel walls in fire depends on the time-temperature development in
the steel studs, the stud profiles and their sizes and the elevated temperature
mechanical properties of cold-formed steels. Kesawan and Mahendran (2016)
found that the stud section profiles do not influence the thermal performance
if they are of the same thickness and flange width. They showed that there is
no significant difference between the FRL of non-load-bearing walls made
of sections with similar overall sizes and thickness. The  structural perfor-
mance of LSF walls in fire can depend on the thermal performance (tempera-
ture development) within the wall, steel sections (profiles and sizes) used and
elevated temperature mechanical property reduction factors of cold-formed
steels. However, fire tests and numerical studies (Kesawan and Mahendran
2016, 2017a) have shown that wall and floor systems made of studs/joists with
 2  •  Applications of Thin-Walled Steel Structures  23

different profiles have the same FRL for a given load ratio if their depth and
flange widths are the same. Increase in stud thickness is likely to improve the
FRL of LSF walls. Effects of web depth on the FRL of walls are dependent on
the type of failure mode of studs (section yielding or local buckling or major
axis buckling) and thermal bowing deflections. In summary, for a chosen wall/
floor depth, FRL cannot be improved by stud/joist section profile. However,
elevated temperature mechanical property reduction factors vary depending
on the type/grade of steel used. Ariyanayagam and Mahendran (2018) showed
that the use of steels of varying strength grades significantly influenced the
FRL of load-bearing walls. Hence developing and using cold-formed steels
with higher elevated temperature mechanical property reduction factors, espe-
cially in the range of 400°C–700°C, can provide higher FRLs for thin-walled
steel wall and floor systems.

2.3.4  Steel Sheathing

Thin-walled steel stud walls lined with steel sheathing are increasingly used in
seismic regions. The use of steel sheathing together with plasterboard sheeting
provides higher FRLs for non-load-bearing systems because it restricts plas-
terboard fall-off and retains vaporized water. However, its effect on the FRL
of load-bearing walls was found to be minimal for a given load ratio (Dias
and Mahendran 2019a), in which a structurally efficient web-stiffened section
was used by eliminating local and distortional buckling effects. This pushed
up the critical mode of failure to be major axis flexural buckling and the use
of steel sheathing gave an increased wall capacity at ambient temperature and
also under fire conditions without compromising the FRL (similar FRL for a
given load ratio). Hence, choosing structurally efficient stud/joist profiles in
combination with sheathing can provide enhanced FRLs.

2.4 SUMMARY
The last section has briefly explained the effects of key parameters on the fire
resistance of thin-walled steel construction systems and suggested methods to
enhance fire resistance. The next chapter will provide further details of fire
performance characteristics of different types of thin-walled construction sys-
tems based on fire resistance tests.
Fire Resistance
Tests 3
Because of complexity of detailing of thin-walled steel structural systems,
most fire design standards worldwide use the so-called standard fire tests to
determine their fire resistance level (FRL). Standard fire resistance tests are
also used by researchers to determine and investigate the FRL of the con-
structed system as a function of key influential parameters. This chapter first
presents the standard fire resistance test in general, and then describes the
fire resistance characteristics of different types of thin-walled steel wall and
floor systems.

3.1  STANDARD FIRE RESISTANCE TEST

When exposed to fire, thin-walled steel structural elements and assemblies
heat up quickly, due to their high section factor, and lose their strength and
stiffness rapidly. Hence, they are normally protected by fire-rated boards
to delay the temperature rise. For  example, light gauge steel-framed (LSF)
wall systems made of lipped channel section studs as vertical members and
unlipped channel section tracks as top and bottom horizontal members are
lined with single or double layers of fire rated boards such as gypsum plaster-
board, calcium silicate board, fibre cement board or other equivalent boards
of thicknesses in the range of 6–25 mm and with or without cavity or external
insulation using rock fibre, glass fibre or cellulose fibre in different thicknesses
and densities (Figure 3.1a). Depending on whether they are used as external
or internal walls or load-bearing or non-load-bearing (partitions only) walls,
such arrangements and sizes will vary significantly among them. Noggings
and bracings are also provided for lateral resistance and construction pur-
poses. LSF floor systems are similar to LSF wall systems, however, larger
lipped channel or Z-sections are used with fire rated boards attached to the
ceiling side (bottom surface) and plywood or particle board on the unexposed

25
26  Fire Performance of Thin-Walled Steel Structures

FIGURE  3.1  Protected thin-walled steel structural elements and assemblies.

(a) LSF wall system, (b) LSF floor system (Promatect), (c) protected beam/column.
(Courtesy of Promatech, Moorestown, NJ.)

side (top surface) (Figure 3.1b). Similarly, thin-walled steel structural elements

such as beams and columns can also be protected (box-protection) using fire
rated boards as shown in Figure 3.1c.
Fire resistance of thin-walled steel structural elements and wall and floor
systems depends on many parameters such as the type of steel and thickness,
member size, type of board and insulation and their thickness and density, wall
and floor configuration and load ratio. Different countries and regions have their
own standard fire resistance test standards; however, they are all very similar.
In standard fire resistance tests of LSF systems, full size walls/floors (minimum
size of 3 m × 3 m for wall panels and 3 m × 4 m for floor panels) are exposed to
the standard fire time-temperature curve (Figure 3.2) on one side using a gas or
oil furnace. If they are load-bearing, they are subjected to a pre-determined load
based on the chosen load ratio during the entire fire test. The load ratio is defined
as the applied load in a standard fire test divided by the ambient temperature
capacity. Full size panels are used in order to include the effects of construc-
tion method used (board joints, connections, etc), thermal expansion, shrinkage,
cracking, fall-off, ablation and localized deformations and damage.
The origins of the standard fire time-temperature curve date back as far as
1903 based on wood fuel-burning furnaces. It is considered to simulate a fully
developed compartment fire. The international standard ISO 834 test fire curve
(ISO 1999) is defined by Eq. (3.1).

T f  20  345 log10  8 t  1 (3.1)

where:
T f is the average furnace temperature (°C) at time t
t is the time elapsed (min)
 3  •  Fire Resistance Tests  27

FIGURE 3.2  Standard fire time-temperature curve. (From ISO 834-1, Fire resis-
tance tests  – elements of building construction, Part 1: General requirements.
International Organization for Standardization, Geneva, Switzerland, 1999.)

Depending on the intended function of the tested system, the fire resistance
level (FRL) of the tested system is the time when one of the following three
fire resistance criteria is reached:

• Load-bearing – Structural failure due to rapid loss of load-bearing

capacity or deflection/deflection rate exceeding their corresponding
limits.
• Integrity  – Passage of hot gases/flames through the construction
assembly due to crack formation. Integrity failure is determined by
using a cotton pad and gap gauges.
• Insulation – The unexposed side temperature exceeds the ambient
temperature by 140°C on average or by 180°C at any point on the
unexposed side.

Figure 3.3 shows a typical fire test set-up used to assess a cold-formed LSF
wall system where pre-determined loads are applied to the studs via the bot-
tom track. Fire tests of LSF floor systems are conducted in a horizontal fur-
nace with fire exposure from underneath after applying the pre-determined
vertical loads to the floor boards.
Fire testing using the standard fire time-temperature curve gives good
comparative results for building components tested under identical condi-
tions, and also valuable basic test data. They are commonly used by product
28  Fire Performance of Thin-Walled Steel Structures

FIGURE 3.3  Fire test set-up for LSF wall systems.

manufacturers, industry associations and some building codes to develop

generic or proprietary FRLs for a range of thin-walled steel wall and floor
systems. Although there is growing criticism that the standard fire time-
temperature curve does not represent fires in real situations, standard fire tests
are used regularly by researchers worldwide to investigate the effects of key
parameters on the fire resistance of thin-walled steel construction systems, to
compare their FRLs and to develop new construction systems with enhanced
fire resistance. The  next three sections will discuss the fire resistance tests
of different types of thin-walled steel construction systems (walls, floors and
members) and their fire performance characteristics. Although product manu-
facturers have carried out numerous standard fire resistance tests, their results
are not publicly available. Therefore, the fire resistance test results in this book
are sourced from publicly available research literature.

3.2  FIRE RESISTANCE TESTS OF WALLS

3.2.1  General Behaviour of LSF Walls in Fire

Gerlich et al. (1996) conducted three standard fire tests of channel section stud
walls vertically lined with one layer of gypsum plasterboard on both sides
under combined axial loading and bending. Studs were made of 1.15  mm
 3  •  Fire Resistance Tests  29

Grade 300 steel (2.85 m height) and 1.0 mm Grade 450 steel (3.6 m height) and
used at 600 mm spacing with a central row of noggings without any interior
insulation. The tests were conducted on horizontally placed panels by loading
via the tracks. Major axis flexural buckling associated with local buckling of
mid-height cold flanges occurred in two tests while in the third test flexural
torsional buckling occurred due to the loss of lateral restraint to the compres-
sion flange by the thinner (9.5 mm only) ambient side plasterboards. With the
use of 12.5 and 9.5 mm boards, 32 min fire resistance was reached while using
16 mm boards gave 72 min fire resistance.
Kodur and Sultan’s (2006) experimental study included 14 LSF wall pan-
els with 0.84 and 0.912 mm thick studs spaced at 406 mm (610 mm in one test)
and lined with one or two 12.7 mm gypsum boards on both sides (15.9 mm
boards in two tests). Glass fibre, rock fibre and cellulose fibre cavity insula-
tions were used. Local buckling of studs was the dominant structural failure
mode except in two tests where overall buckling was observed. Fire resistance
was reduced when cavity insulation was used, with cellulose fibre performing
better than others.
Alfawakhiri (2001) conducted six standard fire resistance tests of load-
bearing LSF walls made of 0.912 mm thick 90 mm studs (low yield strength of
228 MPa). Major axis flexural buckling of studs occurred towards the furnace
with compressive failure of the cold flange near mid-height. When cavity insu-
lation was used, stud hot flange temperatures were increased while cold flange
temperatures were reduced, and the wall moved away from the furnace and
failed by hot flange compression failure. Using cavity insulation reduced fire
resistance. The use of noggings and ambient side cross-bracing was considered
to have eliminated flexural-torsional and minor axis flexural buckling of the
studs. Local compressive failures of studs occurred at one of the four holes
located along the length. The use of resilient channels slightly reduced the fire
resistance times.
Feng and Wang (2005) conducted six standard fire tests of load-bearing
LSF wall panels of 2.2 × 2.0 m lined with 12.5 mm gypsum board lining and
cavity insulation. Lipped channel studs of 100 × 54 × 15 mm made of 1.2 and
2.0 mm thick S350 steels were used with two holes near the ends. Increasing
the load ratio (0.2, 0.4 and 0.7) decreased the stud failure time significantly.
Except for one test (local buckling at the hole), the studs failed by flexural
torsional buckling. This was considered to be due to the inability of fireside
plasterboard to restrain the studs although the ambient side plasterboard was
able to prevent minor axis flexural buckling. Their study noted that plaster-
board fall-off could have been triggered by stud failure, instead of plasterboard
fall-off causing stud failure.
Zhao et al. (2005) conducted 29 tests of 1.2 × 2.8 m LSF walls with two studs
spaced at 600 mm with gypsum plasterboards on one or both sides, and with
30  Fire Performance of Thin-Walled Steel Structures

or without cavity insulation. Studs of varying sizes were used and loaded con-
centrically or eccentrically for load ratios from 0.2 to 0.6. Test results showed
that stud failure is essentially governed by the maximum temperature in fire.
Steel studs supported by plasterboards on both sides failed by major axis flex-
ural buckling with local buckling of the cold flange. This failure mode was
observed with eccentric loading towards the fireside. In some other tests, the
same failure mode occurred but with movement away from the fire and com-
pressive failure of the hot flange. These observations show the presence of
complex stud failure modes due to combined thermal and structural effects
and are similar to those observed by other researchers as discussed above.
Gunalan et  al. (2013) conducted 11 full-scale fire tests of load-bearing
walls made of 1.15 mm high-strength G500 cold-formed steels. Single or dou-
ble gypsum plasterboard lining of 16 mm thickness was used with or without
rock fibre, glass fibre and cellulosic fibre cavity insulation. Typical plaster-
board and stud time-temperature curves from their standard fire tests of double
plasterboard–lined walls are shown in Figure 3.4. They are similar to those
observed by earlier researchers and show that gypsum plasterboards are able to
keep the cavity temperatures low and delay steel stud temperature rise for a
long time. The figure also shows presence of non-uniform temperature distri-
butions (temperature gradient) across the wall/stud depth. Many researchers
(Kaitila 2002, Feng et al. 2003a, Gunalan et al. 2013) have confirmed that it is
acceptable to assume uniform temperatures in the flanges and lips and a linear
web temperature distribution as shown in Figure 3.5.

FIGURE  3.4  Time-temperature curves from a standard fire test of LSF wall.
(Reprinted from Thin-Walled Struct., 65, Gunalan, S. et al., Experimental study of
cold-formed steel wall systems under fire conditions, 72–92, Copyright 2013, with
permission from Elsevier.)
 3  •  Fire Resistance Tests  31

Hot Flange

Linear Web
variation

Cold Flange

FIGURE 3.5  Simplified stud temperature distribution.

The  development of non-uniform temperature distributions across the

stud depth causes thermal bowing towards the fireside as illustrated in
Figure  3.6 (original centroid X moves to Y at mid-height: e∆T ) while the
centre of resistance of the stud shifts towards the ambient side due to the
loss of stiffness of the hotter side of the steel stud, giving a neutral axis
shift (e∆E ). This results in a net eccentricity of e = e∆T − e∆E . Hence the
wall studs are subjected to a combined loading of axial compression and
bending moment in fire. The net loading eccentricity (e) in Figure 3.6 will
vary depending on the values of e∆T and e∆E, which depend on the tem-
perature distributions of the stud cross-section. The net e value is normally
positive and hence the wall failure occurs by movement towards the fur-
nace. However, in some cases, the wall can fail by a sudden movement away
from the furnace near the end (Figure 3.7) depending on the net e value and
the relative yield strengths of hot and cold flanges near the failure time as
explained in Gunalan et al. (2013).

FIGURE  3.6  Effective centroid of studs in fire. (Reprinted from Thin-Walled

Struct., 65, Gunalan, S. et  al., Experimental study of load bearing cold-formed
steel wall systems under fire conditions, 72–92, Copyright 2013, with permission
from Elsevier.)
32  Fire Performance of Thin-Walled Steel Structures

FIGURE 3.7  LSF wall panels after failure. (Reprinted from Thin-Walled Struct., 65,
Gunalan, S. et al., Experimental study of load bearing cold-formed steel wall sys-
tems under fire conditions, 72–92, Copyright 2013, with permission from Elsevier.)

