Fire safety in Timber buildings

Fire safety
in timber buildings

Technical guideline

Second draft for comments

December 2009.

Major changes from first draft are marked in colours

Not to be referenced

Comments latest by 15 January 2010

to national contact or directly to [email protected]

preferably on the attached template for comments
 Second draft for comments December 2009 – Not to be referenced

Background

The design guide has been developed in the European research project FireInTimber (Fire
Resistance of Innovative Timber structures). Project partners are

Country Partners
Sweden SP Trätek Birgit Östman, Project coordinator
Jürgen König, Joachim Schmid
Finland VTT Esko Mikkola, Tuula Hakkarainen
Germany TUM Technische Universität München René Stein, Norman Werther, Stefan Winter
DGfH Matthias Krolak
France BPU Blaise Pascal University Abdelhamid Bouchair
CSTB Dhionis Dhima
Norway TreSenteret, Wood Centre Harald Landrø
UK BRE Building Research Establishment Julie Bregulla
Austria HFA Holzforschung Austria Martin Teibinger
UIBK Innsbruck University Hans Hartl
TUW Technische Universität Wien Karin Hofstetter
Switzerland ETH Zurich Andrea Frangi
Estonia Resand Alar Just
European BWW Building With Wood Dieter Lechner, Die Holzindustrie
industry Jan Lagerström, Swedish Forest Industries
Federation, Project Chairman

The supporting public funding organisations are

Vinnova and Formas (SE), Tekes (FI), Forschungszentrum Jülich (DE), Ministère de l´Agriculture
(FR), Norges forskningsråd (NO), Forestry Commission (UK), FFG (AT) and EMTL (EE).

2
 Second draft for comments December 2009 – Not to be referenced

Preface

This design guide presents information for architects, engineers, educators, regulatory
authorities and building industry for the fire safe use of timber structures and wood products
in buildings. It aims at providing the highest scientific knowledge with regard to fire safety on
the European level. The guidance covers the use of design codes (such as Eurocode 5),
European standards, practical guidance and examples for fire safe design and principles of
performance based design.
The design guide is focusing on structural fire protection by proving latest detailed guidance
on separating and load-bearing functions of timber structures under standard fire exposure. It
includes information on reaction to fire performance of wood products according to the new
European standards. The importance of proper detailing in building design is stressed by
practical solutions. Active measures of fire protection are presented as important means in
fulfilling fire safety objectives.
The design guide has been developed in the European research project FireInTimber (Fire
Resistance of Innovative Timber structures). Leading experts and researchers from ten
European countries have been participating and guarantee its quality and relevance. The
project has been sponsored by national funding organisations within the WoodWisdom-Net
framework and by industry through the European network BWW Building With Wood.

Birgit Östman
Project coordinator

3
Second draft for comments December 2009 – Not to be referenced

4
 Second draft for comments December 2009 – Not to be referenced

Content
1. Timber buildings 7
2. Fire safety in buildings 9
2.1 Fire action and fire load
2.2 Fire scenarios
2.3 Fire safety objectives
2.4 Means of fire protection and evacuation safety
2.5 Means to fulfil fire safety objectives
3. European requirements 21
3.1 Essential requirements and construction products directive/regulations
3.2 European harmonisation of fire classification systems
3.3 Structural Eurocodes
3.4 Implemetation of essential requirements for fire safety in national regulations 29
4. Wood products as linings, floorings, claddings and facades
4.1 Reaction to fire classification in end use conditions
4.2 Improved reaction to fire performance
4.3 Wood claddings with fire protection ability
4.4 Wooden facades
5. Separating timber structures 49
5.1 General
5.2 Basic requirements on fire compartmentation
5.3 Calculations methods
5.4 The EN 1995-1-2 design method
5.5 Improved design method for separationg function of timber constructions
5.6 Examples
6. Load bearing timber structures 67
6.1 General
6.2 Structural stability
6.3 Materials
6.4 Charring of timber and wood-based panels
6.5 Mechanical resistance
6.6 Structural elements
6.7 Calculation vs full-scale testing
7. Timber connections 121
7.1 General
7.2 Design method according to EN 1995-1-2
7.3 Other methods proposed for the design of timer connections
8. Fire stops, service installations and detailing in timber structures 133
8.1 Detailing in timber structures
8.2 Fire stops and element joints
8.3 Building service installtions
9. Novel products and their implementation 155
9.1 Performance requirements and potential solutions
9.2 Procedure for introducing novel products
10. Active fire protection 167
10.1 Fire detection and fire alarm systems
10.2 Fixed fire suppression systems
10.3 Smoke control
10.4 Fire safety design with sprinklers
11. Performance based design 177
11.1 Performance based requirements
11.2 Fire risk assessment
11.3 Example on probabilistic approach to assess fire safety
12. Execution and control 195
12.1 Control of workmanship
12.2 Fire safety at building sites
5
 Second draft for comments December 2009 – Not to be referenced

Description of terms 197

List of terms in different languages 199
Additional information and references 200

6
Chapter 1 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

1 Timber buildings
This chapter gives a short introduction to the long term use of timber buildings and their
renaissance during recent years as a sustainable solution to reach environmental goals.
It also introduces effects of history and traditions and new possibilities to build multi
storey timber buildings based on new knowledge for fire safety design.

Timber has been a favourite construction material from the dawn of civilization because of its
abundance, high stiffness and strength-to-weight ratios, and the relative simplicity with which
it can be fashioned to shape. Because of the low energy use and the low level of pollution
associated with the manufacture of timber structures, the environmental impact is generally
small. With proper management, there is a potential for a continuous and sustainable supply
and use of timber as a building material in the future. It is therefore not surprising that timber
structures are becoming an important element in sustainable and economic development, and
have attracted worldwide attention in recent years.

But what of fire? As a combustible material, timber burns on the surface, releases energy and
thus contributes to fire propagation and the development of smoke in the event of fire. The
combustibility of timber is one of the main reasons for most building standards strongly to
restrict the use of timber as a building material. Fire safety is an important contribution to
feeling safe, and an important criterion for the choice of materials for buildings. The main
precondition for increased use of timber for buildings is adequate fire safety. World-wide,
several research projects on the fire behaviour of timber structures have been conducted over
about the last decade, aimed at providing basic data and information on the safe use of timber.
Novel fire design concepts and models have been developed, based on extensive testing. The
current better knowledge in the area of fire design of timber structures, combined with
technical measures, especially sprinkler and smoke detection systems, and well-equipped fire
services, allow a safe use of timber in a wide field of application. As a result, many countries
have started to revise fire regulations, thus permitting greater use of timber.

Fire test and classification methods have recently been harmonised in Europe, but regulatory
requirements applicable to building types and end uses remain on national bases. Although
these European standards exist on the technical level, fire safety is governed by national
legislation, and is thus on the political level. National fire regulations will therefore remain,
but the new European harmonisation of standards will hopefully speed up also the regulations.

Major differences between European countries have been identified, both in terms of the
number of storeys permitted in timber structures, and of the types and/or amounts of visible
wood surfaces in interior and exterior applications /1.1/. Several countries have no specific
regulations, or do not limit the number of storeys in timber buildings. However, eight storeys
are often used as a practical and economic limit for the use of timber structures. This limit may
be higher for facades, linings and floorings, since these applications may also be used in, for
example, concrete structures.

7
Chapter 1 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 1.1
Restrictions of the use of timber structures for higher buildings, set by national prescriptive
regulations, have been eased in Europe over the last decades. A further increase in permitted
use is expected.

Reference
1.1 Östman B, Rydholm D: National fire regulations in relation to the use of wood in European and
some other countries, Trätek Publication 0212044, 2002.

8
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

2 Fire safety in buildings

Note the highlight colours:
yellow: editorial/small change
green: new text, not language checked
This chapter gives an overview of the basic concepts for fire safety in buildings. It
presents information on fire actions, fire loads, fire scenarios and fire safety objectives.
Means to fulfil the fire safety objectives are given to be used in all buildings and as a
basis for the design solutions in this guideline.

Knowledge of the basic behaviour of fire, occupants and buildings during a fire is an
important precondition for the development of a successful fire safety strategy. Fire safety
strategies need to address life and property hazards in buildings and neighbourhoods,
including environmental aspects. In practice, the strategies chosen may vary considerably
depending on such considerations as emphasis on passive or active fire protection, type and
use of the building, etc.

To implement strategies, guidance on tools to be used and how to apply them are needed. The
guidance is given in national regulations, which present rules in terms of fire classes and
numerical values to assess fire safety and/or performance requirements. In addition, tools for
calculating such aspects as performance of structures (Eurocodes) and simulating the spread of
fire and evacuation of occupants are needed.

2.1 Fire action and fire load

Figure 1 shows the development of a fire in a typical room [1]. After ignition, fires can grow
very rapidly, very slowly (as a smouldering fire), or even self-extinguish, depending on the
arrangement of combustible materials in the vicinity of the ignition source, on the type and
amount of combustible material, and on the geometry, dimensions and ventilation of the room.
Growing fires not controlled by fire-fighting actions may lead to rapidly rising temperatures
and to flashover, where all unprotected combustible materials burn.

Figure 2.1. Typical stages of fire development [1]

9
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In all buildings movables (furniture, equipment, goods, etc.) form the essential fire load at
least for the early, developing phase of fire when safety of life is most emphasized. The
combustion process releases energy, gases and smoke. While gases and smoke are the main
killers in a fire (approx. 80 % of fire fatalities are due to toxic fire effluents), heat is the
primary reason for the damage to the structures of a building. As the mechanical and thermal
properties of building materials change with increasing temperatures, knowledge of the time-
temperature development during a fire in a building is most important for the structural fire
analysis.

Fire load is defined as the energy (total heat release) per square metre of floor area of the
combustible material present within the internal bounding surfaces of a room, compartment or
building. Thus it consists of the energy of the structural building elements and the energy of
the contents. National regulations may also categorise building types according to expected
fire load levels. In principle, fire protection by controlling the amount of fire load is possible,
but in practice the difficulties depend on the energy of the contents of the building, which is
mostly the main source of fire load, and the magnitude of which is not controlled by
regulations (only assumed categories may be used). Furthermore, it should be taken into
account that the energy from the structural elements is usually released much more slowly
than energy from the contents of the building, because of the characteristics of the combustible
building elements (e.g. charring of timber extends the duration of energy release to a long
period), or because the building elements are protected for defined periods.

In practice, the ignition of timber building elements can be retarded by two possible measures:
- Encapsulation of building elements (the most often-used method), e.g. by non-combustible
lining materials
- Impregnation or coating of building elements with fireproof agents (analogous to wood
preservatives).

Treatment and working with fireproof impregnating and coating systems requires expert
knowledge and must be carried out carefully. The long-term behaviour and durability of
impregnating and coating systems (especially for exterior exposure) are under investigation
and development.

Fire protection of timber by means of encapsulation of building elements has proved

successful. The building elements and other combustible materials (e.g. insulation) are
protected with fireproof claddings, which encapsulate the building elements and protect them
from potential ignition. The charring temperature of structural timber of commercial quality is
about 300 oC. The encapsulating effect of fireproof claddings is tested in accordance with
EN 14135 [2]. A typical fireproof cladding is that of gypsum plaster boards. The fire
protection performance of coverings (K1 or K2), as specified in accordance with EN 13501-2
[3] is effective as long as the average temperature rise over the whole exposed surface of the
protected building element is limited to 250 K, and the maximum temperature rise at any point
on that surface does not exceed 270 K.

2.2 Fire scenarios

To describe fire action with simplified models, several nominal fire curves have been
proposed in codes for use in the design process. Nominal fire curves provide a simple
relationship for the temperature of the gases in a compartment as a function of time. However,

10
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

they represent the phase of the fully developed fire, and the significant amount of time that
sometimes elapses from the beginning of a fire until the fully developed fire state is not
included in the curves. In addition, the cooling or decay phase of the fire is not explicitly taken
into account, and the nominal fire curves increase continuously with time. (The effect of the
cooling phase may, however, implicitly be taken into account by fire resistance requirements).
The most frequently applied fire curve is the ISO 834 [4] standard fire curve, which is also
used as the main exposure condition in European EN standards for fire resistance.

More realistic models of fires are given by the parametric fire curves, which take into account
the most important parameters for temperature development, namely:
- the fire load (amount, type and arrangement of combustible material)
- the ventilation conditions in the room
- the thermal properties of the enclosures, and
- the fire fighting action.

2.2.1 Nominal temperature-time curves

For design by classification, nominal temperature-time curves are used for a specified period
of time, without any cooling phase. The period of time, normally specified by national
regulations, is assumed implicitly to take into account features such as risk assessment,
measures of active fire protection and the cooling phase. Three nominal temperature-time
curves exist (see EN 1991-1-2 [5] Clause 3.2, and Figure 2.2):
• Standard temperature-time curve
• External fire curve
• Hydrocarbon curve
For fire design of timber structures, mainly the standard temperature-time curve has hitherto
been used; and so almost all knowledge is based on this fire curve.

Figure 2.2. Nominal temperature-time curves as given in EN 1991-1-2 [5]

11
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

2.2.2 Parametric fire curves

As an alternative to nominal temperature-time curves, fire models have been used in the fire
design of steel and concrete structures for temperature and structural analysis for the full
duration of the fire, including the cooling phase. A simplified fire model for compartment fires
is based on parametric fire curves. These are calculated with simplified formulas developed
for limited boundary conditions, as given (for example) in [5]. For a more comprehensive and
more detailed analysis, computer simulations may be used, e.g. multi-room zone models or
computational fluid dynamics models. With such simulations, more complicated environments
and more detailed parameters can be taken into account than with the simplified parametrical
fire curves.

A natural fire scenario describes the complete development of a fire, including the cooling or
decay phase, as a relationship between the gas temperature in the fire compartment and time as
a function of fire load density, ventilation conditions (opening factor) and thermal properties
of the enclosing walls and floors: see EN 1995-1-2 [6], Annex A. Annex E of EN 1991-1-2 [5]
gives the design value of the fire load density, dependent on various parameters such as the
risk of fire activation due to the size of compartment and type of occupancy, as well as various
measures of active fire protection such as sprinklers, detection, automatic fire alarm or fire-
fighting action. The application of these parametric fire scenarios may be limited to specific
European countries in accordance with relevant national annexes to EN 1991-1-2 [5].

Parametric fire curves can be used in building design utilising fire safety engineering.
However, only limited information is available concerning the performance of timber
structures in natural fires. Charring of glued laminated timber was studied by Hadvig [7].
Charring and resistance in bending of glued laminated timber was studied by Toft Hansen and
Bolonius Olesen [8]. The rules for charring and strength given in EN 1995-1-2, Annex A, are
based on the results of this research. As an alternative, EN 1995-1-2 [6] permits the use of
advanced calculation methods. However, thermal properties of timber given in the code are
valid only for standard fire exposure, which means that advanced calculation methods cannot
be used for parametric fire curves, unless the thermal properties of timber are modified
accordingly [9].

2.3 Fire safety objectives

In European countries, the main principles forming the basis for fire safety regulations (both
prescriptive and performance-based codes) of buildings are as follows:
• Occupants shall be able to leave the buildings or be rescued
• Safety of rescue teams shall be taken into account.
• Load-bearing structures shall resist fire for the required minimum duration of time
• Generation and spread of fire and smoke shall be limited
• Spread of fire to neighbouring buildings shall be limited

These principles have been formulated in order to be able to fulfil the following objectives:
• Safety of life
• Safety of occupants
• Safety of rescue teams
• Loss prevention
• Structures of building

12
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

• Contents of building
• Uninterrupted operation
• Public image
• Environmental protection
• Release of hazardous substances.

The different levels (concepts, objectives, strategies, systems and components) of establishing
fire safety are illustrated in Figure 2.3.

Fire Safety
Concept

Life safety Protection of

Objective the business

Evacuation Suppression Containment

Strategy

Systems Alarm and Sprinkler Building

detection element

Alarm Detectors Sprinkler Building

Components Sounders Heads Products

Figure 2.3. Fire safety from concepts to components

Achieving absolute safety is impossible, regardless of the construction materials and systems
involved. The level of acceptance must be quantified by the authorities or with regard to
financial losses with the owner or the insurance companies. These objectives can be reached
with different fire safety strategies that take into account the type and use of the building. Fire
safety strategies consist of comprehensive structural (passive), technical (active) and
organisational measures to fulfil the predefined fire safety objectives and acceptance criteria.
The most efficient fire safety concept is arrived at by comparing different options. The criteria
for the evaluation of different options that fulfil the predefined fire safety objectives at the
same level are the total cost of fire safety measures and further aspects such as flexibility,
limitation of use, architectural design etc. Different types of strategies are employed to meet
these objectives. Tools available for use to prove/demonstrate the function of a building in
case of fire are briefly described below.

13
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

2.4 Means of fire protection and evacuation safety

This design guide provides detailed design information on passive fire protection by providing
the latest detailed guidance on load-bearing and separating functions of timber structures
under standard fire exposure, and by providing information on the effects of fire loads.
However, active means of fire protection are also very essential in fulfilling fire safety
objectives, especially concerning the safety of life. The main means of passive and active
protection are summarised in the following.

2.4.1 Passive fire protection

Structural (passive) fire protection is provided by building elements which are designed and
constructed in such a way as to maintain their load-bearing and/or separating function for the
required minimum duration of fire exposure. Building elements with separating function form
the boundaries of fire compartments, and must prevent the spread of fire to or from another
part of the same building, or an adjoining building. The generally employed principle of
compartmentation is to subdivide buildings into areas of manageable risk, to give enough time
for escape and rescue, and to provide fire separation for adjoining buildings. In a fully
developed fire within the room of origin, the safety of occupants is no longer an issue, and a
total loss of the contents is accepted.
Fire rarely spreads from one area of a building to another by destroying or burning through a
wall or a floor. It usually spreads via common concealed spaces in walls or ceilings; heating,
ventilation and air conditioning duct systems; holes that have been cut through walls for
electrical or communications cables, doors left open, etc. Thus, in order to ensure the integrity
of a fire compartment, it is necessary that every opening or penetration through its walls, floor
and ceiling, large openings such as doorways and small service penetrations (wiring, piping,
and ducts), is protected to resist the spread of fire. Fire can also spread from the compartment
through windows and along facades and/or to eaves and roof structures. Accurate detailing
and quality assurance during construction are therefore essential for an effective structural
passive fire protection.
Fire resistance periods laid down in national regulations generally range from 30 to 240
minutes, depending on the height and use of the building, and the location of the building
element within the building. Traditionally, the fire resistance of an element has been
determined by testing, i.e. by subjecting the building element to a “fire resistance” test. This is
a furnace test which exposes a representative piece of structure to a prescribed heat exposure,
with the building element being required to remain structurally load-bearing (stable), to shield
heat (insulation) and to prevent the passage of smoke and hot gases (integrity). As an
alternative to fire testing, verification by calculation is increasingly becoming more important:
see, for example, EN 1995-1-2 [6]. However, there are no calculation methods for the integrity
criterion.

Summarising, passive fire protection measures include:

• Adequate compartmentation to inhibit the spread of heat, smoke and gases
• Limitation of fire compartment size
• Control of flammability/combustibility of wall linings – particularly on escape routes
• Control of the spread of smoke
• Provision of protected escape routes, and

14
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

• Provision of adequate thermal insulation, stability and structural performance.

The design of timber elements to provide a separating function is described in detail in
Chapter 5, Load-Bearing Timber Elements and Connections in Chapters 6 and 7. Detailing in
Timber Structures (fire stops, penetrations, service installations, etc.) is described in
Chapter 8.

2.4.2 Active fire protection

As a complement to passive fire protection, active fire protection systems can be used to
increase the fire safety level in buildings for special groups (life safety), e.g. elderly or
handicapped persons, for industrial or public buildings, or for fulfilling insurance
requirements. Active fire protection systems can also be used in some cases as means of
alternative fire design of buildings, allowing for a compensation of some passive fire
protection requirements.

Fire detection and alarm systems are used to provide means of detecting fire and giving
warning. It is essential that an outbreak of fire should be detected at an early phase so that the
occupants are alerted and the emergency fire action plan is implemented as soon as possible.

Portable fire extinguishers and hose reels can be used for initial fire-fighting. If staff have been
trained to use them, the risks to persons from fire can be significantly reduced.

Water sprinkler and high pressure systems are widely used in the process industries for
protection of storage vessels, process plant, loading installations and warehouses. The purpose
of an automatic fire-fighting system is to extinguish the fire or at least control it until the fire
service can take over. Particular attention is paid to the potentials of residential sprinklers, i.e.
simplified sprinkler systems that can be connected to the tap water system.

Smoke extraction is used to improve visibility and reduce temperature. The simplest means are
to use windows and doorways, but more effective results can be obtained by using special
smoke vents which can be either manually or automatically operated.

Effectiveness of fire service actions will depend on such factors as response time (dependent
on distance and complexity of the site) and resources available.

In summary, active fire protection measures include:

• Provision of alarm systems – audible, visual etc.
• Provision of automatic fire detection
• Indication of the location of a fire on a visual display, mimic diagram etc.
• Provision of smoke/heat venting & exhaust control systems to extend escape time
• Closing of smoke dampers in ducts, closing of doors to provide fire compartments etc.
• Release of access control systems on doors, gates, turnstiles etc. to facilitate escape
• Provision of automatic fire suppression systems

The positive effect of active fire protection measures may be taken into account in the
requirements, e.g. in the fire resistance time according to the standard fire curve. Means of
active fire protection are described in more detail in Chapter 10.

15
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

2.4.3 Evacuation safety

The distinction between active and passive fire protection measures is not always clear-cut, as
some measures are a combination of both, such as smoke control doors which remain open
under normal circumstances but are automatically closed on receiving a signal from a
detecting device, or protective coatings which intumesce and become effective only when a
fire occurs. The doors and the coatings in the two cases remain part of the passive defence
system, and it is only the detection and operating mechanism that is an active measure.
Evacuation safety is also based on both passive and active elements. The main principles are
briefly described below.

Emergency evacuation means the immediate and rapid movement of people away from the
threat or actual occurrence of a hazard such as fire. Evacuation plans are developed to ensure
safe and efficient evacuation of all occupants of a building. Proper planning will use multiple
exits and alarm/guidance systems to ensure complete evacuation. These may include alarm
signals that use both aural and visual alerts. Personal situations or factors that may affect an
individual's ability to evacuate must also be considered.

The main requirement is that safe evacuation of a building must be possible in the event of fire or
other emergency. There must be a sufficient number of appropriately located exits which are
sufficiently spacious, easily passable and clearly marked, so that the time to evacuate the
building does not cause danger. The exits must lead outside to ground level or to a place which
is safe in case of fire.

Summarising, the following types of requirements exist in national regulations:

• The greatest permitted distances to the nearest exits are regulated at national levels
• Number of exits - the main principle is that there must be at least two separate, appropriately located
exits from each evacuation area
• Dimensions of exits - the minimum width and height are specified
• Fire-separation and constructions of exits - an exit is often considered as a fire
compartment of its own, and the use of combustible materials on surfaces is controlled
• In general, doors must open in the escape direction of the passage, must be easy to open in
an emergency situation, and exits must be provided with emergency/exit lighting
• The exits and the access to them shall be marked with symbol signs and evacuation floor
plans can be used to show exit routes.

Traditionally, the evacuation capacity of a building is designed according to a set of relatively

simple design criteria (as described above) concerning the required width and length of
evacuation routes, complemented by detailed requirements in the national building codes.

The simplest methods for evacuation time estimation are manual calculation methods, in
which pre-movement and movement times are taken from table values. However, the
increased size and complexity of buildings make new demands on the methods used for fire
and evacuation simulations. It would not, for example, be possible to build large multi-purpose
centres using simple design rules. Instead, such places are usually designed using the
performance-based design method, in which the safety of the design is studied as an entire
system. The evacuation simulations must be able to model the dynamics of large and high
density crowds. The models should also consider the evacuees’ decision-making processes
behind the reaction times, pre-movement times and the exit route selection during the actual

16
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

evacuation. For an individual person, escape and evacuation from a building during a fire is
always an extraordinary situation.

It is recommended that the design of safe evacuation of a building should be entrusted to

professionals who can first choose the level of complexity of the method needed for the particular
case, and secondly are capable of using the more advanced methods (if needed).

2.5 Means of fulfilling fire safety objectives

Prescriptive fire regulations give detailed rules leading to standard concepts. Standard
concepts are considered to give an acceptable level of fire safety. The use of standard fire
safety concepts does not require any further verification of the fire risk, and guarantees
acceptance by the fire authorities. However, performance-based design concepts using detailed
probabilistic and engineering-based analysis optimise the fire safety measures [10, 11],
especially for unusual buildings for which no experience exists.

National fire safety requirements do differ a lot in details and to some extent also in safety levels.
In most countries, there are two alternative ways of satisfying the national requirements:
– The building is designed and built in accordance with standard concepts by applying detailed
rules with regard to fire classes and criteria provided by the regulations and guidelines
(prescriptive code);
– The building is designed and built based on design fire scenarios, which must cover the
conditions that are likely to occur in the building (performance-based code).
Note that, at national levels, these principles may be applied to parts of a building. In some
countries (e.g. UK) regulations specify the level of performance expected, e.g. 60 minutes fire
resistance, but do not prescribe ways of meeting this requirement. The regulations provide a
"deemed to satisfy" solution that provides this performance level, but the user can use other
solutions if they so wish.

2.5.1 Prescriptive codes

The use of standard fire safety concepts as set out in prescriptive regulations does not require
any further verification of the fire risk and guarantees acceptance by the fire authorities. For
most small and medium-sized standard residential and office buildings it is cost-effective to
base design on prescriptive fire regulations. Structural and product requirements are simply taken
from tabulated values and performance of the planned structures and products are taken from
manufacturer’s data sheets, approvals, tabular values or are calculated according to the Eurocodes
(as modified by national Annexes).

2.5.2 Performance-based codes and design tools

A performance-based approach to fire safety design relies on the use of fire safety engineering
principles, calculations and/or appropriate modelling tools to satisfy building regulations.
Building practitioners have added flexibility in the application of fire safety by having a
choice of using the performance-based approach, the prescriptive approach or a combination

17
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

of both. Fire safety engineering is more important in designing as buildings get larger and/or
risks for people and/or property increase.

The rapid development of modern building technology in recent decades has resulted in
unconventional structures and design solutions, with the physical size of buildings continually
increasing. The interior design of many buildings with large light shafts, patios and covered
atriums inside buildings, connected to horizontal corridors or malls, introduces new risk
factors concerning the spread of smoke and fire. Past experience or historical precedents
(which form the basis of current prescriptive building codes and regulations) rarely provide
the guidance necessary to deal with fire hazards in unconventional buildings [12, 13].

Instead of prescribing exactly which protective measures are required (such as prescribing a
number of exits for evacuation purposes), the performance of the overall system is presented
against a specified set of design objectives (such as stating that satisfactory escape must be
effected in the event of fire). Fire and evacuation modelling, together with experimental
evidence, is used to assess the effectiveness of the protective measures proposed in the fire
safety design of a building.

Quantitative analysis methods of fire safety are either probabilistic or deterministic. In a

probabilistic analysis, the analysis multiplies the probabilities for each sequence of events and
ends up with the quantity of risk for individual scenarios. This method of analysis requires a
lot of knowledge and resources. However, its use is increasing, especially in large and high-
risk applications. In a deterministic analysis, a detailed analysis is made only for one or a
small number of scenarios, in order to assess possible consequences.

Acceptance criteria may be comparative as well as absolute. In using comparative criteria, the
level of safety is based on prescriptive regulations. The results of analysis are compared with
the prescribed acceptable solutions. Absolute criteria are, for example, factors such as the
threshold values for maximum heat flux and maximum temperatures to which humans could
be exposed. Statistics, experimental data, experience, etc. are needed to determine the criteria.

The increase in understanding fire processes and their interaction with buildings and humans
has resulted in development of a variety of models. The enclosure fire models can roughly be
divided into three categories; CFD models; zone models and manual-calculation models.
Similarly, models for simulating the escape of persons from buildings range from simple
manual-calculation models to complex interactive computer programs.

More guidance on performance-based design and fire risk analysis are given in Chapter 11.

References
1. Buchanan, A.H. Fire performance of timber construction, Progress in Structural
Engineering and Materials, Volume 2, 2000.
2. EN 14135. Coverings — Determination of fire protection ability. European Committee for
Standardization, Brussels, 2004.
3. EN 13501-2. Fire classification of construction products and building elements - Part 2:
classification using data from fire resistance tests, excluding ventilation services.
European Committee for Standardization, Brussels, 2007.

18
Chapter 2 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

4. ISO 834-1: Fire-Resistance Tests - Elements of Building Construction - Part 1: General

Requirements, International Organization for Standardization, 1999.
5. EN 1991-1-2: Eurocode 1: Action on structures Part 1-2: General action – Actions on
structures exposed to fire, European Committee for Standardization, Brussels, 2002.
6. EN 1995-1-2:2004 Eurocode 5: Design of timber structures – Part 1-2: General –
Structural fire design. European Committee for Standardization, Brussels, 2004.
7. Hadvig, S., Charring of wood in building fires. Technical University of Denmark, Lyngby,
1981.
8. Toft Hansen, F., Bolonius Olesen, F., Full-scale tests on loaded glulam beams exposed to
natural fires. Department of Building Technology and Structural Engineering, Aalborg
University, Aalborg, Denmark, 1992.
9. König, J, Effective thermal actions and thermal properties of timber members in natural
fires. Fire Mater.2006; 30:51-63.
10. Fontana, M., Maag, T. Fire risk assessment based on Bayesian networks, Proceedings of
the 5th International Conference on Performance Based Codes and Fire Safety Design
Methods, Luxembourg, 6-8 October 2004, Society of Fire Protection Engineers, Bethesda,
MD, USA.
11. Joyeux, D.; Bonnot, S.; Hietaniemi, J.; Korhonen, T. Risk-based fire safety engineering
approach applied to a public building, Proceedings of the 5th International Conference on
Performance Based Codes and Fire Safety Design Methods, Luxembourg, 6-8 October
2004, Society of Fire Protection Engineers, Bethesda, MD, USA.
12. Thureson, P., Sundström, B., Mikkola, E., Bluhme, D., Steen Hansen, A. & Karlsson, B.
The use of fire classification in the Nordic countries – Proposals for harmonisation. Borås:
SP Technical Research Institute of Sweden, 2008. 73 p. (SP Report 2008:29.)
13. Dehne, M., Pape, H., Kruse, D., Krolak, M., Brandschutzkonzepte für mehrgeschossige
Gebäude und Aufstockungen. INFORMATIONSDIENST HOLZ, Bonn (2005).

19
Chapter 3 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

3. European requirements
This chapter gives an overview of the new European requirements on fire safety in
buildings based on the Construction Products Directive (CPD) and its essential
requirements. These requirements are mandatory to be used in all countries. They
include the classification systems for reaction to fire of building products, fire resistance
of structural elements, external fire performance of roofs, fire protection ability of
claddings and structural Eurocodes. These requirements are applied to wood products
and timber structures in the next chapters.

To assure fire safety in buildings, a new European system including product standards,
performance classes in case of fire, testing and calculation standards for fire performance and
national fire regulations has been introduced. The European standards for fire safety in
buildings deal mainly with harmonised methods for verification of performance.

3.1 Essential requirements and construction products

directive/regulation
Building regulations are generally being altered towards functional or performance criteria,
rather than being prescriptive. This development was accelerated by the Construction Products
Directive (CPD), which was adopted in 1988. It contains six essential requirements, one of
which is safety in the case of fire. The CPD requirements on fire safety are that structures must
be designed and built such that, in the case of fire:
- load-bearing capacity can be assumed to be maintained for a specific period of time;
- the generation and spread of fire and smoke is limited;
- the spread of fire to neighbouring structures is limited;
- occupants can leave the building or be rescued by other means;
- the safety of rescue teams is taken into consideration.
These essential requirements are implemented and further detailed in European
standardisation: see Figure 3.1.

CPD and its essential requirement on fire safety in the case of fire
Interpretative documents (ID) with classes for fire performance

Testing and classification Calculation rules

CEN TC 127 CEN TC 250 / SC 5
Fire safety in buildings Structural Eurocodes / Timber

Structures Products

Figure 3.1 Principles for European fire standards for building products

21
Chapter 3 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

3.2 European harmonisation of fire classification systems

3.2.1 Reaction-to-fire performance - Building products
A new classification system for the reaction-to-fire performance of building construction
products has been introduced by the CPD. It is often called the Euroclass system, and consists
of two sub-systems, one for construction products excluding flooring materials, i.e. mainly
wall and ceiling surface linings, see Table 3.1, and another similar system for flooring
materials. Both sub-systems have classes A to F, of which classes A1 and A2 are non-
combustible products. The new system will replace the present national classification systems,
which have created obstacles to trade. The new European classification system for reaction-to-
fire performance is based on a set of EN standards for different test methods (EN 13823, EN
ISO 11925-2 and EN ISO 9239-1) and for classification systems (EN 13501-1). Three test
methods are used for determining the classes of combustible building products, see Table 3.2.
The methods are illustrated in Figure 3.2. Methods EN ISO 1182 (non-combustibility) and EN
ISO 1716 (calorific potential) are also used for class levels A1 and A2.

Table 3.1 Overview of the European reaction-to-fire classes for building products excl. floorings
Euro Smoke Burning Requirements FIGRA Product examples
class class droplets class according to
Non SBI Small W/s
comb flame
A1 − − x − − − Stone, glass
A2 s1, s2 or s3 d0, d1 or d2 x x − ≤ 120 Gypsum boards (thin paper),
mineral wool
B s1, s2 or s3 d0, d1 or d2 − x x ≤ 120 Gypsum boards (thick
paper), fire retardant wood
C s1, s2 or s3 d0, d1 or d2 − x x ≤ 250 Coverings on gypsum
boards, fire retardant wood
D s1, s2 or s3 d0, d1 or d2 − x x ≤ 750 Wood, wood-based panels
E − − or d2 − − x − Some synthetic polymers
F − − − − − − No performance determined
SBI = Single Burning Item, EN 13823, main test for the reaction-to-fire classes for building products;
FIGRA = Fire Growth Rate, main parameter for the main fire class according to the SBI test.

Table 3.2 European test methods used for determining the reaction-to-fire classes of combustible
building products
Test method Construction products Floorings Main fire performance measured
excl. floorings and used for the classification
Small flame test X X Flame spread within 60 or 20
EN ISO 11925-2 seconds
Single Burning Item X − - FIGRA, FIre Growth RAte;
test, SBI EN 13823 - SMOGRA, SMOke Growth RAte;
- Flaming droplets or particles
Radiant panel test − X - CHF, Critical Heat Flux;
EN ISO 9239-1 - Smoke production

22
Chapter 3 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 3.2. Three main test methods for the reaction-to-fire performance relevant for wood products:
the SBI Test (Single Burning Item Test), EN 1382, Small Flame Test, EN ISO 11925-2 and the Radiant
Panel Test for floorings, EN ISO 9239-1.

3.2.2 Fire resistance - Structural fire performance

Fire resistance means that structural elements, e.g. wall elements, must withstand a fully developed fire
and fulfil certain performance requirements. If the fire exposure is in accordance with the so called
standard time-temperature curve, the performance requirements are load bearing capacity (R), integrity
(E) and insulation (I), see Figure 3.3. The standard fire curve describes a fully developed fire defined in
EN 1363-1 and ISO 834-1 and is referred to in almost all national building codes. The building
elements are expected to withstand the fire exposure for a specified period of time, e.g. 60 minutes.
Timber structures can achieve high fire resistance, e g REI 60, REI 90 or even higher.

Figure 3.3. Performance criteria

for fire resistance. They are used
together with a time value, e.g.
REI 60 for an element that
maintains its load bearing and
separating functions for
60 minutes.

Fire resistance of a building element can be verified by testing or using calculation methods. Several
European test standards for fire resistance have been developed in recent years and are more detailed
than the earlier ones. Standard EN 1363-1 presents the general requirements for fire resistance tests and
the requirements appropriate to each element e.g. a wall, a floor etc. are given in the specific test
methods. Standard EN 1363-2 includes other fire exposures than the standard temperature-time curve.
Fire resistance classes of construction products and building elements are defined in the classification
standard EN 13501-2 and are related to the standard temperature-time curve. An overview of
performance classes according to EN 13501-2 is given in Table 3.3. Most classes may be achieved by
timber structures.

23
Chapter 3 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 3.3 Overview of the European performance classes for the fire resistance of building elements
according to EN 13501-2.
Building Load bearing, Separating, Insulating, Time, Test method
element R E I min
EN 1363-1, EN 1364-1 or
Wall elements x x x 15-360
EN 1365-1
EN 1363-1, EN 1364-2 or
Floor elements x x x 15-360
EN 1365-2
Beams x - - 15-360 EN 1363-1, EN 1365-3
Columns x - - 15-360 EN 1363-1, EN 1365-4
Balconies and
x - - 15-360 EN 1363-1, EN 1365-5
walkways
Stairs x - - 15-360 EN 1363-1, EN 1365-6
Doors and
shutter - x x 15-240 EN 1634-1, EN 1634-3
assemblies

For doors and shutter assemblies, there are particular performance characteristics (I1 and I2).
There are also additional performance characteristics e g for mechanical action (M), Radiation
(W), self-closing (S) and smoke leakage (S), see EN 13501-2.

3.2.3 External fire performance – Roofs

There are four alternatives for testing and verifying the external fire performance of roofs in
the new European system, see Figure 3.4:
1. Method with burning brands
2. Method with burning brands and wind
3. Method with burning brands, wind and supplementary radiant heat
4. Two-stage method incorporating brands, wind and supplementary radiant heat

The four test methods originate from different European countries, and no harmonisation has
been possible so far. One or both of the following hazard conditions are considered:
- Fire spread at the surface and/or immediately below the roof covering
- Penetration of fire through the roof.
The classification of external fire performance of roofs is specified in EN 13501-5. An
overview is presented in Table 3.4.
When using wood products in roofs, the main concerns are the performance requirements for
the underlayers of the roof covering if wood products are not used as the outermost layer.

24
Chapter 3 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Test 1 (with burning brands) Test 2 (with burning brands and wind)

Test 3 (with burning brands, wind and Test 4 (two stages, with burning brands, wind and
radiant heat) radiant heat)

Figure 3.4. The four test methods for the fire performance of roofs according to ENV 1187.

Table 3.4 Overview of the European performance classes and requirements for roof elements in
accordance with EN 13501-5
Class Performance criteria Test method
Fire spread, Damaged Fire spread Penetration,
burning, length and flame spread
glowing etc penetration etc
BROOF (t1) x - - - ENV 1187-1
FROOF (t1) - - - - -“-
BROOF (t2) - x - - ENV 1187-2
FROOF (t2) - - - - -“-
BROOF (t3) - - x - ENV 1187-3
CROOF (t3) - - x - -“-
DROOF (t3) - - x - -“-
FROOF (t3) - - - - -“-
BROOF (t4) - - - x ENV 1187-4
CROOF (t4) - - - x -“-
DROOF (t4) - - - x -“-
EROOF (t4) - - - x -“-
FROOF (t4) - - - - -“-

25
Chapter 3 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

3.2.4 Fire protection ability

A new system, with K classes, has recently been introduced in Europe and is defined in EN
13501-2 Fire Classification of Construction Products and Building Elements – Part 2:
Classification Using Data from Fire Resistance Tests. The new classes are based on fire
resistance testing, and the main parameter is the temperature behind the panel after different
time intervals (10, 30 and 60 minutes). The principles are illustrated in Figure 3.5.
The K classes originate from the Nordic countries, where they have been used mainly for
gypsum plasterboards, since the Nordic criteria also include reaction-to-fire requirements.
However, in the European system, only fire resistance criteria prevail, so this is a great
opportunity for wood products to demonstrate their fire protection abilities.

Figure 3.5. Principle for testing fire

protection ability according to EN 14135.

Table 3.5 Overview of European performance classes for K classes according to EN 13501-2
Class Test conditions Performance criteria Test method
Substrate Air gap Temp. No No Time,
behind or not behind, collapse burning on min
behind C substrate
Low-
K1 10 density X < 250 x x 10 EN 14135
product
Standard
K2 10 X < 250 x x 10 -“-
product
K2 30 -“- X < 250 x x 30 -“-
K2 60 -“- X < 250 x x 60 -“-

26
Chapter 3 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

3.3 Structural Eurocodes

As part of European Union’s desire to remove technical barriers to trade, a set of European codes of
practice in the field of civil and structural engineering has been published. These European design
standards of components and structures for buildings, known as Eurocodes, will help to standardise
design rules within Europe.

The Eurocodes aim at:

− providing common design criteria and calculation methods to merge necessary requirements
for mechanical resistance, stability and resistance to fire, considering of aspect of durability
and economy,
− establishing common understanding of the design of structures between owner, operators and
users, designers and manufactures of construction products,
− enabling the exchange of construction services between members of the European Union, and
facilitating the use and marketing of building products, materials and prefabricated structures
or elements through the use of comparable design properties in the design calculation,
− providing a common basis for research and development in the construction industry,
− increasing the competitiveness of European civil engineers, architects and product
manufactures in their world-wide activities,
− contributing significantly to single market activities within the European Union.

The Eurocodes have to be implemented by the national standard committees in all European countries.
The national committees develop national annexes with specific rules and values to maintain the level
of security prevailing in the respective countries. The national annexes will be essential documents to
enable Eurocodes to be used. The following appropriate information must be included:
− values or classes where alternatives are given in the Eurocode
− values to be used where only a symbol is given in the Eurocode
− specific data e.g. for material properties, wind or snow load
− the procedure to be used when alternative procedures are given in the Eurocode
− decision on the application of informative annexes.

The Eurocodes comprise ten parts:

Eurocode Basis of structural design
Eurocode 1 Actions on structures
Eurocode 2 Design of concrete structures
Eurocode 3 Design of steel structures
Eurocode 4 Design of composite steel and concrete structures
Eurocode 5 Design of timber structures
Eurocode 6 Design of masonry structures
Eurocode 7 Geotechnical design
Eurocode 8 Design of structures for earthquake resistance
Eurocode 9 Design of aluminium structures

All the Eurocodes relating to materials, i.e. 2, 3, 4, 5, 6 and 9, have a part 1 e g EN 1995-1-1, which
covers the design of civil engineering works and buildings and another part 1-2, e g EN 1995-1-2,
which deals with the structural fire design.

Eurocodes allow the calculation and verification of load-bearing capacity of components and structures
for different materials, based on the semi-probabilistic design concept with partial safety coefficients.
It is therefore also possible to design structures or components for desired behaviour in the case of fire,
based on tabular values and simplified or general calculation methods, and to optimise the design of
fire protection. Application of the Eurocodes fire parts permits the integration of parametric
temperature-time curves and natural fire curves to represent real fire scenarios as an alternative to the
standard time/temperature curve in evaluating the fire resistance of components, which can be useful

27
Chapter 3 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

especially in the evaluation of existing structures. However, the use of extended methods requires an
increased level of expertise from the user. Over the next years, Eurocodes will progressively replace
the national design rules in all European countries.

As an extension to the design of load-bearing capacity when exposed to fire, part 1-2 of Eurocode 5
(EN 1995-1-2) offers the alternative of determining fire resistance in respect to the integrity and
insulation of room compartment structures such as walls and floors, if the method is allowed in the
national annex. Classification of building elements uses the performance criteria of EN 13501-2 (see
chapter 3.2.2).

References
COMMISSION DECISION of 8 February 2000 implementing Council Directive 89/106/EEC as
regards the classification of the reaction-to-fire performance of construction products. Official Journal
of the European Communities 23.2.2000.
EN 13823:2002, Reaction-to-fire tests for building products – Building products excluding floorings –
exposed to the thermal attack by a single burning item, SBI test.
EN ISO 11925-2:2002, Reaction-to-fire tests for building products – Ignitability of building products
subjected to direct impingement of flame - Part 2: Single-flame source test.
EN ISO 9239-1:2002, Reaction-to-fire tests for floor coverings – Part 1: Determination of the burning
behaviour using a radiant heat source.
ENV 1187:2005, Test methods for external fire exposure to roofs, 2002. Amendment A1.
EN 14135:2004, Coverings – Determination of fire protection ability.
EN 13501-1:2007, Fire classification of construction products and building elements - Part 1:
Classification using test data from reaction-to-fire tests.
EN 13501-2:2007 Fire classification of construction products and building elements - Part 2:
Classification using data from fire resistance tests, excluding ventilation services.
EN 13501-5:2005, Fire classification of construction products and building elements - Part 5:
Classification using data from external fire exposure roof tests.
EN 1363-1:1999, Fire resistance tests - Part 1: General requirements
EN 1363-2:1999, Fire resistance tests - Part 2: Alternative and additional procedures
EN 1364-1:1999, Fire resistance tests for non-load bearing elements - Part 1: Walls
EN 1364-2:1999, Fire resistance tests for non-load bearing elements - Part 2: Ceilings
EN 1364-3:2006, Fire resistance tests for non-load bearing elements - Part 3: Curtain walling - Full
configuration
EN 1364-4:2007, Fire resistance tests for non-load bearing elements - Part 4: Curtain walling - Part
configuration
EN 1365-1:1999, Fire resistance tests for load bearing elements - Part 1: Walls
EN 1365-2:1999, Fire resistance tests for load bearing elements - Part 2: Floors and roofs
EN 1365-3:1999, Fire resistance tests for load bearing elements - Part 3: Beams
EN 1365-4:1999, Fire resistance tests for load bearing elements - Part 4: Columns
EN 1365-5:2004, Fire resistance tests for load bearing elements - Part 5: Balconies and walkways
EN 1365-6:2004, Fire resistance tests for load bearing elements - Part 6: Stairs
EN 1991-1-2: 2002, Eurocode 1: Actions on structures - Part 1-2: General actions - Actions on
structures exposed to fire
EN1995-1-2:2004, Eurocode 5: Design of timber structures - Part 1-2: General – Structural fire design
ISO 834-1:1999, Fire resistance tests - Elements of building construction - Part 1: General
requirements

28
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

4 Wood products as linings, floorings, claddings and

facades

This chapter presents the reaction-to-fire performance of wood products according to

the new European classification system. Products included are wood-based panels,
structural timber, glued laminated timber, solid wood panelling and wood flooring.
A new system for the durability of the reaction-to-fire performance of wood products is
introduced. Wood claddings fulfilling the new European K classes for fire protective
ability are summarised.
For façade claddings, there is at present no European harmonised solution on their fire
performance. Fire scenarios for facades are presented together with an overview of the
situation in European countries having façade requirements.

Wood surfaces are appreciated and increasingly specified among designers and architects.
Substantial new information on their fire performance according to the new European systems
has been gained during recent years.
The new classification systems for the reaction-to-fire properties of construction products and
for fire protection ability are relevant for linings, floorings and claddings, see chapter 3.

4.1 Reaction-to-fire classification in end-use conditions

The new European system for reaction-to-fire performance must used for all construction
products in order to get the CE- mark, which is the official mandatory mark to be used for all
construction products on the European market. Different product properties have to be
declared, and may vary for different products, but the reaction-to-fire properties are mandatory
for all construction products. The normal route is that each manufacturer tests and declares its
own products individually. An initiative from the EC permits products with known and stable
performance to be classified as group: it is called Classification Without Further Testing,
CWFT, and can be used for wood products that have a fairly predictable fire performance.
Properties such as density, thickness, joints and type of end use application may influence the
classification.
Application of the new European system differs among the various wood products. End use
applications of the products are essential for deciding the reaction-to-fire classification, e.g.
substrates or air gaps behind the wood product, joints and surface profiles. Classification is
always related to harmonised product standards, in which a CWFT reaction-to-fire
classification table is included. Mounting and fixing conditions must be specified.
The CWFT approach has been applied to five different types of wood products:
- Wood-based panels – e.g. particle board
- Structural timber
- Glued laminated timber
- Solid wood panelling and claddings
- Wood floorings
Classifications are listed in Tables 4.1-4.5. Examples of end use applications are illustrated in
Figures 4.1-4.5.

29
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Note that in some cases better performance can be proven by testing, since the CWFT
classifications include a safety margin.
The reaction-to-fire performances in Tables 4.1-4.4 are valid for uncoated wood products.
Coatings may change the fire performance and their influence must be verified in each case.
For wood-based panels, detailed information is given in Table 4.1a and simplified information
in Table 4.1b.

Table 4.1 a. Classes of reaction-to-fire performance for wood-based panels 1)

Official Journal 23 May 2007 (extension of the decision published 18 January 2003)
EN Min. Min. Class Class
6 (excl.
Product product End use condition density thickness (floorings)
standard 3 floorings)
kg/m mm
Cement-bonded no air gap behind the B-s1, d0 Bfl-s1
EN 634-2 1000 10
particle board 1 panel
Fibreboard, no air gap behind the 900 6 D-s2, d0 Dfl-s1
EN 622-2
hard 1 wood-based panel
with a closed air gap not
Fibreboard, 900 6 D-s2, d2 -
EN 622-2 more than 22 mm behind
hard 3
the wood-based panel
Particle board 1, 2, 5 EN 312
Fibreboard, hard and EN 622-2 no air gap behind the
medium 1, 2, 5 EN 622-3 wood-based panel 600 9 D-s2, d0 Dfl-s1
MDF 1, 2, 5 EN 622-5
OSB 1, 2, 5 EN 300
Plywood 1, 2, 5 EN 636 -“- 400 9 D-s2, d0 Dfl-s1
Solid wood panel 1, 2, 5 EN 13 353 12
Flaxboard 1, 2, 5 EN 15 197 -“- 450 15 D-s2, d0 Dfl-s1
Particle board 3, 5 EN 312
Fibreboard, hard and EN 622-2 with a closed or open air
medium 3, 5 EN 622-3 gap not more than 22 mm 600 9 D-s2, d2 -
MDF 3, 5 EN 622-5 behind the wood-based
OSB 3, 5 EN 300 panel
Plywood 3, 5 EN 636 -“- 400 9 D-s2, d2 -
Solid wood panel 3, 5 EN 13 353 12
Particle board 4, 5 EN 312
Fibreboard, medium4, 5 EN 622-3 with a closed air gap 600 15 D-s2, d0 Dfl-s1
MDF 4, 5 EN 622-5 behind the wood-based
OSB 4, 5 EN 300 panel
Plywood 4, 5 EN 636 -“- 400 15 D-s2, d1 Dfl-s1
Solid wood panel 4, 5 EN 13 353 D-s2, d0
Flaxboard 4, 5 EN 15 197 -“- 450 15 D-s2, d0 Dfl-s1
Particle board 4, 5 EN 312 with an open air gap
Fibreboard, medium4, 5 EN 622-3 behind the wood-based 600 18 D-s2, d0 Dfl-s1
MDF 4, 5 EN 622-5 panel
OSB 4, 5 EN 300
Plywood 4, 5, EN 636 -“- 400 18 D-s2, d0 Dfl-s1
Solid wood panel 4, 5 EN 13 353
Flaxboard 4, 5 EN 15 197 -“- 450 18 D-s2, d0 Dfl-s1
Particle board 5 EN 312 any 600 3 E Efl
OSB 5 EN 300
MDF 5 EN 622-5 -“- 400 3 E Efl
250 9 E Efl
Plywood 5 EN 636 -“- 400 3 E Efl
Fibreboard, hard 5 EN 622-2 -“- 900 3 E Efl
Fibreboard, medium 5 EN 622-3 -“- 400 9 E Efl
Fibreboard, soft EN 622-4 -“- 250 9 E Efl

30
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

1
Mounted without an air gap directly against Class A1 or A2-s1,d0 products with minimum density of 10 kg/m3, or at
least Class D-s2,d2 products with minimum density of 400 kg/m3.
2
A substrate of cellulose insulation material of at least Class E may be included if mounted directly against the wood-
based panel, but not for floorings
3
Mounted with an air gap behind. The reverse face of the cavity shall be at least Class A2-s1, d0 products with
minimum density of 10 kg/m3.
4
Mounted with an air gap behind. The reverse face of the cavity shall be at least Class D-s2, d2 products with
minimum density of 400 kg/m3.
5
Veneered, phenol-faced and melamine-faced panels are included for this class excluding floorings.
6
A vapour barrier with a thickness up to 0,4 mm and a mass of up to 200 g/m2 can be mounted between the wood-
based panel and a substrate if there are no air gaps in between.

Table 4.1 b. Classes of reaction-to-fire performance for wood-based panels. Simplified version.
Product End use condition Min. Min. Class
density thick- (excl.
kg/m3 ness floorings)
mm
Fibreboard, hard without an air gap behind the wood-based panel 900 6 D-s2, d0
Wood-based panels without an air gap behind the wood-based panel 600 9 D-s2, d0
with a closed or an open air gap not more than 22
Wood-based panels 600 9 D-s2, d2
mm behind the wood-based panel
Wood-based panels with a closed air gap behind the wood-based panel 600 15 D-s2, d0
with or without an open air gap behind the wood-
Wood-based panels based panel 600 18 D-s2, d0
Wood-based panels All 600 3 E
Fibreboard, soft -“- 250 9 E

Table 4.2.Classes of reaction-to-fire performance for structural timber1)

(Commission Decision August 2003).
Minimum mean Minimum overall
Material Product detail density thickness Class
kg/m3) (mm)
Structural Visual and machine-graded structural
D-s2, d0
timber timber with rectangular cross-sections 350 22
shaped by sawing, planing or other
methods, or with round cross-sections
accordance with EN 14081.
1) Applies to all species covered by the product standards.

Table 4.3. Classes of reaction-to-fire performance for Glulam1) (Commission Decision August 2005).
Minimum mean Minimum overall
Material Product detail density thickness Class
(kg/m3) (mm)
Glulam Glued laminated timber products in D-s2, d0
accordance with EN 14080. 380 40
1) Applies to all species and glues covered by the product standards.

31
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 4.4. Classes of reaction-to-fire performance for solid wood panelling and cladding (EN 14915)
(Commission Decision March 2006)
Minimum Minimum
Product 11) Product detail 5) mean thicknesses, End-use Class 3)
density 6) total / condition 4)
(kg/m3) minimum 7)
(mm)
Panelling and Wood pieces with or without Without air gap,
cladding 1) tongue and groove, and with or 390 9/6 or with closed air D-s2, d2
without profiled surface gap behind

-“- -“- 390 12 / 8 -“- D-s2, d0

Panelling and With open air gap

cladding 2) -“- 390 9/6 ≤ 20 mm behind D-s2, d0
Without air gap,
-“- -“- 390 18 / 12 or with open D-s2, d0
air gap behind
Wood ribbon Wood pieces mounted on a Surrounded by
elements 8) support frame 9) 390 18 open air on all D-s2, d0
sides 10)
1) Mounted mechanically on a wood batten support frame, with the gap closed or filled with a substrate of at
least Class A2-s1,d0 with minimum density of 10 kg/m3, or filled with a substrate of cellulose insulation
material of at least Class E and with or without a vapour barrier behind. The wood product must be designed
to be mounted without open joints.
2) Mounted mechanically on a wood batten support frame, with or without an open air gap behind. The wood
product must be designed to be mounted without open joints.
3) Class as provided for in Commission Decision 2000/147/EC, Annex Table 1. This decision is currently
under review in respect of façade applications.
4) An open air gap may include possibility for ventilation behind the product, while a closed air gap will
exclude such ventilation. The substrate behind the air gap must be of at least Class A2-s1,d0 with a
minimum density of 10 kg/m3. Behind a closed air gap of maximum 20 mm and with vertical wood pieces,
the substrate must be of at least Class D-s2,d0.
5) Joints include all types of joints, e.g. butt joints and tongue and groove joints.
6) Conditioned according to EN 13238.
7) As illustrated in Figure a below. Profiled area of the exposed side of the panel not more than 20 % of the plane
area, or 25 % if measured on both exposed and unexposed sides of the panel. For butt joints, the larger
thickness applies at the joint interface.
8) Rectangular wood pieces, with or without rounded corners, mounted horizontally or vertically on a support
frame and surrounded by air on all sides, mainly used close to other building elements, both in interior and
exterior applications.
9) Maximum exposed area (all sides of rectangular wood pieces and wood support frame) not more than 110 %
of the total plane area, see Figure b below.
10) Other building elements closer than 100 mm from the wood ribbon element (excluding its support frame)
must be of at least Class A2-s1,d0, at distances 100-300 mm of at least Class B-s1,d0 and at distances more
than 300 mm of at least Class D-s2,d0.
11) Applies also to stairs.

32
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure a
(to Footnote 7)
Total thickness Minimum thickness at
any point of the profile

Profiles for solid wood panelling and cladding.

Figure b t w
(to Footnote 9) t
w

1.00 m 1.00 m 1.00 m

w t

Maximum exposed area of wood ribbon element: 2n (t+w) + a ≤ 1,10

n is the number of wood pieces per meter
t is the thickness of each wood piece, in meter
w is the width of each wood piece, in meter
a is the exposed area of wood support frame (if any), in m2,
per m2 of wood ribbon element

33
 Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 4.5 a. Classes of reaction-to-fire performance for surface coated wood flooring (EN 14342)
(Commission Decision March 2006)
Product 1,7) Product detail 4) Minimum Minimum End use condition Class 3)
mean overall for
density 5) thickness floorings
(kg/m3) (mm)
Wood Solid flooring of oak or beech with Beech: 680 8 Glued to substrate Cfl-s1
6)
flooring and surface coating Oak: 650
parquet
Solid flooring of oak, beech or spruce Beech: 680 20 With or without air Cfl-s1
-“- and with surface coating Oak: 650 gap underneath
Spruce: 450
Solid wood flooring with surface 390 8 Without air gap Dfl-s1
-“- coating and not specified above underneath
390 20 With or without air Dfl-s1
-“- -“- gap underneath
Wood Multilayer parquet with a top layer of 650 10 Glued to substrate Cfl-s1
6)
parquet oak of at least 5 mm thickness and (top layer)
with surface coating
650 14 2) With or without air Cfl-s1
-“- -“- (top layer) gap underneath
Multilayer parquet with surface 500 8 Glued to substrate Dfl-s1
-“- coating and not specified above
500 10 Without air gap Dfl-s1
-“- -“- underneath
500 14 2) With or without air Dfl-s1
-“- -“- gap underneath
Veneered Veneered floor covering with surface 800 6 2) Without air gap Dfl-s1
floor coating underneath
covering
1) Mounted in accordance with EN ISO 9239-1, on a substrate of at least Class D-s2,d0 and with minimum density of 400 kg/m3
or with an air gap underneath.
2) An interlayer of at least Class E and with maximum thickness 3 mm may be included in applications without an air gap, for
parquet products with 14 mm thickness or more, and for veneered floor coverings.
3) Class as provided for in Commission Decision 2000/147/EC, Annex Table 2.
4) Type and quantity of surface coatings included are acrylic, polyurethane or soap, 50-100 g/m2, and oil, 20-60 g/m2.
5) Conditioned according to EN 13238 (50 % RH 23 oC)
6) Substrate at least Class A2-s1,d0.
7) Applies also to steps of stairs.

Table 4.5 b. Classes of reaction-to-fire performance for uncoated wood flooring. (ongoing, to be amended)
Minimum Minimum Class for
Product 1,2) Product detail 3) density 4) thickness End use condition floorings
5)
(kg/m³) (mm)
Wood flooring Solid flooring of spruce 450 14 6) Without air gap underneath [Cfl-s1]
Wood flooring Solid flooring of pine 450 14 6) Without air gap underneath [Dfl-s1]
With or without air gap
Wood flooring Solid flooring of spruce 450 20 6) Cfl-s1
underneath
Wood flooring Solid flooring ash, maple, walnut tbd 20 6) -“- [C/D]
Wood flooring Solid flooring beech, oak, pine tbd 20 6) -“- Dfl-s1
Wood parquet Solid (one layer) parquet of [tbd] tbd 8 6) Glued to substrate [C/D]
Wood parquet Solid (one layer) parquet of [tbd] tbd 14 6) Without air gap underneath [C/D]
Multilayer parquet with oak top
Wood parquet 700 (top) 15 6) -“- [C/D]
layer, at least 3,5 mm
Wood flooring Solid flooring and parquet not
450 Efl
and parquet specified above

34
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 4.1.
Examples of end uses of structural timber as studs (cross-sectional view), beam and free-standing
elements

Figure 4.2.
Examples of end uses of glued laminated timber as beams and columns

Figure 4.3.
Examples of end uses of solid wood panelling as exterior wood claddings with either another material,
e.g. insulation, or an air gap behind

Figure 4.4.
Examples of end uses of solid wood panelling as free-standing wood lamellas (ribbon elements)

a) b) c) d)

Figure 4.5.
Examples of end uses for wood flooring. a) Flooring directly on a substrate; b) with an interlayer
between the flooring and the substrate; c) with a closed air gap in between, and d) with an open air
gap without anything underneath (e g as in an open staircase)

35
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

4.2 Improved reaction-to-fire performance

Fire retardant treatments of wood products e.g. by chemical modification, may considerably
improve their reaction-to-fire properties, to the extent that the highest fire classifications for
combustible products can be reached, e.g. Euroclass B. This allows a wider use of visible
wood, both as interior wall and ceilings linings and as exterior claddings, e.g. in facades.

4.2.1 Durability of the reaction-to-fire performance

However, the durability of the fire retardant treatments needs to be addressed. Two cases of
effects or mechanisms that can affect the durability of the fire retardant treatment of wood-
based products can be identified. One is the risk for high moisture content and migration of the
fire-retardant chemicals within the wood product and salt crystallisation on the product
surface. The other case is the risk for decreased fire performance due to loss of the fire
retardant chemicals by leaching or other mechanisms. This risk is mainly pronounced for
exterior applications and is the main challenge to the development of new fire-retardant-
treated (FRT) wood products.
A new European standard, prEN 15912, with Durability of Reaction-to-Fire performance
(DRF) classes has been developed in order to guide potential users to find suitable FRT wood-
based products, and encourage producers to supply competitive products. The system is
summarised in Table 4.6. It consists of a control system for the durability properties of FRT
wood-based products and suitable test procedures. The European system is based on an
existing Nordic system, Nordtest Method NT Fire 054, and on experience from North
America.

Table 4.6 Requirements for DRF (Durability of Reaction-to-Fire performance) classes of FRT wood-
based products in interior and exterior end-use applications according to prEN 15912.
DRF class Existing fire Additional performance requirements for different end uses
requirements of fire-retardant wood-based products a)
Intended use Reaction-to-fire Hygroscopic properties b) Reaction-to-fire performance
class, initial after weather exposure
ST Short term Relevant fire class - -
INT Interior, dry - Moisture content < [20] % -
1 applications -"- - Minimum visible salt with
no increase at surface
- No exudation of liquid
INT Interior, - Moisture content < [28] % -
2 humid -"- - Minimum visible salt with
applications no increase at surface
- No exudation of liquid
EXT Exterior Maintained reaction-to-fire
applications -"- -"- performance c, d) after
- Accelerated ageing or
- Natural weathering
- Other referenced ageing method
a) to be fulfilled using material produced using the same manufacturing process and having a similar retention level as
for the reaction-to-fire performance.
b) for INT1 at (70 ± 3) % RH and (25 ± 2) °C and for INT2 at( 90 ± 3) % RH and (27 ± 2) °C according to Annex A 1.
c) criteria for fire testing according to 5.2.2.2 after weather exposure: (Rate of Heat Release) RHR ≤ 100 kW/m2 during
1200 s testing time or (Total Heat Release) THR600 s not increased more than 20 % compared to testing before weather
exposure.
d) for DRF Class EXT, the Durability of Reaction-to-fire performance classification is valid only for the type of coating
system to be verified.

36
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

FRT wood products fulfilling both the reaction-to-fire and the durability of reaction-to-fire
requirements are available on the European market.

4.3 Wood claddings with fire protective ability

Wood-based panels and wood panelling fulfil the new European K classes for fire protection
ability. Three levels are defined: K 10, K 30 and K 60. The criteria for the classification are
based mainly on panel thickness.
The new system with K classes has recently been introduced in Europe and is defined in EN
13501-2 Fire Classification of Construction Products and Building Elements – Part 2:
Classification Using Data from Fire Resistance Tests. The new classes are based on fire
resistance testing, and the main parameter is the temperature behind the panel after different
time intervals (10, 30 and 60 minutes). See Chapter 3.2.4.

Table 4.7. European K classes for wood-based panels and solid wood panelling
(to be updated and verified)
Product Product Min. mean Min. overall End use K
details density thickness condition Class
(kg/m3) (mm)
Wood-based Wall and ceiling
- 600 [10] K2 10
panels covering
Tongue and
-“- 600 [25] -“- K2 30
groove
Tongue and
-“- 600 [50] -“- K2 60
groove
Solid wood Tongue and Wall and ceiling
400 [15] K2 10
panelling groove covering
-“- -“- 400 [30] -“- K2 30
-“- -“- 400 [50 / 60] -“- K2 60

37
 Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

4.4 Wooden facades

For façade claddings, there is at present no European harmonised solution on their fire
performance. A draft European technical guideline ETAG has been presented, but no
agreement have been achieved yet. Fire scenarios for facades are presented in this section
together with an overview of the situation in European countries having façade requirements.
Structural fire protection is a fundamental requirement when facades are used as the outer
surface of external walls of multi-storey buildings (regardless of the used material). The goal
is to prevent uncontrolled fire spread on the surface and in ventilation voids (if present) of the
external wall for a required period of time.
Depending on the type and/or material of the façades, additional structural fire protection
measures may be are essential to reduce the spread of fire. Thus timber and wood-based
products can be used without reducing the required fire safety level. However, there is at
present no European harmonised solution for the fire performance of facades.

4.4.1 Fire scenarios of facades

The spread of a fire on the facade of a building is generally influenced by character, intensity
and location of the initial fire.
In principle, the facade can be exposed to the following three fire scenarios:
• Scenario A: fire in a building located next to the facade
• Scenario B: fire outside the building in front of the facade
• Scenario C: fire inside the building in a room next to the exterior wall, with at
least one opening in the façade

Radiation

Scenario A Scenario B Scenario C

Figure 4.6. Fire scenarios for exterior walls (reference: Kotthoff)

Several fire tests have shown that the scenario with a room fire has the most serious effect on
the façade, and causes the highest release of energy in front of the facade surface.
For fire scenarios B and C, conservative values for testing and evaluation of facades can be
determined as:
• release of energy in front of the facade 1 – 1,5 MW
• average extent of flames 2,5 – 3,0 m (max. 6,0 m)
• total time of thermal exposure 15 – 20 minutes
• total time of exposure to blazing fire 10 – 15 minutes

38
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 4.7 shows the maximum temperatures during a fire test. The fire load in the test is
based on temperature measurements in real fire tests.
Scale

sketch

Figure 4.7. Isotherms with maximum temperatures in front of façade

surface at a distance of 5 mm and with a fire load of a 25 kg wooden crib

Real fires and fire tests show that all parts of a facade react to direct flame impact in the area
of the plume, dependent on their flammability. Windows or openings in the facade in this area,
and providing no fire protection, are weak spots, whether open or closed.

4.4.2 Fire behaviour of wooden facades

Timber and timber-derived products must be tested in accordance with EN 13501-1 in order to
arrive at a classification:
• Small flame test according to EN ISO 11925-2
• Single Burning Item test (SBI) according to EN 13823
The SBI test enables building materials and façade systems to be tested. Testing requires the
entire system, i.e. with cladding, substructure, insulation materials, fasteners and joints, to be
determined and tested. The following table shows a selection of façade systems.