Gunalan et  al.’s (2013) fire tests demonstrated integrity of the inner
p­ lasterboard until failure (Figure 3.7) and showed that there was sufficient
lateral restraint to the studs from the plasterboards until failure. In fact, in
all the fire tests of load-bearing walls with load ratios of 0.2 and above, steel
stud failure occurred before integrity or insulation failure. This  is despite
plasterboard fall-off that started to occur when the inner plasterboard sur-
face temperature reached about 900°C. Other researchers (Alfawakhiri
2001, Feng et al. 2003b, 2003c, Zhao et al. 2005) also suggested that it would
be appropriate to assume sufficient plasterboard lateral restraints to prevent
minor axis flexural buckling and flexural-torsional buckling failures of the
steel studs.
From the above standard fire tests of load-bearing LSF walls conducted in
many different countries, the following useful general observations are made:

• LSF walls are generally made of thin-walled steel-lipped channel

studs of thicknesses in the range of 0.84–2.0 mm and made of steels
with yield strengths in the range of 228–550 MPa, and are used as
both non-load-bearing and load-bearing walls.
• In fire, LSF walls are subject to a combined loading of axial com-
pression and bending, and their failure modes and directions are
complicated with thermal bowing and neutral axis shift being
important factors.
• The number, thickness and type of plasterboard lining are influen-
tial factors for the FRL of LSF walls. Protecting them with one or
two layers of 9.5, 12.5 and 16 mm fire rated gypsum plasterboard
has been successful in providing FRLs up to 120 min.
 3  •  Fire Resistance Tests  33

• Stud failure is governed mostly by the hot flange temperature

reached in fire. Stud depth and thickness and the type of steel used
can influence the fire resistance of walls.
• LSF wall failure is usually due to local buckling of the stud element
but can also be caused by major axis flexural buckling or a combina-
tion of local and major axis flexural buckling (Figure 3.7). A single
layer of 9.5  mm plasterboard was found to be unable to provide
minor axis and twist restraints to the studs. Double layers of plas-
terboard were able to provide sufficient lateral restraint to the studs
until failure. The use of noggings and bracing will provide adequate
restraint against minor axis and flexural-torsional buckling.
• Plasterboard fall-off can allow heat inside cavity, leading to earlier
failures and reduce FRLs of LSF walls. Its detrimental effects are
more critical for a single layer of plasterboard-lined walls.
• Glass fibre, rock fibre and cellulosic insulation materials are com-
monly used as cavity insulation to provide sound insulation, thermal
comfort and energy savings. However, they reduce the FRL of load-
bearing walls.
• In all the fire tests of load-bearing walls with load ratios of 0.2 and
above, stud load-bearing failure occurred before integrity or insula-
tion failure. Increasing load ratios led to lower FRLs but the time-
temperature curves are not altered by the load ratios.

Insulation and integrity-based fire resistance of non-load-bearing LSF

walls was investigated by Feng et al. (2003c), Kolarkar and Mahendran
(2012) and Ariyanayagam and Mahendran (2018) using both small-scale
(300  ×  300  mm to 1280  ×  1015  mm) and full-scale (3  ×  3  m) standard
fire tests. Several numerical studies based on small-scale tests were also
undertaken by Feng et al. (2003c) and Sultan (1996). Although small-scale
fire tests cannot capture many important effects of full-scale fire tests, such
as thermal bowing, large lateral deflections, cracking, ablation, fall-off of
plasterboards and integrity failures, they are adequate in providing compara-
tive results and the general trend. These studies showed that insulation-based
FRL was improved by more than 15 min for 92 × 1.15 mm stud walls lined
with one layer of 16 mm plasterboard when glass fibre insulation was used,
by more than 30 min when the plasterboard thickness was increased from
13 to 16 mm, and by more than 100 min when two layers of 16 mm plaster-
board were used. Fire tests of 3 m high walls showed that integrity failure
did not occur when gypsum plasterboard linings were used. Although plas-
terboard joints play an important role in affecting the FRLs of load-bearing
walls (see Section 3.2.5), their effects on the FRLs of non-load-bearing walls
appear to be small.
34  Fire Performance of Thin-Walled Steel Structures

The following subsections will discuss the effects of different influential

factors.

3.2.2 Effects of Cavity Insulation

on Fire Resistance
Ariyanayagam and Mahendran (2019) investigated the effect of cavity insu-
lation on non-load-bearing and load-bearing of 92 mm stud walls. Four fire
tests of 3 × 3 m walls and parametric finite element studies were conducted on
LSF walls lined with one 16 mm gypsum board with and without glass fibre
cavity insulation. Cavity insulation prevents heat transfer across the cavity by
radiation and convection, thereby increasing temperatures of the fireside plas-
terboard and hot flanges while decreasing temperatures of the ambient side
plasterboard surface and cold flanges, resulting in larger temperature gradients
across the cavity and the stud than in uninsulated walls (Figure 3.8a). This led
to higher thermal bowing deformations, and together with higher hot flange
temperatures, resulted in considerably lower FRLs (stud failure times) of
­cavity-insulated LSF walls as shown in Figure 3.8b. Alfawakhiri et al. (1999),
Kodur and Sultan (2006) and Gunalan et al. (2013) also confirmed the detri-
mental effects of cavity insulation on the FRL of load-bearing walls.
When cavity insulation becomes ineffective, the differences in temper-
atures between LSF walls with and without cavity insulation decrease, and
therefore the adverse effect of cavity insulation on FRLs of load-bearing LSF
walls diminishes. For example, the results in Figure 3.8a show that after glass

FIGURE 3.8  Fire test results of cavity-insulated LSF walls: (a) time-temperature

curves of plasterboard surfaces and (b) load ratio versus failure time curves.
(Reprinted from Constr. Build. Mater., 203, Ariyanayagam, A. and Mahendran,
M., Influence of cavity insulation on the fire resistance of light gauge steel framed
walls, 687–710, Copyright 2019, with permission from Elsevier.)
 3  •  Fire Resistance Tests  35

fibre becomes ineffective at about 650°C, the temperature differences between

insulated and uninsulated walls started to decrease. Kodur and Sultan’s (2006)
fire tests showed that when cellulosic insulation was used, the load-bearing
wall failed at 71 min, whereas the same wall with rock fibre insulation failed
at 59 min. This is due to the difference in temperatures at which these cavity
insulation materials become ineffective.
Load-bearing steel stud walls fail when the hot flange temperature is in
the range of 400°C–600°C for common load ratios of 0.4–0.7. Hence, if the
temperature at which cavity insulation becomes ineffective is below 300°C,
it will not reduce the FRL of load-bearing LSF walls. Since cavity insulation
is needed for other purposes such as sound insulation, thermal comfort and
energy savings, such an approach should be pursued.
On the other hand, the use of cavity insulation reduced the ambient sur-
face temperatures and thus increased the insulation-based FRL (Figure 3.8a).
This implies that for non-load-bearing walls the use of rock fibre insulation
with a higher effective temperature will enhance their FRLs. However, higher
thermal deformations leading to large mid-height lateral deflections can trig-
ger premature fall-off of already softened gypsum plasterboard on the fireside
and thus reduce insulation-based FRLs.

3.2.3 Effects of External Insulation

on Fire Resistance
To eliminate the large temperature gradient in LSF walls with cavity insulation
which causes reductions to the FRLs of load-bearing LSF walls while still meet-
ing the requirements for acoustic performance, thermal comfort and energy
conservation, Kolarkar and Mahendran (2008) proposed the use of external
insulation sandwiched between two plasterboards as shown in Figure 3.9a.

FIGURE  3.9  Composite panel: (a) with external insulation and (b) stud time-
temperature curves (HF-hot flange, CF-cold flange).
36  Fire Performance of Thin-Walled Steel Structures

The fire performance of the new LSF wall system was investigated using
small-scale and full-scale fire tests and numerical studies by Kolarkar and
Mahendran (2012), Gunalan et al. (2013), Kesawan and Mahendran (2015) and
Chen and Ye (2014). The use of external insulation resulted in reduced temper-
atures of the inner plasterboard layer, and so its calcination and deterioration
were delayed. It  also allowed direct heat transfer from the fireside plaster-
board to the ambient side plasterboard. Hence in contrast to cavity-insulated
walls (Figure 3.8a), the stud hot flange temperatures and associated tempera-
ture gradients were significantly reduced as shown in Figure 3.9b. Fire tests
showed that the FRLs of load-bearing lipped channel stud walls increased by
about 25%, with rock fibre insulation providing the best outcome. Kesawan
and Mahendran (2015) showed that similar improvements can be obtained
for LSF walls made of other stud sections such as hollow flange channels.
In terms of insulation performance, the corresponding FRLs are likely to be
the same for both cavity and externally insulated panels based on both test and
numerical results.
Furthermore, Gunalan et al.’s (2013) test results showed that the vertical
plasterboard joints on the inner boards were protected by the external insulation
layer even after the outer layer fall-off, thus preventing localised temperature
rise in the stud hot flange. Therefore, plasterboard joints in LSF walls with
external insulation are not  as detrimental as those in cavity-­insulated or
uninsulated walls.

3.2.4 Effects of New Stud Sections

on Fire Resistance
Kesawan and Mahendran (2015) investigated the possibility of increasing
the fire resistance of LSF walls by using a new welded hollow flange chan-
nel (HFC) section of dimensions 150 x 45 x 15 x 1.6 mm as studs as shown
in Figure 3.10. When the results of five fire tests of wall panels lined with
two layers of 16 mm plasterboard were compared with those from Gunalan
et al. (2013) who used conventional lipped channels, there were significant
improvements to fire resistance times. However, their subsequent numerical
studies showed that such improvements were not due to the new HFC stud
geometry. Instead, they benefited from the enhanced elevated temperature
mechanical properties of HFC due to the simultaneous cold-forming and
electric resistance welding fabrication process used. These results demon-
strate the importance of elevated temperature mechanical properties and
how they may be improved to enhance fire resistance.
 3  •  Fire Resistance Tests  37

FIGURE  3.10  HFC stud sections. (Reprinted from Thin-Walled Struct., 98(A),
Kesawan, S. and Mahendran, M., Predicting the performance of LSF walls made
of hollow flange sections in fire, 111–126, Copyright 2019, with permission from
Elsevier.)

3.2.5 Effects of Plasterboard Joints

on Fire Resistance
Plasterboard joints are located along the studs (vertically) in single plas-
terboard–lined walls. Although the recessed edges of the plasterboards are
filled with joint fillers and sealed, plasterboards become soft, brittle and
shrink and cause joint opening-up (Figure  3.11) in fire, especially due to
their dehydration reactions (free and chemically bound water evaporates),
and joints also become weaker and crack, and split due to short edge dis-
tances (10 mm). Such opening-up of joints on the fire side exposes the studs
to direct fire attack and leads to rapid rise in the hot flange temperature
of studs (Figure 3.11) and accelerate their structural failure (Ariyanayagam
et  al. 2016). Although the use of horizontal plasterboard joints along nog-
gings could reduce such adverse effects, vertical joints along the studs still
have to be used for longer walls. Other studies by Chen et al. (2012, 2013a,
2013b) have also shown the detrimental effects of plasterboard joints on fire
resistance of LSF walls.
To mitigate against this shortcoming, Ariyanayagam et al. (2016) suggested
a back-blocking joint arrangement in which the joints are located between
the studs with 150  mm wide plasterboards as back-blocks (Figure  3.12).
They showed that such joint arrangements increased the failure time of single
plasterboard-lined load-bearing walls by about 25% as the stud hot flanges
38  Fire Performance of Thin-Walled Steel Structures

FIGURE 3.11  Cracking and opening-up of joints and their effects on stud time-
temperature curves. (Reprinted from Ariyanayagam, A.D. et  al., Thin-Walled
Struct., 107, Detrimental effects of plasterboard joints on the fire resistance of
light gauge steel frame walls, 597–611, Copyright 2016, with permission from
Elsevier.)

FIGURE  3.12  Plasterboard joints with back-blocks. (Reprinted from

Ariyanayagam, A.D. et al., Thin-Walled Struct., 107, Detrimental effects of plas-
terboard joints on the fire resistance of light gauge steel frame walls, 597–611,
Copyright 2016, with permission from Elsevier.)

were not  directly exposed to fire through the opened-up joints. They  also
showed that since the joint effect is localized by only affecting the stud hot
flange and fireside plasterboard temperatures, its effect on the insulation-based
FRL of non-load-bearing walls is negligible. Innovative plasterboard joint
arrangements can also be used along the studs instead of back-blocking to
reduce construction time/cost.
 3  •  Fire Resistance Tests  39

3.2.6 Effects of Other Types of Boards

on Fire Resistance
Other types of fire protective boards are also being used in LSF walls, for
example, calcium silicate boards, because of their improved properties relating
to impact and moisture resistance. Calcium silicate boards are manufactured
using calcium silicate hydroxide (Ca6Si6O17(OH)2) also known as Xonotlite,
mineral binder and additives.
Chen et  al. (2012) investigated the use of two layers of 12  mm thick
­calcium silicate boards, and their fire tests showed explosive spalling at high
temperatures, resulting in lower FRLs (58 min integrity failure) compared
with gypsum plasterboard–lined walls. When a combination of 12 mm cal-
cium silicate and 12.5 mm gypsum boards was used with the calcium sili-
cate board on the inner side, FRL was improved to 92 min as the integrity
failure associated with cracking was eliminated. Wang et  al. (2015) also
conducted full-scale tests of LSF walls lined with 9  mm calcium silicate
board and observed cracking and spalling of the boards. Ariyanayagam and
Mahendran (2017) conducted full-scale fire tests of non-load-bearing LSF
walls with one 20 mm calcium silicate board lining in one test and then with
a combination of 20 mm calcium silicate board and 16 mm gypsum plaster-
board in the second test. No integrity failure associated with large cracks,
spalling and fall-off occurred in their tests with an insulation-based failure
after 130  min (Figure  3.13) compared to 94  min obtained for the 16  mm
plasterboard-lined wall. In the second test, the average ambient surface tem-
perature was 79°C even after 250 min, in comparison with 197 min for LSF
walls lined with two layers of 16 mm plasterboard. The higher thickness of
calcium silicate board helped to reduce the temperatures on the unexposed
side and prevent integrity failure. The enhanced performance also benefited
from glossy and smooth surface of calcium silicate boards (which have lower
emissivity) and the presence of fibrous material throughout the calcium sili-
cate board thickness.
Magnesium oxide board is another board that is being used as fire protection
for LSF walls. Its main constituents are magnesium oxide (MgO) (40%–55%)
and magnesium chloride (MgCl2) (20%–35%) with small percentages of perlite,
woodchip and fibreglass. Thermal property tests of MgO boards by Rusthi et al.
(2017) showed that they have almost 40%–50% mass loss during their dehydra-
tion reactions, compared to 20% for gypsum plasterboard (Figure 3.14).
Rusthi et al. (2017) conducted three full-scale tests of non-load-bearing
LSF walls lined with one layer of two types of 10  mm MgO boards with
differing percentages of MgO and MgCl2. The first test panel had noggings
40  Fire Performance of Thin-Walled Steel Structures

FIGURE  3.13  Calcium silicate board–lined wall. (Reprinted from Thin-Walled

Struct., 115, Ariyanayagam, A.D. and Mahendran, M., Fire tests of non-load bear-
ing light gauge steel frame walls lined with calcium silicate boards and gypsum
plasterboards, 86–99, Copyright 2017, with permission from Elsevier.)