Table 4.8. Classes of reaction-to-fire performance for wooden façade systems

Façade systems Wood Orientation Joint Ventilation Insulation Class
species width void thickness
(mm) (mm) [mm]
Tongue/groove-cladding, d = 19 mm Spruce Vertical 30 D-s1, d0
Spruce Vertical 100 D-s1,d0
Tongue/groove-cladding, d =9 mm Larch Horizontal 20 30 D-s2, d0
Tongue/groove-cladding, d = 19 mm Spruce Horizontal 20 30 D-s2, d0
Tongue/groove-cladding, d = 19 mm, Spruce Horizontal 40 60 D-s2, d0
battens 100 mm
3-layered board, d = 19 mm Spruce - 20 30 D-s2, d0
3-layered board, d = 19 mm Spruce - 100 D-s2, d0

39
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In some cases and countries, wooden façade systems must also be tested in a specific façade
test rig. This test method is necessary in order to be able to evaluate the spread of fire through
a window burnout based on a fully developed room fire. This means that, beside the
combustibility of timber and timber-derived products in facades, the spread of fire over one or
two storeys above the storey of fire origin must be considered.
Wooden facades must fulfil requirements on the spread of fire and the safety of fire brigades
during fire fighting, and evacuation safety of occupants. Table 4.9 lists some façade fire test
methods and national requirements. Evaluation criteria include:
• no fire outside the initial fire area
• fulfilling temperature criteria in the rear ventilation spaces and ventilation voids
• no collapse of larger parts

Table 4.9. Fire performance requirements on facades in Europe (further information to be added)
Country Type of fire requirements for facades Fire scenario Method Main set up Exposure Measure- Storeys Others; Comments
for levels, kW/m2 ments allowed
verificatio with
n wooden
facades
Upward Falling Fire Others
flame parts resistanc
spread e
Austria x x - Flames out ÖNORM Corner test ca 40 during Temp. 5 Timber constructions
of a window B 3800-5 with façade 20 minutes damage (ÖNORM ÖNORM B 2332
wall B 3806)
1) 2)
Finland - - 2 – 43)
France
Germany x x - Façade E DIN Fire room 20 – 65 Flames, 3 Test standard (draft),
corner with 4120-20 with façade (350 – 400 temp, damage additional evaluation
flames out wall kW), 20 min and glowing criteria DIBT (not
of window by gas burner combustion published)
Norway x In some Reaction 4 Guideline
cases to fire
Switzerland
Sweden x x - - Flames out SP Fire Fire room 15-75 during Temp, heat 2 No testing required for
of a window 105 with façade 15-20 minutes flux, damage two stories
wall
UK
More
countries?
Others:
ISO ISO Not used in Europe
13785-2
Notes for Finland:
1)
If distance between building < 8 m the external wall may be subject to requirements (e.g. EI xx)
2)
Reaction to fire requirements for surfaces of external walls and ventilation gaps
3)
3-4 storeys for defined building types with sprinklers

The following conclusions concerning the use of timber or timber-derived materials can be
drawn from fire tests that have been performed:
- In real fire scenarios (using a practice-orientated set up), non-load-bearing cladding for
external walls, ventilated at the rear, react noticeably better than expected from the
experimental results of small fire tests.
- The self-protection effect of timber through carbonisation of the surface inhibits the rapid
vertical spread of fire. No significant lateral spread occurs, or occurs to only a very limited
and acceptable degree.
- The spread of fire behind cladding, in the rear ventilated area, is controllable by simple
means.
- In particular cases, the flammability of timber cladding systems for external walls is
dependent on the influence of various parameters which, if the worst comes to the worst,
can interfere with each other.
- Problems when fire fighting using water were not noticed in the full-scale tests. In
addition, no extensive drop-off of parts of the structure of the exterior wall, endangering
persons, was observed.

40
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

- Flammable exterior insulation can accelerate the spread of fire.

- The tested standard coatings have no significant effect on the flammability of the cladding
of the exterior wall.

Basically, the behaviour of facades exposed to a fire load is dependent on

- the type and architectural features of the facade (facade with recessed ribbon glazing,
perforated facade, angled facade),
- the type and alignment of the cladding
- the substructure, including the rear ventilation opening

Parameter of Influence to fire behaviour

influence Best Good Critical
Type of cladding

Figure 4.8. Classification of type of cladding depending on the influence to fire behaviour by
multi-storey buildings (reference: Lignum Dokumentation Brandschutz)

4.4.3 Components of wooden façades and their influence on the fire behaviour

Figure 4.9 shows a façade system with allocated layers and areas of an external wall assembly.

External Wall
Fs Ws Areas
Ws Wall structure
Ss Fs Façade system
O V I Ss Substructure
I Insulation
V (Rear) Ventilation void
O Outer panelling

Layers
Insulation outside
Membrane
Battens vertical
Battens horizontal
Cladding material

Coating

Figure 4.9. Areas and layers of an external wall structure [reference: pro:holz: Workbook
8/06]

41
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The influence of single facade layers on fire behaviour is explained in some detail below.
a) Outer panelling
The fire behaviour of cladding materials made of timber and timber-derived products will be
influenced mainly by the geometry. General statements are:
− Façade cladding with space between the boards will increase the fire exposure.
− If a greater surface area is involved in the fire, the fire will be more intense.
− Vertical façade claddings burn significantly more intensely than horizontal claddings.
− The surface (sanded, planed etc.) of façade claddings has an influence on the fire behaviour.
− The ageing of wood (cracks, etc.) has an influence on the fire behaviour.
− Considering cladding boards relieving joints charring along this joints must be avoided.

b) Substructure
The types of substructure, as well as the thickness of battens, have a significant influence on
fire behaviour. General statements are:
− In rear ventilation voids with vertical battens less than 30 mm a lower fire exposure with
unimportant horizontal fire spread may be expected.
− Using cross-battens in the rear ventilation void the horizontal fire spread has to be hindered
with fire stops between each window or at least in distances of 5 m.
− Foils, e.g. for protection against moisture have an insignificant influence on the fire
behaviour of the whole façade system.

c) Insulation
For multi-storey buildings with more than three storeys non-combustible insulation materials
can be required, e.g. in Germany or Austria, in order to avoid ignition and uncontrollable fire
spread within the insulation layer as a result of exposure to external fire. In other European
countries it might be possible to do with combustible materials.

d) External wall structure

For multi-storey buildings with more than three storeys non-combustible outer panelling can
be required, e.g. in Germany or Austria, in order to avoid ignition of the supporting structure
as a result of exposure to external fire. In other European countries it might be possible to do
with combustible materials.

4.4.4 Structural measures for facades

Structural measures for fire stops for facades are described below: general effectiveness,
positions of fire stops in different types of façades, and finally some examples of fire stops in
facades.
General effectiveness of fire stops
Overhanging fire stops and fire stops behind the panelling help to restrict the spread of fire on
and behind façades, reducing the spread of fire over several storeys. Fire stops in facades have
the following effects:
• prevention of stack effect in the rear ventilation void
• deflection of flames from the façade surface
• reduction of fire input inside the rear ventilation void

42
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

a) b) c)

d) e)

Figure 4.10. Basic principles of fire stops in facades; a) rear ventilation voids, b) rear
ventilation voids with fire non-combustible fire stop, c) rear ventilation voids with fire
combustible fire stop, d) ventilation void, e) without ventilation voids [Lignum
Dokumentation Brandschutz: 7.1 Außenwände]

The main threat from a fire that spreads in the ventilation gap of a façade is that the fire can
spread via the façade to other spaces, such as apartments or attic/roof void spaces. The latter of
these two events is the more likely, unless it has been prevented by a special eaves structure.
Fire can spread to apartments, for example, through gaps in the windows or open windows.

The following concentrates on façades, where the outer surface of the façade and the
ventilation void has been made from Class D building materials (= wood). The threat is
external ignition. The most important technical problem in incorporating fire stops in the
ventilation voids behind façades is that the structure must retain its moisture protection
functionality. Creating a functional fire sealing solution for the ventilation void, requires a
compromise between the fire protection and moisture protection requirements.

Two fire stop solutions are described above: a perforated steel profile (Figure 4.10 b) and a
fire stop made from two horizontal wooden battens (Figure 4.10 c). In the steel profile fire
stop, the diameter of the holes was 18 mm and the spacing was 140 mm. On one side of the
fire stop, there were six holes with a total area of 15 cm², which equals 5 % of the cross-
sectional area of the fire stop. According to fluid dynamics, this is sufficient for moisture
protection, as long as no significant amounts of rain water can flow into the ventilation cavity.
A plain tongued and grooved wooden wall, for example, was sufficiently raintight to meet the
requirements. The air flow-restricting type of seal was made of wooden battens of a size to
leave a 7 mm wide gap, equal to about 20 % of the cross-sectional area of the cavity. In
practice, the gap of the wooden batten fire stop cannot be made much narrower because it
would encroach too much on allowances for dimensional accuracy, moisture-related
dimensional changes, litter/dust and other similar matters.

43
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Position of fire stops in facades

Fire protection measures for facades are dependent from the type of façade.

I) Punched window facades II) Horizontal ribbon window facades

Type Type
a) windows lying upon each other a) continuous ribbon windows
b) staggered windows b) windows lying upon each other, in the lining area
c) windows with smaller openings between window with non combustible materials
d) loggia c) horizontal staggered windows, in the lining area
between window with non combustible materials
Only certain panelling systems can be used! d) continuous lining area non-combustible

Only certain panelling systems can be used!

III) Facades with single balconies IV) Facades with ribbon balconies

Type Type
a) Single balconies supported from columns a) Ribbon balconies supported from columns
b) Single balconies as cantilever b) Ribbon balconies as cantilever

Requirements Requirements
Cantilever x ≥ 1,0 m Cantilever x ≥ 0,6 m
Width of balcony y ≥ 2,0 m
Distance between balconies z ≥ 2,0 m Each panelling system can be used!

Only certain panelling systems can be used!

44
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

V) Vertical ribbon window facades VI) Facades on staggered storeys

Type
a) Vertical ribbon glazing Type
b) Vertical band and areas with non-combustible a) x ≥ 1,0 m
panelling b) x ≥ 0,6 m
c) x ≥ 0,2 m

Requirements
Staggered storeys more than 0,2 m fulfil the
requirements of a horizontal fire protection measure.
By staggered storeys more than 0,6 m combustible
balustrades and non-combustible flooring systems can
be realised.
By staggered storeys more than 1,0 m both
balustrades and flooring systems can be realised with
combustible materials.

Each panelling system can be used!

legends
fire protection measure
lining

Figure 4.11. Different types of facades with position of fire stops [Lignum Dokumentation
Brandschutz: 7.1 Außenwände]

45
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Design of fire stops in facades

The examples in Figures 4.12 - 4.14 are based on different performance criteria. Figures 4.12
and 4.13 show examples in situations where a spread of fire over two storeys can be accepted.
Figure 4.14 shows the dimensions of fire stops, if the spread of fire must be limited to the first
fire stop above.

1 batten for erection ≥ 30 x 30 1 steel sheet, thickness t ≥ 0,5 mm

2 batten with compressed mineral wool, melting point 2 timber batten, width ≥ 50 mm
≥ 1000°C, density ≥ 40 kg/m³ spacing of fasteners on battens ≤ 400 mm
3 mineral wool, melting point ≥ 1000°C [DIN 4102-17], 3 timber batten, grooved, width ≥ 70 mm
density ≥ 60 kg/m³ spacing of fasteners on battens ≤ 400 mm
4 local cantilever 4 Leakproof connection between batten and cladding
5 fire stop ≥ 40 mm inside the insulation strip 5 Mineral wool
Melting point ≥ 1000°C [DIN 4102-17]
density ≥ 40 kg/m³
width ≥ 150 mm, rear ventilation void leakproof filled
fixing against slipping (e.g. screws)

Figure 4.12. Fire stop measures in rear Figure 4.13: Fire stops in ventilation voids; 1 –
ventilation voids; 1 - Screwed steel sheet apron, timber batten with steel sheet, 2 – grooved timber
2 – Screwed timber board, 3 – Screwed steel batten, 3 – timber batten with leakproof
sheet apron on uneven surface, 4 – Continuous connection to the cladding, 4 – mineral wool
apron on local cantilever
[Lignum Dokumentation Brandschutz: 7.1 Außenwände]

46
Chapter 4 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

a) b) c) d) e)

Overhang 200 mm: Overhang 100 mm:

Suitable for all cladding materials in connection with all façade systems, Suitable for all closed claddings, e.g. tongue and
including cladding boards with open spaces, vertical and horizontal grooved cladding, OSB, 3-layered board
Requirements:
- Board b x d ≥ 20 mm x 70 mm
- Joint ≤ 10 mm

Figure 4.14. Fire stop measures; a) cantilever non-combustible, b) timber or timber derived board in
connection with covering of non-combustible materials, c) timber or timber derived board, d and e)
smaller design of fire stops in comparison to a), b) and c) [reference: pro:holz: Workbook 8/06]

4.4.5 Conclusions
In the event of a fire on the outer surface of a building, the fire must not spread upwards to
more than [two] storeys above the initial fire before active fire-fighting starts.
The following requirements must be fulfilled in order to achieve the goals:
• Wooden facades in low rise buildings are allowed in all European countries.
• Using timber or timber-derived products for claddings in higher buildings increases the
fire load of enclosing walls, so care must be taken to ensure maintenance of the defined
fire protection target.
• Claddings for external walls are only to be used for buildings up to a height which can
be reached by fire–fighting from outside the building. Additionally, the application of
such claddings is limited to certain building classes in most countries.
• Within a defined time, the fire must not spread beyond an accepted area. (The accepted
area or fire spread upwards have to be discussed with the partners!)
• Ventilation voids behind façade claddings must be sealed in higher buildings, see
chapter 8.2.3.

References
Chapter 4.4
Lignum Dokumentation Brandschutz: 7.1 Außenwände, Konstruktion und Bekleidung
(http://www.lignum.ch)
Kotthoff, I.: Brandschutz im Holzbau, 9. DGfH-Brandschutztagung, Würzburg Okt. 2001
Schober, K. P.; Matzinger, I.: Brandschutztechnische Ausführung von Holzfassaden, proHolz
Austria (Hrsg.) Arbeitsheft 8/06

47
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

5. Separating timber structures

Separating structures are used to limit the fire spread from one fire cell, e g a compartment, to
another. This chapter presents the basic requirements, calculation methods and the design
method according to Eurocode 5. It also presents an improved design method from recent
research and several examples on how to use the method.

5.1 General
This section gives guidance for the fire design of separating timber structures. Reference is made to EN
1995-1-2 [1], which gives a calculation method for the verification of the separation function of timber
structures (see annex E of EN 1995-1-2). Further, it presents a new design method developed from recent
research. The calculation method given in annex E of EN 1995-1-2 is informative and may not be applicable
in all European states. Depending on national regulations, the new design method given in the following may
need agreement by the Competent Authority. Hence the content of this section should be regarded as the
state of the art as of and the new design method as potential input for future revisions of EN 1995-1-2 [1].

5.2 Basic requirements for fire compartmentation

The main objective of structural fire safety measures is to limit the spread of fire to the room of origin by
guaranteeing the load-carrying capacity of the structure (Requirement on Mechanical Resistance R) and the
separating function of walls and floors (Requirement on Insulation I and Integrity E) for the required period
of time. The required period of time is normally expressed in terms of fire resistance, using fire exposure of
the standard temperature-time curve, and is specified by the building regulations. While fire tests are still
widely used for the verification of the fire resistance of timber structures, calculation models are becoming
more and more common.
Concerning the basic requirements for fire compartmentation, EN 1995-1-2 [1] states:
“Where fire compartmentation is required, the elements forming the boundaries of the fire compartment shall
be designed and constructed in such a way that they maintain their separating function during the relevant
fire exposure. This shall include, when relevant, ensuring that:
• integrity failure does not occur (Criterion E),
• insulation failure does not occur (Criterion I), and
• thermal radiation from the unexposed side is limited
Criterion I (insulation) may be assumed to be satisfied where the average temperature rise over the whole of
the non-exposed surface is limited to 140 K, and the maximum temperature rise at any point on that surface
does not exceed 180 K (for fire exposure of the standard temperature-time curve), thus preventing ignition of
objects in the neighbouring compartment. Criterion E (integrity) may be assumed to be satisfied when no
sustained flaming or hot gases to ignite a cotton pad on the fire unexposed side of the construction occur or
no cracks or openings in excess of certain dimensions exist. There is no risk of fire spread due to thermal
radiation when the temperature on the fire unexposed side is below 300°C.
Criterion I (insulation) is clearly defined, and so could be verified by heat transfer calculations instead of by
testing if thermal material properties could be found in function of temperature (conductivity and heat
transfer). On the other hand, Criterion E (integrity) is mostly determined by testing, because calculations are
still quite impossible (crack formation, dynamics of hot gases, etc.). For example, premature integrity failure
may occur due to sudden failure of claddings or opening of gaps, which is often dependent on construction
details such as fixings. However, extensive experience of full-scale testing of wall and floor assemblies made
it possible to define some rules about detailing of wall and floor assemblies, that have been included (for
example) in EN 1995-1-2 [1]. Thus EN 1995-1-2 assumes that Criterion E (integrity) is satisfied where
Criterion I (insulation) has been satisfied and panels remain fixed to the timber structure on the fire
unexposed side.

49
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

5.3 Calculation methods

In timber buildings, walls and floors are mostly built up by adding different layers to form an assembly. For
the calculation of fire resistance with regard to the separation function of timber assemblies, component
additive methods can be used. These methods are thus called because they determine the fire resistance of a
layered construction by adding the contribution of the different layers to obtain the fire resistance. Reference
[2] presents and reviews calculation methods for verification of the separating performance of wall and floor
assemblies as used in United Kingdom [3], Canada [4] and Sweden [5, 6], as well as according to ENV
1995-1-2 [7]. The Swedish component additive method builds upon that described in ENV 1995-1-2 and the
North America method by taking into account the influence of adjacent materials on the fire performance of
each layer, and therefore describes the real fire performance more accurately.
The EN 1995-1-2 (Annex E) design method is based on modification of the Swedish component additive
method by extending it to floors, including the effect of joints in claddings that are not backed by members,
battens or panels, and applying some of the position coefficients to further test results that became available
during the drafting of EN 1995-1-2. The EN 1995-1-2 design method is capable of considering claddings
made of one or two layers of wood-based panels and gypsum plasterboard, and also voids or insulation-filled
cavities. The insulation may be made of mineral wool.

5.4 The EN 1995-1-2 design method

Analysis method for the separating function of wall and floor assemblies is presented in EN 1995-1-2 Annex
E which is informative. This means that the method shall be used according to the National Annex in the
country concerned.
EN 1995-1-2 requires verification that the time (tins) that it takes for the temperature to increase (starting
from room temperature) by 140 K / 180 K on the side of the member that is not exposed to fire is equal to or
greater than the required fire resistance period (treq) for the separating function of the member.

t ins ≥ t req [min] (1)

The insulation time tins depends on the fire behaviour of the layers used in the assemblies, as well as the
positions and joint configurations of the layers. For simplicity, the time tins can be calculated as the sum of
the contributions of the individual layers to fire resistance, considering different heat transfer paths (see
Figure 5.1).

i=n
t ins = ∑ t ins,i [min] (2)
i =1

Figure 5.1 Illustration of heat transfer paths through separating multiple layered construction
These contributions firstly depend on the inherent insulation property of each layer, as given by the basic
insulation value, and secondly on the position of the respective layer and the materials backing or preceding
that layer (in the direction of the heat flux), as given by the position coefficient. Further, a joint coefficient is

50
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

used in order to take into account the influence of joint configurations on the insulation time of layers with
joints. Thus the contribution of each layer tins,i is calculated using the basic insulation value (tins,0,i), the
position coefficient (kpos,i) and the joint coefficient (kj,i).

t ins,i = t ins,0,i ⋅ k pos,i ⋅ k j,i [min] (3)

The basic insulation value corresponds to the contribution of a single layer to fire resistance without the
influence of adjacent materials, and depends on the material and the thickness of the layer. EN 1995-1-2
gives equations for calculating the basic insulation values for the following materials:
Panels:
• Plywood (ρ ≥ 450 kg/m3)
• Particleboard and fibreboard (ρ ≥ 600 kg/m3)
• Wood panelling (ρ ≥ 400 kg/m3)
• Gypsum plasterboard, types A, F, R and H
Cavity insulations:
• Stone wool (26 kg/m3 ≤ ρ ≤ 50 kg/m3)
• Glass wool (15 kg/m3 ≤ ρ ≤ 26 kg/m3)
The position coefficient considers the position of the layer within the assembly (in direction of the heat flux),
because the layers preceding and backing the layer under consideration have an influence on its fire
behaviour. EN 1995-1-2 gives tabulated data for position coefficients for wall and floor assemblies with
claddings made of one or two layers of wood-based panels and gypsum plasterboards, and void or insulation-
filled cavities. The position coefficients were determined based on testing of non-load bearing wall
assemblies, both in full and small scale. This means that the position coefficients that are given are limited to
a small number of timber constructions.

5.5 Improved design method for separating function of timber constructions

5.5.1 Introduction
The design method for the verification of the separating function of wall and floor assemblies that is given in
EN 1995-1-2 is based on input data that were deduced from a limited number of fire tests of wall assemblies,
and covers therefore only a limited area of timber structures. For this reason, a research project on the
separating function of timber assemblies was conducted and recently completed at ETH Zurich. As a final
result, an improved design method for determining the separating function of timber structures has been
developed, based on an extensive experimental results and finite-element thermal analysis [8, 9]. The design
method is capable of considering timber assemblies with an unlimited number of layers made of gypsum
plasterboards, wood panels or combinations thereof. The cavity may be void or filled with glass wool or
stone wool insulation. The design method considers the following materials:
• Solid timber with characteristic density ≥ 290 kg/m3
• Cross-laminated timber with characteristic density ≥ 290 kg/m3
• Laminated Veneer Lumber (LVL) with characteristic density ≥ 480 kg/m3
• Oriented Strand Board (OSB) according to EN 300 [10] with characteristic density ≥ 550 kg/m3
• Particleboards according to EN 312 [11] with characteristic density ≥ 500 kg/m3
• Fibreboards according to EN 622-2 [12], EN 622-3 [13] or EN 622-5 [14]
with characteristic density ≥ 500 kg/m3
• Plywood according to EN 636 [15] with characteristic density ≥ 400 kg/m3
• Gypsum plasterboards Type A, H and F according to EN 520 [16]
• Gypsum fibre boards according to EN 15283-2 [17]
• Mineral wool insulation according to EN 13162 [18]

51
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The developed design method is based on the additive component method given in EN 1995-1-2. The total
fire resistance is therefore taken as the sum of the contributions from the different layers (claddings, void or
insulated cavities), considering different heat transfer paths (see Figure 5.1) and according to their function
and interaction (see Figure 5.2):

i = n −1
t ins = ∑ t prot,i + t ins,n (4)
i =1

with i = n −1 Sum of the protection times tprot,i of the layers (in the direction of the heat flux)
∑ t prot,i preceding the last layer of the assembly on the side not exposed to fire [min]
i =1

tins,n Insulation time tins,n of the last layer of the assembly on the side not exposed to fire
[min]

Figure 5.2 Timber frame wall-and-floor assemblies: numbering and function of the different layers
Protection and insulation times of the layers can be calculated according to following general equations,
taking into account the basic values of the layers, the coefficients for the position of the layers in the
assembly and the coefficients for the joint configurations:

t prot ,i = ( t prot ,0,i ⋅ k pos,exp,i ⋅ k pos, un exp,i + Δt i ) ⋅ k j,i (5)

t ins, n = ( t ins,0, n ⋅ k pos,exp, n + Δt n ) ⋅ k j, n (6)

with tprot,0,i Basic protection value [min] of Layer i (see Figure 5.2 and Table 5.1)
tins,0,n Basic insulation value [min] of the last layer n of the assembly on the fire-
unexposed side (see Figure 5.2 and Table 5.1)
kpos,exp,i, kpos,exp,n Position coefficient that takes into account the influence of layers preceding
the layer considered (see Table 5.2)
kpos,unexp,i Position coefficient that takes into account the influence of layers backing
the layer considered (see Table 5.3)
Δt i , Δt n Correction time [min] for layers protected by Type F gypsum plasterboards
as well as gypsum fibreboards (see Table 5.4)
kj,i, kj,n Joint coefficient (see Table 5.5)
The coefficients of the design method (basic values, correction time and position coefficients) were
calculated by extensive finite-element thermal simulations based on physical models for the heat transfer
through separating multi-layered construction [9,19]. The coefficients given by general equations permit

52
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

replacement of the tabulated data given in EN 1995-1-2. The material properties used for the finite element
thermal simulations were calibrated and validated by fire tests performed on unloaded specimens at Empa
(Swiss Laboratories for Materials Testing and Research) in Duebendorf, using fire exposure of the standard
temperature-time curve. The design method was verified with full-scale fire tests [20-24], as well as in
addition with 27 full-scale fire tests recently performed in Austria [25] on slabs and walls using current EN
testing standards. The comparison between test results and the design method shows that the improved
model is able to predict the fire resistance of timber assembly safely. All details with regard to the
development and validation of the design method can be found in [9]. The developed design method
significantly improves the EN 1995-1-2 design method, and permits verification the separating function of a
large number of common timber assemblies. The coefficients of the design method are explained below.

5.5.2 Basic values

The basic insulation value tins,0 corresponds to the fire resistance of a single layer without the influence of
adjacent materials, i.e. the average temperature rise over the whole of the non-exposed surface is limited to
140 K, and the maximum temperature rise at any point of that surface does not exceed 180 K (for fire
exposure of the standard temperature-time curve). These temperature criteria safely prevent ignition of
objects in the neighbouring compartment. The temperature of the layer at the beginning of the fire on the side
exposed to fire, as well as on the non- exposed side, is assumed to be 20 °C. The basic insulation value can
be assessed by tests in accordance with (for example) EN 1364-1 [26] or FE thermal analysis. It should be
noted that, for FE thermal analysis, only the 160 °C average temperature criterion is used (see Figure 5.3).

Figure 5.3. Definition of the basic insulation value tins,0 according to EN 1995-1-2
Single-layer wall and floor assemblies are only a limited application area for timber assemblies. Most
structures consist of assemblies having two or more layers. The contribution of each preceding layer to the
separating function of the construction is mainly protection of the following layers. It therefore seems more
appropriate to introduce a basic protection value tprot,0, defined as the time until loss of the fire protective
function, in a similar manner as for evaluation of fire-protective claddings of load-bearing timber structures
in accordance with EN 13501-2 [27]. The testing method for fire-protective claddings in accordance with EN
14135 [28] is performed with 19 mm particleboard backing for the studied layer. The contribution to the fire
protection of the cladding may be assumed to be satisfied where the average temperature rise over the whole
exposed surface of the particleboard is limited to 250 K, and the maximum temperature rise at any point on
that surface does not exceed 270 K. In the same way as for EN 13501-2, the definition of the basic protection
value tprot,0 is illustrated in Figure 5.4. The temperature of the layer at the beginning of the analysis on the
fire-exposed side as well as on the unexposed side is assumed to be 20 °C. For FE thermal analysis, only the
average temperature criterion of 270 °C is used. It should be noted that EN 1995-1-2 gives rules for
calculation of the start of charring tch of timber surfaces protected by fire-protective claddings made of wood-
based panels or wood panelling, as well as gypsum plasterboards, by assuming that charring starts at a
temperature of 300 °C. Although the EN 13501-2 temperature criteria of 270 / 290 °C are slightly lower than
300 °C (i.e. conservative), the tprot,0 basic protection value has the same significance as the tch start of charring
defined by EN 1995-1-2.

53
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 5.4. Definition of the basic protection value tprot,0 according to EN 13501-2 [24]
Table 5.1 gives the equations for calculation of the basic insulation value tins,0 as well as the basic protection
value tprot,0 for different materials that were systematically calculated using finite-element numerical
simulations and verified with fire tests [8, 9]. Only the basic protection value tprot,0,i is given for stone wool
and glass wool insulation, as wall and floor assemblies with the insulation as the last layer of the assembly
are rarely used in buildings.

Table 5.1. Basic insulation value tins,0 and basic protection value tprot,0 for different materials

Basic insulation
Material Basic protection value tprot,0,i [min]
value tins,0,n [min]
1.4 1.2
Gypsum plasterboard, ⎛h ⎞ ⎛h ⎞
24 ⋅ ⎜ i ⎟ 30 ⋅ ⎜ i ⎟
gypsum fibre board ⎝ 15 ⎠ ⎝ 15 ⎠
1.4 1.1
Solid timber, cross-laminated ⎛h ⎞ ⎛h ⎞ hi
19 ⋅ ⎜ i ⎟ 30 ⋅ ⎜ i ⎟ ≤
timber, LVL ⎝ 20 ⎠ ⎝ 20 ⎠ β0
1.4 1.1
⎛h ⎞ ⎛h ⎞ hi
Particleboard, fibreboard 22 ⋅ ⎜ i ⎟ 33 ⋅ ⎜ i ⎟ ≤
⎝ 20 ⎠ ⎝ 20 ⎠ β0
1.4 1.1
⎛h ⎞ ⎛h ⎞ hi
OSB, plywood 16 ⋅ ⎜ i ⎟ 23 ⋅ ⎜ i ⎟ ≤
⎝ 20 ⎠ ⎝ 20 ⎠ β0
Stone wool insulation
0 0.3 ⋅ h i ( 0.75⋅log(ρ i ) − ρ i 400)
with ρ ≥ 26 kg/m3

for hi < 40 mm: 0

Glass wool insulation 0 for hi ≥ 40 mm:
with ρ ≥ 15 kg/m3
(0.0007 ⋅ ρ i + 0.046) ⋅ h i + 13 ≤ 30

Where hi: Thickness of the layer considered [mm]

ρi: Density of the layer considered [kg/m3]
β0: One-dimensional charring rate (β0 = 0.65 mm/min)
The fire behaviour of cross-laminated timber panels is influenced by the behaviour of the adhesive used for
bonding the cross-laminated timber panels [29]. For cross-laminated timber panels with gaps less than 2 mm,
where the char layer does not fall off when the char front has reached a bonded connection, the fire
resistance with regard to the separation function can be calculated in the same way as for solid timber panels,
considering only the total thickness of the cross-laminated timber panels. If the char layer falls off when the
char front has reached a bonded connection, then an increased charring rate must be considered (see Section
6.3.4). In this case, for simplicity, the fire resistance with regard to the separation function can be calculated
considering the single layers of the cross-laminated timber panels.

54
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

5.5.3 Position coefficients

The position coefficient considers the position of the layer within the assembly (in the direction of the heat
flux), because the layers preceding and backing the layer under consideration affect its fire behaviour. The
physical meaning of the position coefficient can be explained by looking at the fire behaviour of the timber
assembly with three layers as shown in Figure 5.5. For simplicity, it is assumed that each layer has the same
thickness and density, and the influence of joints is neglected. In this case, the basic protection value for each
layer is the same (tprot,0,1 = tprot,0,2 = tprot,0,3). The first layer is directly exposed to fire and backed by the second
layer. When the fire starts, the temperature of both sides of all layers is 20 °C (see 5.5a). The contribution of
the first layer to the total fire resistance is defined as tprot,1. The position coefficient kpos,1 of the first layer can
be described as the ratio tprot,1 to tprot,0,1, and depends on the layer backing the first layer.
The second layer is protected by the first layer. It is conservatively assumed that, after failure of protection
by the first layer (temperature of 270 °C at the interface) at time t = tprot,1, the second layer is directly exposed
to fire. The main difference in comparison with the initially unprotected first layer is that the temperature of
the second layer on the fire-exposed side is 270 °C (as defined previously), and the temperature on the
unexposed side is equal to or greater than 20 °C, depending on the thickness of the second layer and the
material preceding and backing the layer (see Figure 5.5b). In addition, the temperature in the fire
compartment is already at a high level, while no protective layer exists to reduce the effect of the temperature.
For these reasons, the contribution of the second layer to the total fire resistance is lower than the
contribution of the first layer, i.e. tprot,2 < tprot,1. The position coefficient kpos,2 of the second layer can be
described as the ratio tprot,2 to tprot,0,2 and is < 1.0. For the same physical reasons, EN 1995-1-2 assumes that,
after failure of a cladding, charring of initially unprotected surfaces takes place at an increased rate. The third
layer is the last layer in the assembly. For this layer, the 140 K / 180 K temperature criteria should therefore
be applied, and an insulation value (tins,3) must be calculated.

20˚C > 20˚C >

_ 20˚C
temperature distribution

temperature distribution

temperature distribution
_
h 3. WP h 3. WP h 3. WP
20˚C >
_ 20˚C
h 2. WP h 2. WP (2. layer falls off) 270˚C
20˚C
h 1. WP (1. layer falls off) 270˚C
20˚C

a) Start of fire: b) Second layer exposed to fire: c) Third layer exposed to fire:
t=0 t = tprot,1 t = tprot,1 + tprot,2

Figure 5.5. Temperature distribution of timber assembly with three layers at different times

The influence of the layers preceding and backing the layer considered was analysed separately. The position
coefficient kpos,exp considers the influence of the layer preceding the layer studied, while the influence of the
layer backing the layer studied is considered by kpos,unexp.

Finite-element thermal simulations showed that the influence of preheating is small. The position coefficient
kpos,exp is mainly influenced by the time when the layer considered is exposed directly to fire and the material
and thickness of the layer considered. It was therefore possible to determine the position coefficient kpos,exp as
a function of the sum of the protection times of the layers preceding the layer considered ( t prot ,i −1 ), and ∑
the basic protection value tprot,0,i or basic insulation value tins,0,n as relevant for the layer considered, making
calculation of the position coefficient kpos,exp easier for the designer (see Table 5.2).

55
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Results of fire tests supported by finite-element thermal simulations showed that the influence of the layer
backing the layer under consideration is small if the backing layer is made of gypsum or wood. Thus, for
simplicity, it is assumed that kpos,unexp = 1.0 for these cases (see Table 5.3). On the other hand, insulating batts
backing the layer caused the layer to heat up more rapidly, reducing the protection time of the layer. For the
different materials, this effect is allowed for by introducing the position coefficient kpos,unexp (see Table 5.3).

Table 5.2 and 5.3 give the position coefficient kpos,exp and kpos,unexp, that were systemically calculated using
finite-element numerical simulations and verified with fire tests [8, 9]. For the finite-element numerical
simulations, it was assumed that a layer fails (i.e. falls off) when the temperature on the unexposed side of
the layer reaches 270 °C.

Table 5.2. Position coefficient kpos,exp,i and kpos,exp,n. For stone wool and glass wool insulation, only the
position coefficient kpos,exp,i for the protection value tprot,i is given, as wall and floor assemblies with
the insulation as last layer of the assembly are rarely used in buildings.
Material Position coefficient kpos,exp,i and kpos,exp,n
kpos,exp,n for tins,n

1 − 0.6 ⋅
∑ t prot,n −1 for ∑ t prot,n −1 ≤
t ins,0, n
t ins,0, n 2

t ins,0, n t ins,0, n
0.5 ⋅
∑ t prot,n −1
for ∑ t prot,n −1 > 2
Cladding (gypsum,
timber) kpos,exp,i for tprot,i

1 − 0 .6 ⋅
∑ t prot,i −1 for ∑ t prot,i−1 ≤
t prot ,0,i
t prot ,0,i 2

t prot,0,i t prot ,0,i

0.5 ⋅
∑ t prot,i−1
for ∑ t prot,i−1 > 2

kpos,exp,i for tprot,i

1 − 0 .6 ⋅
∑ t prot,i −1 for ∑ t prot,i −1 ≤
t prot ,0,i
Stone wool insulation t prot ,0,i 2

t prot ,0,i t prot ,0,i

0.5 ⋅
∑ t prot,i −1
for ∑ t prot,i −1 > 2

kpos,exp,i for tprot,i

1 − 0 .8 ⋅
∑ t prot,i −1 for ∑ t prot,i −1 ≤
t prot ,0,i
Glass wool insulation t prot ,0,i 4
for hi ≥ 40 mm ( 0.75− 0.002⋅ρ i )
⎡ t prot ,0,i ⎤ t prot ,0,i
(0.001 ⋅ ρ i + 0.27) ⋅ ⎢ ⎥ for ∑ t prot,i−1 >
∑
⎢⎣ t prot ,i −1 ⎥⎦ 4

With ρi: Density of the layer considered [kg/m3]

56
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 5.3. Position coefficient kpos,unexp

kpos,unexp,i for layers backed by
kpos,unexp,i for layers backed by
Material of the layer considered cladding made of gypsum or
insulation
timber
Gypsum plasterboard,
0.5 ⋅ h i 0.15
gypsum fibre board
Solid timber, cross-laminated
0.35 ⋅ h i 0.21
timber, LVL

Particleboard, fibreboard 0.41 ⋅ h i 0.18

1.0
OSB, plywood 0.5 ⋅ h i 0.15

Stone wool insulation 0.18 ⋅ h i ( 0.001⋅ρi + 0.08)

hi2
Glass wool insulation 0.01 ⋅ h i − + ρi 0.09 − 1.3
30000
With hi: Thickness of the layer considered [mm]

57
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The position coefficients kpos,exp given in Table 5.2 were calculated assuming that the layers fall off when the
temperature of 270 °C is reached on the unexposed side of the layers. Fire tests showed that this assumption
is conservative for Type F gypsum plasterboards and gypsum fibre boards [9]. The protection or insulation
times of layers protected by Type F gypsum plasterboards or gypsum fibre board can therefore be increased
using respective correction times of ∆ti and ∆tn. Table 5.4 shows the values of these correction times that
were systemically calculated using finite-element numerical simulations. It was assumed that, for floor
assemblies, the Type F gypsum plasterboards or gypsum fibre boards do not fall off until the temperature on
the unexposed side of the board reaches 400 °C, while the corresponding temperature for wall assemblies
was assumed to be a temperature of 600 °C [9].

Table 5.4. Correction time ∆ti and ∆tn of protection and insulation times tprot,i and tins,n of layers protected by
Type F gypsum plasterboards and gypsum fibre boards

Material Floor assemblies Wall assemblies

Δtn for tins,n [min]
0.06 ⋅ t prot, n −1 + 1.1 ⋅ t ins,0, n − 5 0.03 ⋅ t prot, n −1 + 0.9 ⋅ t ins,0, n − 2.3
for t ins,0, n < 8 min for t ins,0, n < 12 min

0.1 ⋅ t prot, n −1 − 0.035 ⋅ t ins,0, n + 1.2 0.22 ⋅ t prot, n −1 − 0.1 ⋅ t ins,0, n + 4.7

Cladding for t ins,0, n ≥ 8 min for t ins,0, n ≥ 12 min

(gypsum,
timber) Δti for tprot,i [min]
0.06 ⋅ t prot, i −1 + 1.1 ⋅ t prot,0, i − 5 0.03 ⋅ t prot, i −1 + 0.9 ⋅ t prot,0, i − 2.3
for t prot,0, i < 8 min for t prot,0,i < 12 min

0.1 ⋅ t prot, i −1 − 0.035 ⋅ t prot,0, i + 1.2 0.22 ⋅ t prot, i −1 − 0.1 ⋅ t prot,0, i + 4.7
for t prot,0, i ≥ 8 min for t prot,0,i ≥ 12 min

Δti for tprot,i [min]

0.1 ⋅ t prot, i −1 + t prot,0, i − 1.0
Insulation
(stone wool for t prot,0, i < 6 min
and glass
wool 0.1 ⋅ t prot, i −1 − 0.035 ⋅ t prot,0, i
insulation) 0.22 ⋅ t prot, i −1 − 0.1 ⋅ t prot,0, i + 3.5
for t prot,0, i ≥ 6 min

58
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

5.5.4 Joint coefficient

The joint coefficient considers the influence of joints in panels (claddings) not backed by battens or
structural members or panels, and their influence on the protection and insulation time of these layers.
According to EN 1995-1-2, joints with a width greater than 2 mm are not allowed. Results of the fire tests
showed that the influence of joints with a width less than 2 mm, and backed by a layer, is small. Thus, for
simplicity, the design method considers the influence of joints only for the last layer of the assembly on the
unexposed side and for the layer preceding a void cavity (see Table 5.5). For all other layers, it is assumed
that kj,i = 1.0.

Table 5.5. Joint coefficient kj,i

kj,i for tprot,i

Material Joint type kj,n for tins,n Layer backed by battens

Layer backed by a
or panels or structural
void cavity
members or insulation

<
_ 2mm

0.3 0.3 1.0

>
_ 30mm

<
_ 2mm <
_ 2mm

0.4 0.4 1.0

Cladding (timber) >
_ 15mm >
_ 30mm

<
_ 2mm <
_ 2mm

0.6 0.6 1.0

>
_ 15mm >
_ 30mm

no joint 1.0 1.0 1.0

Gypsum 0.8 0.8 1.0

plasterboard,
gypsum fibre board
no joint 1.0 1.0 1.0

Insulation (stone - 0.8 1.0

wool and glass
wool insulation)
no joint - 1.0 1.0

59
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

5.5.5 Void cavities

The influence of void cavities between two layers is considered in the design method by modifying the
position coefficient kpos,exp for the layer on the side of the cavity not exposed to fire and the position
coefficient kpos,unexp for the layer on the side of the cavity that is exposed to fire (see Table 5.6).

Table 5.6. Modification of position coefficient kpos,exp and kpos,unexp for void cavities

Layer on the exposed side of

Material Layer on the unexposed side of the cavity
the cavity

kpos,unexp,i 3 x Δti (or 3 x Δtn)

Cladding (gypsum,
timber) according to according to
Table 5.3, column 3 1.6 x kpos,exp,i Table 5.4
according to
Insulation (stone wool Table 5.2 Δti (or Δtn)
and glass wool kpos,unexp,i = 1.0 according to
insulation) Table 5.4

5.5.6 Detailing
The same EN 1995-1-2 rules must be satisfied in order to avoid a premature failure (i.e. fall off) of cladding
and insulation. Insulating layers that are taken into account in the calculation should be tightly fitted and
fixed to the timber frame such that premature failure or slumping is prevented. Edge distances strongly
influence the fire behaviour of cladding. Increasing the edge distances in comparison with those specified in
the rules for normal temperature design is beneficial with respect to failure of claddings. For Type F gypsum
plasterboards and gypsum fibre boards, the penetration length of fasteners into the residual cross-section
should not be less than 10 mm. In addition, it must be verified that the failure time of Type F gypsum
plasterboards and gypsum fibre boards with respect to pull-out failure of fasteners exceeds the protection
time calculated according to Equation 5.

5.5.7 Other materials

Coefficients of the design method (basic values, correction time and position coefficients) for specific
products can be evaluated by means of fire tests and FE-thermal simulations (see e.g. [30]) and found in data
sheets from the producers.

5.6 Examples
5.6.1 Worked example 1
A timber floor structure consists of joists and claddings made of timber boards and gypsum plasterboards
(see Figure 5.6). The cavities of the timber floor are completely filled with stone wool insulation, with a
density of 30 kg/m3. The joints of the decking (20 mm thick solid timber panels) are single tongued and
grooved with a maximum gap width of 2 mm. The required fire resistance is EI 30.

60
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 5.6. Cross-section of the timber frame floor assembly

The insulation time tins of the timber floor should be calculated considering different heat transfer paths.
However, for this example, only Heat Transfer Path 3 will be analysed (see Figure 5.6).
An insulation time must be calculated for the last layer of the floor assembly on the unexposed side (solid
timber panel), while for the other layers with protective function a protection time must be calculated.
Protection time of gypsum plasterboard Type A (Layer 1)
1.2 1.2
⎛h ⎞ ⎛ 12.5 ⎞
t prot,0,1 = 30 ⋅ ⎜ 1 ⎟ = 30 ⋅ ⎜ ⎟ = 24.1 min
⎝ 15 ⎠ ⎝ 15 ⎠
k pos,exp,1 = 1.0 (no layer preceding the gypsum plasterboard)

k pos, un exp,1 = 1.0 (OSB backing the gypsum plasterboard)

k j,1 = 1.0 (OSB backing the gypsum plasterboard)

t prot ,1 = ( t prot , 0 ,1 ⋅ k pos , exp, 1 ⋅ k pos , un exp, 1 + Δ t 1 ) ⋅ k j,1 = (24 . 1 ⋅ 1 . 0 ⋅ 1 . 0 + 0 ) ⋅ 1 . 0 = 24 . 1min

Protection time of OSB (Layer 2)

1.1 1.1
⎛h ⎞ ⎛ 12 ⎞
t prot,0,2 = 23 ⋅ ⎜ 2 ⎟ = 23 ⋅ ⎜ ⎟ = 13.1 min
⎝ 20 ⎠ ⎝ 20 ⎠
i =1 t prot ,0,2 t prot ,0,2 13.1 13.1
∑ t prot,i > 2
→ k pos,exp, 2 = 0.5 ⋅ i =1
⇒ 24.1 >
2
→ k pos,exp, 2 = 0.5 ⋅
24.1
= 0.37
i =1
∑ t prot,i
i =1

k pos, un exp,2 = 0.5 ⋅ h 02.15 = 0.5 ⋅ 120.15 = 0.73 (insulation backing the OSB)

k j,2 = 1.0 (insulation backing the OSB)

t prot , 2 = ( t prot ,0, 2 ⋅ k pos , exp, 2 ⋅ k pos , un exp, 2 + Δt 2 ) ⋅ k j, 2 = (13.1 ⋅ 0.37 ⋅ 0.73 + 0 ) ⋅ 1.0 = 3.5min

Protection time of stone wool insulation (Layer 3)

30
ρ3 0.75⋅log(30 ) −
t prot,0,3 = 0.3 ⋅ h 30.75⋅log(ρ 3 ) − 400 = 0.3 ⋅ 80 400 = 27.7 min

61
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

i=2 t prot ,0,3 t prot ,0,3 27.7 27.7

∑ t prot,i > 2
→ k pos,exp, 3 = 0.5 ⋅ i=2
⇒ 24.1 + 3.5 = 27.6 >
2
→ k pos,exp,3 = 0.5 ⋅
27.6
= 0 .5
i =1
∑ t prot,i
i =1
k pos , un exp, 3 = 1.0 (solid timber panel backing the insulation)

k j,3 = 1.0 (solid timber panel backing the insulation)

t prot ,3 = ( t prot ,0,3 ⋅ k pos, exp, 3 ⋅ k pos, un exp, 3 + Δt 3 ) ⋅ k j,3 = (27.7 ⋅ 0.5 ⋅ 1.0 + 0 ) ⋅ 1.0 = 13.9min

It should be noted that the calculated protection time of stone wool insulation can be taken into account in
the calculation only if the stone wool insulation is tightly fitted and fixed to the timber frame such that
premature failure (i.e. fall off) or slumping is prevented.

Insulation time of solid timber panel (Layer 4, last layer)

1.4 1.4
⎛h ⎞ ⎛ 20 ⎞
t ins,0,4 = 19 ⋅ ⎜ 4 ⎟ = 19 ⋅ ⎜ ⎟ = 19.0 min
⎝ 20 ⎠ ⎝ 20 ⎠
i=3
t ins,0,4 t ins,0,4 19.0 19.0
∑ t prot,i > 2
→ k pos,exp, 4 = 0.5 ⋅ i =3
⇒ 24.1 + 3.5 + 13.9 = 41.5 >
2
→ k pos,exp, 4 = 0.5 ⋅
41.5
= 0.34
i =1
∑ t prot,i
i =1
k j, 4 = 0.4 (solid timber panel with single tongued and grooved joints)

t ins , 4 = ( t ins , 0 , 4 ⋅ k pos , exp, 4 + Δ t 4 ) ⋅ k j, 4 = (19 . 0 ⋅ 0 . 34 ⋅ 1 . 0 + 0 ) ⋅ 0 . 4 = 2.6min

Insulation time tins (fire resistance) of the timber floor

i =3
t ins = ∑ t prot,i + t ins,4 = t prot,1 + t prot,2 + t prot,3 + t ins,4 = 24.1 + 3.5 + 13.9 + 2.6 = 44.1 min ≥ t req = 30 min
i =1

5.6.2 Worked example 2

A timber wall consists of studs and claddings made of Type F gypsum plasterboards (see Figure 5.7). The
cavities of the timber wall are voids. The required fire resistance is EI 60.

Figure 5.7. Cross-section of the timber frame wall assembly

The insulation time tins of the timber wall should be calculated for different heat transfer paths. However, for
this example, only Heat Transfer Path 2 will be analysed (see Figure 5.7).
An insulation time must be calculated for the last layer of the wall assembly on the unexposed side, while for
the first layer with protective function a protection time must be calculated.

62
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Protection time of gypsum plasterboard type F (Layer 1)

1.2 1.2
⎛h ⎞ ⎛ 15 ⎞
t prot,0,1 = 30 ⋅ ⎜ 1 ⎟ = 30 ⋅ ⎜ ⎟ = 30.0 min
⎝ 15 ⎠ ⎝ 15 ⎠
k pos, exp,1 = 1.0 (no layer preceding the gypsum plasterboard)

Δt1 = 0 min

k pos, un exp,1 = 0.5 ⋅ h10.15 = 0.5 ⋅ 150.15 = 0.75 (void cavity backing the gypsum plasterboard)

k j,1 = 1.0 (gypsum plasterboard without joints)

t prot ,1 = ( t prot ,0,1 ⋅ k pos ,exp,1 ⋅ k pos , un exp,1 + Δt1 ) ⋅ k j,1 = (30.0 ⋅ 1.0 ⋅ 0.75 + 0 ) ⋅ 1.0 = 22.5min

Insulation time of gypsum plasterboard type F (Layer 2, last layer)

1.4 1.4
⎛h ⎞ ⎛ 15 ⎞
t ins,0,2 = 24 ⋅ ⎜ 2 ⎟ = 24 ⋅ ⎜ ⎟ = 24.0 min
⎝ 15 ⎠ ⎝ 15 ⎠
i =1
t ins,0,2 t ins,0,2 24.0 24.0
∑ t prot,i > 2
→ k pos,exp, 2 = 1.6 ⋅ 0.5 ⋅ i =1
⇒ 18 >
2
→ k pos,exp, 2 = 1.6 ⋅ 0.5 ⋅
22.5
= 0.83
i =1
∑ t prot,i
i =1

( )
t ins ,0, 2 ≥ 12 min → 3 ⋅ Δt 2 = 3 ⋅ 0.22 ⋅ t prot ,1 − 0.1 ⋅ t ins ,0, 2 + 4.7 = 3 ⋅ (0.22 ⋅ 22.5 − 0.1 ⋅ 24.0 + 4.7 ) = 21.8 min

k j, 2 = 1.0 (gypsum plasterboard without joints)

t ins, 2 = ( t ins,0,1 ⋅ k pos,exp, 2 + 3 ⋅ Δt 2 ) ⋅ k j, 2 = (24.0 ⋅ 0.83 ⋅ 1.0 + 21.8 ) ⋅ 1.0 = 41.7min

Insulation time tins (fire resistance) of the timber wall

i =1
t ins = ∑ t prot,i + t ins,2 = t prot,1 + t ins,2 = 22.5 + 41.7 = 64.2 min ≥ t req = 60 min
i =1

5.6.3 Worked example 3

A timber floor consists of joists and claddings made of timber boards and gypsum plasterboards (see Figure
5.8). The cavities of the timber floor are completely filled with glass wool insulation with a density of
20 kg/m3. The joints of the decking (20 mm thick solid timber panels) are single tongued and grooved with a
maximum gap width of 2 mm. The required fire resistance is EI 30.

Figure 5 8. Cross-section of the timber frame floor assembly

63
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The insulation time tins of the timber floor should be calculated for different heat transfer paths. However, for
this example, only Heat Transfer Path 3 will be analysed (see Figure 5.8).
An insulation time must be calculated for the last layer of the floor assembly on the unexposed side (solid
timber panel), while for the other layers with protective functions a protection time must be calculated.
Protection time of Type F gypsum plasterboard (Layer 1)
1.2 1.2
⎛h ⎞ ⎛ 15 ⎞
t prot,0,1 = 30 ⋅ ⎜ 1 ⎟ = 30 ⋅ ⎜ ⎟ = 30 min
⎝ 15 ⎠ ⎝ 15 ⎠
k pos,exp,1 = 1.0 (no layer preceding the gypsum plasterboard)

k pos, un exp,1 = 1.0 (Type A gypsum plasterboard backing the Type F gypsum plasterboard )

k j,1 = 1.0 (Type A gypsum plasterboard backing the Type F gypsum plasterboard)

t prot ,1 = ( t prot ,0,1 ⋅ k pos ,exp,1 ⋅ k pos , un exp,1 + Δt 1 ) ⋅ k j,1 = (30 ⋅ 1.0 ⋅ 1.0 + 0 ) ⋅ 1.0 = 30 min

Protection time of gypsum plasterboard Type A (Layer 2)

1.2 1.2
⎛h ⎞ ⎛ 12.5 ⎞
t prot,0,2 = 30 ⋅ ⎜ 2 ⎟ = 30 ⋅ ⎜ ⎟ = 24.1 min
⎝ 15 ⎠ ⎝ 15 ⎠
i =1 t prot ,0,2 t prot ,0,2 24.1 24.1
∑ t prot,i > 2
→ k pos,exp, 2 = 0.5 ⋅ i =1
⇒ 30 >
2
→ k pos,exp, 2 = 0.5 ⋅
30
= 0.45
i =1
∑ t prot,i
i =1

k pos,un exp,2 = 0.5 ⋅ h 02.15 = 0.5 ⋅ 12.50.15 = 0.73 (insulation backing the Type A gypsum plasterboard)

t prot ,0, 2 ≥ 8 min → Δt 2 = 0.1 ⋅ t prot ,1 − 0.035 ⋅ t prot ,0, 2 + 1.2 = 0.1 ⋅ 30 − 0.035 ⋅ 24 .1 + 1.2 = 3.3 min

k j,2 = 1.0 (insulation backing the Type A gypsum plasterboard)

t prot , 2 = ( t prot ,0, 2 ⋅ k pos,exp, 2 ⋅ k pos, un exp, 2 + Δt 2 ) ⋅ k j, 2 = (24.1 ⋅ 0.45 ⋅ 0.73 + 3.3) ⋅ 1.0 = 11.2min

Protection time of glass wool insulation (Layer 3)

t prot,0,3 = (0.0007 ⋅ ρ 3 + 0.046 ) ⋅ h 3 + 13 = (0.0007 ⋅ 20 + 0.046) ⋅ 80 + 13 = 17.8 min
( 0 .75 − 0 .002 ⋅ρ 3 )
⎡ ⎤
i=2 ⎢ ⎥
t prot , 0 , 3 ⎢ t prot , 0 , 3 ⎥
∑ t prot ,i > 4
→ k pos ,exp, 3 = ( 0 .001 ⋅ ρ 3 + 0 .27 ) ⋅ i = 2
⎢ ⎥
i =1 ⎢
⎢⎣ i =1
∑t prot ,i ⎥
⎥⎦
( 0 .75 − 0 .002 ⋅20 )
17 .8 ⎡ 17 .8 ⎤
⇒ 30 + 11 .2 = 41 .2 > → k pos ,exp, 3 = ( 0 .001 ⋅ 20 + 0 .27 ) ⋅ ⎢ ⎥ = 0 .16
4 ⎣ 41 .2 ⎦
k pos , un exp, 3 = 1.0 (solid timber panel backing the insulation)

k j,3 = 1.0 (solid timber panel backing the insulation)

t prot ,3 = ( t prot ,0,3 ⋅ k pos,exp, 3 ⋅ k pos, un exp, 3 + Δt 3 ) ⋅ k j,3 = (17.8 ⋅ 0.16 ⋅ 1.0 + 0 ) ⋅ 1.0 = 2.8min

64
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Insulation time of solid timber panel (Layer 4, last layer)

1.4 1.4
⎛h ⎞ ⎛ 20 ⎞
t ins,0, 4 = 19 ⋅ ⎜ 4 ⎟ = 19 ⋅ ⎜ ⎟ = 19 min
⎝ 20 ⎠ ⎝ 20 ⎠
i=3 t ins,0,4 t ins,0,4 19 19
∑ t prot,i > 2
→ k pos,exp, 4 = 0.5 ⋅ i=3
⇒ 30 + 11.2 + 2.8 = 44 >
2
→ k pos,exp, 4 = 0.5 ⋅
44
= 0.33
i =1
∑ t prot,i
i =1
k j, 4 = 0.4 (solid timber panel with single tongued and grooved joints)

t ins , 4 = ( t ins ,0, 4 ⋅ k pos,exp, 4 + Δt 4 ) ⋅ k j, 4 = (19.0 ⋅ 0.33 ⋅ 1.0 + 0 ) ⋅ 0.4 = 2.5min

Insulation time tins (fire resistance) of the timber floor

i =3
t ins = ∑ t prot,i + tins,4 = t prot,1 + t prot,2 + t prot,3 + tins,4 = 30 + 11.2 + 2.8 + 2.5 = 46.5 min ≥ t req = 30 min
i =1

5.7 References
[1] EN 1995-1-2, Eurocode 5 – Design of timber structures, Part 1-2: General – Structural fire design,
European Standard, CEN, Brussels, 2004.
[2] König J., Oksanen T., Towler K., A review of component additive methods used for the determination
of fire resistance of separating light timber frame construction, International Council for Research and
Innovation in Building and Construction, Working Commission W18 – Timber Structures, Delft,
Netherlands, CIB-W18/33-16-3, 2000.
[3] British Standard BS 5268-4: Structural use of timber - Section 4.2. Recommendations for calculating
fire resistance of timber stud walls and joisted floor constructions, BSI, 1990.
[4] National Building Code of Canada NBCC - Volume 2, Canadian Commission on Building and Fire
Codes, Institute for Research in Construction, National Research Council of Canada, Ottawa, Canada,
2005.
[5] Norén J., Additionsmetoden – Beräkning av brandmotstånd hos avskiljande väggar (Addition method
– Calculation of fire resistance for separating wood frame walls), Trätek – Swedish Institute for Wood
Technology Research, Report I 9312070, 1994.
[6] Östman B., König J., Norén J., Contribution to fire resistance of timber frame assemblies by means of
fire protective boards, Proceedings of the 3rd International Fire and Materials Conference, Washington
D.C., 1994.
[7] ENV 1995-1-2, Eurocode 5 – Design of timber structures, Part 1-2: General rules – Structural fire
design, European Prestandard, CEN, Brussels, 1994.
[8] Schleifer V., Frangi A., Fontana M., Experimentelle Untersuchungen zum Brandverhalten von
Plattenelementen, Institute of Structural Engineering IBK, ETH Zurich, IBK-report No. 302, ISBN
978-3-7281-3149-2, May 2007, vdf Hochschulverlag AG.
[9] Schleifer V., Zum Verhalten von raumabschliessenden mehrschichtigen Holzbauteilen im Brandfall,
PhD Thesis ETH No. 18156, ETH Zurich, 2009.
[10] EN 300, Oriented Strand Boards (OSB) - Definitions, classification and specifications, European
Standard, CEN, Brussels, 2006.
[11] EN 312, Particleboards - Specifications, European Standard, CEN, Brussels, 2003.

65
Chapter 5 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

[12] EN 622-2, Fibreboards - Specifications - Part 2: Requirements for hardboards, European Standard,
CEN, Brussels, 2004.
[13] EN 622-3, Fibreboards - Specifications - Part 3: Requirements for medium boards, European Standard,
CEN, Brussels, 2004.
[14] EN 622-5, Fibreboards - Specifications - Part 5: Requirements for dry process boards (MDF),
European Standard, CEN, Brussels, 2006.
[15] EN 636: Plywood - Specifications, European Standard, CEN, Brussels, 2003.
[16] EN 520, Gypsum plasterboards - Definitions, requirements and test methods, European Standard, CEN,
Brussels, 2004.
[17] EN 15283-2, Gypsum boards with fibrous reinforcement - Definitions, requirements and test methods
- Part 2: Gypsum fibre boards, European Standard, CEN, Brussels, 2008.
[18] EN 13162, Thermal insulation products for buildings - Factory made mineral wool (MW) products –
Specification, European Standard, CEN, Brussels, 2001.
[19] Frangi A., Schleifer V., Fontana M., Hugi E., Experimental and numerical analysis of gypsum
plasterboards in fire, Fire Technology, Paper in press.
[20] König J., Fire resistance of timber joists and load bearing wall frames, Trätek, Rapport I 99412071,
Stockholm, 1995.
[21] König J., Norén J., Timber frame assemblies exposed to standard and parametric fires - Part 1: Fire
tests, Trätek, Rapport I 9702015, Stockholm, February 1997.
[22] Sultan M.A., Séguin Y.P., Leroux P., Results of fire resistance tests on full-scale floor assemblies,
Internal Report No. 764, Institute for Research in Construction, National Research Council of Canada,
Ottawa, Canada, Mai 1998.
[23] Sultan M.A., Lougheed G. D., Results of fire resistance tests on full scale gypsum board wall
assemblies, IR-833, Institute for Research in Construction, National Research Council of Canada,
August 2002.
[24] Collier P.C.R., Buchanan A.H., Fire resistance of lightweight timber framed walls, Fire Technology,
38, 125-145, 2002.
[25] Teibinger M., Matzinger I., Charwat-Pessler J., Grundlagen zur Bewertung des Feuerwiderstandes
von Holzrahmenkonstruktionen, Endbericht, Holzforschung Austria, 2010.
[26] EN 1364-1, Fire resistance tests for non-loadbearing elements – Part 1: Walls, European Standard,
CEN, Brussels, 1999.
[27] EN 13501-2, Fire classification of construction products and building elements – Part 2: Classification
using data from fire resistance tests, excluding ventilation services, European Standard, CEN, Brussels,
2003.
[28] EN 14135, Coverings – Determination of fire protection ability, European Standard, CEN, Brussels,
2004.
[29] Frangi A., Fontana M., Hugi E., Jöbstl R., Experimental analysis of cross-laminated timber panels in
fire, Fire Safety Journal 2009; 44: 1078–1087.
[30] Schleifer V., Frangi A., Untersuchungen zum Raumabschluss von Bauteilen mit Isoresist 1000,
Institute of Structural Engineering IBK, ETH Zurich, Research report, April 2009.

66
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6 Load-bearing timber structures 68

6.1 General 68
6.2 Structural stability 68
6.3 Materials 69
6.3.1 Timber and wood-based materials 69
6.3.2 Gypsum plasterboard 69
6.3.3 Insulation materials 69
6.3.3.1 Mineral wool 69
6.3.4 Adhesives 70
6.4 Charring of timber and wood-based panels 70
6.4.1 General 70
6.4.2 One-dimensional charring 70
6.4.3 Two-dimensional charring 71
6.4.4 Effect of protection 73
6.4.4.1 Large cross-sections 73
6.4.4.2 Small-sized timber frame members 76
6.4.4.3 Determination of start of charring and failure times according to EN
1995-1-2 78
6.4.4.4 Start of charring and failure according to component additive method
given in Chapter 5 79
6.4.4.5 Start of charring and failure times from data base 79
6.4.5 Effect of bonded joints 82
6.5 Mechanical resistance 82
6.5.1 Simplified methods for strength and stiffness parameters 82
6.5.2 Advanced calculation methods 84
6.6 Structural elements 87
6.6.1 Beams and columns exposed on three or four sides 87
6.6.1.1 Unprotected members 87
6.6.2 Solid timber decks and walls 89
6.6.2.1 Tongued and grooved timber decking 89
6.6.2.2 Glued-laminated and stress-laminated timber plates 90
6.6.2.3 Nail-laminated timber plates 91
6.6.2.4 Cross-laminated timber plates (CLT) 92
6.6.3 Timber frame floor and wall assemblies 97
6.6.3.1 General 97
6.6.3.2 Assemblies with cavities completely or partially filled with rock wool
insulation 99
6.6.3.3 Assemblies with void cavities 102
6.6.3.4 Cavities completely filled with glass wool insulation 104
6.6.3.5 Cavities filled with insulation other than glass or rock wool 106
6.6.3.6 Worked examples 106
6.6.4 Light weight timber frame floors with I-joists 110
6.6.4.1 General 110
6.6.4.2 Cross-section 110
6.6.4.3 Failure of cladding fasteners 112
6.6.4.4 Strength parameters 112
6.6.4.5 Finger-joint strength in flanges 113
6.6.4.6 Reduced cross-section method for bending strength 113
6.6.5 Other structural elements 113
6.6.5.1 Timber decks made of hollow core elements 113
6.6.5.2 Timber-concrete composite slabs 115
6.7 Calculation vs. full-scale testing 116
6.8 References 117

67
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6 Load-bearing timber structures

This chapter presents design methods for the verification of structural stability of timber
structures in the event of fire, applying the classification for criterion R for fire
resistance (load-bearing function). Reference is made to EN 1995-1-2 with respect to
charring and strength and stiffness parameters. Alternative design models are presented
as well as new design methods for timber structures not being dealt with in the standard.

Drafting note:
Text in blue: Amendments in October version (language checked).
Text in green: Amendments in November version.
Text market yellow: Incomplete

6.1 General
This section gives guidance for structural fire design of timber structures. Reference is
made to EN 1995-1-2 [1], including Corrigenda [2][3] and other parts of Structural
Eurocodes and, where new knowledge is available, to other references. Some of the
design rules given in informative annexes to Eurocodes may not be applicable in all
European States, see National annexes to Eurocodes. Depending on national regulations,
some of the new design methods given in the following may need agreement by the
Competent Authority. Hence the content of this section should be regarded as the state of
the art, and new items as potential input for future revisions of EN 1995-1-2 [1].

6.2 Structural stability

The model of the structural system adopted for the design must reflect the performance of
the structure in a fire situation. EN 1995-1-2 [1] provides the following alternatives for
verification of the structural performance of the building:
• Member analysis
• Analysis of parts of the structure
• Global structural analysis.

The structural system may be different in a fire situation, e.g. where a structural member
is braced at ambient temperature and the bracing fails in the fire situation, the member
must be regarded as unbraced in the structural fire design, see also 6.6.3.1, last paragraph.
Elements that are used for the stabilisation of the building, e.g. wood-based panels or
gypsum plasterboard in wall or floor diaphragms, often lose their racking resistance in a
fire situation unless they are protected from the fire. This effect on the global structural
system must therefore be taken into account. In redundant structural systems it may be
advantageous to allow for premature failure if an alternative load path is possible, e.g. a
column in a fire compartment.

Unlike steel and concrete, thermal expansion of timber need not be taken into account.

68
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.3 Materials
6.3.1 Timber and wood-based materials
The main properties of timber and wood-based materials relevant in structural fire design
are charring and the reduction of strength and stiffness properties due to elevated
temperature. For simplified design, it is sufficient to consider charring, see 6.4 and
simplified model strength verification, see 6.5.1. For advanced calculations (see 6.5.2),
thermal thermo-mechanical properties of softwoods (solid timber, glued-laminated timber
and LVL) are given in EN 1995-1-2 Annex B. Thermal properties of OSB, plywood and
wood fibreboard are given in 6.5.2.

6.3.2 Gypsum plasterboard

Some properties of gypsum plasterboards and gypsum fibreboards are given in EN 520
[4] and EN 15283-2[5] respectively. With respect to performance in fire of these boards,
these European standards give insufficient information, such as thermal properties for
heat transfer calculations and mechanical properties in fire. The latter are important with
respect to failure of gypsum plasterboard claddings due to thermal degradation. Failure
times (i.e. fall-off times) of gypsum plasterboards are given in EN 1995-1-2 [1] for
gypsum plasterboards Type A and H. For Type F, they must be determined on the basis
of tests. Some data on failure times of gypsum plasterboards are given in [6], see 6.4.4.5.

For thermal properties of gypsum plasterboards, see 6.5.2.

6.3.3 Insulation materials

6.3.3.1 Mineral wool

EN 13162 [7] gives product characteristics of mineral wool (i.e. rock wool and glass
wool) such as thermal conductivity, density and other. However, this European standard
does not classify mineral wool in terms needed for structural fire design. The designer
knows, for example, that rock wool performs better than glass fibre when directly
exposed to fire, however no relevant test method exists to quantify the difference in terms
of product properties. It is well-known that glass wool resists a gas temperature of about
550°C and rock wool a gas temperature of about 1000°C or more without collapse of the
structure of the wool. Both for rock wool and glass wool there is a weak relationship
between thermal conductivity and density. Density is therefore not sufficient to
characterise the performance of the insulation. The terminology used below is therefore
traditional and recognized by the designer. Where density requirements are given, they
refer to the requirements given in EN 1995-1-2 [1]. For better specification of mineral
wool, a new classification of mineral wool is needed with respect to its performance in
fire.

For thermal properties of mineral wool, see 6.5.2.

6.3.3.2 Cellulose insulation

For cellulose insulation no European standards exists. Their performance in fire must be
determined by fire testing. For their use in timber structures, of special interest is the
degree of their capability of providing protection of timber members, to resist
smouldering, shrinkage and surface recession. See also 9.1.2.3.