FIGURE 3.14  Mass loss of MgO boards. (Reprinted from Fire Saf. J., 90, Rusthi,
M. et al., Fire tests of magnesium oxide board lined light gauge steel frame wall
systems, 15–27, Copyright 2017, with permission from Elsevier.)
 3  •  Fire Resistance Tests  41

at 600  mm spacing while the third one had glass fibre cavity insulation.
The cavity-insulated wall panel exhibited higher hot flange temperatures and
associated high temperature gradients and larger lateral deflections of the wall
as observed for gypsum plasterboard–lined walls. All three wall panels suf-
fered severe board cracking on the ambient side (Figure 3.15) and failed by
integrity criterion, giving only a 30 min FRL. The high mass loss caused board
cracking in all three tests.
The use of noggings led to excessive board cracking as it restricted ther-
mal expansion and bowing of the boards. When noggings were removed, joint
opening and cracking due to excessive shrinkage and bowing of the boards
caused panel failure (Figure  3.15). Fire rated sealants between boards and
studs and in the joints could not withstand the excessive movement caused by
board bowing and shrinkage. Strengthening the board joints with high tem-
perature mortar also did not prevent joint cracking. Fibre mesh appeared to
have been removed along the recessed board edges, which also affected the
board strength in fire. This study has shown that mass loss characteristics of
boards, joint detailing and stud spacing are important factors affecting the
FRLs of LSF walls.
Hanna et al.’s (2015) fire tests of load-bearing walls also showed similar
observations. In their tests, ambient side cracks were seen at 40 and 50 min,
and they were wide open allowing the passage of hot gases. To eliminate the
problems with MgO board, Chen et al. (2013a) suggested the use of MgO board
as the inner layer and gypsum plasterboard as the outer layer. These studies
seem to indicate that MgO boards are not an effective solution in improving
the fire resistance of LSF walls.

FIGURE  3.15  Fire tests of MgO board-lined LSF walls. (Reprinted from Fire
Saf. J., 90, Rusthi, M. et  al., Fire tests of magnesium oxide board lined light
gauge steel frame wall systems, 15–27, Copyright 2017, with permission from
Elsevier.)
42  Fire Performance of Thin-Walled Steel Structures

New types of fire protective boards are being regularly introduced into
the building sector. It is essential that full-scale fire tests are conducted to
demonstrate that they can provide thin-walled steel structures with ade-
quate FRLs. Furthermore, to facilitate advanced analyses, it is important
that manufacturers and suppliers provide temperature-dependent thermal
properties of their products such as specific heat, thermal conductivity and
mass loss.

3.2.7 Effects of Steel Sheathing

on Fire Resistance
Dias et al. (2019a) investigated the effects of using 0.55 mm thick G300 steel
sheathing together with two layers of 16 mm gypsum boards on the FRLs of
LSF walls made of 1.15 mm G500 steel web-stiffened studs. Three standard
fire tests (Table 3.1) were conducted under a load ratio of 0.4, with no steel
sheathing (Test No. 1), internal steel sheathing (Test No. 2) and external steel
sheathing (Test No. 3).
The web-stiffened stud was designed to eliminate local, distortional and
flexural-torsional buckling when used as 3  m wall studs with rigid sheath-
ing on both sides (Dias et  al. 2018). Hence the dominant failure mode was
major axis flexural buckling of the studs. Small-scale non-load-bearing fire
tests showed that when steel sheathing was used either internally or externally,
the ambient side temperature was reduced and the insulation-based FRL was
enhanced by about 16%. This was found to be due to retention, cyclic evapora-
tion and recondensation of plasterboard moisture as a result of confinement
provided by the steel sheathing (Dias et al. 2019b). In contrast, the addition
of steel sheathing only provided marginal improvement to the FRL of load-
bearing walls (Table 3.1) because both plasterboard and steel sheathing joints
opened up and caused localised stud temperature rise, leading to structural
failure, as seen in Figure 3.16.
Table 3.1 indicates that internally sheathed walls performed slightly bet-
ter than externally sheathed walls (118 versus 111 min) because the internal
sheathing joints are concealed and confined between the studs and plaster-
boards. However, slender steel–sheathing buckles prematurely between the
screws when exposed to heat, exposing the plasterboard joints to high tem-
peratures. In general, at the same load ratio (ratio of fire test load to ambient
temperature capacity), significant FRL improvements are unlikely with inter-
nal or external steel sheathing.
TABLE 3.1  LSF wall configurations with steel sheathing and test results

SHEATHING ARRANGEMENT AMBIENT

WALL TEMPERATURE FIRE TEST
TEST NO. CONFIGURATION PLASTERBOARD STEEL SHEET CAPACITY (KN) LOAD (KN) FRL (MIN) HFMAX (°C)

1 Double – 118 47.2 110 417

2 Double Internal 130 52.0 118 452

3 Double External 130 52.0 111 414

Source: Reprinted from Thin-Walled Struct., 137, Dias, Y. et al., Full-scale fire tests of steel- and plasterboard-sheathed web-stiffened stud
walls, 81–93, Copyright 2019, with permission from Elsevier.
3  •  Fire Resistance Tests  43
44  Fire Performance of Thin-Walled Steel Structures

FIGURE 3.16  Failures of steel-sheathed LSF walls. (a) LSF wall with internal steel
sheathing, (b) LSF wall with external steel sheathing. (Reprinted from Thin-Walled
Struct., 137, Dias, Y. et al., Full-scale fire tests of steel and plasterboard sheathed
web-stiffened stud walls, 81–93, Copyright 2019, with permission from Elsevier.)

3.3  FIRE RESISTANCE TESTS OF FLOORS

3.3.1  General Behaviour of LSF Floors in Fire

Sultan et  al. (1998), Alfawakhiri and Sultan (2001) and Alfawakhiri (2001)
conducted five standard fire tests of 4.85 × 3.95 m floors lined with 15.9 mm
thick plywood (in one test a layer of 51 mm thick concrete was added to the
plywood) and one or two layers of 12.7 mm thick gypsum plasterboard under
an applied loading in the range of 1.8–2 kN/m2. Four of them had glass or rock
fibre insulation in the cavity. Their tests showed that the number, thickness
and type of fire protective boards and cavity insulation significantly influenced
the FRLs of LSF floors whereas joist spacing (406 vs 610 mm) or type of sub-
floor did not influence the FRL. As observed for LSF walls and for the same
 3  •  Fire Resistance Tests  45

FIGURE  3.17  Failures of LSF floor panels. (From Alfawakhiri, F. and Sultan,
M.A., Loadbearing capacity of cold-formed steel joists subjected to severe
heating, Proceedings of the 9th International Conference Proceedings, Interflam
2001, Edinburgh, UK, Vol. 1, pp. 431–442, 2001. With permission from National
Research Council of Canada.)

reasons, the FRL of load-bearing floors was significantly reduced by the use of
cavity insulation. Integrity and insulation failures were not observed, and the
floor panel collapsed due to section failure of the joists (Figure 3.17).
Sakumoto et  al. (2003) conducted six fire tests of 4.26  ×  2.95  m LSF
floors made of back-to-back lipped channel joists lined with composite boards
including plywood on the top and one or two layers of gypsum plasterboards
of thicknesses 9.5, 12.5 and 15  mm. Their tests also showed that the FRL
increased significantly with increasing thickness and number of boards, but
cavity insulation did not reduce the FRL as expected. Plasterboard fall-off was
identified as the key factor influencing the FRL.
Zhao et al. (2005) conducted full-scale fire tests of two floor systems of
5.5 × 2.99 m made of 250 × 2.5/2.0 mm lipped channel joists at 600 mm spac-
ing and lined with two layers of gypsum board and cavity insulation. With
fire on one side, non-uniform temperature distribution was developed, which
caused thermal bowing deformations in addition to deflections caused by the
applied load. With increasing temperatures on the fireside and plasterboard
fall-off, the joists failed in bending due to reduced mechanical properties in
fire. Localised web crippling failures were also observed at the joist to track
supports.
Baleshan and Mahendran (2016) conducted three fire tests of LSF floors
made of 180 × 1.15 mm lipped channel joists with two layers of 16 mm plaster-
board on the fireside under a load ratio of 0.4. Rock fibre insulation was used in
46  Fire Performance of Thin-Walled Steel Structures

the cavity in the second test while it was used between the two plasterboards as
external insulation in the third test, following the recommendation of Kolarkar
and Mahendran (2008, 2012). In all three tests, the joists moved towards the
fireside and failed locally by web crippling near the supports. The use of cavity
insulation reduced the failure time from 107 to 99 min due to higher hot flange
temperatures and larger lateral deflections caused by greater temperature gra-
dients across the floor depth. On the other hand, the use of external insulation
increased the failure time to 139 min. These observations are similar to those
observed for LSF walls as discussed in Sections 3.2.2 and 3.2.3.
From the aforementioned standard fire tests of LSF floors, the following
useful observations are made:

• LSF floors are generally made of single or back-to-back thin-walled

steel-lipped channel joists of thicknesses in the range of 1.15–2.5 mm
and made of steels with yield strengths up to 550 MPa.
• Protecting them with one or two layers of fire rated gypsum plas-
terboard with thicknesses in the range of 9.5–16 mm has been suc-
cessful in providing the required FRLs. Significant improvement to
FRLs can be achieved by using thicker plasterboards and/or more
than one layer.
• Plasterboard fall-off can significantly reduce the FRLs of LSF
floors. This is more critical for LSF floors than for LSF walls.
• Cavity insulation is commonly used, but it will reduce the FRL of
load-bearing LSF floors.
• Joist failure is governed mostly by the average joist temperature.
It is usually based on a combination of local buckling of the com-
pression cold flanges and yielding of the hot flanges.
• Increasing load ratios leads to lower FRLs as for LSF walls.

3.3.2 Effects of New Joist Sections

on Fire Resistance
Jatheeshan and Mahendran (2016b) investigated the possibility of increasing
the fire resistance of LSF floors by using a new welded hollow flange chan-
nel (HFC) section of dimensions 200  ×  45  ×  15  ×  1.6  mm as joists. When
the results of full-scale fire tests of floor panels lined with two 16 mm plas-
terboards with and without cavity insulation were compared with those from
Baleshan and Mahendran (2016), significant improvements to fire resistance
times were noted. However, as with LSF walls, the improvement was a result
of the enhanced elevated temperature mechanical properties of HFC due to
 3  •  Fire Resistance Tests  47

FIGURE 3.18  Failure modes of LSF floors made of HFC joists.

the simultaneous cold-forming and dual electric resistance welding fabrication

process used and the possible improved connectivity of joist to plasterboards
through both flanges. Cavity insulation adversely affected the failure times as
expected. Plasterboard joint and fall-off (Figure  3.18) affected the tempera-
ture development in the joists and influenced the failure times significantly.
The failure of HFC joist was due to local buckling of the compression cold
flanges and significant yielding of the hot flanges with large lateral deflections
as seen in Figure 3.18.

3.4  OTHER FIRE RESISTANCE TESTS

ON THIN-WALLED STEEL STRUCTURES
Thin-walled steel structural elements and assemblies are normally protected
by fire rated boards and are thus subject to non-uniform temperature distribu-
tions when exposed to fire conditions as discussed in the previous sections.
However, in some applications such as columns surrounded by fire from all
sides, they can experience nearly uniform temperature distributions. Their
behaviour under uniform temperature condition also forms the basis of devel-
oping understanding for the more complex condition of non-uniform temper-
ature distribution. For  these reasons, many experimental studies have been
undertaken on unprotected thin-walled steel columns and beams under uni-
form elevated temperature conditions. This  section presents brief details of
selected experimental studies and their results.
Feng et al. (2003a, 2003b) experimentally and numerically investigated
both local and distortional buckling capacities of short lipped and unlipped
channel columns at uniform elevated temperatures including the effects of
web holes. Gunalan et al. (2015) investigated local buckling behaviour of short
48  Fire Performance of Thin-Walled Steel Structures

thin-walled steel columns under uniform elevated temperatures up to 700°C.

Both unlipped and lipped channel sections made of low- and high-strength
steels were considered. Similarly, Ranawaka and Mahendran (2009a) inves-
tigated the distortional buckling behaviour of two types of lipped channel
columns (with and without additional lips) made of low- and high-strength
steels exposed to uniform elevated temperatures up to 800°C. Figure  3.19
shows the typical test set-up used in their tests. The test results clearly showed
reductions in the local and distortional buckling capacities with increasing
temperatures.
These results also indicated that the current ambient temperature design
equations based on the effective width method or the direct strength method
(DSM) in AISI S100 (AISI 2016), Eurocode 3 Part 1.3 (CEN 2006a) and
AS/NZS 4600 (SA  2018) can be used to conservatively predict the local
buckling capacities by using appropriately reduced mechanical properties

FIGURE  3.19  Uniform elevated temperature tests and failure modes of short
thin-walled steel columns: (a) electrical furnace, (b) distortional buckling, (c) local
buckling of SHS, (d) local web buckling of lipped channel and (e) local flange
buckling of unlipped channel. (Reprinted from J. Constr. Steel Res., 65, Ranawaka,
T.  and Mahendran, M., Distortional buckling tests of cold-formed steel com-
pression members at elevated temperatures, 249–259, Copyright 2009, with
­permission from Elsevier.)
 3  •  Fire Resistance Tests  49

in all the ­calculations. The DSM-based equations for distortional buckling

at ambient temperature were also able to predict the elevated temperature
capacity reasonably well if appropriately reduced mechanical properties
were used.
Bandula Heva and Mahendran (2013) and Gunalan et al. (2014) inves-
tigated the flexural-torsional buckling behaviour of long thin-walled lipped
channel columns under uniform elevated temperatures up to 700°C using
an electric furnace as shown in Figure 3.20. In general, ambient tempera-
ture design rules from AS/NZS 4600 (SA 2018), AISI S100 (AISI 2016)
and Eurocode 3 Part 1.3 (CEN 2006a), with a suitable modification of
the buckling curves, accurately predicted the reduced elevated temper-
ature capacities of pin-ended slender lipped channel columns, provided
appropriately reduced mechanical properties are used. For  fix-ended
columns, AISI S100 (AISI 2016) and AS/NZS 4600 (SA  2018) design
equations were modified by Gunalan et al. (2013) to include the beneficial
effect of warping restraint and were then used for elevated temperature
conditions successfully.

FIGURE  3.20  Uniform elevated temperature tests of long thin-walled steel

columns. (Reprinted from Eng. Struct., 79, Gunalan, S. et al., Flexural-torsional
buckling behaviour and design of cold-formed steel compression members
at elevated temperatures, 149–168, Copyright 2014, with permission from
Elsevier.)
50  Fire Performance of Thin-Walled Steel Structures

3.5  FINAL COMMENTS

This chapter has provided the details of a large number of standard fire tests
on LSF walls and floors and uniform elevated temperature tests on thin-walled
steel sections. The test results were used to explain their behaviour and how
key parameters influence their fire resistance. An important focus of this chap-
ter was to identify methods of enhancing the fire resistance of load-bearing
and non-load-bearing LSF structures. The next chapter will describe how to
numerically evaluate the fire resistance of LSF structures.
Numerical
Modelling of
Fire Resistance
4
of Thin-Walled
Steel Structures
In  order to evaluate the fire resistance of thin-walled structures at elevated
temperatures, it is necessary to understand the fire behaviour and to perform
heat transfer analysis to obtain the temperatures of the fire exposed ­structures.
It  is also necessary in elevated temperature structural analysis to include
reduced mechanical properties and additional thermal effects on structural
performance such as thermal expansion and thermal bowing. This  chapter
summarises the key features of numerically performing fire resistance evalua-
tion and provides some simplified solutions.