69
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.3.3.3 Other insulation materials

For more information, see 9.1.2

6.3.4 Adhesives
EN 1995-1-2 Clause 5.2 requires that “adhesives for structural purposes shall produce
joints of such strength and durability that the integrity of the bond is maintained in the
assigned fire resistance period. For bonding of wood to wood, wood to wood-based
materials or wood-based materials to wood-based materials, adhesives of phenol-
formaldehyde and aminoplastic type 1 adhesive according to EN 301 [8] may be used.
For plywood and LVL, adhesives according to EN 314 may be used”.

Since publication of EN 15425 [9], the second sentence above also applies to one-
component polyurethane adhesives. According to EN 301 [8] and EN 15425 [9], the most
severe scenario at elevated temperatures are tests at 70°C, being held over two weeks
under constant loading of the specimens. It has not been shown that this scenario is
sufficient for approval in a fire scenario. Therefore, it has been stressed that there is a
need to establish a new classification for adhesives with respect to their performance in
fire [10] and to develop relevant test methods.

For effect on charring in laminated members, see 6.4.5, for effect on finger joint strength,
see 6.6.4.5. More information is given in 9.1.1.

6.4 Charring of timber and wood-based panels

6.4.1 General
Timber members exposed to fire exhibit charring unless they are protected during the
relevant time of fire exposure. For calculation of the resistance of timber members, the
original cross-section is reduced by the charring depth.

In the following, charring of timber members is divided into:

• One-dimensional charring as a physical property for a specific species, or timber of
specific density or strength class, see 6.4.2 below.
• Two-dimensional charring, including the effects of cross-sectional dimensions and
other effects, see 6.4.3 below.

The charring rates are applicable for any orientation of fire-exposed surfaces and
direction of fire exposure, i.e. there is no distinction between vertical or horizontal
surfaces. For example, for surfaces on floors with fire exposure from above, the same
charring rates apply as for surfaces with fire exposure from below. For fire exposure from
above, fall-off of fire protective claddings supported by a decking are not relevant need
not be assumed.

6.4.2 One-dimensional charring

As a basic value, the one-dimensional charring rate β0 is the charring rate observed for
one-dimensional heat transfer under standard fire exposure of an unprotected semi-
infinite timber slab. The conditions are similar in a slab of limited thickness, or in wide
timber cross-sections remote from corners.

70
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The one dimensional charring depth dchar,0 is expressed as

d char,0 = β 0 t (6.1)

where t is the time of fire exposure and β0 is the one-dimensional charring rate
perpendicular to the grain, as shown in Table 3.1 of EN 1995-1-2 [1]. For charring in the
direction of the grain, these charring rates should be doubled. The one-dimensional
charring rate given for softwoods is valid for European species (0,65 mm/min); it may
also be applicable to other species, e.g. radiata pine, while the charring rates of several
North American softwoods may considerably deviate, see Schaffer [11]. The influence of
density within European strength classes for softwoods (solid timber, glued-laminated
timber and LVL) is small and therefore neglected.

Figure 6.1– One-dimensional charring of wide cross-section

The one-dimensional charring rate of wood panelling and wood-based panels is given for
a panel thickness of 20 mm and a density of 450 kg/m3. For other thicknesses and
densities, β0 should be multiplied by factors kh and kρ respectively, see EN 1995-1-2
Subclause 3.4.2. The charring rates for wood panelling and wood-based panels as given
in EN 1995-1-2 do not take into account the fact that the panels or wood panelling burn
through much more quickly at joints. In addition, the charring rates given in Table 3.1 of
[1] are not consistent with respect to the distinction made between wood panelling and
timber. For example, tongued and grooved timber decking more than 30 mm thick would
imply a charring rate for timber rather than for wood panelling, 0,65 mm/min rather than
0,9 mm/min. Here it would be more appropriate to calculate the burn-through time, or
basic protection value tprot,0 as described in Section 5.

6.4.3 Two-dimensional charring

Near corners of, for example, rectangular cross-sections, the heat flux is typically two-
dimensional, giving rise to a rounded shape of the residual cross-section near corners. At
first, the radius of the arris rounding is about equal to the one-dimensional charring depth,
see Figure 6.2a. Finally, due to the superposition of rounding of the two opposite arrises,
the charring depth on the narrow side of a rectangular cross section increases more than it
does on the wide side, see Figure 6.2b. For timber members with rectangular cross
sections exposed on three or four sides, and with normal load ratios relevant for structural
fire design, increased charring of the narrow side has a very limited influence on the
resistance of timber members and can therefore be neglected [12]. Consequently, this
effect need not be taken into account.

For some specific cases, e.g. timber studs or joists protected by cavity insulation on their
wide sides, increased charring of the narrow side must be taken into account, see 6.6.3.2.

71
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

For simplicity, the residual cross-section as shown in Figure 6.2 is normally replaced by a
rectangular cross-section, replacing the one-dimensional charring depth and arris
rounding with an equivalent notional charring depth (see Figure 6.3), calculated as
d char,n = β n t (6.2)

where βn is the notional charring rate. EN 1995-1-2, Table 3.1 gives notional charring
rates for timber members with rectangular cross-sections exposed to fire on three or four
sides as
β n = 0,7 mm/min for glued-laminated timber and LVL (softwood)
β n = 0,8 mm/min for solid timber (softwood).
The notional charring rates given in EN 1995-1-2, Table 3.1 also take into account the
effects of fissures. For this reason, different values are assumed for glued-laminated
timber and solid timber respectively.
The notional charring rate can be expressed as
β n = kn β 0 (6.3)
With β0 = 0,65 mm/min, we get kn = 1,08 and kn = 1,23 for glued laminated and solid
timber respectively.

d char,1 d char,1
dchar,2

dchar,2

d char,2 = d char,1 d char,2 > d char,1

a) b)
Figure 6.2 – Effect of arris rounding on charring on wide and narrow sides of cross-
section

d char,n

d char,0

Figure 6.3 – Replacing the one-dimensional charring depth and corner rounding with an
equivalent (notional) charring depth

72
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

As an alternative to the simplification of using notional charring depths, it is possible to

consider a residual cross-section with linear and rounded boundaries. Calculation of
cross-sectional properties will become more complicated, but normally it is not
worthwhile to consider it since the difference is negligible.

For other cases, see 6.6.2, 6.6.3 and 6.6.5.

6.4.4 Effect of protection

6.4.4.1 Large cross-sections

The formation of a char layer may provide effective protection against heat flux,
especially in large cross-sections. If the structure also incorporates applied protection, e.g.
in the form of wood-based panels, gypsum plasterboard, rock wool batt-type insulation or
other materials, the start of charring is delayed and, where the protection remains in place
after the start of charring, the rate of charring is slowed down in comparison with the
charring rate for initially unprotected cross-sections. Simplified relationships of charring
phases with start of charring, charring rates and failure times of protection are illustrated
in Figure 6.4 to 6.6 [1].

Since the charring rate immediately after failure of the fire protection – i.e. after the
protection has fallen off – is much greater than for initially unprotected timber (due to the
combination of high temperature and absence of, or insufficient protection by, the char
layer), some of the fire protection effect is lost some time after failure. Effective
protection provided by the char layer requires a char layer thickness of about 25 mm.
When the char layer has grown to that depth, the charring rate falls to the rate for initially
unprotected surfaces. A lasting protection effect is therefore only possible when a char
layer thickness of 25 mm can be built up during the phase of increased charring rate
immediately after failure of the fire protection, see Curve 2a of Figure 6.4. With rapid
failure of the protection there is some delay before the start of charring, but no lasting fire
protection effect: see Figure 6.5. Applied protection remaining in place after some
considerable time provides the most effective fire protection, especially for protection
materials with low thermal conductivities at high temperature, e.g. some gypsum
plasterboards type F [3] exhibiting long failure times. Fall-off of claddings protecting
surfaces fire-exposed from above is not a relevant scenario to be taken into consideration.

73
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

40

3b
Charring 30
depth
d char,0
20 1 d char,0 = 25 mm
or or
d char,n
d char,n = 25 mm
[mm] 10
3a

0
t ch = t f ta
Time t

Key:
1 Relationship for members unprotected throughout the
time of fire exposure for charring rate βn (or β0).
3a, 3b Relationship for initially protected members after failure
of the fire protection:
3a After the fire protection has fallen off, charring
starts at an increased rate
3b After the char depth exceeds 25 mm, the charring
rate falls to the rate for initially unprotected
members.

Figure 6.4 – Charring depth vs. time when charring starts at the time of failure (tch = tf)

74
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

40

Charring 30
depth
d char,0
20 1
or
d char,n
[mm] 10

3a
0
tf ta
Time t

Key:
1 Relationship for members unprotected throughout the
time of fire exposure for charring rate βn (or β0).
3a Relationship for initially protected members with rapid
failure times of the fire protection tf.

Figure 6.5 – Charring depth vs. time for protection with rapid failure time.

75
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

40

Charring 30 1 3b
depth
d char,0
20 d char,0 = 25 mm
or 3a or
d char,n
d char,n = 25 mm
[mm] 10 2

0
t ch tf ta
Time t
Key:
1 Relationship for members unprotected throughout
the time of fire exposure for charring rate βn (or β0).
2, 3a, 3b Relationship for initially protected members where
charring starts before failure of the fire protection:
2 Charring starts at tch at a reduced rate when the
fire protection is still in place
3a After the fire protection has fallen off, charring
starts at increased rate
3b After the char depth exceeds 25 mm, the
charring rate reduces to the rate for initially
unprotected members.

Figure 6.6 – Charring depth vs. time when charring takes place behind the fire protection
(tch < tf)

6.4.4.2 Small-sized timber frame members

In small-sized timber frame members, e.g. floor joists or wall studs in assemblies with
void cavities (see Figure 6.7), increased charring takes place after failure of the cladding.
However, the timber member will normally collapse before reaching the consolidation
phase with a char depth of 25 mm. Such conditions are described in EN 1995-1-2 [1]
Annex D. See also 6.6.3. (below).

For small-sized timber frame members in assemblies with cavity insulation, charring
mainly takes place on the narrow, fire-exposed side, see 6.6.3.2. Since there is a
considerable heat flux through the insulation to the sides of the member during the stage
after failure of the lining (provided that the cavity insulation remains in place), the effect
of increasing arris rounding becomes dominant and no consolidation of the charring rate
is possible.

76
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Key:
1. Narrow side of timber beam initially
protected then exposed to fire
2. Wide side of timber beam facing the cavity
3. Fire protective cladding (lining) on exposed
side of timber frame floor assembly
4. Fire protective cladding (lining) on side of
timber frame floor assembly not exposed to
fire
Figure 6.7 – Example of section of timber frame assembly

1
Charring
depth
d char,0
or 3a
d char,n
[mm]
2

t ch tf
Time t
Key:
1 Relationship for members unprotected throughout the
time of fire exposure for charring rate βn (or β0).
2, 3a Relationship for initially protected members where
charring starts before failure of the fire protection:
2 Charring starts at tch at a reduced rate when the fire
protection is still in place
3a After the fire protection has fallen off, charring
starts at an increased rate without reaching the
consolidation phase.

Figure 6.8 – Charring depth vs. time without consolidation of charring rate for timber
frame members in wall and floor assemblies with cavity insulation.

77
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.4.4.3 Determination of start of charring and failure times according to

EN 1995-1-2

EN 1995-1-2 provides limited information on determination of the start of charring

behind applied protection (claddings, linings), and on failure times of protective layers. In
general, the standard states that, unless expressions are given to calculate the start of
charring, failure times of protective layers and charring rates of wood behind the
protection where relevant, these must be determined by testing. For some specific
materials, expressions are given for the calculation of:
• start of charring (wood-based panels and wood panelling, regular gypsum
plasterboard type A, H and F [3], rock wool insulation)
• failure times (wood-based panels and wood panelling, regular gypsum plasterboard
type A [3])
• charring rates of timber behind the protection (gypsum plasterboard type F [3], batt-
type rock wool insulation).

Since gypsum plasterboard of types E, D, R and I have equal or better thermal and
mechanical properties than gypsum plasterboard of type A and H, the expressions for the
calculation of start of charring of gypsum plasterboard type A and H may be
conservatively used for those types. Although not explicitly stated, the same applies to
gypsum plasterboard type F.

EN 1995-1-2 also provides information on the start of charring where two layers of
gypsum plasterboard are attached to the timber member. Where both layers are of Type A
or Type H, the contribution of the inner layer is reduced by taking into account only 50 %
of its thickness since, after failure of the outer layer, the inner layer is already preheated
and has partially calcined and is exposed to a higher temperature.

Where two layers of different quality, e.g. Type F and Type A are attached to the timber
member, it is important that the better quality (Type F, in this example) is used as the
outer layer, while the contribution of the inner layer (Type A or H) is reduced by taking
into account only 80 % of its thickness. If the outer layer is of Type A or H, and the inner
layer of Type F, it should conservatively be assumed that both layers are of Type A or H.

Since the thermo-mechanical properties of gypsum plasterboard Type F are not part of the
classification given in [3], failure times of different makes may vary considerably. No
generic failure times for gypsum plasterboard are known; and so it is expected that the
producer should declare failure times determined on the basis of tests, including
information on spacing of joists, studs, battens etc. and edge distances and spacing of
fasteners. Conservative values based on evaluation of a large number of full-scale fire
tests are given in 6.4.4.5. For the time being, the European system of CE-marking does
not include such information. It is important that the failure times of gypsum plasterboard
should be related to thermo-mechanical degradation of the boards, i.e. issues such as
position (horizontal or vertical), span and edge distances of fixings (screws, nails,
staples). Pull-out failure of fasteners due to charring behind the cladding should be
verified by the designer; expressions for this failure type are only given for screws: it is
required that the minimum penetration length into uncharred wood is 10 mm.

Failure times of wood-based panels and wood panelling depend on the field of
application. For beams and columns protected by these claddings, it is assumed that the
cladding falls off at the time of start of charring tch. For walls and floors, however,
normally with greater distances between the fixings (the distance on centres of supporting

78
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

studs, joists or resilient channels is normally 400 to 600 mm or more), it is assumed that
the cladding falls off four minutes before the panel has burned through.

For gypsum plasterboard cladding, there is no corresponding distinction between beams

and columns on the one hand and timber frame assemblies on the other hand. More
appropriate values can be obtained from fire testing.

6.4.4.4 Start of charring and failure according to component additive

method given in Chapter 5

Differing from the components additive method given by EN 1995-1-2 [1][2], the new
components additive method presented in Section 5 is consistent with the needs of the
designer to determine the start of charring and failure times of protective layers. The
method given in Section 5 is therefore applicable and offers more precise solutions for a
greater variety of materials. The start of charring can therefore, by modification of
Expression (5.4), be calculated as:
tch = ∑ tprot,i (6.4)
i.e. the sum of protection times of i layers protecting the timber member, where tprot,i is
calculated according to Section 5, taking into account relevant position coefficients.

6.4.4.5 Start of charring and failure times from data base

Evaluating more than 320 full-scale fire test reports, failure times for gypsum
plasterboard claddings and start of charring of timber studs or joists were, when recorded,
collated in a data base [6]. The tested constructions were either timber frame assemblies,
the great majority with solid timber members and some with I-joists, or in a few cases
lightweight steel members. Below pessimized expressions are given which give
conservative results, see Table 6.2. The spread of data in the data base may be due to one
or several of the following:
• Variation of mechanical properties (between manufacturers or batches);
• Variation of thermal properties;
• Insulated or void cavity behind cladding;
• Timber or steel studs;
• Fixing of boards with respect to edge distance and spacing;
• Fixing of boards with respect screw length;
• Distance between battens or resilient steel channels fixed to floor joists.

In some tests, premature failure of the claddings may have been caused by too short
screws leading to fall-off of boards due to pull-out failure rather than thermal degradation
of the boards.

The failure times of gypsum plasterboards attached to large timber members such as
glued-laminated beams and columns or solid wood panels such as CLT may be
considerably greater, especially when edge distances of screws are greater than possible
in timber frame construction.

79
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 6.1 – Start of charring behind gypsum plasterboards tch in minutes with board
thickness hp and total board thickness hp,tot in millimetres.(incomplete)
Cladding Walls Floors

Type F, one layer

Type F, two layers

Type F + Type Aa

Type A, one layer

Type A, two layers

Type A, three layers

a
Outer layer Type F, inner layer type A

Table 6.2 – Failure times of gypsum plasterboards tf in minutes with board thickness hp
and total board thickness hp,tot in millimetres .(incomplete)
Cladding Walls Floors
4,5 hp − 24 hp + 10
Type F, 9 mm ≤ hp ≤ 18 mm 12,5 mm ≤ hp ≤ 16 mm
(6.5) (6.6)
one layer
57 hp > 18 mm 26 hp > 16 mm
4,5 hp,tot − 40 2 h p,tot − 3
Type F, 25 mm ≤ hp,tot ≤ 31 mm 25 mm ≤ hp,tot ≤ 32 mm
two layers (6.7) (6.8)
100 hp,tot ≥ 31 mm 61 hp,tot ≥ 32 mm
Type F +
81 hp ≥ 15 mmb 59 hp ≥ 15 mmb
Type Aa
1,9 hp − 7 1,8 hp − 7
Type A, 9 mm ≤ hp ≤ 15 mm 12,5 mm ≤ hp ≤ 15 mm
one layer (6.9) (6.10)
21,5 hp > 15 mm 20 hp > 15 mm
2,1 hp,tot − 14 c
Type A, 25 mm ≤ hp,tot ≤ 30 mm
two layers (6.11)
49 hp,tot ≥ 30 mm
Type A,
three 55 hp,tot ≥ 37,5 mm –d
layers
GF, 4,5 hp − 24
10 mm ≤ hp ≤ 12,5 mm –d
one layer (6.12)
a
Outer layer Type F, inner layer type A
b
Thickness of first layer (Type F)
c
Same as EN 1995-1-2 Clause 3.4.3.3(3)
d
No data available.

80
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.4.4.5.1 Worked example: Calculation of start of charring of timber stud

Key:
1 Solid timber joist
2 12 mm OSB
3 15 mm gypsum plasterboard Type F

1 2 3
Figure 6.9 – Timber frame wall assemby

a. Calculation in accordance with EN 1995-1-2 [1]

Material properties for gypsum plasterboard Type F:
Since no failure times of gypsum plasterboard Type F are given in EN 1995-1-2 [1], they must be
determined by testing, see [1] 3.4.3.1(2). The value of failure time of gypsum plasterboard Type F
attached to floors given by the producer is tf = 65 minutes.
Material properties for OSB: ρk = 550 N/mm2 in accordance with EN 12369-1 [13].

Below, following Expressions of EN 1995-1-2 [1] are applied: (3.4), (3.5), (3.6), (3.7) and (3.11).

Start of charring of OSB:

tchar = 2,8 hp − 14 = 2,8 × 15 − 14 = 28 mm
Charring rate of unprotected OSB with β0 = 0,9 mm/min from Table 3.1 in [1]:
450 20 450 20
β 0, ρ ,t = β 0 kρ kt = β 0 = 0,9 × = 0,94 mm/min
ρk hp 550 15
With
k2 = 1 − 0, 018 hp = 1 − 0, 018 × 15 = 0, 73 for 28 min ≤ t ≤ 65 min
the charring rate of OSB is
β OSB = k2 β0, ρ ,t = 0, 73 × 0,94 = 0,87 mm/min
Start of charring of timber stud:
h 12
tch = tch,OSB + OSB = 28 + = 45,5 min
β OSB 0,87

b. Calculation in accordance with Chapter 5

Protection time of gypsum plasterboard Type F (layer 1):

1,2 1,2
⎛h ⎞ ⎛ 15 ⎞
tprot,0,1 = 30 ⎜ 1 ⎟ = 30 × ⎜ ⎟ = 30 min
⎝ 15 ⎠ ⎝ 15 ⎠
kpos,exp,1 = 1, 0 (no preceding layer)
kpos,unexp,1 = 1, 0 (backed by OSB)
k j,1 = 1, 0 (backed by OSB)
Δt1 = 0 (no preceding layer)
( )
tprot,1 = tprot,0,1 kpos,exp,1 kpos,unexp,1 + Δt1 k j,1 = ( 30 × 1, 0 × 1, 0 + 0 ) × 1, 0 = 30 min

Protection time of OSB (layer 2):

81
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

⎧ ⎛ h ⎞1,1 ⎛ 12 ⎞
1,2
⎪23 ⎜ 2 ⎟ = 23 × ⎜ ⎟ = 13,1 min
⎪ ⎝ 20 ⎠ ⎝ 20 ⎠
tprot,0,2 = min ⎨
⎪ h2 = 12 = 12,8 min
⎪ β 0,ρ ,h 0,94
⎩
tprot,0,2 12,8 tprot,0,2 12,8
kpos,exp,2 = 0,5 = 0,5 × = 0,33 min (since tprot,1 > = = 6, 4 )
tprot,1 30 2 2
kpos,unexp,2 = 1, 0 (backed by timber stud)
Δt2 = 0, 22 tprot,1 − 0,1tprot,0,2 + 4, 7 = 0, 22 × 30 − 0,1× 12,8 = 10, 0 min
since tprot,1 = 30 min > 12 min

( )
tprot,2 = tprot,0,2 kpos,exp,2 kpos,unexp,2 + Δt2 k j,2 = (12,8 × 0,33 × 1, 0 + 10, 0 ) × 1, 0 = 14, 2 min

Start of charring of timber stud, see Expression (6.4):

tch = tprot,1 + tprot,2 = 30 + 14, 2 = 44, 2 min

6.4.5 Effect of bonded joints

From several experimental investigations, e.g. [14], it is known that some novel adhesives
soften at relatively low temperatures, causing the char layer to fall off the timber member
when the char front has reached a bonded joint, e.g. in layered composites such as cross-
laminated or glued-laminated timber. Since the char layer normally provides effective
protection of the residual cross-section against heat, this ablative behaviour of the char
layer has the same adverse effect as failure of a fire protective panel, see Figure 6.4 and
6.5. Therefore, for cross-laminated and glued-laminated timber bonded with adhesives
showing this effect, once a lamination has burnt through, the charring rate for those
surfaces in question should be doubled until a char layer of 25 mm thickness has built up.
For unprotected cross-laminated timber (CLT) there may be a considerable effect of char-
layer ablation on charring depth, especially when layer thickness is small. This is
illustrated in 6.6.2.4.4. For CLT protected by a fire protective cladding such as gypsum
plasterboard Type F, the char layer remains in place during the protection phase until
failure of the fire protective cladding. However, char ablation may take place after the fire
protective cladding has fallen off. Since fire protective claddings delay the start of
charring and reduce the charring rate, the effect of char ablation is very small. This is
illustrated in 6.6.2.4.5.

No observations are known about similar effects in glued-laminated timber. Since

lamination thicknesses are considerably greater than in CLT, burn-through of the outer
lamellae does normally not occur, or takes place at a very late stage. It is therefore
recommended to disregard this effect.

6.5 Mechanical resistance

6.5.1 Simplified methods for strength and stiffness parameters

Since a zone of about 35 to 40 mm depth below to the char-line of the residual cross-
section, although unburned, is heated above normal temperature, strength and modulus of
elasticity in this zone are reduced. Therefore, for structural fire design, the strength and
stiffness parameters of the timber must be reduced as shown below.

In general, EN 1995-1-2 [1] gives the design strength of timber members as

82
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

f 20
f d,fi = kmod,fi
γ M,fi
(6.13)
With
f 20 = f k kfi (6.14)
we get
fk
f d,fi = kmod,fi kfi (6.15)
γ M,fi
where
fd,fi is the strength property in the fire situation, e.g. bending strength;
f20 is the 20 % fractile of the strength property;
fk is the characteristic strength property, e.g. as given in EN 338 [17];
kmod,fi is the modification factor for fire expressing the reduction of strength in the fire
situation;
γM,fi is the partial safety factor for timber in fire.

The recommended value of the partial factor for fire is 1; information on a specific
national value may be given in the National annex. Values of kfi are given in Table 2.1 of
EN 1995-1-2 [1]; e.g. kfi = 1,25 for solid timber, kfi = 1,15 for glued laminated timber and
kfi = 1,1 for LVL.

It should be noted that the modification factor kmod, reducing the design strength taking
into account the duration of load and moisture content at normal temperature, as given in
EN 1995-1-1 [15], is not relevant in the fire situation and therefore not included in
Expression (6.13).

In the same way as for Expression (6.13), the design value of a stiffness property, i.e. the
modulus of elasticity and shear modulus, is given as
S
Sd,fi = kmod,fi 20 (6.16)
γ M,fi
With
S 20 = S05 kfi (6.17)
we get
S05
Sd,fi = kmod,fi kfi (6.18)
γ M,fi
For a linear (1st order) ultimate limit state structural analysis, it is generally assumed that
the internal force distribution is not influenced by the stiffness properties, unless the
timber is combined with other materials in statically indeterminate structures. Thus,
according to EN 1995-1-1, mean values of stiffness properties must be used for a 1st order
structural analysis of a structure, where the distribution of internal forces is not affected
by the stiffness distribution within the structure. For a non-linear (2nd order) ultimate limit
state structural analysis, stiffness properties are taken into account in order to assess
higher order effects caused by deformation. In this case, design values of a stiffness
property as defined previously must be used. It should be noted that, contrary to EN
1995-1-1 [15], where the design value of a stiffness property Ed or Gd is given by

83
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Emean Gmean
Ed = and Gd = respectively, the above Expression (6.16) relates to the 20 %
γM γM
value of the stiffness property1.

For consideration of strength and stiffness reduction of members exposed on three or four
sides, EN 1995-1-2 [1] offers two methods:
• the reduced cross-section method, which is the recommended method;
• the reduced properties method.
The National annex may give information about the national choice. These two methods
were discussed elsewhere [16]. In the following only the reduced cross-section method is
applied, except for specific applications, see 6.6.3.2 and 6.6.4.

According to the reduced cross-section method, the reduction of strength and stiffness
parameters is taken into account by assuming normal temperature properties of timber
multiplied by kfi. However a zero-strength layer of depth d0 is subtracted from the residual
cross-section (or, in other words, the charring depth is increased by d0). Although as
much as 35 to 40 mm below the char layer are affected by elevated temperature, the depth
of the zero-strength layer is only 7 mm for normal cross-sections exposed to fire on three
or four sides [1]. For specific cross-sections and partial protection, d0 may be
considerably larger, see 6.6.2 to 6.6.5.

For calculation of the resistance of a fire-exposed member, the original cross-section is

therefore reduced by the notional charring depth and the zero-strength layer d0. The
effective charring depth is given as:
d ef = d char,n + k0 d 0 (6.19)

where k0 takes into account the fact that the zero-strength layer is not fully effective
during the first twenty minutes of initially unprotected members, or until the start of
charring of protected members. EN 1995-1-2 [1] assumes linear increase from 0 to 1
during these time intervals, see [1] Figure 4.2.
Therefore:
⎧ t
⎪⎪ 20 for unprotected members
k0 = ⎨ (6.20)
⎪t for protected members
⎪⎩ tch
As a consequence, at the time of start of charring of small-sized protected cross-sections,
a considerable reduction of resistance may occur although the cross-section is still
uncharred.

6.5.2 Advanced calculation methods

For determination of the mechanical resistance of structural timber members, an
advanced calculation, e.g. using finite-element modelling of fire-exposed structural
timber members, comprises several steps:
• Determination of temperatures in the timber member including the charring depth

1
This inconsistency is due to a late change of the final draft of [15] which was not taken into
account accordingly in [1].

84
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

• Determination of the resistance of cross-sections using the temperature field in

the timber member and the temperature-dependent reduction of strength and
stiffness at each location of the cross-section
• Determination of resistance of the structure (beam, column, frame etc.)

A general outline of the procedure is given in EN 1995-1-2 [1], Annex B. However, for
the thermal analysis, it is the thermal properties only of wood that are given. Thermal
properties for other materials often used together with timber, e.g. gypsum plasterboard,
insulation materials and others are expected to be found in other sources. A problem is
that the data from various sources may vary considerably. Since available commercial
software for heat transfer calculations does not explicitly take into account mass transfer
(water, steam, gases), its effect must be accounted for by using effective conductivity
values rather than real ones [18], [19]. This also applies to the formation of cracks, e.g. in
the char layer or gypsum plasterboards, causing increasing heat flux which is taken into
account by increased conductivity values. For the char layer, this has not been considered
in some sources, which give considerably lower conductivity values than 1995-1-2 [1],
Annex B. Since the protection provided by boards and insulation is often important for
the performance of the timber member, the software should be capable of taking into
account sudden failure (fall-off) of applied protection. Examples of commercial software
including this option are SAFIR [20] and ANSYS [21].

Thermal properties of gypsum plasterboard can be found in [22], see Table 6.3.

Table 6.3 – Temperature-dependent thermal properties of gypsum plasterboard and

gypsum fibreboard [22].
Gypsum plasterboard Gypsum fibreboard
T λ c ρ/ρ20 λ c ρ/ρ20
°C W/mK kJ/kgK – W/mK kJ/kgK –
20 0,40 0,96 100 0,40 0,96 100
70 0,40 0,96 100 0,40 0,96 100
100 0,27 0,96 100 0,27 0,96 100
130 0,13 14,9 92,6 0,13 9,17 92,6
140 0,13 25,2 90,2 0,13 17,55 90,2
150 0,13 21,7 87,7 0,13 16,66 87,7
170 0,13 0,96 82,8 0,13 0,96 82,8
600 0,13 0,96 82,7 0,13 0,96 82,7
720 0,33 4,36 82,6 0,39 4,359 82,6
750 0,38 0,96 77,6 0,46 0,96 77,6
1000 0,80 0,96 77,6 1,00 0,96 77,6
1200 2,37 0,96 77,6 2,37 0,96 77,6

Thermal properties of wood-based panels may be taken as follows. For particle board
and wood fibreboard, thermal properties of wood apply. However, due to premature fall-
off of charcoal, the thermal conductivity of OSB and plywood should be taken from
Table 6.4.

85
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 6.4 – Temperature-dependent thermal properties OSB, plywood and wood

fibreboard [22]
T λ
°C W/mK
20 0,12
100 0,30
110 0,23
120 0,15
200 0,18
275 0,14
350 0,09
500 023
800 0,74
1200 4,2

Thermal properties of insulation materials can be found in [22], see Table 6.5 and Table
6.6. For rock wool, the conductivity values are determined as
⎧⎪ ρ0 for T ≤ 100°C
λ (ρ) = ⎨ (6.21)
⎩ (
⎪λ ( ρ0 ) × 11e
−0,05 ρ
)
+ 1,9 for T > 100°C
where ρ0 is given in Table 6.5.

Table 6.5 – Temperature-dependent thermal properties of batt-type rock wool insulation

[22].
T λ(ρ0) T c ρ/ρ20
°C W/mK °C kJ/kgK –
20 0,036 20 0,880 1,00
100 0,047 100 1,040 1,00
400 0,090 200 1,160 0,980
600 0,150 400 1,280 0,977
800 0,230 600 1,355 0,973
925 0,300 800 1,430 0,970
1200 0,450 925 1,477 0,960
1200 1,580 0,887

Table 6.6 – Temperature-dependent thermal properties of batt-type glass wool insulation

[22].
T λ(ρ0) c ρ/ρ20
°C W/mK kJ/kgK –
20 0,035 1,200 1,00
100 0,068 1,340 0,983
200 0,110 1,380 0,961
300 0,151 1,382 0,940
400 0,192 1,384 0,940
510 0,238 1,386 0,940
660 0,300 1,389 0,940
1200 100 1,400 0,940

86
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

For calculation of mechanical resistance, EN 1995-1-2 [1], Annex B gives temperature-

dependent reduction factors kΘ for compressive, tensile and shear strength as well as for
the modulus of elasticity parallel to the grain for timber. The reduction factor kΘ should
not be confused with the modification factor for fire, kmod,fi, see 6.5.1, which gives the
reduction of a strength or stiffness property with the assumption of a linear relationship
between stress and strain ([23] Clause 3.1.2).

For timber members, it is sufficient to assume ideal elastic-plastic behaviour for

compression and purely elastic behaviour for tension, see Figure 6.10, relationship for
20°C. The figure also shows the corresponding relationships for some other temperatures
obtained by multiplication by kΘ.

σ
ft
20°C
60°C
100°C
ε
200°C
fc

Figure 6.10 – Temperature-dependent strain-stress relationships.

Using values for fc and ft in accordance with EN 338 [17] will never permit plastic flow
on the compression side of a member in bending at ambient temperature, since fc > ft.
Plastic flow would only occur at locations with a large rise of temperature. It would,
however, be more correct to use the local strength values, see [24]. The compressive
strength should be determined from
f c = 0,114 ρ0,12 − 9 (6.22)

where ρ0,12 is the dry density in kg/m3.

Replacing dry density with the density of wood with a moisture content of 12 %, we get
f c = 0,1 ρ − 9 (6.23)

Expression (6.22) was derived from data given in [25]. Although the values were
determined for Swedish-grown Scotch pine wood, it is reasonable to apply them to
timbers used for strength classes up to C40 and GL36 in accordance with [17] and [26].

6.6 Structural elements

6.6.1 Beams and columns exposed on three or four sides
6.6.1.1 Unprotected members

6.6.1.1.1 Charring
The notional charring rate to be used is (see 6.4.3):
βn = 0,7 mm/min for glued laminated timber

87
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

βn = 0,8 mm/min for solid timber.

The corresponding charring depths are calculated as (see Expression (6.2)

d char,n = β n t

where t is the time of fire exposure.

6.6.1.1.2 Strength and stiffness properties

The reduction of strength and stiffness properties (e.g. bending strength or modulus of
elasticity) is taken into account by increasing the charring depth by a zero-strength layer
of depth d0 ≤ 7 mm. The effective charring depth is, see Expression (6.19):
d ef = d char,n + k0 d 0

with k0 according to Expression (6.20).