4.1  FIRE BEHAVIOUR

The behaviour of fires in compartments is complex. For structural fire engi-
neering applications, simplifying assumptions have to be made. These simpli-
fying assumptions lead to gas/fire temperature–time relationships to be used
in thermal boundary conditions for heat transfer analysis.
Thin-walled steel structures are typically used in buildings to enclose
relatively small spaces. Therefore, it is acceptable to treat flashover enclosure

51
52  Fire Performance of Thin-Walled Steel Structures

(1) (3)
Growth phase Decay phase
(2)
Temperature

Fully developed phase

Ignition Time

FIGURE 4.1  Enclosure post-flashover fire temperature–time relationship.

fires as uniformly distributed. If uninterrupted, such an enclosure fire goes

through three phases as illustrated in Figure 4.1.

1. Growth phase until flashover: Immediately after ignition the fire

begins to grow. In this phase, the fire is localised in the enclosure
with cooler air near the floor of the compartment and smoke and
hotter air accumulating beneath the ceiling. This lasts until the hot
smoke and fire radiates sufficient heat to pyrolysis all other combus-
tible materials to cause ignition within a very short period of time,
this event being termed flashover.
2. Fully developed (post-flashover) phase: After flashover, the heat
release rate of the fire reaches the maximum, either limited by the
supply of oxygen through the openings or available fuel in the com-
partment. This phase is the most critical stage to structural stability
because very high temperatures are involved.
3. Decay or cooling phase: When nearly all combustible materi-
als are consumed, the fire enters the decay phase with decreasing
temperature. This  phase may still be important for the structure
because cooling may induce tensile forces in some structural mem-
bers if they are restrained. However, this book focuses on isolated
structural members, as is commonly practised. The effects of struc-
tural restraints are not considered.
 4  •  Numerical Modelling of Fire Resistance of Thin-Walled  53

For  assessment of the behaviour of structures in fire, the growth phase is

ignored as their temperatures are usually low.

4.1.1  Nominal Fires

Precise quantification of enclosure fire behaviour remains a scientific chal-
lenge. For practical fire resistance design, simplified analytical models are
available to calculate the fire temperature–time relationships. The most com-
mon analytical models are based on nominal fire curves and parametric fire
curves.
The  nominal fire curves define a single temperature–time relationship
independent of realistic fire conditions. They were developed for grading fire
resistance of construction elements and are adopted in design standards world-
wide with some minor differences. However, the nominal fire curves have a
number of shortcomings, including:

1. They do not represent real fire behaviour.

2. They do not always give the most severe fire condition.
3. They do not have a cooling phase.

Table  4.1 lists the most commonly used nominal fire conditions and their
parameters.
The nominal fire temperature–time curves are:

• Standard fire curve (ISO 1999, CEN 2002, SA 2005):

T f  20  345 log10  8t  1 (4.1)

• External fire curve (CEN 2002):

 
T f  660 1  0.687 e 0.32 t  0.313 e 3.8 t  20 (4.2)

• Hydrocarbon curve (CEN 2002):

 
T f  1080 1  0.325 e 0.167 t  0.675 e 2.5 t  20 (4.3)

where:
T f is the gas/fire temperature in °C
t is time in min

Figure 4.2 compares the different nominal fire temperature–time curves.

54  Fire Performance of Thin-Walled Steel Structures

TABLE 4.1  Nominal fire curves and their applications

COEFFICIENT OF HEAT
TRANSFER BY
FIRE TYPE APPLICATION CONVECTION (W / m2K)
Standard fire Fully developed cellulosic fire 25
inside fire enclosure
External fire Fully developed cellulosic fire 25
outside fire enclosure
Hydrocarbon fire Fully developed hydrocarbon fire 50
inside enclosure
Source: European Committee for Standardization (CEN), Eurocode 1: Actions on structures –
Part 1–2: General actions – Actions on structures exposed to fire, Brussels, Belgium,
2002.

1200

1000
Temperature (θ, ºC)

800

600

400

200

Standard fire External fire Hydrocarbon

0
0 20 40 60 80 100 120
Time (t, min)

FIGURE 4.2  Comparison of nominal fire temperature–time curves.

4.1.2  Parametric Fires

A  more realistic representation of fires in enclosures is the parametric fire
curve according to EN 1991-1-2 (CEN 2002). The parametric fire temperature-
time curve is a function of fire load, ventilation condition and compart-
ment characteristics. Figure 4.3 shows an example of parametric fire curve,
consisting of a heating phase until the maximum temperature followed by a
 4  •  Numerical Modelling of Fire Resistance of Thin-Walled  55

θmax
Temperature

Heating phase Cooling phase Residual phase

t*max Time

FIGURE 4.3  An example of parametric fire temperature–time curve.

linearly decreasing cooling phase until the ambient temperature. Complete

details of how to quantify a parametric fire curve are given in EN 1991-1-2
(CEN 2002).

4.1.3  Travelling Fires

Recently, a considerable amount of research studies have been devoted to trav-
elling fires, as they have been observed in large compartment fire tests and
fire accidents. A travelling fire consists of a near field (flame) and a far field
(smoke), with different temperatures (Stern-Gottfried and Rein 2012), as illus-
trated in Figure 4.4. The size of the near field can change and can be considered
a design variable. Due to non-uniformity of heating across a compartment, the
behaviour and failure mechanism of structural elements can be very complex.
For applications of thin-walled steel structures in relatively small enclosures
where the fire temperature can be considered uniform, it is not necessary to
consider travelling fires.

4.1.4  Computer Programs

Should it become necessary, a number of computer programs are available for
modelling compartment fires. However, using computer programs to model
56  Fire Performance of Thin-Walled Steel Structures

FIGURE 4.4  Travelling fire. (From Stern-Gottfried, J. and Rein, G., Fire Saf. J., 54,
96–112, 2012. With permission.)

enclosure fire behaviour requires special expertise and detailed fire dynamics
knowledge. They are rarely used by structural engineers.
Enclosure fires may be modelled using zone models, where the enclosure
is divided into a very small (typically 2) number of zones, each having uniform
properties or computational fluid dynamics (CFD) models. Popular zone and
CFD models for modelling enclosure fires include the following (Wang et al.
2012):

• COMPF2 (Babrauskas 1979): ‘C0MPF2 is a computer program

for calculating the characteristics of a post-flashover fire in a single
building compartment (single zone), based on fire-induced ventilation
through a single door or window. It is intended both for performing
design calculations and for the analysis of experimental burn data.’
• OZone V2 (Cadorin et al. 2003): The OZone V2 software calculates
the evolution of gas temperature in a compartment under fire and
evaluates the fire resistance of single steel elements. It is based on
one- or two-zone models, with an automatic shift from two to single
zone model under certain circumstances.
• SFIRE-4 (Magnusson and Thelandersson 1970): The model takes
into consideration the amount of water discharged by sprinklers,
etc. The bounding structures of the enclosure can be composed of
 4  •  Numerical Modelling of Fire Resistance of Thin-Walled  57

up to three different wall constructions or systems. This program is

based on single zone model.
• CCFM (Cooper and Forney 1990): This program was developed for
multi-room fire behaviour based on two-zone models. CCFM refers
to Consolidated Compartment Fire Model (CCFM).
• CFAST (Peacock et al. 2017): CFAST is a two-zone fire model used
to calculate the evolving distributions of smoke, fire gases and tem-
perature throughout compartments of a building during fire.
• Fire Dynamics Simulator (FDS) and Smokeview (SMV) (McGrattan
et al. 2018): FDS is a CFD model of fire-driven fluid flow. The software
solves numerically a form of the Navier-Stokes equations appropriate
to low-speed, thermally driven flow, with an emphasis on smoke and
heat transport from fires. FDS is commonly used by fire protection
engineers.

4.2  HEAT TRANSFER MODELLING

The  theory of heat transfer is well established. The  main issues are to
acquire reliable thermal material properties and to develop simplified mod-
els for practical design. Chapter  5 presents relevant material properties.
This  section describes the fundamental heat transfer equations  and their
implementation in a simplified heat transfer model for thin-walled steel
structural panels.

4.2.1  Basics of Heat Transfer

The basic equation for conductive heat transfer equation is:

dT
Q cond  kA (4.4)
dx

where, referring to Figure 4.5:

A is the area across which heat is transferred [m2]
k is the thermal conductivity of the material [W/m K]
T is temperature [K]
x is the distance normal to the area A [m]
dT/dx is temperature gradient
58  Fire Performance of Thin-Walled Steel Structures

FIGURE 4.5  Heat transfer within a member.

For  transient heat transfer in three-dimensions in the absence of internal

generation of heat, the conductive heat transfer equation  is (Cengel and
Ghajar 2014):

  T    T    T  T
 k   k  k    Cp (4.5)
x  x  y  y  z  z  t

where:
t is time [sec]
ρ is the density of the material [kg/m3]
Cp is the specific heat of the material [J/kg K]

4.2.2 Thermal Boundary Conditions

for Heat Transfer Modelling
In order to solve the general conductive heat transfer Eq. (4.5), thermal bound-
ary conditions are required. For  evaluating structure fire resistance, the
 4  •  Numerical Modelling of Fire Resistance of Thin-Walled  59

thermal boundary conditions consist of convective heat transfer and radiant

heat transfer boundary conditions.
The convective heat transfer boundary condition is expressed as (Cengel
and Ghajar 2014):

Q conv  hconv As T f  Ts  (4.6)

where:
Q conv is the rate of convective heat transfer (W)
hconv is the convective heat transfer coefficient in W/m2°C
As is the surface area through which convective heat transfer takes place
Ts is the surface temperature [°C]
T f is the temperature of the gas sufficiently far from the surface [°C] (fire
temperature)

The radiation heat transfer boundary condition is (CEN 2002):

 4

Q rad   As T f  273.15   Ts  273.15 
4
(4.7)

where:
Q rad is the rate of radiant heat transfer (W)
ε is the emissivity of the surface. The property emissivity, whose value is
in the range between 0 and 1, is a measure of how closely a surface
approximates a blackbody surface, which has an emissivity value
of 1 (  1)
σ is the Stefan–Boltzmann constant (  5.67  10 8 W / m 2  K 4 )
As is the surface area through which radiant heat transfer takes place
Ts is the surface temperature
T f is the surrounding air (fire) temperature

Figure  4.6 illustrates different modes of heat transfer for a thin-walled steel
structural panel with fire exposure from one side (Shahbazian and Wang
2014b).

4.2.3  Computer Programs

In general, there is no analytical solution to three-dimensional transient heat
transfer problems. Therefore, numerical simulation methods are used. Any
general finite element package, such as ABAQUS and ANSYS, will be able to
60  Fire Performance of Thin-Walled Steel Structures

FIGURE 4.6  Thermal boundary conditions.

perform heat transfer simulations. There are also a few specialist heat transfer
programs that have been specifically developed for performing heat transfer
calculations for structures under fire conditions, although some were developed
a long time ago and are not maintained. They include FIRES-T3 (Iding et al.
1977), TASEF (Gerlich et al. 1996), WALL2D (Takeda and Mehaffey 1999)
and SAFIR (Franssen and Gerney 2017).

4.2.4 A Simplified Heat Transfer Model

for Thin-Walled Steel Structural Panels
Whilst the above-mentioned general finite element packages and special-
ist heat transfer programs are powerful and flexible modelling tools, their
applications require specialist expertise. To facilitate rapid and accurate
calculations of heat transfer in thin-walled steel structural panels subjected
to fire from one side, Shahbazian and Wang (2013) developed a simplified
procedure that is suitable for implementation in Excel or other programs (An
example is available on MATLAB File Exchange (2013)). This procedure is
described next.
 4  •  Numerical Modelling of Fire Resistance of Thin-Walled  61

FIGURE 4.7  Schematic view of a thin-walled steel structural panel.

Figure 4.7 shows the problem under consideration. In the simplified proce-

dure, the two-dimensional heat transfer problem is converted to a one-dimensional
problem by using equivalent thermal resistances for the combined steel section
and cavity insulation. This is possible because the research studies by Feng et al.
(2003d) and Shahbazian and Wang (2013) have concluded that the steel section
temperature distribution can be assumed to be linear through the depth of
the panel and the average flange temperatures can be used (see Figure 3.5).
Figure 4.8 shows a schematic view of the simplified model including ther-
mal boundary conditions and the thermal network for heat transfer and calculat-
ing heat capacity. For this scheme, the gypsum board on each side is divided into
five layers (1–5, 11–15) and the steel section/insulation is also divided into five
layers (6–10). In the simplified method, heat transfer in the wall panel containing
a steel cross-section is assumed to be one dimensional through the depth of the
panel. For one layer, the following heat balance equation may be written:

 heat   heat   heat in the layer 

     
 entering    leaving    to change the layer 
(4.8)
 the layer   the layer   temperature 
     

Over a small time interval, the heat transfer between any two layers can be
generally written as:

T
Q  (4.9)
R

where:
∆T is the temperature difference between these two layers
ΣR is the total thermal resistance in the heat transfer path
62  Fire Performance of Thin-Walled Steel Structures

FIGURE 4.8  Schematic view of the simplified heat transfer model.

The amount of heat required to change the temperature of a layer is:

R
dT
Q cap  cap (4.10)
dt

where:
t is time
ΣRcap is the total heat capacitance (mass times specific heat) of the layer
 4  •  Numerical Modelling of Fire Resistance of Thin-Walled  63

In  this simplified method, the thermal resistance is calculated using the
weighted average of the materials within the heat transfer path. Details are as
follows:

• For heat transfer between the fire and the slice of gypsum board on
the fire exposed side (point 1), the thermal resistance is the total of
the thermal boundary layer and ½ of the gypsum slice.
• For  heat conduction between two adjacent gypsum slices (points
2–3, 3–4, 4–5, 10–11, 11–12 and 12–13), the thermal resistance is
the total of the two halves of each slice.
• For heat transfer between the slice of gypsum board on the air side
with the air layer, the thermal resistance is the total of the air layer
and ½ of the gypsum slice (point 14).
• For heat transfer between a gypsum slice and a steel slice or between
two adjacent steel slices, calculation of the total thermal resistance
should include different materials that form the slices. For  each
slice, the thermal resistances are in parallel.
• To calculate the heat capacitance of each slice, all the materials
within that slice should be included.

To calculate the thermal resistance and heat capacitance, it is necessary

to determine an effective width of the panel that should be included in the
calculations (for one-dimensional assumption). The  effective width of the
panel (We) used for calculating the weighted average of thermal resistance is
approximately:

We  45  0.85b f  in mm  (4.11)

where bf is the flange width of the thin-walled steel section.

4.3  MODELLING BEHAVIOUR OF

THIN-WALLED STEEL STRUCTURES
AT ELEVATED TEMPERATURES
The  calculation of stresses and failure modes in thin-walled steel structural
members is a complex procedure even at ambient temperature. Only very brief
guidance is provided in this section.
64  Fire Performance of Thin-Walled Steel Structures

When a thin-walled structure is subjected to compressive stress, different

modes of buckling may occur. Elastic buckling is characterised by a sudden
out-of-plane deflection and it occurs well below the yield stress of the material.
In modelling of thin-walled structures, the elastic buckling modes, which can
be obtained by performing eigenvalue analysis using finite element software
such as ABAQUS or finite strip programs such as CUFSM, are used to specify
initial geometric imperfections for nonlinear analysis. In order to apply initial
geometric imperfection, eigenvalue analysis should be performed first in order
to find the lowest buckling modes (local, distortional or global) and any inter-
action between different modes. Then based on sensitivity study and numeri-
cal validation, imperfection scale/magnitude should be defined in nonlinear
numerical analysis. However, in general, the local initial imperfection scale,
global initial imperfection scale and distortional initial imperfection scale
may be assumed to be equal to the thickness of the plate (Feng 2003), L/500
(where L is the column length) (Kaitila 2002) and half of the thickness of the
plate (Shahbazian and Wang 2012), respectively.
When conducting numerical simulation of structural behaviour in fire,
steady state or transient state analysis may be used. In steady state analysis, the
temperatures of structure are increased to the pre-determined level and this is
then followed by increasing the mechanical load. In transient state analysis, the
mechanical load is applied first and maintained, followed by increasing the tem-
peratures of structures. The steady state analysis is easier to implement. However,
for modelling structural behaviour in fire, the transient state analysis should be
used because this mimics the situation of structural behaviour in fire.

4.4 SUMMARY
The  distribution of temperatures in thin-walled steel sections as part of a
panel exposed to fire on one side can be complex. However, for evaluations
of the load carrying capacity of LSF structures, the temperature profile can
be simplified. Furthermore, the two-dimensional heat transfer problem can
be approximately reduced to a simple one-dimensional problem by using the
weighted average of thermal properties of the different constituent parts of the
thin-walled steel structural panel.
Elevated
Temperature
Properties of
5
Materials

5.1 INTRODUCTION
In fire engineering calculations of structural resistance in fire, it is necessary
to have accurate and reliable information on thermal and mechanical prop-
erties of materials of the structure. Compared to hot-rolled steel structures,
there are more issues for the material properties of cold-formed thin-walled
steel structures. Cold-formed steel structures are made by cold working of
thin-walled cold-rolled steel strips into structural shapes. This  introduces
strain hardening around corners, and the amount of strain hardening may be
affected by the thickness of the structure. Lightweight board and insulation
materials, such as gypsum plasterboard and mineral fibres, are used in thin-
walled steel construction systems. The thermal properties, in particular ther-
mal conductivity, of these materials are not  constant. Since temperatures
attained in steel sections of thin-walled construction are critically depen-
dent on insulation p­ roperties, thermal properties of insulation materials and
gypsum plasterboards should be provided. This  chapter will review avail-
able data from major sources of relevant information in literature and make
recommendations.

65
66  Fire Performance of Thin-Walled Steel Structures

5.2  MECHANICAL PROPERTIES

OF COLD-FORMED STEELS AT
ELEVATED TEMPERATURES
Cold-forming of steel does not  change its composition, therefore, the ther-
mal properties of cold-formed steel are the same as hot-rolled steel. It is also
expected that the elastic modulus of cold-formed steel is the same as hot-
rolled steel. Results of mechanical property tests by various investigators have
revealed some differences in elevated temperature modulus of elasticity values,
but because the modulus of elasticity value is a sensitive quantity for measure-
ment, these differences can usually be attributed to differences in test setup and
measurement. Therefore, this section will focus on changes in effective yield
strength of cold-formed steel at elevated temperatures.
Chen and Young (2006) reported experimental results of mechanical proper-
ties of cold-formed steel. They compared elevated temperature mechanical proper-
ties of steel coupons taken from the corner and flat parts of steel sections. The steel
was supplied by BlueScope Lysaght (Singapore) and the steel grade was G500
(minimum yield strength of 500 MPa). The elevated temperature tests were carried
out under steady state condition. They found that the corner and flat coupons had
very similar results for elevated temperature mechanical property reduction factors.
Chen and Young (2007) provided further experimental results for
BlueScope Lysaght steel grades G550 (1.0 mm) and G450 (1.9 mm) under both
steady state and transient state conditions. Their results seem to be very differ-
ent from the results of others, with much higher modulus of elasticity but much
lower yield stress at temperatures above about 500°C. Their results are also
quite different from their own test results reported in Chen and Young (2006)
for a slightly different grade of steel (G500). These make it difficult to use their
results to identify any discernible trend.
Ye and Chen (2013) carried out similar tests as Chen and Young (2006)
on Q345 steel, also supplied by BlueScope Lysaght (Shanghai), and came to
the same conclusion with regard to corner and flat plate steels. They compared
the test results under steady and transient state test conditions and found that
transient state testing gave similar results at temperatures not exceeding 400°C
and higher yield strengths at higher temperatures. Similar results have been
presented by Chen and Ye (2012) for G550 steel. However, even their transient
test results of yield strength reduction factors of steel were lower than those
recommended in EN 1993-1-2 (CEN 2005). The EN 1­ 993-1-2 reduction fac-
tors for cold-formed steel at elevated temperatures are the same as the 0.2%
proof stress of hot-rolled steel. They are very close to those of Outinen (1999).
 5  •  Elevated Temperature Properties of Materials  67

Ranawaka and Mahendran (2009b) provided experimental results for three

thicknesses (0.6, 0.8 and 0.95 mm) of two grades of cold-formed steel (G550 and
G250) under steady state condition. Their results indicate that the effect of thick-
ness is minor. However, the two grades of steel have different yield strength reduc-
tion factors, with the values for G550 steel much greater than those of G250 steel
at temperatures lower than 550°C, but much lower at temperatures above 550°C.
Kankanamge and Mahendran (2011) performed further mechanical property
tests of G250 and G450 steels for two different thicknesses (1.50 mm/1.55 mm,
1.90 mm/1.95 mm). They reached similar conclusions as Ranawaka and Mahendran
(2009b) described above and proposed improved predictive equations by combin-
ing with Ranawaka and Mahendran’s (2009b) test results. Both research studies
concluded that the thickness of steel in the range of 0.55–1.95 mm has minor influ-
ence on the mechanical properties of cold-formed steel at elevated temperatures.
A  review paper by Craveiro et  al. (2016) confirmed that the values of
modulus of elasticity of cold-formed steel obtained by different researchers for
different cold-formed steels, except for those of Chen and Young (2007), were
reasonably consistent. The recommendations in EN 1993-1-2 (CEN 2005) tend
to be at the upper bound of various test results at temperatures not exceeding
600°C but at the lower bound of test results at temperatures higher than 600°C.
On the other hand, there are large differences in yield strength results
from different researchers across the whole elevated temperature range of
interest to fire resistance. The recommended yield strength reduction factors
in EN 1993-1-2 (CEN 2005) for cold-formed steel are close to the average of
results by various researchers.
To summarise, test results by different researchers show large differences in
yield strength reduction factors of cold-formed steels at elevated temperatures.
However, these differences are not caused by steel thickness or location (corner/
flat) of steel in the cross-section. The results obtained under the steady and tran-
sient state conditions can be different for the same steel, with transient state results
seeming to be higher than steady state test results. The  most influential factor
seems to be steel grade/manufacturer. This makes it difficult to propose a univer-
sally acceptable set of yield strength reduction factor – temperature relationship
for different cold-formed steels. Ideally, specific elevated temperature mechanical
property tests should be carried out to obtain data for the specific grade and type
of steel. For example, the relationship in AS/NZS 4600 (SA 2018) should only
be applied to Australian/NZ cold-formed steels. Nevertheless, the recommended
values in EN 1993-1-2 represent approximately the average of results from differ-
ent researchers for different grades and types of steel. Therefore, where specific
information is not available, this book recommends using the EN 1993-1-2 values.
The recommended reduction factors for modulus of elasticity and effec-
tive yield strength of cold-formed steel at elevated temperatures are listed in
Table 5.1.
TABLE 5.1  Reduction factors for yield strength and modulus of elasticity of cold-formed steel at elevated temperatures
TEMPERATURE (°C) 20 100 200 300 400 500 600 700 800 900 1000 1100 1200
Yield strength 1 1 0.89 0.78 0.65 0.53 0.30 0.13 0.07 0.05 0.03 0.02 0.0
Modulus of elasticity 1 1 0.9 0.8 0.7 0.6 0.31 0.13 0.09 Linear interpolation 0.0
Source: European Committee for Standardization (CEN), EN 1993-1-2, Eurocode 3: Design of steel structures – Part 1.2: General rules –
structural fire design, Brussels, Belgium, 2005.
68  Fire Performance of Thin-Walled Steel Structures
 5  •  Elevated Temperature Properties of Materials  69

5.3  THERMAL PROPERTIES OF

FIRE PROTECTION MATERIALS AT
ELEVATED TEMPERATURES

5.3.1 General
In  cold-formed thin-walled steel panel construction for walls and floors,
lightweight fire protection and interior/exterior insulation materials are used
to provide sufficient fire resistance and thermal/sound insulation. Accurate
data of thermal properties of these materials are critical for safe and reliable
calculation of fire resistance of these systems. In fact, since the mechanical
properties of cold-formed steel decrease sharply at elevated temperatures
in the region of interest, small errors in the calculation of steel temperature
results can result in very large errors of calculating structural resistance.
For example, when the steel temperature is increased from 500°C to 600°C
(20% increase), according to the EN 1993-1-2 recommendation of reduc-
tion factors of steel yield strength and modulus of elasticity in Table 5.1, the
yield strength and modulus of elasticity of steel decrease by 43% and 48%,
respectively.
The thermal properties of a material include density (ρ), specific heat (Cp)
and thermal conductivity (k). The product of density and specific heat (ρ Cp) is
thermal capacitance of the material, measuring the amount of heat (energy)
required to raise the temperature of per unit volume of material by 1 degree.
Since fire protection and insulation materials used in cold-formed thin-walled
steel structures are lightweight, the accuracy of calculating steel temperatures
is insensitive to even large changes in thermal capacitance of fire protection/
insulation materials. On the other hand, fire protection and insulation materi-
als have low thermal conductivities, therefore, temperatures attained in the
steel structure vary almost linearly with variations in thermal conductivity of
protection/insulation materials.
The thermal conductivities of fire protection and insulation materials at
ambient temperature can be expected to be provided by the material manu-
facturers because they have to provide such information for evaluating the
thermal insulation performance of the panels. However, the ambient tempera-
ture values should not be used in fire engineering calculations. This is because
the thermal conductivity of lightweight materials increases with temperature.
70  Fire Performance of Thin-Walled Steel Structures

As mentioned above, small errors of underestimation of steel temperature

can lead to large errors of overestimation of steel mechanical properties, thus
resulting in grossly unsafe design.
In  lightweight materials, air voids exist. Therefore, void radiation at
high temperatures accelerates heat transfer through the material. Since
the power of radiation is related to T 4, the rate of heat transfer (thermal
conductivity) through the material is related to T 3. Therefore, the thermal
conductivity–temperature relationship of a lightweight material may be
expressed as follows:

k  k0  Const  T 3 (5.1)

where:
k and k0 are thermal conductivity of the material at temperature T and
ambient temperature, respectively
T is the material temperature

Wang et al. (2012) carried out an assessment of publicly available informa-

tion of thermal properties, in particular, thermal conductivities of some
fire protection materials at elevated temperatures. They recommended the
thermal property models in Table 5.2. It can be seen in Table 5.2 that the
thermal conductivities of calcium silicate and vermiculite depend on their
densities. This is because the higher the density, the lower the air void, thus
the lower the void radiation. In fact, when their densities reach their com-
plete solid densities of about 2540 kg/m 3 and 1000 kg/m 3, respectively, heat
transfer due to void radiation stops because there is no void in the material.
On the other hand, since the ambient temperature thermal conductivity of
air is negligible compared to that of the material, the thermal conductivity
of the material at ambient temperature k 0 increases linearly with increasing
density.

5.3.2 Gypsum
Gypsum plasterboards are commonly used in cold-formed thin-walled steel
panels as fire protection. Due to the presence of a large amount of free and
chemically bound water, their thermal properties are affected by water move-
ment and evaporation.
The total amount of water is about 24% by weight, consisting of about 21%
of chemically bound water and 3% of free water. Water evaporation occurs in
two stages, the first stage starting at about 100°C, during which about 75% of
the chemically bound water and free water are lost. The  temperature range
over which the second stage of water evaporation occurs is still subject to
 5  •  Elevated Temperature Properties of Materials  71

TABLE 5.2  Thermal property models for some common types of insulation
material
BASE VALUE OF
DENSITY SPECIFIC HEAT THERMAL CONDUCTIVITY
MATERIAL ρ (kg/m3) (J/kg·K) (W/m·K)
3
 T 
Rock fibre 155–180 900 k  0.022  0.1475  
 1000 
3
 T 
Mineral wool 165 840 k  0.03  0.2438  
 1000 
3
 T 
Calcium silicate Various 900 k  k0  Const  
 1000 

k0  0.23
1000
2540  
Const  0.08
2540
3
 T 
Vermiculite Various 900 k  k0  Const  
 1000 

k0  0.27
1000
1000  
Const  0.18
1000
Source: With kind permission from Taylor & Francis: Performance-Based Fire Engineering of
Structures, 2012, Wang, Y.C. et al.

debate, with some stating immediately following completion of the first stage
at about 200°C and others saying as late as 600°C. Because the amount of
water involved in the second stage is relatively low, it is acceptable to assume
that water evaporation is complete in the temperature range between 100°C
and 200°C.
Based on the above assumption of water evaporation, the thermal properties
of gypsum plasterboard can be quantified as follows.

5.3.2.1 Density
The density of gypsum plasterboard is 100% of its ambient density until 100°C,
then linearly changing to 76% (100%–24%) at 200°C and maintaining at this
value at higher temperatures.
72  Fire Performance of Thin-Walled Steel Structures

5.3.2.2  Specific heat

Dehydration of gypsum is an endothermic process which consumes heat
(energy). Exact evaluation of the temperature field in gypsum plasterboard,
and hence the steel structure, will require modelling of combined heat
and mass transfer. An alternative and simplified approach is to deal with
only the heat transfer process and using an equivalent (increased) specific
heat to take into consideration the additional heat required in dehydra-
tion. The additional heat of dehydration of gypsum plaster consists of three
parts: the amount of heat required to break the chemical bond to release
the chemically bond water, heat required to drive water from inside gyp-
sum to the surface and the latent heat of water for evaporation. The latent
heat of water is 2260 MJ/(kg·C). If the total amount of water is ‘e’, then the
total additional heat to be added to the base value of specific heat of dry
(dehydrated) gypsum can be calculated as 2260 × e × f , where ‘f ’ is to take
into consideration the effects of water movement and breakage of chemi-
cal bonds to release water. The value of ‘f’ is approximately 1.8 according
to Ang and Wang (2009). The average increase in specific heat of gypsum
can then calculated as 2260  e  f /T where ΔT is the temperature range
during which dehydration happens, which can be taken as 100°C starting
at 100°C and ending at 200°C. The distribution of this additional specific
heat has little influence on steel temperatures of interest, although a trian-
gular distribution may be used for convenience. Therefore, the additional
specific heat of gypsum due to dehydration is 0 at 100°C, ­l inearly increas-
ing to 2  2260  e  f /T , where T  100C at 150°C and then ­l inearly
decreasing to 0 at 200°C.