6.6.1.1.3 Worked example 1

A timber floor consists of joists and tongued and grooved decking. The dimensions of the joists of
Strength Class C18 are b = 45 mm, h = 195 mm and their spacing is c = 600 mm on centres. The
thickness (depth) of the decking is h1 = 28 mm. The joints of the decking are double tongued and
grooved with a maximum gap width of 2 mm. The required fire resistance is REI 15. Determine
the design moment resistance of the joists for a) fire under the floor, b) fire above the floor. (For
fire resistance of the decking, see worked example in 6.6.2.1.)
h

h1

c
b

Figure 6.11 – Dimensions of timber floor

a. Joists fire exposed from below

Since there is no lining fixed to the joists, they are exposed and unprotected on three sides.
Determination of effective charring depth:

t 15
d ef = β n t + d 0 = 0,8 × 15 + × 7 = 12 + 5, 25 = 17, 25 mm
20 20
The section modulus is:
(b − 2 d ef ) × (h − d ef ) 2 (45 − 2 × 17, 25) × (195 − 17, 25) 2
Wef = = = 55291 mm3
6 6
The design bending resistance is:
f k W 18 × 1, 25 × 55291× 10−3
M d,fi = m,k fi ef = = 1244 Nm
γ M,fi 1

b. Joists fire exposed from above

The joists are protected by the decking against fire exposure. Since drying gaps at each tongued
and grooved joint normally exist, the influence of increased charring should be taken into account
by using notional charring rates rather than the one-dimensional values. The charring depth in the
tongued and grooved decking at time t = 15 min is
d char,n = β n t = 0,8 ×15 = 12 mm
Alternatively, a check of protection time can be made, see Table 5.1:

88
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

1,1 1,1
⎛h ⎞ ⎛ 28 ⎞ h 28
tprot,0,1 = 30 × ⎜ i ⎟ = 30 × ⎜ ⎟ = 43, 4 min ≤ i = = 43,1 min
⎝ 20 ⎠ ⎝ 20 ⎠ 0, 65 0, 65
tpos,exp = tpos,unexp = 1
tch = tprot,1 = 43,1min
Check of insulation criterion EI 15:
1,4 1,4
⎛h ⎞ ⎛ 28 ⎞
tins,0,1 = 19 × ⎜ 1 ⎟ = 19 × ⎜ ⎟ = 30, 4 min
⎝ 20 ⎠ ⎝ 20 ⎠
tpos,exp = tpos,unexp = 1
k j,1 = 0, 6
tins,1 = tins,0,1 tpos,exp,1 tpos,unexp,1 k j,1 = 30, 4 × 1× 1× 0, 6 = 18, 2 min

Conclusion 1: There is no risk of charring of the joists. Although there might be a slight increase
of temperature in the joist, this effect can be neglected.
Conclusion 2: For bending resistance of joists, fire exposure from below the floor is decisive.
Note: For verification of the load resistance of the decking, see 6.6.2.1c.

6.6.1.1.4 Worked example 2

(Timber column fire exposed on four sides, protected by a single layer of gypsum
plasterboard, requirement R 30)

6.6.2 Solid timber decks and walls

6.6.2.1 Tongued and grooved timber decking

6.6.2.1.1 Charring

Unlike the constructions shown in the following subclauses, a tongued and grooved
timber deck is normally supported by a number of timber frame joists. EN 1995-1-2 [1]
gives no specific information on charring rates of tongue and groove timber decking.
Since timber decking is load-bearing, charring rates should be taken from Table 3.1 in [1]
rows a to c, but not d (wood panelling). Since there normally exist drying gaps at each
joint, the influence of increased charring should be taken into account by using notional
charring rates, given in Table 3.1 in [1], rather than the one-dimensional values. Since
tongued and grooved joints are “weak zones” with respect to heat transfer, the protection
time tprot for determination of burn-through time should be determined and be multiplied
by the joint coefficient kj, see Table 5.5.

6.6.2.1.2 Strength and stiffness properties

The reduction of strength and stiffness properties (e.g. bending strength or modulus of
elasticity) is taken into account by increasing the charring depth by a zero-strength layer
of depth d0. It has been shown [27] that d0 for timber deck plates may be greater than
7 mm. For plate thicknesses of up to 35 mm, d 0 = 7 mm should be a reasonable value.
Therefore (see Expression (6.19))
d ef = d char,n + k0 d 0

with k0 according to Expression (6.20).

6.6.2.1.3 Worked example

See 6.6.1.1, worked example. Check mechanical resistance of decking.

89
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

a. Decking exposed to fire from below

Decking: C18. Dimension of planks 36 mm ×150 mm.
According to EN 1991-1-1 [28] the recommended actions on the floor for Category A are
qk = 2 kN/m2 and Qk = 2 kN. For local action effects, only the concentrated load needs to be taken
into account, acting on a surface area of 50 mm × 50 mm. It is assumed that the concentrated load
is taken by one plank (no load distribution to other planks).
Combination factor for quasi-permanent action:ψ 2,1 = 0,3 (see EN 1991-1-1 [28])
Partial factor for leading variable action: γ Q,1 = 1,5 (see EN 1990 [29], Table A1.2(A))
Neglecting the self-weight of the decking, the design action in the fire situation is :
Ed,fi = ηfi Ed (see EN 1995-1-2 [1], Expression (2.8))
Gk + ψ 2,1 Qk,1 0 + 0,3 × 2
ηfi = = = 0, 2 (see EN 1995-1-2 [1], Expression (2.9))
γ G Gk + γ Q,1 Qk,1 0 + 1,5 × 2

Design value of action at normal temperature: Qd = γ Q Qk = 1,5 × 2 = 3 kN

Maximum bending moment: M d,fi = 0,175 × Qd ηfi A = 0,175 × 3000 × 0, 2 × 0, 6 = 63 Nm

Since there are gaps of maximum 2 mm width in the decking, a notional charring rate of
0,8 mm/min should be used.
Charring depth at 15 minutes: dchar = β n tfi = 0,8 × 15 = 12 mm
15
Effective charring depth: def = dchar + d0 k0 = 12 + × 7 = 17,3 mm
20
Effective deck thickness: hef = h − def = 36 − 17, 25 = 18, 7 mm
50 × 18, 752
Effective section modulus: Wef = = 2930 mm3
6
M d,fi 63000
Design value of maximum bending stress: σ d,fi = = = 21,5 N/mm 2
Wef 2930
Design value of bending strength in fire: f m,d,fi = kfi f m,k = 1, 25 × 18 = 22,5 N/mm 2

6.6.2.2 Glued-laminated and stress-laminated timber plates

This clause deals with timber deck or wall plates made of edgewise-arranged laminations
of solid timber, held together either by adhesive bonding or by pre-stressing, see Figure
6.12.

Key:
1 Tendon
a. b.
Figure 6.12 – Examples of deck plates: a. Glued-laminated, b. Pre-stressed

The charring depth should be calculated in accordance with EN 1995-1-2 [1] Clause 3.4,
using the one-dimensional charring rate β0.

90
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

For verification of mechanical resistance, the reduced cross-section method given in EN

1995-1-2 [1] Clause 4.2.2 should be used, but with the values for the zero-strength layer
d0 [27] given in Table 6.7:

Table 6.7 – Zero-strength layer d0 in mm where h is in mm

Floors Walls
Exposure on
unprotected protected unprotected protected
h
+ 8,5
Tension side 55 … –b –b
(6.24)
h h
Compression +9 +9
20 … 20
side
(6.25) (6.26)
a
Values also to be used for t > tf
b
Not relevant
20
1x

15 Floors, tension
d 0 [mm]

Floors, compression
10
Walls
5

0
80 130 180
h [mm]

Figure 6.13 – Zero-strength layers d0 for timber plates.

6.6.2.3 Nail-laminated timber plates

This clause deals with timber deck or wall plates made of edgewise-arranged laminations
made of solid timber, held together by nailing, see Figure 6.14.

Key:
1 Nail

Figure 6.14 – Example of nail-laminated timber plate.

The charring depth should be calculated in accordance with EN 1995-1-2 [1] Clause 3.4
using the notional charring rate βn charring rate for solid timber [30]. For verification of
mechanical resistance, see 6.6.2.2.

91
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.6.2.4 Cross-laminated timber plates (CLT)

6.6.2.4.1 General

This clause deals with cross-laminated timber panels (solid timber panels) consisting of
an odd number of layers, each with a minimum thickness of 15 mm, bonded together,
supported on two opposite edges, e.g. one-span floors and axially loaded walls, or
supported on several supports acting as a continuous floor, see Figure 6.15. The direction
of the grain of outer layers and every second layer (layers No. 1, 3, 5, etc.) is orientated in
the (main) load-bearing direction, while layers 2, 4, etc. are orientated in the transverse
direction, see Figure 6.16a. In the design model presented here, the layers in the
transverse direction are not regarded as load-bearing in the longitudinal direction; they
contribute to load resistance by taking shear forces between the layers in the longitudinal
direction. In the following design, expressions for fire exposure on one side are given, see
Figure 6.16b.

Figure 6.15 – Example of cross-laminated deck or wall plate.

The design model [31] follows the general outline of 6.4.2 and 6.5.1 (reduced cross-
section method) and should not be used for fire durations of more than two hours. The
fire protective effect of claddings (boards and insulation batts on the fire-exposed side) is
taken into account in accordance with 6.4.4.). If the residual depth of a charred layer is
less than 3 mm, it should not be taken into account when deriving hef.
≥15mm

n=i
hef

n = i-1
d0 ≥3mm

n=3
h

n=2
dchar

n=1

a. b.
Figure 6.16 – Cross-section of cross-laminated timber and definitions:
a. Cross-section at ambient temperature,
b. Residual cross-section, char layer and zero-strength layer of cross-section
exposed to fire on one side.

92
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.6.2.4.2 Charring

Where a layer of a cross-laminated timber panel consists of boards bonded together along
their edges, or if the edge width between two boards is not greater than 2 mm, the one-
dimensional charring rate (see 6.4.2) should be applied. When the gap width is greater
than 2 mm but not more than 6 mm, a notional charring rate should be applied according
to Expression (6.3) where kn = 1,2. When the gap width is grater than 6 mm, each board
should be regarded as exposed on three sides. Charring of protected members should be
calculated in accordance with 6.4.4.1.
Some novel adhesives may cause premature fall-off of the char layer once the char front
has reached a bondline, and so the charring rate should be increased accordingly, see
6.4.5.

6.6.2.4.3 Strength and stiffness properties

The reduced cross-section method should be used for the design of cross-laminated
timber panels; see 6.4. In general, d0 for cross-laminated timber panels depends on:
• the number of layers;
• the depth of the cross-laminated member;
• the state of stress (tension or compression) on the fire-exposed side;
• the temperature gradient below the char layer (i.e. whether the member is
unprotected or protected.

The depth of the zero-strength layer should be taken from Table 6.8 to 6.9.

Table 6.8 – Zero-strength layer d0 in mm for CLT with 3 layers where h is in mm

Floors Walls
Exposure on unprotected protecteda unprotected protecteda
h
+ 3, 7
Tension side 30 10 –b –b
(6.27)

⎧13,5 ⎧14,5
h ⎪ h ⎪
+ 4,5 min ⎨ h + 3,95 min ⎨ h
Compression 25 ⎪12,5 + 7 12,5 ⎪12,5 + 7
side (6.28) ⎩ ⎩
(6.30)
(6.29) (6.31)
a
Values also to be used for t > tf
b
Not relevant

93
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 6.9 – Zero-strength layer d0 in mm for CLT with 5 layers where h is in mm

Floors Walls
Exposure on unprotected protecteda unprotected protecteda
For 75 mm ≤ h ≤
100 mm:
h
h − + 34
+ 10 4
Tension side 100 (6.33) –b –b
(6.32) For h > 100 mm:
h
+6
35
(6.34)
h h
Compression + 11 18 + 10,5 20
side 20 15
(6.35) (6.36)
a
Values also to be used for t > tf
b
Not relevant

Table 6.10 – Zero-strength layer d0 in mm for CLT with 7 layers where h is in mm

incomplete
Floors Walls
Exposure on unprotected protecteda unprotected protecteda
⎧9,8
⎪
min ⎨ h
Tension side ⎪⎩ 50 + 6 ….. –b –b

(6.37)
⎧13, 0 ⎧15,8
⎪ ⎪
Compression min ⎨ h min ⎨ h
⎪⎩ 20 + 3,5 ….. ⎪⎩17 + 4 …..
side
(6.38) (6.39)
a
Values also to be used for t > tf
b
Not relevant

Table 6.11 – Zero-strength layer d0 in mm for CLT with 9 layers where h is in mm

incomplete
Floors Walls
Exposure on unprotected protecteda unprotected protecteda
Tension side ….. –b –b
Compression
….. …..
side
a
Values also to be used for t > tf
b
Not relevant

The expressions given above are illustrated in Figure 6.17 to Figure 6.20.

94
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

20
18 3x
Floors, unprotected, tension
16
14 Floors, protected, tension
12
d 0 [mm]

Floors, unprotected, compr.

10
Floors, protected, compr.
8
6 Walls, unprotected
4
Walls, protected
2
0
40 50 60 70 80 90 100 110 120 130 140
h [mm]

Figure 6.17 – Zero-strength layers d0 for CLT with three layers.

30
5x
Floors, unprotected, tension
25
Floors, protected, tension
20
d 0 [mm]

Floors, unprotected, compr.

15
Floors, protected, compr.
10
Walls, unprotected
5 Walls, protected

0
50 100 150 200 250
h [mm]

Figure 6.18 – Zero-strength layers d0 for CLT with five layers.

35
7x
30

25 Floors, tension
d 0 [mm]

20 Floors, compression
15 Walls
10

0
100 150 200 250 300 350
h [mm]

Figure 6.19 – Zero-strength layers d0 for CLT with seven layers.

To be added To be added

95
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 6.20 – Zero-strength layers for CLT with nine layers.

6.6.2.4.4 Worked example1 – Determination of charring depth and effective

depth of an unprotected CLT deck plate

An unprotected cross-laminated timber plate (CLT) consisting of seven 19 mm thick layers is

exposed to fire on one side. There are no gaps between the boards of each lamination. Determine
the effective plate depth at 60 minutes.

a. No char ablation
The adhesive used for bonding between laminations is fully effective in fire.
Charring depth at 60 minutes: dchar = β 0 treq = 0, 65 × 60 = 39 mm
⎧9,8 ⎧9,8
⎪ ⎪
Zero-strength layer: d0 = min ⎨ h = min ⎨133 = 8, 7 mm
⎪⎩ 50 + 6 ⎪⎩ 50 + 6
Effective depth: hef = h − dchar − d0 = 133 − 39 − 8, 7 = 85 mm

b. Calculation with ablation of char layer (see 6.4.5)

The failure time of layer 1 is equal to the start of charring of layer 2:
h 19
tch = tf,1 = 1 = = 29, 2 min
β 0 0, 65
Charring depth at 60 minutes:
( )
d char = h1 + treq − tf β 0 k3 = 19 + ( 60 − 29, 2 ) × 0, 65 × 2 = 59 mm
Zero-strength layer as above: d 0 = 8, 7 mm
Effective depth: hef = h − d char − d 0 = 133 − 59, 2 − 8, 7 = 65,3 mm

6.6.2.4.5 Worked example 2 – Determination of charring depth and effective

depth of a protected CLT deck plate

The same CLT as above in 6.6.2.4.4 is protected by a 12,5 mm thick layer of gypsum plasterboard
Type F. Determine the charring depth at 60 minutes. The failure time of gypsum plasterboard
Type F is given by the producer as tf = 45 minutes.

a. No char ablation
The adhesive used for bonding between laminations is fully effective in fire.
Start of charring:
tch = 2,8 hp − 14 = 2,8 × 12,5 − 14 = 21 min
Protection factor according to EN 1995-1-2 [1] Expression (3.7):
k2 = 1 − 0, 018 hp = 1 − 0, 018 × 12,5 = 0, 775
Post-protection factor according to EN 1995-1-2 [1] Paragraph 3.4.3.2(4):
k3 = 2
Calculation of ta according to EN 1995-1-2 [1] Expression (3.9), see Figure 6.6:
25 − ( tf − tch ) k2 β 0 25 − ( 45 − 21) × 0, 775 × 0, 65
ta = + tf = + 45 = 54,9 min
k3 β 0 2 × 0, 65
Charring depth at 60 minutes:
( )
dch = 25 + treq − ta β 0 = 25 + ( 60 − 54,9 ) × 0, 65 = 28,3 mm
Zero-strength layer:
Effective depth:

b. Calculation with ablation of char layer (see 6.4.5)

Charring depth at time of failure (tf = 45 min):
d ch,f = ( tf − tch ) β 0 k2 = ( 45 − 21) × 0, 65 × 0, 775 = 12,1 mm

96
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Since the char layer falls off the CLT-plate when the charring depth is equal to 25 mm, no value of
ta exists, i.e. charring continues at a rate of k3 β0.
Charring depth at 60 minutes:
( )
dch = d ch,f + treq − tf k3 β 0 = 12,1 + ( 60 − 45 ) = 31, 6 mm
Zero-strength layer:
Effective depth:

6.6.3 Timber frame floor and wall assemblies

6.6.3.1 General

Timber frame assemblies are normally built up of the timber frame (floor joists or wall
studs) and a cladding attached to each side of the timber frame (the cladding may be a
lining or, in the case of floors, the decking or a sub-floor and additional layers). The
cavities may be void or partially or completely filled with insulation. Since the timber
frame is sensitive to fire exposure, it must be effectively protected against fire.

In the design and optimisation of a timber frame assembly with respect to maximising fire
resistance, the following rules are important:
• There exists a hierarchy of contribution to fire resistance of various layers of the
assembly;
• The greatest contribution to fire resistance is obtained from the membrane (layer)
on the fire-exposed side first directly exposed to the fire, both with respect to
insulation and failure (fall-off) of the membrane.
• In general, it is difficult to compensate for poor fire protection performance of
the first membrane by improved fire protection performance of the following
layers.
• Cavity insulation improves the fire resistance of the timber frame. The best
protection against fire is achieved when the insulation effectively protects the
sides of the timber member facing the cavity against the fire.

For the stage before failure of the cladding (protection phase t ≤ tf), or more precisely,
failure of the layer of the cladding next to the insulation, both batt-type and loose-fill
mineral wool (rock or glass wool) insulation perform approximately equally. However,
once the cladding has fallen off and the insulation is directly exposed to the fire (post-
protection phase t ≥ tf), glass wool insulation will undergo decomposition, gradually
losing its protecting effect for the timber member by surface recession. Rock wool
insulation, provided that it remains in place, will continue to protect the sides of the
timber member facing the cavity. During this post-protection stage, loose-fill insulation
should not be used. Batt-type mineral wool insulation should always be secured
mechanically, e.g. by resilient steel channels or battens. The steel channels must be fixed
with screws of sufficient length to prevent pull-out failure due to extensive charring of the
joists. In the case of wall assemblies, mineral wool batts are normally fixed by oversizing
the width of the batts. When the thickness of the batts is insufficient, they tend
prematurely to fall off a wall assembly; and so batts less than 120 mm thick should be
mechanically secured, e.g. by wires or chicken net fixed to the studs, with the wires or
chicken net in turn being secured by staples of sufficient length to prevent pull-out failure
due to extensive charring of the studs.

The design model presented in 6.6.3.2 is valid for cavities filled with rock wool. Where
glass wool is used, the model is valid until failure (fall-off) of the cladding. Immediate
failure of the assembly is a conservative assumption. From full-scale wall tests it is

97
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

known that it will take some time until the glass wool insulation has completely recessed
once it has been directly exposed to the fire. A model of cavities filled with batt-type
glass wool insulation has been developed, describing the surface recession of the glass
wool insulation [32]. A new design model for the post-protection stage of assemblies with
glass wool cavity insulation can be found in 6.6.3.4.

Where the wall is placed between two floors (platform frame construction), each
providing a horizontal support, it may be advantageous to take into account partial
restraint at the supports for the calculation of the axial load capacity of the wall with
respect to buckling of the studs perpendicular to the wall. The partial restraint is due to
the movement of the reaction forces towards the unexposed edge of a stud as its ends
rotate, thus reducing the eccentricity of the axial loading. This positive effect was studied
in [33] and may be taken into account assuming a buckling length of
A y = 0,7 A wall
where
Ay is the buckling length with respect to buckling about the y-axis (out-of-plane
buckling)
Aw is the height of the wall including sole and head plates.

For best protection against noise, walls separating two dwellings are normally built as
two separate timber frame walls with studs either being staggered or placed opposite to
each other. It is normally not possible or advantageous to attach sheeting panels to the
unexposed side of the studs which would act as a bracing with respect to in-plane
buckling of the studs. Once the fire protective cladding attached to the exposed side of
the studs has lost its bracing function, the stud is unbraced. Therefore, in order to reduce
the buckling length with respect to in-plane buckling of the studs, noggins should be
inserted between the studs. In order to prevent in-plane buckling of all studs into the same
direction, a horizontal support is necessary, e.g. by a wall in transverse direction, or by
attaching diagonal steel straps to the unexposed side of the studs (the latter is not possible
where the studs are staggered).

Since the design models given below are only valid under the assumption that the
insulation remains in place, Table 6.12 gives an overview of design models and
recommendations regarding the need for mechanical fixing of the insulation.

Table 6.12 – Design models for floor and wall assemblies with insulated cavities and the
need for mechanical fixing of the insulation.
Rock wool Glass wool
Insulation
Batt-typea Loose fill Batt-typea Loose fill
Floors
t ≤ tf 6.6.3.2 6.6.3.2 6.6.3.2 6.6.3.2
Mechanical
t ≥ tf fixing – – –
6.6.3.2
Walls
t ≤ tf 6.6.3.2 6.6.3.2 6.6.3.2 6.6.3.2
Mechanical Mechanical
fixing for fixing for
t ≥ tf – –
hins < 120 mm. hins < 120 mm.
6.6.3.2 6.6.3.4
a
For floors also mat-type.

98
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.6.3.2 Assemblies with cavities completely or partially filled with rock

wool insulation

6.6.3.2.1 General
A design model is given in EN 1995-1-2 [1] Annex C (informative). The model was
originally published in [34] and [35], and is based on research results reported in [36] and
[37]. According to EN 1995-1-2 [1], the model should not be used for durations
exceeding one hour, while [34] and [35] do not express such a limit, since several wall
tests performed in accordance with the model have lasted for more than 90 minutes. The
design model was developed for assemblies with cavities that are completely filled with
insulation, see Figure 6.21a. Thermal analyses have shown that the model may be used
for partially filled cavities where the insulation is placed on the fire-exposed side of the
cavity and the insulation thickness is not less than 100 mm. The model does not explicitly
take into account air gaps between the insulation and the lining caused by the installation
of resilient channels or battens. Although the influence of these air gaps theoretically is
slightly positive, it is known from full-scale testing that their influence is actually
negative, since once of the fire penetrates through the lining the fire and hot gases can
easily spread across the entire width of the assembly; see [38].

6.6.3.2.2 Charring

Although the timber members are protected by insulation batts on their wide sides, it is
not only on their fire-exposed narrow side that charring takes place. Due to the heat flux
through the insulation, the timber members also char on their wide sides, giving rise to
extensive arris roundings, see Figure 6.21b. Therefore no consolidation of the charring
rate is possible, see Figure 6.8. For simplicity, the irregular residual cross-section is
replaced by an equivalent rectangular cross-section, replacing the charring depth dchar and
arris roundings with the notional (or equivalent) charring depth dchar,n, see Figure 6.21c.

2 3 1

4
6
h

2
dchar,n

5 7
dchar

b b

a) b) c)
Key:
1 Solid timber member (stud or joist)
2 Cladding
3 Insulation
4 Residual cross-section (real shape)
5 Char layer (real shape)
6 Equivalent residual cross-section
7 Char layer with notional charring depth

Figure 6.21 – Charring of timber frame member (stud or joist): a. Section through
assembly. b. Real residual cross-section and char layer. c. Notional charring depth and
equivalent residual cross-section.

99
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The notional charring rate is given as

β n = β 0 ks k n k p (6.40)

The coefficients ks, kn and kp are explained as follows:

The cross-section factor ks takes into account the effect of the width of the original cross-
section. In [34] and [35] it given as
⎧0,000167 b 2 − 0,029 b + 2, 27 for 38 mm ≤ b ≤ 90 mm
ks = ⎨ (6.41)
⎩1 for b > 90 mm

while [1] gives only a table with values for specific widths b.

Expression (6.41) assumes a linear relationship between dchar and time, which is slightly
conservative for dchar < 30 mm and non-conservative for dchar > 30 mm (for dchar > 30 mm
the load resistance is normally exhausted).

New results from [39] show that rock wool from various producers provides somewhat
less fire protection of the wide sides of the timber member than reported in [35][36][37].
Expression (6.41) should therefore be replaced with
⎧0,00023 b 2 − 0,045 b + 3,19 for 38 mm ≤ b ≤ 90 mm
=
ks ⎨ (6.42)
⎩1 for b > 90 mm

This expression can be used for rock wool insulation of minimum density of 26 kg/m2.

The coefficient kn converts the irregular charring depth into a notional charring depth, see
Figure 6.21 b and c. Strictly speaking, it depends on time, cross-section dimensions and
the cross-section property in question (area, section modulus or second moment of area).
The value kn = 1,5 given by [1] is a reasonable approximation for the notional charring
depth that would be relevant for a relative resistance between 0,2 and 0,4. For lower
values of relative resistance, and member widths of b > 60 mm, it is more conservative; kn
= 1,25 would be more appropriate..

The coefficient kp expresses the effect of the protection by claddings, where kp = k2 for
charring before the cladding has fallen off ( tch < tf ) and kp = k3 for charring after the
cladding has fallen off ( t ≥ tf ), see EN 1995-1-2 Annex C Expressions (C.3), (C.4) and
(C.5). Expression (C.4) should only be used for timber members over joints of gypsum
plasterboards Type F with tapered edges (this is not explicitly said in EN 1995-1-2).

The expressions for the notional charring depth, dchar,n, are given as:
d char,n = β n k2 ( t − tch ) for tch ≤ t ≤ tf (6.43)
d char,n = β n ⎡⎣ k2 ( tf − tch ) + k3 ( t − tf ) ⎤⎦ for tf ≤ t (6.44)

6.6.3.2.3 Strength and stiffness properties using the reduced properties method

According to EN 1995-1-2 [1] and [34], the strength and stiffness properties of the timber
members are determined using modification factors for fire kmod,fi, see Expressions (6.15)
to (6.18). The expressions for kmod,fi are given as:
d
kmod,fm,fi = a0 + a1 char,n (6.45)
h

100
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

d char,n
kmod,E,fi = b0 + b1 (6.46)
h
for strength and stiffness properties respectively, with parameters a0, a1, b0, and b1 given
for specific cross-sections in a number of tables. For other cross-sections, these
parameters can be determined by linear interpolation.

The number of cross-sections in EN 1995-1-2 [1] was reduced in comparison with [34].

6.6.3.2.4 Strength and stiffness properties using the reduced cross-section method

In order to simplify the calculations, zero-strength layers d0 were derived for application
of the reduced cross-section method, see [40]. The effective charring depth is calculated
as (see Figure 6.22)
d ef = d char,n + d 0 (6.47)

The following values of d0 should be used for members in bending (floor joists) with
b ≥ 38 mm and h ≥ 95 mm:
− For members with the fire-exposed side in tension
d 0 = 13,5 + 0,1h (6.48)
− For members with the fire-exposed side in compression
d 0 = 21,5 + 0,1h (6.49)
h

d0
def
dchar,n
dchar

Key:
1 Residual cross-section
2 Char layer
3 Notional (equivalent) cross-section
4 Notional char layer
5 Effective cross-section
6 Zero-strength layer below char layer

Figure 6.22 – Definition of charring depth, notional charring depth, effective charring
depth and zero-strength layer.

The values of d0 given in Table 6.13 should be used for axially loaded members (wall
studs).

101
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 6.13 – Values for d0 for wall studs with h in mm

Construction Exposure Buckling about Limitations d0

b ≥ 38 mm 13,5 + 0,1h (6.50)

Wall studs One-sided Stiff axis y-y
h ≥ 95 mm
b ≥ 38 mm 17 + 0, 25 h (6.51)
Wall studs One-sided Weak axis x-x
h ≥ 95 mm
b ≥38 mm
Wall studs Two-sided Stiff axis y-y 25 mm
h = 145 mm
b ≥ 38 mm
Wall studs Two-sided Weak axis x-x 44 mm
h = 145 mm

6.6.3.3 Assemblies with void cavities

Timber frame floors consist of solid timber joists, a decking made of wood-based panels
or timber and cladding directly fixed to the joints, resilient channels or other, see Figure
6.23. Timber frame walls consist of solid timber studs and claddings (linings) fixed
directly to the studs.

2
1
h

c
b
3 4

Key:
1 Solid timber joist
2 Decking
3 Lining
4 Void cavity

Figure 6.23 – Timber frame floor with void cavities and the cladding directly fixed to the
joists

a. Charring
Two design models for the determination of charring depths are available:

1. A design model according to EN 1995-1-2 [1]:

Information on charring of the joists or studs is given in Annex D (informative) of [1],

while the reduction of strength properties is calculated in accordance with (for example)
the reduced cross section method given in Clause 4.2 of [1]; see 6.5.1. Failure times of
wood-based panels and gypsum plasterboard used as claddings are given.

2. A design model according to [41] [42]:

Compared to the charring phases shown in EN 1995-1-2 [1] (Figure 6.4 - 6.6, [41] and
[42]), similar relationships apply with following assumptions; see Figure 6.24:
• Failure of the lining takes place at the time of start of charring, i.e.
tf = tch
• The charring rate has consolidated at time t*, given as

102
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

t * = tf + 5 min (6.52)
At that time, the charring depth is about 10 mm.

For charring before the lining has fallen off (tch < tf, Charring Phase 2 in Figure 6.6),
charring on the narrow side of the cross-section must be taken into account with kp,2a = k2
according to EN 1995-1-2.

The charring rates during the first phase, 3a, and second phase, 3b, after failure of the
cladding (see Figure 6.23) depend on the failure time of the cladding (i.e. implicitly on
the gas temperature in the fire compartment). On the narrow side of the cross-section the
charring rate during the second phase, 3b, after consolidation is also dependent on the
width of the joists or studs, i.e. on the degree of two-dimensional heat flux giving rise to
extensive influence of arris roundings. The set of expressions for the calculation of
charring depths is given in the following, with definitions where dchar,1 is the charring
depth on the wide sides and dchar,2 is the charring depth on the narrow side, see Figure 6.2:
Charring depth d char

3b

3a

t f t*
Time t

Key:
1 Relationship for initially unprotected members
3a Increased charring rate after failure of ceiling lining
3b Charring after consolidation at time t*

Figure 6.24 – Charring depth vs. time according to [42]

For charring phase 3a with tf ≤ t ≤ t *

d char,1 = d char,2 = kp,3a β 0 t (6.53)

For charring phase 3b with t ≥ t *

d char,1 = kp,3a β 0 ( t * − tf ) + kp,3b β 0 ( t − t * ) (6.54)

d char,2 = kp,3a β 0 ( t * − tf ) + kp,3b β 0 ( t − t * ) (6.55)

where
β 0 = 0,65 mm/min for softwood (solid and glued laminated timber and LVL)
8
kp,3a = 1 + tf for 0 ≤ tf ≤ 15 min (6.56)
75

103
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

7
kp,3a = 1,9 + tf for 15 min ≤ tf ≤ 60 min (6.57)
150
2
kp,3b = 1 + tf for 0 ≤ tf ≤ 60 min (6.58)
225
The corresponding notional charring depths are calculated as
d char,1,n = d char,1 (6.59)

d char,2,n = d char,2 kn (6.60)

where
kn = 1, 25

Note 1: [42] gives limits of validity for these expressions. These limits are normally not of
practical importance, since the joist would fail much earlier (to be checked for some cases).
Note 2: The model from [42] is valid for b ≥ 60 mm. Proposal: for b < 60 mm, only Stage 3a, but
unlimited, should be used.

b. Strength and stiffness properties

Strength and stiffness properties can be determined according to the reduced cross-section
method given in EN 1995-1-2 [1], Clause 4.2.

6.6.3.4 Cavities completely filled with glass wool insulation

a. Charring
The different charring phases are illustrated in Figure 6.25. For the time before failure of
the cladding, charring takes place as described in 6.6.3.2, see Figure 6.25a. Once the
cladding has fallen off at time t = tf, surface recession of the glass wool insulation takes
place due to thermal decomposition, so that the wide sides of the timber member are
increasingly exposed to the fire and start to char, see Figure 6.25b. When surface
recession of the glass wool insulation has reached the unexposed side of the insulation at
t = tf,ins, (Figure 6.25c), charring on the wide sides of the timber member will take place
over the whole depth of the cross-section (Figure 6.25d).

Note: EN 1995-1-2 [1] conservatively assumes that the mechanical resistance of the timber
member is exhausted at the time of failure of the cladding, tf.

104
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

d char,1,n,unexp

hchar,2b

dchar,2,n
d char,1,n d char,1,n d char,1,n
dchar,2,n

dchar,2,n

dchar,2,n
a b c d
Figure 6.25 – Illustration of charring phases:
a. Charring on narrow fire-exposed side before cladding has fallen off (tch ≤ t ≤ tf)
b. Charring on narrow side and wide sides during surface recession of glass wool
insulation (tf ≤ t ≤ tf,ins)
c. Recession of glass wool completed (t = tf,ins)
d. Charring on three sides after failure of glass wool insulation (t ≥ tf,ins)

In the following, it is either assumed that the cladding remains in place after the start of
charring of the timber member, i.e. , tch ≤ tf, or that the cladding falls off at the time of
start of charring, i.e. tch = tf.