5.3.2.3  Thermal conductivity

Water has much higher thermal conductivity than air. Therefore, after
dehydra­tion, the thermal conductivity of gypsum is lower than before dehy-
dration. How­ever, at high temperatures after dehydration, the thermal con-
ductivity of gypsum increases with increasing temperature. There  are two
schools of thought on why the thermal conductivity of gypsum increases
with increasing temperature. Rahmanian and Wang (2012) attributed this to
void radiation after dehydration, while Ghazi Wakili et  al. (2015) thought
that this was induced by a type of sintering process (softening of the contacts
between individual crystal needles), which increases the total contact surface
and hence the conduction of heat through the bulk material.
Whatever the cause, they suggested similar thermal conductivity–temperature
relations for gypsum. The relationship is as follows:
 5  •  Elevated Temperature Properties of Materials  73

• Ambient temperature value maintained until 100°C,

• Linearly decreasing to a value of 0.1 W/(m·C) until 200°C,
• Linearly increasing to a value to that of ambient temperature until
1000°C.

5.4  ADDITIONAL ISSUES

When loaded thin-walled steel panels with gypsum plasterboard facing and
interior insulation are exposed to fire attack, the gypsum plasterboard may
crack and fall-off and the interior insulation may shrink or detach from each
other to form holes, as shown in Figure 5.1. This will have effects on the tem-
perature developments in the steel structure. At present, this phenomenon has

FIGURE  5.1  Behaviour of gypsum and interior insulation after a fire test.
(Reprinted from Fire Saf. J., 40, Feng, M. and Wang, Y.C., An experimental study
of loaded full-scale cold-formed thin-walled steel structural panels under fire
conditions, 43–63, Copyright 2005, with permission from Elsevier.)
74  Fire Performance of Thin-Walled Steel Structures

been observed by many researchers, but it is still not possible to predict when

this would happen. This usually happens later in fire tests and it is possible that
this occurrence is induced by the failure of the steel structure. Should this be
the case, then fire resistance of the construction is not affected.

5.5 SUMMARY
The accuracy of any analysis depends on the accuracy of input of the material
properties. For  the evaluation of fire resistance of structures, accurate data
of both thermal and mechanical properties of the materials are necessary.
For thin-walled steel structures, this chapter has recommended temperature-
dependent thermal and mechanical property values for some protective mate-
rials and steel, respectively. However, significant gaps in knowledge on the
material properties of thin-walled steel structure still exist and further research
studies are necessary.
Performance-
Based Design
Methods of
6
Thin-Walled
Steel Members
at Elevated
Temperatures
Thin-walled steel members usually form part of a panel system, either as part
of walls or floors. When designing for fire resistance, these panels are assumed
to be exposed to fire from one side. Therefore, temperature distributions in
thin-walled steel members are non-uniform, with steep gradients. Thin-walled
steel members are also prone to different modes of buckling, including local
buckling, distortional buckling and global buckling. Therefore, the behaviour
of thin-walled steel members in fire is very complicated. It is not surprising that
so far there is no widely accepted design calculation method for t­ hin-walled
steel members in fire.
This chapter will present a number of different methods that are either in
design codes and standards or were a result of extensive research, and explain
their scopes of application. The  authors hope that they become the basis of
further assessment by interested readers who may wish to develop a design
method for their applications.

75
76  Fire Performance of Thin-Walled Steel Structures

6.1  THIN-WALLED STEEL MEMBERS WITH

UNIFORM TEMPERATURE DISTRIBUTION
As mentioned above, thin-walled steel members experience non-uniform
temperature distributions in their sections in the majority of applications.
Therefore, this chapter will primarily focus on this case.
In the case that the temperature distribution in the thin-walled steel sec-
tion is uniform, the existing methods of design for thin-walled steel mem-
bers at ambient temperature, such as the effective width method in Eurocode
(EN 1993-1-3, CEN 2006a) or the direct strength method in North American
Specification AISI S100-16 (AISI 2016) and Australia/New Zealand Standard
AS/NZS 4600 (SA  2018), may be used, provided the ambient temperature
mechanical properties of steel are replaced by those at elevated temperatures.
Alternatively for simple estimations, the limiting temperature method may
be used. The results of a numerical investigation by Maia et al. (2016) suggest
that limiting temperatures of 600°C, 550°C, 500°C and 400°C may be used for
load ratios of 0.3, 0.4, 0.5 and 0.6, respectively, the load ratio being defined as
the applied load in fire to the load carrying capacity of the member at ambient
temperature. This limiting temperature – time relationship is broadly similar to
the temperature-0.2% proof stress (yield stress) reduction factor of cold-formed
steel, given in Chapter 5. These limiting temperatures are higher than the default
value of 350°C, which is the currently recommended value in Eurocode 3 Part
1.2 (CEN 2005) and in Australian standard AS/NZS 4600:2018 (SA 2018).

6.2  SIMPLIFIED METHODS FOR

THIN-WALLED MEMBERS WITH NON-
UNIFORM TEMPERATURE DISTRIBUTION

6.2.1  Limiting Temperature Method

The  limiting temperature method has already been mentioned and exist-
ing research studies indicate that this method is reasonable for uniformly
heated thin-walled steel members. However, for thin-walled steel members
with non-uniform temperature distribution in the cross-section, Feng et  al.
(2003d) found that the limiting temperature method is no longer applicable,
due mainly to the influence of thermal bowing. In EN 1993-1-2 (CEN 2005)
 6  •  Performance-Based Design Methods  77

and Australian/New Zealand standard AS/NZS 4600:2018 (SA  2018), the

recommended limiting temperature is 350°C. The  results of Feng et  al.
(2003d) confirm that this value is safe provided this value is applied to the
maximum temperature of the steel section. The average temperatures in thin-
walled steel members at failure may be lower than 350°C.
However, applying a limiting temperature value of 350°C to the maxi-
mum temperature of the steel cross-section is unduly conservative in many
cases.

6.2.2  Extension of Fire Test Results

The previous discussions have focused on resistance of thin-walled steel mem-
bers at elevated temperatures, which is only part of the process of assessing
their fire resistance. The other aspect is thermal response. Therefore, a com-
plete design calculation method would need to be able to accurately predict
the thermal responses of thin-walled steel members. Such a method has been
described in Chapter  4, but the calculation results using such methods may
not  be sufficiently accurate because thin-walled steel panel structures typi-
cally contain many components and there are uncertainties in their properties,
for example thermal properties and falling off of gypsum boards and ther-
mal properties of the interior or external insulation. Therefore, fire resistance
testing still represents the dominant way of demonstrating acceptance of thin-
walled steel member panel systems in practice.
However, fire resistance tests are time-consuming and expensive to con-
duct and the dimensions of structures are limited by the dimensions of the
fire test furnaces. Therefore, only a limited number of fire resistance tests are
conducted for each system and the construction used in practice may be dif-
ferent from those tested.
To allow different constructions of the same system to be used, an
approach is to extend the scope of application of the fire test results. Provided
the extension is over a limited range, this may be possible and one simple
method has recently been developed by the UK’s Steel Construction Institute
(SCI 2019).

6.2.2.1  Members in walls

For thin-walled steel members in panels under axial load, the starting point of
the extension method is:

N Rd , f  0.6 N b, Rd SRF Tref  (6.1)

78  Fire Performance of Thin-Walled Steel Structures

where:
NRd, f is the load carrying capacity of the thin-walled steel member at the
fire limit state
Nb, Rd is the load carrying capacity of the member in normal conditions

Tref is the reference temperature of the thin-walled steel section, which is taken
as the mean temperature of the section. This may be taken as the average of
the flange temperatures.
SRF(Tref ) is the strength reduction factor for cold-formed steel at Tref
according to BS EN 1993-1-2 (2005). Because the extension method is based
on having availability of a valid fire test, the steel temperatures are from the
fire test. In fire resistance tests on thin-walled steel panels, the thinnest steel
section is usually used. Therefore, the test temperatures can be applied to
thicker steel sections, which will have lower temperatures.
The  multiplication factor of ‘0.6’ is to allow for the effects of ther-
mal bowing due to non-uniform temperature distribution in the steel
cross-section.
This method allows the fire test result to be used for steel members that
are slightly longer (+10% in height), thicker steel members that resist higher
loads at the same fire resistance time, or to allow steel members to resist
higher loads at lower fire resistance times for which temperature data would
be available. However, this method is not applicable to steel members with
reduced load level to achieve longer fire resistance times for which experi-
mental data of temperatures would not be available. Due to gross approxima-
tions, this method should only be extended (extrapolated) over a small range,
e.g. about 10%.

6.2.2.2  Members in floors

Steel members in floor panels will be under bending so the effects of thermal
bowing are negligible. Also because the steel members will be restrained by
panel boards at regular intervals, global buckling behaviour of the steel mem-
bers is controlled by spacing of the restraints. Therefore, it may be assumed
that the steel member resistance to bending is independent of the span.
Consequently, the same fire resistance test result for one span can be used for
others provided the maximum bending moments in the members are the same.
As with the extension method for vertical members, the extension method can
be used for thicker steel sections and for fire resistance times no longer than
the test fire resistance time.
 6  •  Performance-Based Design Methods  79

6.3  THIN-WALLED STEEL COLUMNS

WITH NON-UNIFORM TEMPERATURE
DISTRIBUTION IN THE CROSS-SECTION

6.3.1  Effective Width Method

In Europe, the effective width method is used to calculate the resistances of
thin-walled steel members at ambient temperature. Based on the results of a
relatively limited set of parametric studies, Feng et al. (2003d) evaluated modi-
fication of the effective width method for elevated temperature applications.
Due to non-uniform temperature distribution in the cross-section, a number of
assumptions had to be made, including the following:

• The shear centre and warping constant of the cross-section are the

same as at ambient temperature. The  weighted average stiffness
value of the cross-section is used.
• The effective width of the web is calculated using the weighted aver-
age steel stiffness value. This affects the location of the centroid and
stress distribution of the cross-section.
• The cross-section resistance of the effective section is either governed
by first yield, i.e. whenever the compression or tensile stress reaches
the corresponding elevated temperature yield stress the earliest, or
partial plasticity. Partial plasticity happens when the tensile side
reaches yield and extends the region of plasticity until the maximum
compressive stress reaches yield.
• Due to the change of centre of resistance of the cross-section with
non-uniform temperature distribution and thermal bowing, bend-
ing moments are present in axially loaded members. Therefore, the
thin-walled member should be evaluated under combined compres-
sion and bending. The interaction equations in EN 1993-1-3 (CEN
2006a) may be used.

The results of Feng et al. (2003d) indicate that it is possible to use the effective
width method to calculate the resistances of thin-walled steel members in wall
construction with a reasonable level of accuracy. Using the first yield condition
to calculate the cross-section resistance would give lower resistances than using
the partial plasticity condition, but overall differences are small.
80  Fire Performance of Thin-Walled Steel Structures

This  assessment of the effective method was performed for a limited

number of cases and there was no detailed consideration of different failure
modes (local, distortional, global buckling). For  thin-walled sections with
internal web stiffeners, the approach of using the weighted average stiffness
value of the whole web will not be appropriate, instead, the weighted average
values between adjacent stiffeners should be used. The calculation procedure
would be rather tedious to implement by hand, so a bespoke computer program
should be written.

6.3.2  Direct Strength Method

The direct strength method (DSM) for the design of thin-walled structures is
now widely adopted in design standards of many countries, including North
American (AISI S100-16) and Australia/New Zealand (AS/NZS 4600:2018).
It  uses buckling equations  to combine the cross-sectional resistances of
­thin-walled member with elastic critical loads of the member to calculate the
design resistance, similar to design calculations for steel members with thicker
sections. The elastic critical loads can be easily calculated using a computer
program, one of which is CUFSM which is freely available. When calculat-
ing the plastic cross-section resistance (squash load), the gross cross-section
dimensions are used.
Compared with the effective width method, DSM has advantages
including being able to deal with interactions of different buckling modes
and a streamlined calculation procedure. For  thin-walled members under
compression in wall panel applications, Shahbazian and Wang (2014a) car-
ried out an extensive assessment of DSM. The results of their investigations
confirm that DSM can be applied for thin-walled steel members with both
uniform and non-uniform temperature distributions in the cross-sections for
all three buckling modes. Due to temperature effects, modifications to the
ambient temperature DSM equations  are necessary. This  method will be
summarised below.
For thin-walled members with non-uniform temperature distribution under
compression, additional bending moments will be generated due to thermal
bowing and shift of centre of resistance of the cross-section. To make use of
ambient temperature equations for thin-walled steel members under compres-
sion, Shahbazian and Wang (2014a) treated a thin-walled steel member under
combined compression and bending as an equivalent column under compres-
sion only. The cross-section resistance of the equivalent column is the same
as the axial force in the cross-section plastic compression-bending resistance
 6  •  Performance-Based Design Methods  81

interaction curve under the same bending moment as caused by thermal

bowing (at member centre) or shift of centre of resistance (at m ­ ember ends).
The  bending resistance is taken about the centre of resistance with non-
uniform temperature distribution. At  the member ends, thermal bowing is
zero so there is only shift of centre of resistance. At column mid-height, both
thermal bowing and shift of centre of resistance exist and they act in opposite
direction. The thermal bowing is calculated using Eq. 6.2. This is illustrated in
Figure 6.1 (Shahbazian and Wang 2014a).
The elastic critical loads of a thin-walled member with non-uniform tem-
perature distribution can be calculated using CUFSM, with the different parts
of the cross-section having different Young’s moduli at different elevated tem-
peratures. The shift of centre of resistance is automatically included by input-
ting different Young’s moduli of steel. To include the effect of thermal bowing,

FIGURE 6.1  Determination of effective squash load of effective column. (From

Shahbazian, A. and Wang, Y.C., Struct. Eng., 92, 52–62, 2014.)
82  Fire Performance of Thin-Walled Steel Structures

an additional bending moment is applied to the member, which is equal to the

applied compression force acting at an eccentricity equal to the thermal bow-
ing of the member calculated below:

 L2 T
m  (6.2)
8d

where:
α is the coefficient of thermal expansion of steel
L is the column length (panel height)
ΔT is the temperature difference between the exposed and unexposed
sides
d is depth of the cross-section

The DSM equations recommended by Shahbazian and Wang (2014a) are:

• Global buckling:
for λ e ≤ 1.5:
 2
Pne  0.495 e Py  (6.3)

for λ e > 1.5:
 0.462 
Pne   2  Py (6.4)
 e 
• Local buckling:
for λ l  ≤ 0.776:
Pnl = Pne (6.5)
for λ l  > 0.776:
  Pcrl    Pcrl 
0.75 0.75


Pnl  1  0.22      Pne (6.6)
  Pne    Pne 

• Distortional buckling:
for λ d  ≤ 0.561:
Pnd = Py (6.7)
for λ d  > 0.561:
  P   P 
0.7 0.7

Pnd  0.65 1  0.14  crd    crd  Py (6.8)

  Py    Py 

 6  •  Performance-Based Design Methods  83

In the above formulas Py is the equivalent squash load of the cross-section; Pcre
is the critical elastic global buckling load; Pcrl is the critical elastic local buck-
ling load; Pcrd is the critical elastic distortional buckling load; λ e, λ l and λ d
are the slenderness for global, local and distortional buckling modes, respec-
tively; Pne, Pnl and Pnd are the column axial strength for global, local and distor-
tional buckling modes, respectively; Pn is the final design strength under axial
compression.
The DSM equations above have been demonstrated to be applicable for
thin-walled members in wall panels under both the standard and parametric
design fire curves of EN 1991-1-2 (2005).