For tch ≤ t ≤ tf (Figure 6.25a) the design model described in 6.6.3.2 applies:
d char,n = β n k2 ( tf − tch ) (6.61)

For tf ≤ t ≤ tf,ins (Figure 6.25b) the notional charring depths should be calculated as:
d char,1,n = k3 β 0 ( t − tf ) (6.62)

hchar,2b = vrec,ins (t − tf ) (6.63)

d char,2,n = k2 ks kn β n (tf − tch ) + k3 β n (t − tf ) (6.64)

where
vrec,ins is the surface recession rate for glass wool insulation:
vrec,ins = 30 mm/min
k2 is the insulation factor of the cladding from EN 1995-1-2 [1] expressions (C.3) or
(C.4), given as:
k2 = 1,05 − 0,0073 hp for unjointed claddings (6.65)
k2 = 0,86 − 0,0073 hp for jointed claddings (6.66)
k3 is the post-protection factor, given as
k3 = 0,036 tf + 1 (6.67)

For t ≥ tf,ins (Figure 6.25d) the notional charring depth dchar,1,n should be calculated
according to Expression (6.62) and the charring depths at the unexposed edge of the wide
sides of the timber member should be calculated as:

105
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

d char,1,unexp,n = k3 β 0 (t − tf,ins ) (6.68)

b. Strength and stiffness properties

For tch ≤ t ≤ tf the design model described in 6.6.3.2 applies.

For t > tf, the reduced cross-section method given by EN 1995-1-2 [1], Clause 4.2.2
should be used, i.e. the charring depth should be increased by d0 = 7 mm.
Note: For assemblies with tch ≤ tf, at time t = tf and shortly after, this is not consistent with 6.6.3.2.

6.6.3.5 Cavities filled with insulation other than glass or rock wool

Insulation other than glass or rock wool may exhibit better or poorer protection of the
timber member. In 6.6.3.2, insulation performance may affect the following parameters:
ks see Expression (6.41) or (6.42)
kn normally taken as kn = 1,5
kmod,fm,fi see Expression (6.45)
kmod,E,fi see Expression (6.46)
d0 see Expressions (6.48), (6.49) and Table 6.13.

Where insulation materials behave in a similar manner to glass wool, by thermal

decomposition when a specific critical temperature is exceeded, the surface recession rate
vrec,ins may be different from that for glass wool, see 6.6.3.4. These parameters may be
declared by the producer of the insulation, see Section 9.1.2 for guidance to determine
these parameters.

6.6.3.6 Worked examples

Worked example 1
A timber floor as shown in Figure 6.26 is exposed to fire from below. The dimensions of the
joists are 45 mm × 220 mm and the strength class is C 24. The cavities are filled with 100 mm
thick rock wool batts. The cladding consisting of an outer layer of 12,5 mm thick gypsum
plasterboard Type A (Layer 1), and an inner layer of 12 mm thick OSB (Layer 2) is fixed to
resilient channels of depth 25 mm. The resilient channels are fixed to the joists using screws of
length Af =29 mm. Determine the moment resistance of the joists for a fire resistance of R 30 and
check the screw length with respect to risk of withdrawal failure.

1 2 Key:
1 Joist
2 Decking (sub-floor)
3 25 mm resilient channel, ts = 0,6 mm
4 12 mm OSB
5 12,5 mm gypsum plasterboard Type
3 4 5 6 A
6 100 mm rock wool batts
Figure 6.26 – Cross-section of floor assembly.

Determination of charring depth:

Charring of the joists starts at the failure time of the cladding
tch = tf = tprot,1 + tprot,2 = 24,1 + 3,5 = 27, 6 min
where the protection times of Layers 1 and 2 are determined as in 5.5.1.

106
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The notional charring rate to be used is (EN 1995-1-2 [1], Expression (C.2))
β n = β 0 ks kn k3 = 0, 65 × 1,3 × 1,5 × 2, 0 = 2,54 mm/min
with
ks = 0, 000167 b 2 − 0, 029 b + 2, 27 = 1,3
kn = 1,5
k3 = 0, 036 tf + 1 = 0, 036 × 27, 6 + 1 = 2, 0
β 0 = 0, 65 mm/min

The notional charring depth at tfi = 30 min is

d char,n = β n ( tfi − tf ) = 2,54 × ( 30 − 27, 6 ) = 6,1 mm

Determination of design moment resistance in fire in accordance with EN 1995-1-2 [1],

Annex C:
The section modulus of the notional residual cross-section is
b ( h − d char,n )
2
45 ( 220 − 6,1)
2
Wr = = = 343149 mm3
6 6
The design bending strength in fire is
f m,k
f m,d,fi = kmod,fi kfi , see Expression (6.15)
γ M,fi
With a0 = 0,76 and a1 = 0,51 (see EN 1995-1-2 [1], table C2)
d char,n 6,1
kmod,fm,fi = a0 − a1 = 0, 76 − 0,51× = 0, 746 (see Expression (6.45))
h 220
f m,k = 24 N/mm 2
kfi = 1, 25 (see EN 1995-1-2 [1], Clause 2.3)
γ M,fi = 1, 0
24
f m,d,fi = 0, 746 × 1, 25 × = 22, 4 N/mm 2
1, 0
the design moment resistance in fire is
M d,fi = Wr f m,d,fi = 343.149 × 22, 4 = 7.679.675 Nmm = 7,68 kNm

Determination of design moment resistance in fire by the reduced cross section method, see
6.6.3.2d:
The effective charring depth is calculated using Expressions (6.47) and (6.48)
d0 = 13,5 + 0,1 h = 13,5 + 0,1× 220 = 35,5 mm
def = dchar,n + d0 = 6,1 + 35,5 = 41, 6 mm
The section modulus of the effective residual cross-section is
b ( h − d ef ) 45 ( 220 − 41, 6 )
2 2
Wef = = = 238.699 mm3
6 6
The design moment resistance in fire is
k 1, 25
M d,fi = Wef f m,k fi = 238.699 × 24 × = 7.160.976 Nmm = 7,16 kNm
γ M,fi 1, 0

Check of screw length:

The failure time of the screws, tsf, is calculated in accordance with EN 1995-1-2 [1], Expression
(C.12) as
A f − A a,min − ks k2 kn β 0 − ts 29 − 10 − 0 − 0, 6
tsf = tf + = 27, 6 + = 34,9 min > treq = 30 min
ks k3 kn β 0 1,3 × 2, 0 × 1,5 × 0, 65
Conclusion: There is no risk of premature failure of the resilient channels.

107
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Worked example 2
A wall assembly with a total height of 2800 mm, including top and sole plates, is shown in Figure
6.27. The dimensions of the studs are 45 mm × 145 mm. The strength class of the studs is C 24.
The cavities are completely filled with 145 mm batt-type rock wool. The cladding on the fire-
exposed side consists of one layer of 15 mm thick plasterboard Type F with a declared failure time
from the producer. The cladding on the unexposed side is made of one layer of gypsum
plasterboard Type H. The length of the screws on the fire-exposed side is Af = 52 mm. Determine
the design value of the axial load resistance of a stud without a cladding joint above it for a fire
resistance of R 60.
1 3 Key:
1 Stud
2 Cladding on fire-exposed side:
15 mm gypsum plasterboard Type F
3 Cladding on unexposed side:
2 4 9 mm gypsum plasterboard Type H
4 145 mm rock wool batts
Figure 6.27 – Cross-section of wall assembly.

Determination of charring depth:

The time of start of charring is given by EN 1995-1-2 [1] Expression (3.11), as
tch = 2,8 hp − 14 = 2,8 × 15 − 14 = 28 min
The insulation factor, k2, is given by EN 1995-1-2 [1], Expression (C.3), as
k2 = 1, 05 − 0, 0073 hp = 1, 05 − 0, 0073 × 15 = 0,94
The failure time of the cladding with respect to thermal degradation is declared by the producer as
tf = 65 min .
The failure time of the cladding with respect to pull-out failure of screws is (see EN 1995-1-2 [1],
Expression (C.9)), calculated as
A f − A a,min − hp 52 − 10 − 15
tf = tch + = 28 + = 50, 7 min < 65 min
ks k 2 k n k j β 0 1,3 × 0,94 × 1,5 × 1, 0 × 0, 65
The post-protection factor, k3, is given by EN 1995-1-2 [1], Expression (C.5), as
k3 = 0, 036 tf + 1 = 0, 036 × 50, 7 + 1 = 2,825
The notional charring depth at 60 minutes is:
dchar,n = ks k2 β 0 ( tf − tch ) + ks k3 β 0
= 1,3 × 0,94 × 0, 65 × ( 50, 7 − 28 ) + 1,3 × 2,825 × 0, 65 × ( 60 − 50, 7 ) = 40, 2 mm
The dimensions of the residual cross-section of the studs are
br = 45 mm
hr = 145 − 40, 2 = 104,8 mm

Determination of design value of axial resistance in fire in accordance with EN 1995-1-2 [1],
Annex C:
With
a0 = 0,55
a1 = 0, 40
b0 = 0, 60
b1 = 0,84
the modification factors for fire (for compressive strength and modulus of elasticity respectively)
are calculated as, see EN 1995-1-2 [1], Expressions (C.13) and (C.14):
d char,n 40, 2
kmod,fc,fi = a0 − a1 = 0,55 − 0, 40 × = 0, 44
h 145

108
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

d char,n 40, 2
kmod,E,fi = b0 − b1 = 0, 60 − 0,84 × = 0,36
h 145
The design values of compressive strength and modulus of elasticity are, see expressions (6.15)
and (6.18):
f 21
f c,d,fi = kmod,fc,fi kfi c,k = 0, 44 × 1, 25 × = 11,55 N/mm 2
γ M,fi 1, 0
E05 7400
Ed,fi = kmod,E,fi kfi = 0,36 × 1, 25 × = 3300 N/mm 2
γ M,fi 1, 0
With EN 1995-1-1 [15], Expressions (6.21) to (6.29), we get
A y A y 12 2800 12
λy = = = = 92,55
i hr 104,8
λy kmod,fc,fi f c,0,k 92,55 0, 44 × 21
λrel,y = = = 1, 73
π kmod,E,fi E0,05 π 0,36 × 7400
β c = 0, 2
( ( 2
k y = 0,5 1 + β c λrel,y − 0,3 + λrel,y ) ) (
= 0,5 × 1 + 0, 2 (1, 73 − 0,3) + 1, 732 = 2,139 )
1 1
kc,y = = = 0, 294
ky + k y2 2
− λrel,y 2,139 + 2,1392 − 1, 732
The design value of the axial load resistance of one stud is
N d,f i = Ar kc,y f c,d,fi = 45 × 104,8 × 0, 294 × 11,55 = 16.014 N = 16, 0 kN

Determination of design value of axial resistance in fire by the reduced cross-section method,
see 6.6.3.2d:
The effective charring depth is calculated as (Expressions (6.47) and (6.50)):
d0 = 13,5 + 0,1 h = 13,5 + 0,1× 145 = 28, 0 mm
d ef = d char,n + d 0 = 40, 2 + 28, 0 = 68, 2 mm
The effective depth of the cross-section is
hr,ef = 145 − 68, 2 = 76,8 mm
With kmod,fc,fi = 1, 0 , the design value of compressive strength in fire is calculated as
f c,k 21
f c,d,fi = kmod,fc,fi kfi = 1, 0 × 1, 25 × = 26, 25 N/mm 2
γ M,fi 1, 0
With EN 1995-1-1 [15], Expressions (6.21) to (6.29), we get
A y A y 12 2800 12
λy = = = = 126,3
i hr,ef 76,8
λy kmod,fc,fi f c,0,k 126,3 1, 0 × 21
λrel,y = = = 2,14
π kmod,E,fi E0,05 π 1, 0 × 7400
β c = 0, 2
( ( 2
k y = 0,5 1 + β c λrel,y − 0,3 + λrel,y ) ) (
= 0,5 × 1 + 0, 2 ( 2,14 − 0,3) + 2,142 = 2,977 )
1 1
kc,y = = = 0,198
ky + k y2 2
− λrel,y 2,977 + 2,977 2 − 2,142
The design value of the axial load resistance of one stud is
N d,f i = Ar kc,y f c,d,fi = 45 × 76,8 × 0,198 × 26, 25 = 17.963 N = 18, 0 kN

109
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.6.4 Light weight timber frame floors with I-joists

6.6.4.1 General

The following design model is taken from [43]. For general aspects, see 6.6.3.1. For
partially filled cavities, see 6.6.3.2.1.

6.6.4.2 Cross-section

The dimensions of an I-joist are shown in Figure 6.28. At ambient temperature, the
characteristic moment resistance is calculated as:
M k = f m,k Wef kh (6.69)

with
2 I ef
Wef = (6.70)
h
Ew
I ef = I f + Iw (6.71)
Ef
where:
fm,k is the characteristic bending strength of the I-joist, see below;
Ef is the mean value of the modulus of elasticity of the flange;
Ew is the mean value of the modulus of elasticity of the web;
If is the contribution of the flanges to the second moment of area;
Iw is the contribution of the web to the second moment of area;
kh is a depth effect, where applicable; see below.
h

bw
dchar,n
hf

b b

a b

Figure 6.28 – Cross-section of I-joist, a: at ambient

conditions, b: in the fire situation
When the bending resistance of the I-joist is calculated according to EN 1995-1-1 [23],
kh = 1 and the characteristic bending strength of the I-joist, fm,k, is taken as the bending
strength of the flanges. Since the mean flange design stresses σf,c,d and σf,t,d should not be

110
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

greater than the design compressive and tensile strength of the flanges, fm,k can be
replaced by
⎧
⎪f
⎪ m,k
⎪⎪ f c,k h
f m,k,ef = min ⎨ (6.72)
⎪ h − hf
⎪ f h
⎪ t,k
⎪⎩ h − hf
where the bending resistance of the I-joist was derived from testing, and the characteristic
bending strength of the I-joist, fm,k, and kh are declared by the producer of the I-joist.

In the fire situation, for I-beams in floor assemblies with cavities that are completely
insulated, the cross-section shown in Figure 6.28b should be used to calculate the
mechanical resistance for the required period of fire exposure t. This also applies for
partially insulated cavities with a minimum insulation thickness of 100 mm.

For failure during Charring Phase 3, – that is, that the cladding has fallen off at time tf, –
the notional charring depth, dchar,n, should be taken as:
dchar,n = β n (t − tf,ef ) (6.73)

where:
β n = β0 kb,ch k3 kn (6.74)

27, 4
kb,ch = +1 (6.75)
b
k3 = 0,0157 tf + 1 (6.76)
tf,ef = 0,9 tf (6.77)

kn = 1, 4
β0 is the one-dimensional charring rate given in EN 1995-1-2 [1], i.e. β0 = 0,65 mm/min
for solid softwood and LVL;
tf is the failure time of the cladding, in mm. It may be given by EN 1995-1-2 [1] or by
the producer, or be determined with respect to withdrawal failure of cladding
fasteners.
tf,ef is the effective failure time of the cladding;
b is the flange width in mm.

For failure during Charring Phase 2, i.e. that failure takes place at or before the time of
failure of the cladding, the notional charring depth, dchar,n, should be taken as:
dchar,n = β n (t − tch ) (6.78)

where:
β n = β 0 kb,ch k2 kn (6.79)

and k2 = 1 .

111
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.6.4.3 Failure of cladding fasteners

The penetration length into unburnt wood of fasteners for fixing claddings or resilient
channels should be at least 10 mm. The charring depth may be taken as the notional
charring depth dchar,n.

6.6.4.4 Strength parameters

For I-joists in bending where the fire-exposed flange is in tension, the modification factor
for bending strength, kmod,fm,fi, should be calculated as:
kmod,fm,fi = 1 − 0,016 d char,n kb,fm khf,fm kh,fm (6.80)

with
11,5
kb,fm = 0,76 + (6.81)
b
68
khf,fm = − 0, 41 (6.82)
hf
80
kh,fm = 1, 4 − (6.83)
h
where the notional charring depth dchar,n, the flange width b, the flange depth hf and cross-
section depth h are in mm.

For the influence of finger joints, see Sub-clause 6.5.4.1.5 below.

For shear strength verification of the web, the maximum temperature of the web should
be calculated as:
160 kb d char,n
Θ w,max = − 47 (6.84)
hf
For wood-based webs, the modification factor for shear strength, kmod,fv,fi, may be
calculated, using the reduction factor for shear strength given in Figure B4 of EN 1995-1-
2 [1], as:
⎧ 48 kb,ch dchar,n
⎪1 for ≤ 20
⎪ hf
kmod,fv,fi =⎨ (6.85)
⎪1, 47 − 1,13 kb,ch d char,n for
48 kb,ch dchar,n
> 20
⎪ hf hf
⎩

For shear strength verification of the glue-line between the web and the flange, the
temperature in degrees Celsius should be taken as:
⎧ kb,ch
⎪666 dchar,n − 12
Θ joint = max ⎨ b hf (6.86)
⎪20
⎩
where dchar,n, b and hf are in mm.

112
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.6.4.5 Finger-joint strength in flanges

Finger joint strength in fire may be dependent on the adhesive being used [10]. Since I-
joists are sensitive to finger joint failure, the bending resistance should be determined for
the design bending strength given as, replacing Expression (6.15):
f
f m,d,fi = kmod,fi kmod,fj,fi kfi k (6.87)
γ M,fi
where kmod,fj,fi is the modification factor for fire, expressing the reduction of finger joint
strength given in Table 6.14.

Table 6.14 – Modification factors for fire for finger joints

Adhesive kmod,fj,fi
PRF (phenolic resorcinol formaldehyde) 1,0
MUF (melamine urea formaldehyde) 0,75
PUR (polyurethane) 0,75

6.6.4.6 Reduced cross-section method for bending strength

For application of the reduced cross-section method, see 6.5.1 or 6.6.3.2, the depth of the
zero-strength layer should be calculated as, see Figure 6.29:
d 0 = 5,3 + 0,165 hf − 0,018 b − 0,0006 hf b (6.88)
dchar,n

d0

def

b b

a b

Figure 6.29 – Definition of notional

charring depth dchar,n, zero-strength
layer d0 and effective charring depth def

6.6.5 Other structural elements

6.6.5.1 Timber decks made of hollow core elements

6.6.5.1.1 General

The following design model is taken from [44]. It was developed based on fire tests of
timber decks made of hollow core elements with either void or insulation filled cavities.
For the fire tests performed with filled cavities, rock wool typically used in Switzerland

113
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

was used. In the design model, strength properties are determined using the reduced
cross-section method.

6.6.5.1.2 Charring

The charring model takes into account two different charring phases as shown in Figure
6.30. The first charring phase is given by the time needed for complete burn-through of
the fire-exposed bottom timber layer, i.e. dchar,n ≤ hu). The second charring phase is
characterised by charring of the vertical timber members (the webs) after the charring
depth has reached the thickness of the fire-exposed bottom timber layer (dchar,n ≥ hu). For
simplicity, linear relationships between charring depth and time are assumed for each
phase. Further, it is assumed that the vertical timber members are not exposed to fire on
three sides during the required fire resistance. Thus the fire-exposed bottom timber layer
is designed such that a fire penetration into the cavities is prevented, or the cavities are
filled with mineral wool batts with a degradation point greater than 1000°C, remaining in
place after the charred fire-exposed bottom timber layer has fallen off.

Figure 6.30 – Charring model for the calculation of the residual cross-section of the
hollow core elements.

The time t1, when the charring depth has reached the thickness of the fire-exposed timber
layer (dchar,n = hu) should be calculated using the notional charring rate β1,n for the first
charring phase as following:
hu
t1 = (6.89)
β1,n
For a required time of fire resistance, treq, the notional charring depth for the vertical
members of the hollow core elements should be calculated as:
d char,n = β1,n treq for 0 ≤ treq ≤ t1 (6.90)
(
dchar,n = hu + β 2,n treq − t1 ) for treq ≥ t1 (6.91)

Since timber decks made of hollow core elements are load-bearing, charring rates should
be taken from EN 1995-1-2 [1] Table 3.1 in rows a to c, but not d (wood-panelling). For
the first charring phase, since there normally exist joints between the hollow core
elements, the influence of increased charring should be taken into account by using
notional charring rates given in table 3.1 in [1] rather than the one-dimensional values,
i.e. β1,n = 0,8 mm/min for hollow core elements made of solid timber. The notional
charring rate β2,n for the second charring phase is mainly influenced by the thickness of
the vertical members and should be calculated as
β 2,n = β 0 ks kn (6.92)

where the coefficient ks should be taken from Expression (6.32) and kn = 1.5. Expression
(6.32) takes into account the protective function of the wide sides of the vertical member
using rock wool from various producers.

114
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

6.6.5.1.3 Strength and stiffness properties

For the design of timber decks made of hollow core elements the reduced cross-section
method should be used, see 6.4. The effective charring depth is calculated according to
Expression (6.93) using the values of d0 as given in Table 6.15.

Table 6.15 – Values of d0

d0
Charring phase
mm
First charring phase: dchar,n ≤ hu 7
Second charring phase: dchar,n > hu 20

6.6.5.2 Timber-concrete composite slabs

6.6.5.2.1 General

The following design model is taken from [45][46]. The design model is based on the
reduced cross-section method and takes into account the temperature-dependent reduction
of stiffness and strength of the connection.

6.6.5.2.2 Charring

The notional charring rate should be used (see 6.4.3) and the corresponding charring
depths are calculated according to Expression (6.2).

6.6.5.2.3 Strength and stiffness properties

For the design of timber-concrete composite slabs the reduced cross-section method
should be used, see 6.4. The effective charring depth is calculated according to
Expression (6.94), (see Figure 6.31).

Figure 6.31 – Determination of the effective cross-section in fire.

The timber board protects the concrete slab from the influence of high temperatures. As
the reduction of stiffness and strength properties of concrete are negligible at temperature

115
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

up to 200°C, the properties of concrete at normal temperature may be assumed for a fire
exposure of t ≤ 60 minutes and a timber board thickness of hs ≥ 20 mm.
The temperature dependent reduction of stiffness and strength of the connection is taken
into account using the modification factor kmod,fi. For connections with screws arranged at
an inclination of ± 45°, the modification factors kmod,fi depend on the side cover x of the
connectors (see Figure 6.32) and are given in Table 6.3. The influence of heat flux from
the bottom and the opposite side of the connector may be neglected if xu ≥ x + 20 mm and
xs ≥ 20 mm.

Table 6.16 – Modification factor kmod,fi for fire taking into account the effects of
temperature on the mechanical properties of the screwed connection, where x is the side
cover in mm as shown in Figure 6.31 and t is the fire duration time in minutes.
Parameter kmod,fi valid for
0 x ≤ 0, 6 t
0, 2 x − 0,12 t
Slip modulus (6.95) 0, 6 t ≤ x ≤ 0,8 t + 3
0, 2 t + 3
of screwed
connection 0,8 x − 0, 6 t + 1,8
(6.96) 0,8 t + 3 ≤ x ≤ t + 24
0, 2 t + 21
1,0 x ≥ t + 24
0 x ≤ 0, 6 t
0, 44 x − 0, 264 t
Strength of (6.97) 0 6, t ≤ x ≤ 0,8 t + 5
0, 2 t + 5
screwed
connection 0,56 x − 0,36 t + 7,32
(6.98) 0,8 t + 5 ≤ x ≤ t + 28
0, 2 t + 23
1,0 x ≥ t + 28

The expressions given in Table 6.16 for the temperature-dependent strength of the
screwed connection are the same as in [1], for axially loaded screws.

The normal stresses in timber and concrete, the shear stress in the timber and the
connection forces can be calculated according to the simplified method (γ-method) for
mechanically jointed beams with flexible elastic connections given in EN 1995-1-1
Annex B using the effective cross-section as shown in Figure 6.33 and the modification
factors kmod,fi for fire according to Table 6.9 taking into account the effects of elevated
temperature on the mechanical properties of the screwed connection.

6.7 Calculation vs. full-scale testing

In structural fire design by calculation, partial factors for actions and materials are chosen
in order to achieve a required reliability level. For timber structures, for example, design
values for resistance are based on the 20 % fractile of strength, see EN 1995-1-2 [1] and
6.5.1 above.

The traditional way of structural fire design of building elements was carried out with the
aid of full-scale fire tests. The test procedure is given in a series of European standards,
such as EN 1365-1 [47]. The results are presented according to a classification standard
EN 13501-2.

116
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In design by testing, the procedures applied at ambient temperature are different from the
procedure for the application of the results from fire tests. Evaluating the results from
tests at ambient temperature the result (resistance) is transformed into a characteristic
resistance with respect to the strength and/or stiffness properties of the specimens in
relation to the corresponding characteristic values (for more information see EN 1990
[29]). The results from fire tests are normally directly applied. For example, when a
timber frame wall assembly is to be tested, the timber is normally randomly selected from
the specific strength class to be used by the client. Since it is most likely that the strength
of the studs scatter around mean values rather than 20-percentile values within the
strength class in question, a direct application of the test results would lead to non-
conservative mechanical resistance in comparison with the reliability level assumed by
the Eurocodes. In the case of fire testing of walls, bending stiffness is the dominating
parameter. It would be sufficient, prior to assembling the wall, to make non-destructive
bending stiffness tests and to calculate the characteristic resistance taking into account the
real stiffness properties of the tested wall. In the case of fire testing of floors, it would be
best to select more joists than needed for the fire tests. A part of the material should be
tested in bending (destructive bending tests) in order to determine (ambient) bending
strength.

In order to eliminate inconsistency between the reliability system of Structural Eurocodes

and application of results from full-scale fire testing, the procedure of selecting materials
should be improved and include methods for determining strength and stiffness properties
of the test specimens to be tested.

6.8 References
[1] EN 1995-1-2:2004 Eurocode 5: Design of timber structures – Part 1-2: General –
Structural fire design. European Standard. European Committee for Standardization,
Brussels, 2004.
[2] EN 1995-1-2:2004/AC:2006 Design of timber structures – Part 1-2: General – Structural
fire design. Corrigendum. European Standard. European Committee for Standardization,
Brussels, 2006.
[3] EN 1995-1-2:2004/AC:2008 Design of timber structures – Part 1-2: General – Structural
fire design. Corrigendum. European Standard. European Committee for Standardization,
Brussels, 2008.
[4] EN 520:2004, Gypsum plasterboards – Definitions, requirements and test methods.
European Committee for Standardization, Brussels, 2004.

[5] EN 15283-2:2008, Gypsum boards with fibrous reinforcement – Definitions,

requirements and test methods - Part 2: Gypsum fibre boards, European Standard, CEN,
Brussels, 2008.
[6] Just, A, Gypsum plasterboards used as fire protective protection – a database. In
preparation.
[7] EN 13162:2008, Thermal insulation products for buildings - Factory made mineral wool
(MW) products – Specification. European Committee for Standardization, Brussels, 2008.
[8] .EN 301:1997, Adhesives, phenolic and aminoplastic for load-bearing timber structures –
Classification and specifications. European Committee for Standardization, Brussels,
1997.
[9] EN 15425:2008, Adhesives - One component polyurethane for load bearing timber
structures - Classification and performance requirements. European Committee for
Standardization, Brussels, 2008.

117
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

[10] König, J, Norén, J, Sterley, M, Effect of adhesives on finger joint performance in fire.
CIB W18, Meeting 41, St. Andrews, Canada. Lehrstuhl für Ingenieurholzbau, University
of Karlsruhe, Karlsruhe, Germany, 2008.
[11] Schaffer, E L, Charring rate of selected woods – transverse to grain. U.S. Department of
Agriculture, Forest Service, Forest Products Laboratory, Research paper FPL 69,
Madison, 1967
[12] Frangi, A, König, J, Effect of increased charring on the narrow side of rectangular cross-
sections exposed on three or four sides (tentative title), in preparation.
[13] EN 12369-1:2001, Wood-based panels – Characteristic values for structural design – Part
1: OSB, particleboards and fibreboards. European Standard. European Committee for
Standardization, Brussels, 2001.
[14] Frangi, A, Fontana, M, Hugi, E, Jöbstl, R, Experimental analysis of cross-laminated
timber panels in fire. Fire Safety Journal, 2009, in press.
[15] EN 1995-1-1:2004 Eurocode 5: Design of timber structures – Part 1-1 – General –
Common rules and rules for buildings. European Standard. European Committee for
Standardization, Brussels, 2004.
[16] König, J, Structural fire design according to Eurocode 5—design rules and their
background. Fire Mater. 2005; 29:147-163.
[17] EN 338:2008, Structural timber – Strength classes, European Standard, CEN, European
Committee for Standardization, Brussels, 2008.
[18] König, J, Effective thermal actions and thermal properties of timber members in natural
fires. Fire Mater.2006; 30:51-63.
[19] Källsner, B, König, J, 2000, Thermal and mechanical properties of timber and some other
materials used in light timber frame construction. CIB W18, Meeting 33, Delft. .
Lehrstuhl für Ingenieurholzbau, University of Karlsruhe, Karlsruhe, Germany, 2000.
[20] Franssen, J M, User’s manual for SAFIR 2007. A computer program for analysis of
structures. University of Liege, Department ArGEnCo, Service Structural Engineering,
2007.
[21] ANSYS Handbuch: FEM für Praktiker – Band 3: Temperaturfelder. 4. Auflage, Expert
Verlag, Renningen, Deutschland, 2001.
[22] Schleifer, V, Zum Verhalten von raumabschliessenden mehrschichtigen Holzbauteilen im
Brandfall. PhD Thesis No. 18156, ETH Zürich, 2009.
[23] EN 1995-1-1:2004 Eurocode 5: Design of timber structures – Part 1-1: General –
Common rules and rules for buildings. European Standard. European Committee for
Standardization, Brussels, 2004.
[24] König, J, Källsner, B, Cross-section properties of fire exposed rectangular timber
members. CIB W18, Meeting 34, Venice. Lehrstuhl für Ingenieurholzbau, University of
Karlsruhe, Karlsruhe, Germany, 2001.
[25] Thunell, B, Hållfasthetsegenskaper hos svenskt furuvirke utan kvistar och defekter. Royal
Swedish Institute for Engineering Research, Proceedings No. 161, Stockholm, 1941.
[26] EN 1194:1999. Glued laminated timber – Strength classes and determination of
characteristic values. CEN, European Committee for Standardization, Brussels, 2008.
[27] König, J, Schmid, J, Bonded timber deck plates in fire. CIB W18, Meeting 40, Bled,
Slovakia. Lehrstuhl für Ingenieurholzbau, University of Karlsruhe, Karlsruhe, Germany,
2007.
[28] EN 1991-1-1:2002 Eurode 1 Actions on structures – Part 1-1: General actions –
Densities, self-weight, imposed loads for buildings. CEN, European Committee for
Standardization, Brussels, 2002.
[29] EN 1990:2002 Eurocode Basis of structural design. CEN, European Committee for
Standardization, Brussels, 2002.
[30] König, J, Basic and notional charring rates. CIB W18, Meeting 35, Kyoto. Lehrstuhl für
Ingenieurholzbau, University of Karlsruhe, Karlsruhe, Germany, 2002.
[31] Schmid, J, König, J, [SP-Report]

118
Chapter 6 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

[32] Just, A, (report on glass wool-filled cavities), in preparation

[33] König, J, The structural behaviour of axially loaded wood studs exposed to fire on one
side. Trätek, Swedish Institute for Wood Technology Research, Report no. 8808057,
Stockholm, 1988.
[34] Östman, B., König, J., Mikkola, E., Stenstad, V., Carlsson, J., and Karlsson, B.,
Brandsäkra trähus. Version 2. Trätek, Swedish Institute for Wood Technology Research,
Publ. no. 0210034, Stockholm, 2002.
[35] König, J, Timber frame assemblies exposed to standard and parametric fires. Part 2: A
design model for standard fire exposure. Trätek, Swedish Institute for Wood Technology
Research, Report no. 0001001, Stockholm, 2000.
[36] König, J, Fire resistance of timber joists and load bearing wall frames. Trätek, Swedish
Institute for Wood Technology Research, Report no. 9412071, Stockholm, 1995.
[37] König, J, Norén, J, Bolonius Olesen, F, Toft Hansen, F, Timber frame assemblies
exposed to standard and parametric fires. Part 1: Fire tests. Trätek, Swedish Institute for
Wood Technology Research, Report. no. 9702015, Stockholm, 1997.
[38] Richardson, L R, Thoughts and observations on fire-endurance tests of wood-frame
assemblies protected by gypsum board. FireMater. 2001; 25:223-239.
[39] Just, A, The effect of insulation on charring of timber frame members. (To be published
as SP Trätek report.)
[40] König, J, The effective cross-section method for light timber frame construction with
solid timber members. SP-Report in preparation.
[41] Frangi, A, Erchinger, C, Design of timber frame floor assemblies in fire. CIB W18,
Meeting 40, Bled, Slovakia. Lehrstuhl für Ingenieurholzbau, University of Karlsruhe,
Karlsruhe, Germany, 2007.
[42] Frangi, A, Erchinger, C, Fontana, M, 2008, Charring model for timber frame floor
assemblies with void cavities. Fire Safety Journal 43 (2008) 551–564.
[43] König, J, 2006, Fire exposed simply supported wooden I-joists in floor assemblies. SP
Report 2006:44.
[44] Frangi, A, Knobloch, M, Fontana, M, Fire design of slabs made of hollow core elements.
Engineering Structures 31 (2009) 150-157.
[45] Frangi, A, Knobloch, M, Fontana, M, Fire design of timber-concrete composite slabs with
screwed connection, Journal of Structural Engineering 2010, Paper in press.
[46] Frangi, A, Brandverhalten von Holz-Beton-Verbunddecken, PhD Thesis No. 14328, ETH
Zurich, 2001.
[47] EN 1365-1:1999, Fire resistance tests for loadbearing elements — Part 1: Walls.
European Committee for Standardization, Brussels, 1999.
[48] EN 13501-2:2009. Fire classification of construction products and building elements –
Part 2: Classification using data from fire resistance tests, excluding ventilation services.
European Committee for Standardization, Brussels, 2009.

119
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

7- Timber connections

This chapter presents basic requirements for timber connections. Calculation methods in
Eurocode 5 are complemented with other design methods from recent research. Both timber-to-
timber and steel-to-timber connections are included. The models are applied and worked
examples are presented.