6.3.3 Simplified Effective Width/Direct

Strength Method
The  only national standard that includes a complete ‘performance’ based
design method for thin-walled steel members is the Australian/New Zealand
standard AS/NZS 4600:2018 (SA  2018). This  standard recommends using
either the effective width method or the direct strength method, as summarised
below.

6.3.3.1  Effective width method

The effective width method is used to calculate the cross-section resistances of
thin-walled steel walls. The effective section is calculated using the elevated
temperature mechanical properties. The effects of non-uniform temperatures
are considered when calculating ‘effective’ average stresses for different cross-
section resistances. For calculating the cross-section squash load, the weighted
average yield stress of the gross cross-section is used. The cross-section bend-
ing resistances are the elastic moment resistances. At the member ends where
the hot flange is under compression due to shift of centre of resistance, the
yield stress of the hot flange is used. At  the member centre where the cold
flange is under compression due to thermal bowing, the yield stress at mid-web
temperature is used.
The  effective width method is combined with second-order member
analysis in which the design bending moment at the member centre is mag-
1
nified using the magnification factor 1 NNEd  , in which NEd is the design axial
cr ,T
compression load in the member and Ncr, T the flexural axial buckling capac-
ity. Ncr, T is calculated using the flexural stiffness of the cross-section taking
into consideration the non-uniform distribution of Young’s modulus of steel
84  Fire Performance of Thin-Walled Steel Structures

in the cross-section due to non-uniform temperature distribution and shift

of neutral axis caused by the variation of Young’s modulus.
A linear moment-axial load interaction relationship is then used to check
whether the member has adequate load carrying capacity.

6.3.3.2  Direct strength method

When using the direct strength method, the axial compression and bending
resistances of the member are calculated using the direct strength method
under individual actions. The direct strength equations are the same as those
in AISI S100-16 (AISI 2016) at ambient temperature but the gross cross-
section resistances (bending and compression) and critical buckling loads
are modified from those at ambient temperature to take into consideration
the effects of elevated temperatures. For  calculating the critical buckling
loads, the critical global flexural buckling resistance is that of the gross
cross-section with ­non-uniform temperature distribution. The critical local
buckling resistance (bending or compression) is modified by using the
Young’s modulus reduction factor at the mid-web temperature. Distortional
buckling is not considered. The critical bending resistances are obtained by
modifying the ambient temperature values of the gross cross-section using
Young’s modulus reduction factors at the reference temperatures. For  the
member ends, the reference temperature is that of the hot flange (compres-
sion flange) and for the member mid-height, the reference temperature is that
of the mid-web.
A  linear interaction equation  between axial compression and bend-
ing is then used to check whether the member has sufficient load carrying
capacity. In  the interaction equation, the design bending moment is cal-
culated in the same way as in the effective width method, by considering
member second-order effects. Compared to the effective width method,
the differences are that the direct strength resistances of the member with
gross cross-section under compression and bending separately are used
in the direct strength method, instead of the resistances of the effective
cross-section.
Compared with the effective width method of Feng et  al. (2003d) and
the direct strength method of Shahbazian and Wang (2014a), the AS/NZS
4600:2018 (SA  2018) method appears to be simpler for implementation by
hand, as this method makes use of only selected reference temperatures to
modify the ambient temperature calculation equations to obtain each respec-
tive elevated temperature value.
 6  •  Performance-Based Design Methods  85

6.4  THIN-WALLED STEEL BEAMS

WITH NON-UNIFORM TEMPERATURE
DISTRIBUTION IN THE CROSS-SECTION
For thin-walled steel members in floors, the steel member is restrained by floor
boards so there is no global lateral torsional buckling. Also there is no axial
load in the member, therefore the shift of centre of resistance and thermal bow-
ing, which dominate the behaviour and resistance of thin-walled members in
wall panels, do not affect the member resistance.
AS/NZS 4600:2018 (SA 2018) appears to be the only standard that includes a
design method for thin-walled members in floors. Again, both the effective width
method and the direct strength method can be used. In both methods, the steel
section with non-uniform temperature distribution is treated as that with uniform
temperature with the mid-web temperature being taken as the reference uniform
temperature of the cross-section. When using the effective width method, the
resistance of the member is that at first yield of the effective cross-section.

6.5  ILLUSTRATIVE DESIGN EXAMPLE

USING DIRECT STRENGTH METHOD
Requirement: calculate the load carrying capacity of a panel at 60 min of the
standard fire exposure time.

Basic input data:

Steel section dimensions: lipped channel 75 × 50 × 15 × 2.5
Interior insulation: ISOWOOL
Gypsum: One layer of 12.5 mm thick Fireline British gypsum on each
side
Panel height: 3 m

The  calculated temperatures of the exposed and unexposed flanges (follow-

ing the procedure described in Section  4.2.4) are 510.42°C and 273.85°C,
respectively.
86  Fire Performance of Thin-Walled Steel Structures

MECHANICAL PROPERTIES OF STEEL AT AMBIENT

TEMPERATURE (N/mm2)
Elastic modulus (E) Yield stress (fy)
205,000 350

Using program CALESL (http://j.mp/CALESL), the centre of plastic resistance

is 31.3 mm from the lower temperature flange side (37.5 mm at ambient tem-
perature). Therefore, the shift of centre of resistance is 37.5 − 31.3 = 6.2 mm.
Using Eq. 6.2 gives a thermal bowing value of 49.6 mm. Therefore, the eccen-
tricities at the top and bottom, and at the centre of the column are 6.2  mm
and ⎮6.2 − 49.6⎮ = 43.4 mm, respectively.
Using program CALESL, the column axial load-bending moment inter-
action curve is obtained, as shown in Figure 6.2. From the two eccentricities
above, the effective squash loads are 98.29 kN (top and bottom) and 49.41 kN
(middle), respectively. Therefore, the effective squash load is min(98.29,
49.41) = 49.41 kN.

FIGURE 6.2  Determination of the effective squash load. (From Shahbazian, A.

and Wang, Y.C., Struct. Eng., 92, 52–62, 2014.)
 6  •  Performance-Based Design Methods  87

Using the CUFSM (http://bit.ly/CUFSM) program, with input of the ele-

vated temperature Young’s modulus values of the steel section, the elastic criti-
cal loads for global, distortional and local buckling are:

Pcre = 68.04 kN
Pcrl = 388.85 kN
Pcrd = 317.7 kN

Using Eqs. 6.3–6.8:

λc = 0.8522; Pne = 29.65 kN (Eq. 6.4)

λl = 0.2761; Pnl = 29.65 kN (Eq. 6.6)
λd = 0.3944; Pnd = 49.41 kN (Eq. 6.8)
Pn = min(29.65, 29.65, 49.41) = 29.65 kN

6.6 SUMMARY
This chapter has presented an outline of a number of methods of calculating
the load carrying capacity of thin-walled steel structures at elevated tempera-
tures. Thin-walled steel structures have non-uniform temperature distributions
and complex modes of failure. Therefore, there is still no universally accepted
design calculation method. Nevertheless, as have been demonstrated in this
chapter, it is possible to use the existing ambient temperature design methods
as the basis for further development.
References
Alfawakhiri, F. (2001). Behaviour of cold-formed steel framed walls and floors in stan-
dard fire resistance tests. PhD thesis, Carleton University, Ottawa, Canada.
Alfawakhiri, F. and Sultan, M.A. (2001). Loadbearing capacity of cold-formed
steel joists subjected to severe heating. Proceedings of the 9th International
Conference Proceedings, Interflam 2001, Edinburgh, UK, Vol. 1, pp. 431–442.
Alfawakhiri, F., Sultan, M.A. and MacKinnon, D.H. (1999). Fire resistance of
load bearing steel-stud walls protected with gypsum board. Journal of Fire
Technology, Vol. 35, pp. 308–335.
American Iron and Steel Institute (AISI) (2016). S100-16: North American specifica-
tion for the design of cold-formed steel structural members, 2016 edition, AISI,
Washington, DC.
Anapayan, T. and Mahendran, M. (2012). Improved design rules for hollow flange sec-
tions subject to lateral distortional buckling. Thin-Walled Structures, Vol.  50,
pp. 128–140.
Ang, C.N. and Wang, Y.C. (2009). Effect of moisture transfer on specific heat of gyp-
sum plasterboard at high temperatures. Construction and Building Materials,
Vol. 23, pp. 675–686.
Ariyanayagam, A. and Mahendran, M. (2018). Fire performance of load bearing
LSF walls made of low strength steel studs. Thin-Walled Structures, Vol. 130,
pp. 487–504.
Ariyanayagam, A. and Mahendran, M. (2019). Influence of cavity insulation on the
fire resistance of light gauge steel framed walls. Construction and Building
Materials, Vol. 203, pp. 687–710.
Ariyanayagam, A.D. and Mahendran, M. (2017). Fire tests of non-load bearing light
gauge steel frame walls lined with calcium silicate boards and gypsum plaster-
boards. Thin-Walled Structures, Vol. 115, pp. 86–99.
Ariyanayagam, A.D., Kesawan, S. and Mahendran, M. (2016). Detrimental effects of
plasterboard joints on the fire resistance of light gauge steel frame walls. Thin-
Walled Structures, Vol. 107, pp. 597–611.
Avery, P., Mahendran, M. and Nasir, A. (2000). Flexural capacity of hollow flange
beams. Journal of Constructional Steel Research, Vol. 53, pp. 201–223.
Babrauskas, V. (1979). COMPF2, a program for calculating post-flashover fire tem-
peratures (Vol. 991). Department of Commerce, National Bureau of Standards.
Baleshan, B. and Mahendran, M. (2016). Experimental study of LSF floor
systems under fire conditions. Advances in Structural Engineering.
doi:10.1177/1369433216653508.
Bandula Heva, Y. and Mahendran, M. (2013). Flexural-torsional buckling tests of
cold-formed steel compression members at elevated temperatures. Steel and
Composite Structures, Vol. 14, No. 3, pp. 205–227.

89
90 References

Baspinar, M. and Kahraman, E. (2011). Modifications in the properties of gypsum

construction element via addition of expanded macroporous silica granules.
Construction and Building Materials, Vol. 25, pp. 3327–3333.
Baux, C., Mélinge, Y. and Lanos, C. (2008). Enhanced gypsum panels for fire protec-
tion. Journal of Materials in Civil Engineering, Vol. 20, pp. 71–77.
The Building Regulations 2010 (BR2010). Approved Document B: Fire safety, 2006
edition incorporating 2010 and 2013 amendments, HM Government, London,
UK, 2013.
Cadorin, J.F., Pintea, D., Dotreppe, J.C. and Franssen, J.M. (2003). A tool to design
steel elements submitted to compartment fires—OZone V2. Part 2: Methodology
and application. Fire Safety Journal, Vol. 38, No. 5, pp. 429–451.
Cengel, Y. and Ghajar A.J. (2014). Heat and Mass Transfer: Fundamentals and
Applications. McGraw-Hill Higher Education, New York.
Chen, J. and Young, B. (2006). Corner properties of cold-formed steel sections at ele-
vated temperatures. Thin-Walled Structures, Vol. 44, pp. 216–223.
Chen, J. and Young, B. (2007). Experimental investigation of cold-formed steel mate-
rial at elevated temperatures. Thin-Walled Structures, Vol. 45, pp. 96–110.
Chen, W. and Ye, J. (2014). Fire resistance prediction of load bearing cold-formed steel
walls lined with gypsum composite panels. Proceedings of 22nd International
Specialty Conference on Cold-formed Steel Structures. St. Louis, MO, pp. 541–555.
Chen, W. and Ye, J.H. (2012). Mechanical properties of G550 cold-formed steel under
transient and steady state conditions. Journal of Constructional Steel Research,
Vol. 73, pp. 1–11.
Chen, W., Ye, J., Bai, Y. and Zhao, X.L. (2012). Full-scale fire experiments on
load-bearing cold-formed steel walls lined with different panels. Journal of
Constructional Steel Research, Vol. 79, pp. 242–254.
Chen, W., Ye, J., Bai, Y. and Zhao, X.L. (2013a). Improved fire resistant performance of
load bearing cold-formed steel interior and exterior wall systems. Thin-Walled
Structures, Vol. 73, pp. 145–157.
Chen, W., Ye, J., Bai, Y. and Zhao, X.L. (2013b). Thermal and mechanical modelling of
load-bearing cold-formed steel wall systems in fire. ASCE Journal of Structural
Engineering, Vol. 140, p. A4013002.
Cooper, L.Y. and Forney, G.P. (1990). The  Consolidated Compartment Fire Model
(CCFM) computer code application CCFM. Vents-Part III: Catalog of algo-
rithms and subroutines. Report NISTIR 4344, National Institute of Standards
and Technology, Gaithersburg, MD.
Craveiro, H.D., Rodrigues, J.P.C., Santiago, A. and Laim, L. (2016). Review of the high
temperature mechanical and thermal properties of the steels used in cold formed
steel structures  – The  case of the S280Gd+Z steel. Thin-Walled Structures,
Vol. 98, pp. 154–168.
Dempsey, R.I. (1990). Structural behaviour and design of hollow flange beams.
National Structural Engineering Conference, Adelaide, Australia.
Dias, Y., Keerthan, P. and Mahendran, M. (2019a). Fire performance of steel and plas-
terboard sheathed non-load bearing LSF walls. Fire Safety Journal, Vol.  103,
pp. 1–18.
Dias, Y., Mahendran, M. and Keerthan, P. (2018). Predicting the fire performance
of LSF walls made of web stiffened channel sections. Engineering Structures,
Vol. 168, pp. 320–332.
 References 91

Dias, Y., Mahendran, M. and Keerthan, P. (2019b). Full-scale fire tests of steel and plas-
terboard sheathed web-stiffened stud walls. Thin-Walled Structures, Vol.  137,
pp. 81–93.
European Committee for Standardization (CEN) (2002). Eurocode 1: Actions on
structures – Part 1–2: General actions – Actions on structures exposed to fire.
European Committee for Standardization, Brussels, Belgium.
European Committee for Standardization (CEN) (2005). EN 1993-1-2, Eurocode 3:
Design of steel structures  – Part 1.2: General rules  – Structural fire design.
European Committee for Standardization, Brussels, Belgium.
European Committee for Standardization (CEN) (2006a). EN 1993-1-3 2006, Eurocode
3: Design of steel structures – Part 1–3: General rules – Supplementary rules for
cold-formed members and sheeting. European Committee for Standardization,
Brussels, Belgium.
European Committee for Standardization (CEN) (2006b). EN 1993-1-5: Eurocode
3: Design of steel structures  – Part 1–5: Plated structural elements. British
Standards Institution, London, UK.
Feng, M. (2003). Numerical and experimental studies of cold-formed thin-walled steel
studs in fire. PhD thesis, University of Manchester.
Feng, M. and Wang, Y.C. (2005). An experimental study of loaded full-scale cold-
formed thin-walled steel structural panels under fire conditions. Fire Safety
Journal, Vol. 40, pp. 43–63.
Feng, M., Wang, Y.C. and Davies, J.M. (2003a). Structural behaviour of cold-formed
thin-walled short steel channel columns at elevated temperatures, part 1:
Experiments. Thin-Walled Structures, Vol. 41, No. 6, pp. 543–570.
Feng, M., Wang, Y.C. and Davies, J.M. (2003b). Structural behaviour of cold-formed
thin-walled short steel channel columns at elevated temperatures, part 2: Design
calculations and numerical analysis. Thin-Walled Structures, Vol.  41, No.  6,
pp. 571–594.
Feng, M., Wang, Y.C. and Davies, J.M. (2003c). Thermal performance of cold-formed
thin-walled steel panel systems in fire. Fire Safety Journal, Vol.  38, No.  4,
pp. 365–394.
Feng, M., Wang, Y.C. and Davies, J.M. (2003d). Axial strength of cold-formed thin-
walled steel channels under non-uniform temperatures in fire. Fire Safety
Journal, Vol. 38, pp. 679–707.
Franssen, J.M. and Gernay, T. (2017). Modeling structures in fire with SAFIR®:
Theoretical background and capabilities. Journal of Structural Fire Engineering,
Vol. 8, No. 3, pp. 300–323.
Gerlich, J.T., Collier, P.C.R. and Buchanan, A.H. (1996). Design of light steel‐framed
walls for fire resistance. Fire and Materials, Vol. 20, No. 2, pp. 79–96.
Ghazi Wakili, K., Koebel, M., Glaettli, T. and Hofer, M. (2015). Thermal conductivity
of gypsum boards beyond dehydration temperature. Fire and Materials, Vol. 39,
pp. 85–94.
Gunalan, S., Bandula Heva, Y. and Mahendran, M. (2014). Flexural-torsional buck-
ling behaviour and design of cold-formed steel compression members at elevated
temperatures. Engineering Structures, Vol. 79, pp. 149–168.
Gunalan, S., Bandula Heva, Y. and Mahendran, M. (2015). Local buckling studies of
cold-formed steel compression members at elevated temperatures. Journal of
Constructional Steel Research, Vol. 108, pp. 31–45.
92 References

Gunalan, S., Kolarkar, P.N. and Mahendran, M. (2013). Experimental study of load
bearing cold-formed steel wall systems under fire conditions. Thin-Walled
Structures, Vol. 65, pp. 72–92.
Ha, T.H., Cho, B.H., Kim, H. and Kim, D.J. (2016). Development of an efficient
steel beam section for modular construction based on six-sigma. Advances
in Material Science and Engineering, Vol.  2016, Article ID 9687078.
doi:10.1155/2016/9687078.
Hanna, M.T., Abreu, J.C.B., Schafer, B.W. and Abu-Hamd, M. (2015). Post-fire buck-
ling strength of CFS walls sheathed with magnesium oxide or ferrocement
boards. Proceedings of the Annual Stability Conference, Structural Stability
Research Council, Nashville, TN, March 24–27, 2015.
Iding, R., Bresler, B. and Nizamuddin, Z. (1977). FIRES-T3, A computer program for
the fire response of structures-thermal. Report No. UCB FRG 77, 15.
ISO 834-1 (1999). Fire resistance tests  – Elements of building construction, Part 1:
General requirements. International Organization for Standardization, Geneva,
Switzerland.
Jatheeshan, V. and Mahendran, M. (2016a). Experimental study of cold-formed steel
floors made of hollow flange channel section joists under fire conditions. ASCE
Journal of Structural Engineering, Vol. 142, No. 2, p. 04015134.
Jatheeshan, V. and Mahendran, M. (2016b). Fire resistance of LSF floors made of hol-
low flange channels. Fire Safety Journal, Vol. 84, pp. 8–24.
Kaitila, O. (2002). Imperfection sensitivity analysis of lipped channel columns at
high temperatures. Journal of Constructional Steel Research, Vol.  58, No.  3,
pp. 333–351.
Kankanamge, N.D. and Mahendran, M. (2011). Mechanical properties of cold-formed
steels at elevated temperatures. Thin-Walled Structures, Vol. 49, pp. 26–44.
Keerthan, P. and Mahendran, M. (2011). New design rules for the shear strength
of LiteSteel beams. Journal of Constructional Steel Research, Vol.  67,
pp. 1050–1063.
Keerthan, P., Mahendran, M. and Frost, R.L. (2013). Fire safety of steel wall systems
using enhanced plasterboards. Proceedings of the 19th International CIB World
Building Congress 2013, Brisbane, Australia, pp. 1–12.
Kesawan, S. and Mahendran, M. (2015). Predicting the performance of LSF walls made
of hollow flange sections in fire. Thin-Walled Structures, Vol. 98(A), pp. 111–126.
Kesawan, S. and Mahendran, M. (2016). Thermal performance of load‐bearing walls
made of cold‐formed hollow flange channel sections in fire. Fire and Materials,
Vol. 40, No. 5, pp. 704–730.
Kesawan, S. and Mahendran, M. (2017a). Fire performance of LSF walls made of hol-
low flange channel studs. Journal of Structural Fire Engineering, Vol. 8, No. 2,
pp. 149–180.
Kesawan, S. and Mahendran, M. (2017b). A review of parameters influencing the fire
performance of light gauge steel framed walls. Fire Technology. doi:10.1007/
s10694-017-0669-8.
Kodur, V.K.R. and Sultan, M.A. (2006). Factors influencing fire resistance of load-
bearing steel stud walls. Fire Technology, Vol. 42, pp. 5–26.
Kolarkar, P.N. and Mahendran, M. (2008). Thermal performance of plasterboard lined
steel stud walls. Proceedings of the 19th International Specialty Conference on
Cold Formed Steel Structures 2008, St. Louis, MO, pp. 517–530.
 References 93

Kolarkar, P.N. and Mahendran, M. (2012). Experimental studies of non-load bearing

steel wall systems under fire conditions. Fire Safety Journal, Vol. 53, pp. 85–104.
Lawson, R.M. and Way, A.G.J. (2016). Value Benefits of Light Steel Construction,
Technical Information Sheet P409. Steel Construction Institute, London, UK.
Leng, J., Li, J., Guest, J.K. and Schafer, B.W. (2014). Shape optimization of cold-formed
steel columns with fabrication and geometric end-use constraints. Thin-Walled
Structures, Vol. 85, pp. 271–290.
Magnusson, S.E. and Thelandersson, S. (1970). Temperature-time curves for the
complete process of fire development. A  theoretical study of wood fuel fires
in enclosed spaces. Acta Polytechnica Scandinavica, Civil Engineering and
Building Construction Series, No. 65, CIB/CTF/72/46, Stockholm 1970.
Maia, E., Couto, C., Vila Real, P. and Lopes, N. (2016). Critical temperatures of class 4
cross-sections. Journal of Constructional Steel Research, Vol. 121, pp. 370–382.
MATLAB FILE EXCHANGE, program code, accessed on 2013, https://www.mathworks​
.com/matlabcentral/fileexchange/41085-calculations-of-temperature-through-the-
wall-panel-assembly-exposed-to-fire-from-one-side-1.
McGrattan, K., Hostikka, S., McDermott, R., Floyd, J., Weinschenk, C. and Overholt,
K. (2018). Fire dynamics simulator technical reference guide volume 1:
Mathematical model. NIST Special Publication.
Outinen, J. (1999). Mechanical properties of structural steel at elevated temperatures.
Licentiate thesis, Helsinki University of Technology, Helsinki, Finland.
Peacock, R.D., McGrattan, K.B., Forney, G.P. and Reneke, P.A. (2017). CFAST–
Consolidated Fire and Smoke Transport (Version 7) Volume 1: Technical Reference
Guide. National Institute of Standards and Technology, Gaithersburg, MD.
Pham, C. and Hancock, G. (2013). Experimental investigation and direct strength design
of high-strength, complex C-sections in pure bending. Journal of Structural
Engineering, American Society of Civil Engineers, Vol. 139, pp. 1842–1852.
Rahmanian, I. and Wang, Y.C. (2012). A  combined experimental and numerical
method for extracting temperature-dependent thermal conductivity of gypsum
boards. Construction and Building Materials, Vol. 26, No. 1, pp. 707–722.
Ranawaka, T. and Mahendran, M. (2009a). Distortional buckling tests of cold-formed
steel compression members at elevated temperatures. Journal of Constructional
Steel Research, Vol. 65, pp. 249–259.
Ranawaka, T. and Mahendran, M. (2009b). Experimental study of the mechanical
properties of light gauge cold-formed steels at elevated temperatures. Fire Safety
Journal, Vol. 44, pp. 219–229.
Rusthi, M., Ariyanayagam, A.D., Mahendran, M. and Keerthan, P. (2017). Fire tests of
magnesium oxide board lined light gauge steel frame wall systems. Fire Safety
Journal, Vol. 90, pp. 15–27.
Sakumoto, Y., Hirakawa, T., Masuda, H. and Nakamura, K. (2003). Fire resis-
tance of walls and floors using light-gauge steel shapes. Journal of Structural
Engineering, Vol. 129, No. 11, pp. 1522–1530.
Shahbazian, A. and Wang, Y.C. (2012). Direct Strength Method for calculating distor-
tional buckling capacity of cold-formed thin-walled steel columns with uniform and
non-uniform elevated temperatures. Thin-Walled Structures, Vol. 53, pp. 188–199.
Shahbazian, A. and Wang, Y.C. (2013). A simplified approach for calculating tempera-
tures in axially loaded cold-formed thin-walled steel studs in wall panel assem-
blies exposed to fire from one side. Thin-Walled Structures, Vol. 64, pp. 60–72.
94 References

Shahbazian, A. and Wang, Y.C. (2014a). A performance-based fire resistance design

method for wall panel assemblies using thin-walled steel sections. The Structural
Engineer, Vol. 92, pp. 52–62.
Shahbazian, A. and Wang, Y.C. (2014b). A  fire resistance design method for thin-
walled steel studs in wall panel constructions exposed to parametric fires. Thin-
Walled Structures, Vol. 77, pp. 67–76.
Siahaan, R., Keerthan, P. and Mahendran, M. 2016. Finite element modelling of rivet
fastened rectangular hollow flange channel beams subject to local buckling.
Engineering Structures, Vol. 126, pp. 311–327.
Standards Australia (SA) (2005). Methods for fire tests on building materials, compo-
nents and structures. Fire-resistance tests of elements of building construction.
AS 1530.4, Sydney, Australia.
Standards Australia (SA) (2018). Australia/New Zealand standard AS/NZS 4600 cold-
formed steel structures. Sydney, Australia.
Steel Construction Institute (2019). Fire resistance of light steel framing, SCI
Publication P424, Steel Construction Institute, Ascot, UK.
Stern-Gottfried, J. and Rein, G. (2012). Travelling fires for structural design-part II:
Design methodology. Fire Safety Journal, Vol. 54, pp. 96–112.
Sultan, A.M. (1996). A  model for predicting heat transfer through non-insulated
unloaded steel-stud gypsum board wall assemblies exposed to fire. Fire
Technology, Vol. 32, No. 3, pp. 239–259.
Sultan, M.A., Seguin, Y.P. and Leroux, P. (1998). Results of fire tests on full-scale floor
assemblies. Internal Report, National Research Council of Canada, Ottawa,
Canada.
Takeda, H. and Mehaffey, J.R. (1999). WALL2D: A model for predicting heat transfer
through wood‐stud walls exposed to fire. Fire and Materials, Vol.  22, No.  4,
pp. 133–140.
Wang, Y., Burgess, I., Wald, F. and Gillie, M. (2012). Performance-Based Fire
Engineering of Structures. CRC Press, London, UK.
Wang, Y., Ying-Ji, C. and Lin, C. (2015). The performance of calcium silicate board
partition fireproof drywall assembly with junction box under fire. Journal of
Advances in Materials Science and Engineering, Vol. 125, pp. 1–12.
Ye, J.H. and Chen, W. (2013). Elevated temperature material degradation of cold-
formed steels under steady- and transient-state conditions. ASCE Journal of
Materials in Civil Engineering, Vol. 25, No. 8, pp. 947–957.
Zhao, B., Kruppa, J., Renaud, C., O’Connor, M., Mecozzi, E., Apiazu, W., Demarco,
T. et al. (2005). Calculation rules of lightweight steel sections in fire situations.
Technical Steel Research, European Union.
Index
A insulation performance, 22
joining methods, 20
applications, 2, 11 magnesium oxide board, 39–41
fire resistance
C calculation methods, 7
cavity insulation effect, 34–35
cavity insulation, effects on fire resistance, columns, 48
34–35 external insulation effect, 35–36
floors, 44–47
D general methods of determination, 6–7
method of testing, 25–28
direct strength method methods of enhancement, 18
Australian/New Zealand standard, 84 plasterboard joints, 37–38
equivalent column method, 80–83 requirements, 4, 27
example, 85–87 stud section, 36
walls, 28–33
E fire safety requirements, 3–5
fire tests
effective width method extension of results, 77–78
Australian/New Zealand standard, 83 standard fire resistance test, 25–28
general, 79
elevated temperatures
mechanical properties, steel, 66–68 G
thermal properties, gypsum, 70–73
gypsum plasterboard
thermal properties, insulation, 71
fall-off, 20, 47, 73
external insulation, effects on fire resistance,
heat transfer modelling, 60–63
35–36
joints, effects on fire resistance,
37–38
F thermal properties, 70–73
failure modes
columns, 48 H
floors, 45
walls, 32 heat transfer
fire dynamics boundary conditions, 58–59
general behaviour, 52 computer programs, 56–57
modelling, 55 general principles, 57–58
nominal temperature-time curves, 53 modelling, 58–59
parametric fires, 54–55 simplified method for panels, 60–63
travelling fires, 55
fire protection boards I
calcium silicate board, 39–41
effects on fire resistance, 20, 39–42 interior insulation, thermal properties of,
gypsum board, 19 71–73

95
96 Index

L scope, 9
uniform temperature distribution, 76
limiting temperature method, 76–77 prefabricated structural units, 16–18

M S
methods of building construction, 2 section profiles
modular construction, 16 effects on fire resistance, 22, 36, 46
modular units, 16–18 hollow flange sections, 16
innovation, 14
N mechanical properties, 66–68
thermal properties, 65
non-uniform temperature thin-walled steel sections, 12
design method, 76 steel
simplification, 30 grade, 1
numerical modelling joists, 22–23
heat transfer, 59 mechanical properties, elevated
structural behaviour, 64 temperatures, 66–68
sheathing, 23, 42
P studs, 22–23
thickness range, 1
panel construction, 3
performance based design methods T
general, 75
non-uniform temperature distribution, thermal bowing, 82
76–78 thickness range, 1