7.1 - General
One of the several important factors in improving the fire safety of timber buildings is a thorough
knowledge of the fire behaviour of connections between members. Among the various structural
components, connections are key elements because of their variety of configurations and. In fire
situation and in normal conditions, they determine the load-carrying capacity of the structure and its
safety. In structural analyses, timber connections are usually considered either fully rigid or fully
hinged, but in reality their behaviour is semi-rigid. Knowledge of the behaviour of timber connections in
fire is necessary in order to perform adequate analysis and modelling of the structures at the Ultimate
Limit State.
This chapter presents two approaches to calculation methods. The first is that of Eurocode 5 (EN 1995-
1-1, EN 1995-1-2) [1, 2], while the second applies connection design methods developed, based on the
results from experimental studies and numerical simulations.
In [3-5] the fire behaviour of multiple-shear steel-to-timber connections using dowels and slotted-in
steel plates was studied based on an extensive experimental and numerical analysis. The work proposes
an analytical model for calculation of the fire resistance of this type of connection for fire resistances up
to 60 minutes.
In [6-10], experimental and numerical studies are presented. They concern the timber connections using
mainly bolts and dowels and loaded in tension parallel to the grain. The available results for dowelled
connections in tension parallel to the grain are compared with the Eurocode formulae that allow an
extension of their domain of validity.
In [11, 12] a component model for dowelled timber-to-timber or timber-to-steel connections has been
presented. It is based on analysis of a steel beam on a continuous foundation (wood in embedment)
using a finite-element model for thermal and mechanical analyses
Other works concerning connections are presented in references [13-15, 18-19] bolted and doweled
types and in reference [16] for nailed types.

The main objective of a timber connection is to guarantee the mechanical resistance (R) of load-bearing
structures for at least a required time in order to allow safe evacuation of the building and to ensure the
safety of fire-fighters.
The required time is normally expressed in terms of fire resistance using the ISO standard fire exposure,
and is specified by the building regulations of each country. As far as timber connections are concerned,
the majority of the available fire test results have been produced during the last ten years. Over the same
period, numerical modelling of timber connections have been developed in some European countries.
The available data concern mainly dowelled and bolted timber-to-timber and timber-to-steel
connections.

121
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

7.2 - Design method according to EN 1995-1-2

EN 1995-1-1 [1] provides a general methodology for the design and calculation of timber connections in
normal conditions, while EN 1995-1-2 [2] provides a corresponding methodology for the design of
timber connections in fire. The formulae (8.6) and (8.7) given in EN1995-1-1, used for the design of
timber connections, are based on Johansen yield theory [17] with the rope effect contribution. They can
be applied to the nails, staples, bolts, dowels and screws per shear plane per fastener. The rope effect is
negligible in fire conditions. These formulae are used in one of the proposed Eurocode methods for fire
situations.
Investigations for an analytical model are based on the Johansen yield theory [17] using an embedment
strength reduction with temperature. The reduction of steel fastener characteristics with temperature is
not significant for the parts of the fastener in contact with the “active” wood in embedment (not charred
parts).
Under normal conditions, the load-bearing capacity of connections can be calculated using the EN
1995-1-1 rules based on a plastic limit state design. This approach uses the bending plastic resistance of
the fastener and the embedment strength of timber. The embedment strength is given by formulae using
two parameters: the fastener diameter, the wood density and, possibly also, the angle of the load to the
grain. The wood density and the fastener diameter are also the two parameters used to calculate the
initial stiffness of the connection (slip modulus, Kser). At the ultimate limit state, the slip modulus of a
connection Ku is related to Kser, and used for the structural analysis (internal loads distribution). The
plastic approach considers that the fastener in bending and the wood in embedment exhibit a rigid
plastic behaviour. This hypothesis is normally checked for connections loaded in the direction parallel
to the grain, and respecting the minimum spacing and edge and end distances as given in Eurocode.
When a force in a connection acts at an angle to the grain, the possibility of splitting caused by the
tension force component perpendicular to the grain, must be taken into account. A formula is available
to take account of the possibility of splitting caused by the tension force component perpendicular to the
grain (accordingly to section 8.1.4 of EN 1995-1-1). In addition, for steel-to-timber connections
comprising multiple dowel-type fasteners subjected to a force component parallel to the grain near the
end of the timber member, the characteristic load-carrying capacity of fracture along the perimeter of
the fastener area (block shear failure and plug shear failure), should be considered. This means that the
load-carrying capacity of timber connections, calculated according to the Johansen theory, can be
reduced by other types of failures depending on the configuration of the connections.
EN 1995-1-2 gives a method for designing timber connections under standard fire conditions for
fire resistance not exceeding R60. In this European standard, design rules are given for connections
made with nails, bolts, dowels, screws, split-ring connectors, shear-plate connectors and toothed-
plate connectors.
For the connections with wood-side members (for more details: see EN 1995-1-2, section 6.2), the
design approaches concern:
¾ Simplified rules for unprotected and protected connections, and some additional rules for
connections with internal steel plates.
¾ A reduced load method for unprotected and protected connections.
For connections with external steel plates (according to EN 1995-1-2, section 6.3), the approach
concerns the design of unprotected or protected connections.
Section 6.4 of EN 1995-1-2 presents simplified rules for axially loaded screws.
Reference [20] presents explicit application details of the calculation method based on EN 1995-1-1
and EN 1995-1-2.

122
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

7.3 - Other methods proposed for the design of timber connections

These “other” methods combine test results, finite-element approaches and analytical formulae related
directly or indirectly to the Eurocode principles. They can be considered as a complement to Eurocode
because they cover types of connections or validity domains which can complement the Eurocode
approaches. In reality, these studies are still at the research stage, and the results can be used as a step to
future improvement or as a complement to the Eurocode approaches. Nevertheless, the methods
proposed can be used as a basis for designing the connections for the dimension ranges defined in this
document. As explained above, as they are validated by experimental results, their application results in
safe design.
In these methods, the calculation approaches apply to the design of timber-to-timber and steel-to-timber
connections with dowel and bolt fasteners. As far as steel-to-timber connections are concerned, the
calculation methods apply to connections with one, two or three slotted-in steel plates.

7.3.1 - Timber-to-timber connection

7.3.1.1 - Introduction
Figure 7.1 is a schematic representation of an experimental realisation of a timber-to-timber connection.
In order to avoid separation of the main and the side members, including connections using dowel
fasteners, bolts are used near the ends of the external members -t1- (see black circles in Figure 7.1). The
connections are designed according to EN 1995-1-1.
The formulae given for these connections are valid for glued laminated members of Strength Class
GL24h and GL28. They could be used for connections composed of other wood species if the charring
rate is equal to or less than the charring rate of the tested glulam members, and if the connections are
designed according to EN 1995-1-1. The same type of formulae as in EN 1995-1-2 is used to calculate
the characteristic load-carrying capacity of the connection in shear (Equation 7.1). However the values
of the parameter k are modified as well as the domain of validity. The conversion factor η is calculated
according to equation 7.2.
Fd , t ,fi = η F v, Rk (7.1)

with:
η = e − k⋅t fi (7.2)
where:
Fd ,t ,fi is the design value of load-bearing capacity of the connection in shear under standard fire,

Fv,Rk is the characteristic lateral load-carrying capacity of the connection with fasteners in shear at
normal temperature (see EN 1995-1-1 section 8).

123
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

mm

Zone Zone
Zone tampon Zone tampon h
d'attache d'attache

Pièce moisée de
Lf
section 50 x t 2

t1 P1

P2 t2

t1 P1

Figure 7.1: Type of timber-to-timber connections

Table 7.1 – parameter k for the timber-to-timber connections

Maximum period of validity for parameter
Fasteners Timber element k
–k- in an unprotected connection
Bolts: d ≥12mm t1 ≥45mm 0,060 30 min
Dowels : d ≥12mm t1≥ 60mm 0,035 60 min

The parameter k is defined in table 7.1,

where:
t1 is the thickness of the timber side members in the connection

d is the fastener diameter,

t fi is the required fire resistance period.

For bolted connections, if t fi required is higher than 30 minutes, the thickness of the timber side
members of the connections must be increased by the following value (equation 7.3).
Δt = ( t fi − 30) ⋅ β n (7.3)

For dowelled connections, if t fi required is higher than 60 minutes, the thickness of the timber side
members of the connections must be increased by the following value (equation 7.4).
Δt = ( t fi − 60) ⋅ β n (7.4)

where:
Δt is the thickness added to each timber side member of the connection,

βn is the design notional charring rate under standard fire exposure (see Table 3.1 of EN 1995-
1-2).

The design model is based on the following assumptions:

• Fv,Rk is the characteristic lateral load-carrying capacity of the connection with fasteners in
shear at normal temperature (see EN 1995-1-1 section 8).
124
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

• a1=7d, a2=4d, a3=7d, a4=3d (see figure 7.3), d is the fastener diameter,
• the a3 and a4 are increased with a fi = β n ⋅ k flux ⋅ (t fi − t d ,fi )

o where :
ƒ βn : is the charring rate for glulam according to table 3.1 of EN1995-1.2 is
equal 0.7,
ƒ k flux :is a coefficient taking into account increased heat flux through the
fastener, equal 1.5,
ƒ t fi : is the required standard fire resistance period,
ƒ t d ,fi : is the fire resistance period of the unprotected connection given in table
6.1,
• the timber density takes into account in the calculations, is the characteristic density of timber
element ρ k .

7.3.2 - Steel-to-timber connection

7.3.2.1 - Steel-to-timber connection with one slotted-in steel plate
Figure 7.2 illustrates a steel-to-timber connection tested experimentally. In order to avoid separation of
the main and side members, including connections with dowel fasteners, bolts are used near the ends of
the external members –t1- (see black circles in Figure 7.2). The connections are designed according to
EN 1995-1-1.
The formulae given for these connections are valid for glued laminated members of Strength Class
GL24h. They could be used for connections composed of LVL and of other wood species if the charring
rate is equal to or less than the charring rate of the tested glulam members, and if the connections are
designed according to EN 1995-1-1.

Zone Zone
h
d'attache d'attache

section h 2t 1 + δ p

δp
t1

+0
2 -1 Piè
Epaisse=6 mm
se

Figure 7.2: Type of steel-to-timber connections

The same formulae as for timber-to-timber connection can be used but with the factor k taken from table
7.2
Table 7.2 – parameter k for the steel-to-timber connections

125
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Maximum period of validity for parameter

Fasteners Timber element k
–k- in an unprotected connection
Bolts: d ≥12mm, t1 ≥45mm 0,070 30 min
Dowels: d ≥12mm, t1≥ 60mm 0,045 60 min

If the required time to failure ( t fi ) is higher than 30 minutes, for the bolted connections, and 60
minutes for the dowelled connections, the thickness of the timber side members of the connections must
be increased by the values given in equations 7.3 and 7.4, respectively.
The thickness of the steel member (plate) in the connection is defined in accordance with EN 1995-1-1.

The design model is based on the same assumptions as the timber to timber model (see §7.3.1).

Worked example
The characteristic load-carrying capacity of steel-to-timber connection (with dowel fastener), given on
the figure 7.3 is design according to EN1995-1.1, from the following expressions:

⎧ f h ,α , k td mode 1
⎪
⎪⎪ ⎡ 4M y, Rk ⎤
Fv, Rk = min ⎨f h , α, k td ⎢ 2 + 2
− 1⎥ mode 2
⎪ ⎢ f dt ⎥
⎣ h ,α, k ⎦
⎪
⎪⎩ 2.3 M y, Rk f h ,α , k d mode 3
e1 a2
hp

a4+afi
a3a+a
3 fi 3 x a1 e1

δp
t1

t2
t1

Figure 7.3: Configuration of the steel-to-timber connection (d=16mm)

The required fire resistance is 60 minutes.

Strength class of timber elements is GL28, with ρ k =410 kg/m³;

The fastener (dowel) diameter is d=16 mm,

126
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

t1=76 mm, h=252 mm, δ p =8 mm and hp=114 mm.

a1=7d, a2=4d, a3=7d, a4=3d and e1=25 mm

f u =400 N/mm².

The a3 and a4 are increased with a fi = β n ⋅ k flux ⋅ (t fi − t d ,fi )

t d,fi : for mixed dowel and bolt fasteners take into account is equal 15.

This connection for mixed is loaded in tension parallel to grain α = 0° and f h ,0,k is calculated according
to the following formula:
f h , 0,k = 0,082(1 − 0,01d ) ρ k
The characteristic values for yield moment is calculated according to the following formula:
M y , Rk = 0,3 f u ,k d 2.6
The characteristic load-carrying capacity of steel-to-timber connection in cold conditions is:
Fv ,Rk =274 kN (defined according to (7.5))

For this steel-to-timber connection η = e − k⋅tfi = e −0,045⋅60 =0.0672

The design value of the load-bearing capacity F d ,t ,fi of this connection will be:

F d , t , fi = 0.0672 ⋅ 274 = 18,4 kN

In order to have a fire résistance R60, for this connection is necessary to applied un load Ed ,fi ≤18,4 kN.

In other words, it is necessary that Ed ,fi ≤ Fd , t ,fi ,

where: E d ,fi is the design effect of actions for the fire situation.

7.3.2.2 - Steel-to-timber connection with two or three slotted-in steel plates and with dowels

7.3.2.2.1 Introduction
An analytical design model for calculation of the load-carrying capacity in fire of unprotected multiple
shear steel-to-timber dowelled connections was developed in analogy with the reduced cross-section
method as given in EN 1995-1-2, and based on a combination of experimental and numerical analysis
[3-5]. The model is based on failure mode I according to the Johansen yield model (i.e. embedment
failure) and takes into account the influence of the steel elements (i.e. steel plates and steel dowels) on
the charring of the connection. The model was developed for a fire resistance up to 60 minutes.
The effective cross-section is calculated by reducing the initial cross-section by the effective charring
depth (def) as shown in Fig. 7.4. The temperature-dependent reduction of strength and stiffness of timber
in the heat-affected zones as well as the effects of corner roundings are considered by adding a further
layer (dred) to the charring depth (dchar). For simplicity, the same value of (dred) is used for charring on
side (index - s) and on top/bottom (index - o).
d ef ,s = d char ,s + d red (7.5)

d ef ,o = d char ,o + d red (7.6)

The design value of the load-carrying capacity Rd,fi of the connection loaded in tension parallel to the

127
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

grain can be calculated as follows:

R d,fi = A ef ⋅ f t ,0, k ⋅ k fi (7.7)

where:
Rd,fi is the design value of the load-carrying capacity of the connection
Aef is the effective cross-section ( A ef = b ef ⋅ h ef )
ft,0,k is the characteristic tensile strength of timber parallel to the grain direction

Figure 7.4: Residual (Ar) and effective cross-section (Aef) for the determination of the load-carrying
capacity of multiple shear steel-to-timber dowelled connections in fire, shown on the example of one
quarter of cross-section of a connection with three slotted-in steel plates (SD = steel dowel)
The design model is based on the following assumptions:
• Unprotected multiple-shear dowelled connections with two or three slotted-in steel plates. Spacings,
edge and end distances of the dowels according to EN 1995-1-1 for ambient-temperature design
(except for the spacing between dowels parallel to grain direction: a1=7d instead of a1=5d, where d is
the diameter of the steel dowels). A comparative numerical analysis showed that the assumption of a
spacing a1=5d leads to a reduction of the load-carrying capacity Rd,fi of the connection in the range of
8 to 10%.
• ISO-fire exposure on four sides.
• Glued laminated timber members with a minimum width of b≥160 mm and a minimum thickness of
the timber side member t1≥35 mm (see Fig. 7.4). The thickness of the timber middle member t2 (see
Fig. 7.4) is 8d as normally required for the design at ambient temperature for failure mode III.
• For one (n=1) or two (n=2) dowels within one row parallel to the load direction at least strength class
GL24h is required. In order to avoid net cross-section timber failure strength class GL36h should be
used for connections with three dowels (n=3) within one row parallel to the load direction (see Fig.
7.5).
• Embedment failure. i.e. failure mode (I) according to the Johansen yield model (because of charring
of the timber side members embedment failure was observed during all fire tests).

128
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 7.5: Minimal requirements on strength class of glued-laminated timber as a function of the
number of dowels n within one row parallel to the load direction

In order to reach a fire resistance of 60 minutes the size of the timber members shall be increased by a
thickness (c) (see Fig. 7.6). If the width (b) of the timber members is smaller than 200 mm, then a
thickness of c=45 mm is required; for b≥200 mm a thickness of c=40 mm is required. For a fire
resistance between 30 and 60 minutes the thickness (c) can be linearly interpolated as follows:
c = 1.5 ⋅ t − 45 for b < 200 mm and 30 ≤ t ≤ 60 minutes (7.8)
c = 4 3 ⋅ t − 40 for b ≥ 200 mm and 30 ≤ t ≤ 60 minutes (7.9)

While the timber size is increased by a thickness (c), the width of the steel plates should not be changed.
Thus, an air gap is created leading to a better protection of the steel plates against heat due to the
insulation effect of the air gap, particularly during the first phase of fire exposure (see Fig. 7.6).
Increasing the overall thickness of the timber members by a thickness (c) would not apply when
protecting adequately the connection with timber boards or gypsum plasterboards.

Figure 7.6: Definition of the timber size increased by a thickness c in order to reach a fire resistance of
more than 30 minutes shown on the example of one quarter of cross-section of a connection with three
slotted-in steel plates (SD = steel dowel)

7.3.2.2.2 Residual cross-section

An extensive FE-thermal analysis on a large number of geometries of multiple shear dowelled
connections commonly used showed that the side charring (dchar,s)is mainly influenced by the thickness
of the timber side member (t1), (see Fig. 7.3). The required minimum thickness of the timber side
member t1≥35 mm and the required increased size of the timber members by a thickness (c) for a fire
resistance between 30 and 60 minutes allowed the development of the simplified charring model for the
side charring as follows:
d char ,s = β 0 ⋅ t for 0 ≤ t ≤ 30 minutes (7.10)

129
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

d char ,s = β 0 ⋅ 30 + 1.5 ⋅ β 0 ⋅ (t − 30) for 30 ≤ t ≤ 60 minutes (7.11)

The charring model is based on the one-dimensional charring rate β0 and is characterised by two
charring phases. For simplicity linear relationships between charring depth and time are assumed for
each phase. The influence of the steel plate on charring leads to an increased charring rate of 1.5.β0
during the second charring phase (30 ≤ t ≤ 60 minutes).
The results of the extensive FE-thermal analysis showed that the top/bottom charring can be calculated
assuming an increased charring rate of (1.1.β0) from beginning up to a fire duration of 60 minutes as
follows:
d char ,o = 1.1 ⋅ β 0 ⋅ t for 0 ≤ t ≤ 60 minutes (7.12)

7.3.2.2.3 Effective cross section

The depth (dred) required for the calculation of the effective cross-section (see Fig. 7.4) depends mainly
on the following three parameters:
- the ratio between initial width (b) of the cross-section and width (br) of the residual cross-section
- the time of fire exposure (t)
- the number of dowels (n) within one row parallel to the load direction (see Fig. 7.5)
As the minimum thickness of the timber middle member (t2) required for the design at ambient
temperature is (8d) and timber cross-sections with a width greater than 300 mm are unusual,
connections with three slotted-in steel plates were analysed only for dowel diameters up to 10 mm (for
30 minutes fire resistance) as well as 8 mm (for 60 minutes fire resistance).
The depth (dred) can be calculated according to equations 7.13 to 7.16 for a fire resistance of 30 minutes
and equations 7.17 to 7.20 for a fire resistance 60 minutes. For a fire resistance between 30 and 60
minutes a linear interpolation can be carried out.

Fire resistance of 30 minutes

- Multiple shear steel-to-timber connection with two slotted-in steel plates:

n = 1: d red = −60(b / b r ) − 0.1d + 126.5 [mm] with: 8 ≤ d ≤ 16 mm (7.13)

n = 2, 3: d red = −40( b / b r ) − n(0.5d − 2) + 94 [mm] with: 8 ≤ d ≤ 16 mm (7.14)

- Multiple shear steel-to-timber connection with three slotted-in steel plates:

n = 1: d red = −60(b / b r ) − 0.4d + 133 [mm] with: 6 ≤ d ≤ 10 mm (7.15)

n = 2, 3: d red = −40(b / b r ) − 0.4d(n + 2) + 101 [mm] with: 6 ≤ d ≤ 10 mm (7.16)

Fire resistance of 60 minutes

- Multiple shear steel-to-timber connection with two slotted-in steel plates:

n = 1: d red = −30(b / b r ) − 0.6d + 117 [mm] with: 8 ≤ d ≤ 12 mm (7.17)

n = 2, 3: d red = −20( b / b r ) − d(0.2n + 1.4) + 101.5 [mm] with: 8 ≤ d ≤ 12 mm (7.18)

- Multiple shear steel-to-timber connection with three slotted-in steel plates:

n = 1: d red = −30(b / b r ) + 115.5 [mm] with: 6 ≤ d ≤ 8 mm (7.19)

130
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

n = 2, 3: d red = −20(b / b r ) − 4n + 94 [mm] with: 6 ≤ d ≤ 8 mm (7.20)

The calculation of dred is based on the strength class GL24h for connections with one or two dowels
within one row parallel to the load direction, while for connections with three dowels within one row
the strength class GL36h is assumed (see Fig. 7.5). For other strength classes the load-carrying capacity
in fire Rd,fi of connections with one or two dowels within one row parallel to the load direction can be
increased with the conversion factors given in Table 1.

Table 2. Conversion factors for the calculation of the load-carrying capacity in fire (Rd,fi) of multiple
shear steel-to-timber connections taking into account different strength classes
Number of dowels within one row [-] GL 24h GL 28h GL 32h GL 36h
n=1 1.00 1.08 1.13 1.18
n=2 1.00 1.08 1.13 1.18
n=3 --- --- --- 1.00

7.4 - References
[1] EN 1995-1-1, Eurocode 5 – Design of timber structures, Part 1-1: General – Commun rules and
rules for buildings, CEN, November, 2005.

[2] EN 1995-1-2, Eurocode 5 – Design of timber structures, Part 1-2: General – Structural fire
design, CEN, Brussels, 2004.

[3] C. Erchinger, Zum Verhalten von mehrfchnittigen Stahl-Holz-Stabdübelverbindungen im

Brandfall, PhD thesis, 2009.

[4] A. Frangi, Carsten Erchinger, M. Fontana - Experimental fire analysis of steel-to-timber

connections using dowels and nails, Fire and Materials, Paper in press

[5] Carsten Erchinger, A. Frangi, M. Fontana - Fire design of steel-to-timber dowelled connections,
Engineering Structures, Paper in press.

[6] D. Dhima - Vérification expérimentale de la résistance au feu des assemblages d’éléments en

bois – Saint Rémy les Chevreuses : CTICM, 1999, 61p. – Ref.INC-99/399-DD/NB.

[7] K. Laplanche, D. Dhima, P. Racher - Predicting the behaviour of dowelled connections in fire :
Fire tests results and heat transfer modelling – Proceedings of the 8th World Conference of
Timber Engineering, June 2004 – RIL / VTT / Wood Focus –Volume II – p. 335-341.

[8] P. Racher, K. Laplanche, D. Dhima - Thermo- mechanical modelling of the timber connection
behaviour under fire, Sif06, Aviero, Portugal.

[9] K. Laplanche, D.Dhima, P. Racher - Thermo-mechanical analysis of the timber connection

under fire using 3D, WCTE06, Portland.

[10] K. Laplanche - Etude du comportement au feu des assemblages de structures bois : Approche
expérimentale et modélisation, PhD Thesis, 19 juillet 2006.

[11] P. B. Cachim, J-M Franssen, Numerical modelling of timber connections under fire loading
using a component model, Fire safety journal, 2009.

[12] P. B. Cachim, J-M Franssen - Timber connections under fire loading, a component model for
numerical modelling, Application of structure Fire Engineering, 19-20 February 2009, Prague,
Czech Republic.

131
Chapter 7 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

[13] J. Fornather, K. Bergmeister, and H. Hartl, - Fire behaviour of steel and fasteners in wood
composites, RILEM Symposium Joints in Timber Structures, Stuttgart, pp 619-628, 2001.

[14] J. H. Fornather - Brandverhalten von Holz unter besonderer Berückscichtigung stiftförmiger

Verbindungsmittel – experimentelle und numerische Untersuchungen, PhD thesis, Wien,
November, 2003.

[15] D. Povel - Tragfähigkeit von Holzverbindungen mit stabförmigen Verbindungsmitteln im

Brandfall – Doktorarbeit von der Fakultät IV Bauingenieurwesen und Angewandte
Geowissenschaften der Technischen Universität Berlin zur Erlangung des akademischen
Grades – 2002 – 209 p. „Load-bearing capacity of timber connections with dowel-type
fasteners in case of fire“.

[16] J. Noren - Load-bearing of nailed joints exposed to fire, Fire and Materials. 1996; Vol. 20, pp
133-143, 1996.

[17] K. W. Johansen - Theory of Timber Connections, Forest Products Journal, Vol.25, N°2, pp 249
-262, 1949.

[18] P.H. Lau - Fire resistance of connections in laminated veneer lumber – PhD thesis Canterbury
University, New Zealand, May 2006

[19] T. Chuo, B. Chung - Fire resistance of connections in laminated veneer lumber – PhD thesis
Canterbury University, New Zealand, May 2006
[20] D. Dhima, A. Bouchair and A. FRANGI - Fire-safe use of wood in building, Timber
connection, 2nd draft, 28 august 2009, Study realised for CEI – Bois roadmap.

132
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

8 Fire stops, service installations and detailing in timber

structures

OBS Further lingual checks will be done before the next draft.
This chapter deals with the need for proper detailing in the building structure to prevent fire
spread within the building elements to other parts of the building. Special attention is paid to
basic principles, fire stops, element joints and building services installations. Several examples
of proper detailing are included.
F

The professional planning and execution of construction works in practice is again and again
critical and can, if inadequately done, lead to a wide range of damage to buildings, which can
often be removed only with considerable financial expense. Details that are particularly
sensitive be part of protection against heat, humidity or noise, the support structure, building
services and fire protection. This applies for all trades, materials and building systems.
Especially the detailing in fire protection is effected therefrom (see Figure 8.1). Thus
guaranteed component properties of buildings may be jeopardised.

facade; 19%

carcass; 21%

roof; 27%

fire doors / smoke control doors; 28%

raised access- / hollow core floors; 29%

drywall construction - walls; 43%

drywall construction - ceilings; 44%

fire dampers / ducts; 45%

penetration seals - pipes; 49%

penetration seals - cables; 54%

Figure 8.1. Frequency of defects or non-compliances in fire protection details in building

constructions [1]

The required quality characteristics of a building structure will determine by the standard of
design of the building, production conditions in the factory and during erection, the
qualifications of the persons involved and by the standard of supervision of construction.

Important aspects of quality assurance are therefore:

• deduction and further training of involved persons, teaching of responsible and self-
monitoring,
• definition of clear workflow processes from the target formulation to the point of
utilization,
• definition of responsibilities and interfaces,
• use of appropriate building systems and components with approved and/or tested details,

133
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

• controlling of performed works in the design and erection including ultimate inspection by
a suitable person (authorised inspector for structural fire protection), building inspection
in the licensing process.

The subsequent references and recommendations of construction details for fire protection are
essential tools for the effective quality improvement in the design and construction of
buildings.

8.1 Detailing in timber structures

8.1.1 Timber constructions
Trying to classify timber construction systems a typology from rod-shaped constructions to solid
cross-sections can be demonstrated. Regarding the chronology the timber construction can be divided
into the following types:
• Log construction
• Framework construction
• Skeleton construction
• Timber-frame construction
• Solid timber construction

a) Log construction b) Framework construction c) Skeleton construction

d) Timber frame construction e) Solid wood construction f) Mixed construction

(Cross-laminated timber, …)

Figure 8.2. Types of timber constructions with design of load-bearing structure [2]

134
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Until the beginning of industrialization log and framework constructions were widespread in European
countries. These types of construction are very traditional and manufactured with craftsmanship.
The development of skeleton and timber-frame constructions is a result of the increasing
industrialisation. The characteristic of skeleton and timber-frame constructions is the high grade of
pre-fabrication. In recent years modern solid wood constructions like cross-laminated timber
constructions become more important. The advantage is the possibility of large pre-fabricated wall-
and floor assemblies.
Modern constructions with timber are corresponded with the requirements of a sustainable
development regarding to the minimization of energy consumption.

The different timber construction principles are shown in Figure 8.2. Components, connections and
building services systems must therefore be appropriately designed to suit the particular type of
structure, in order to withstand the different fire behaviours of the structures.

Log construction:
The building envelop of previous log constructions were built with one layer, which have to take over
separating and load-bearing functions. Because of the requirements to comfort and insulation recent
log constructions will be built with more than one layer. These several layers have to fulfil different
functions.

Framework construction:
Most of the framework constructions are built with a visible load-bearing structure. The load-bearing
elements have larger and quadratic cross-sections. Vertical and horizontal loads are transferred to
foundation from the load-bearing structure consisting of triangular bracings in external and internal
walls. The connections designed with traditional connections such as scarfs and step joints. Can only
transfer compressive forces.
The voids have been filled with masonry or a mixture made from clay and straw in the traditional way.
At present insulation materials will be used to fulfil requirements for heat protection.

Skeleton construction:
The design and position of columns and beams as well as connections of skeleton constructions can be
arranged in different ways. The arrangement of the structure will be affected by the type of design of
the vertical, horizontal and diagonal structural rod-shaped assemblies on one knot. Either the beams or
the columns or the beams and columns passes the knot or the beam and columns will be slotted on the
knot. In Figure 8.2.c are illustrated two-piece beams passes and columns as double tie joints. The
characteristic of such a connection are the length of timber in front of step joints using dowelled
connection. Wall elements can be placed inside, between or outside of the load-bearing structure. To
reduce penetrations of the building envelop the wall or roof elements shall be fixed on the outside of
the skeleton structure.

Timber-frame construction:
Timber-framing is a flexible and cost-effective construction method. Contrary to buildings in skeleton
and framework construction, where the loads will be not transferred from rod-shaped elements, but
from elements loaded in-plane or perpendicular to plane. There is no difference between loadbearing
and separating assemblies.
The studs have to carry mainly vertical loads from the roof and floorings. The panels will allocated the
wind- and stiffening loads.

Solid timber construction:

As a result of additional industrial methods to produce large-area elements new timber
constructions have been developed. Spacial elements with loadbearing and separating
function offer a wide range of application for wall-, floor- und roof elements.
The main part of the structure system is a loadbearing core made from solid timber and timber
derived products. Solid timber element such as cross-laminated timber can be also designed as

135
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

two-way slabs. An additional similarity of solid timber constructions is that the insulation will
be fixed on the outside of the loadbearing structure.

Mixed construction:
To fulfil structural, architectural or economic aspects it can be necessary to use mixed
constructions.

The different constructions mentioned above have different properties, depending on the fire
behaviour of assemblies and connections. It is, for example, easier to prevent hot gases and flames
breaking through or pass connections in solid wood constructions than it is in skeleton constructions.
To improve the connections of wall or ceiling elements the influencing factors of fire spread through
connections must be determined.

8.1.2 Influencing factors for fire safety

Depending on the intended purpose of a building, the fire resistance of walls or ceilings depends in
turn on the number of storeys and the floor areas. Single-family or two-family houses have lower
requirements than multi-storey buildings, because of the number of occupants and shorter escape
routes in the former class.

To achieve the required fire safety level, the fire behaviour of the building construction, service
installations and additional safety measures must be considered in relation to the possible fire
scenarios. The evaluation criteria above are subdivided, for instance the design of joints, building
assemblies and supporting structure by the criteria building construction have to be determined.
The sub-class of building construction are interlinked with those of services and additional measures.
For instance the fire resistance of external or internal wall have to evaluated, considering the assembly
with and without service installations (see Figure 8.3).

Figure 8.3. Evaluation criteria for timber constructions [3]

The spread of fire can be minimised with fire stops inside, or as part of, connections between
assemblies and at interfaces, such as penetration seals where electric cables or heating system pipes or
ducts pass from one fire cell to another. Additional safety measures include preventive structural
measures and the use of specific fire protection systems such as sprinklers or smoke detectors.
Connections of wall, ceiling and roof elements have a significant influence on fire behaviour and the
danger of uncontrolled spread of smoke, hot gases or fire. Poorly designed or constructed connections

136
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

can present a risk to life or safe evacuation through the spread of carbon monoxide to neighbouring
areas, or a risk to property through the spread of fire via voids.
The following basic principles will maintain the fire resistance of assemblies.

8.1.3 Basic principles for detailing

Normal framing methods in construction have voids between insulation and panels. If these voids
connect with those in a ceiling, hot gases and flames can spread into these spaces. Fire stops restrict
the passage of flames, hot gases and smoke and as a consequence an uncontrollable fire spread will be
prevented.

A schematic illustration of fire spread paths is shown in Figure 8.4.

2) 1)

3)

4)
1)

Figure 8.4. Schematic illustration of risks presented by connections between building components,
loadbearing separating wall and floorings [4]

To evaluate the fire behaviour of connections, the fire resistance of assemblies, insulation performance
and failure times of panels must be calculated or tested. Connections between building components
having a separating function must achieve the same fire resistance as the adjacent components. The
load-bearing structure and panels must not be exposed to fire through poorly designed element joints.

Path 1:
The fire resistance of separating walls and floorings can be calculated in accordance with the methods
prescribed in EN 1995-1-2:2006 [7]. This includes consideration of load-bearing capacity, insulation
and integrity as described in Chapters 5 and 6. Calculation methods can also be applied to element
joints, e.g. the reduction of the insulation value of panels through different joint types.

Path 2:
Penetrations or openings for building services must not degrade the fire resistance of the floorings.
Section 8.3 describes methods of sealing electrical cables passing between fire compartments.

Path 3:
The load-bearing structure and inside surface of panels must not be exposed to fire, before the failure
time of interior ceiling panels. Interior wall panels and element joints must be airtight. Airtightness of
the building envelope is a requirement in order to avoid condensation inside exterior walls.

In a fire, loss of the airtightness depends on the position in the construction and the material that has
been used for the airtight layer. Adhesive tapes, for example, will lose their sealing performance at
temperatures above 100 °C.

137
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Path 4:
Continuous element joints in the connection area must be sealed on the narrow side. Hot fire gases
pass these joints and reach the neighbouring area. Such continuous joints exist by solid timber
elements or structural hollow sections. The narrow sides must be protected with fire stops made of
insulation or timber derived boards.
In general, building components having a fire-separating function must be connected as follows:

• Continuous joints must be avoided.

• Panels must be fitted accurately to adjacent components.
• Joints in multi-layered panels must be staggered.
• Connections between walls and adjacent components must be friction-locked.
• The load-bearing function of connections between ceilings and walls must be maintained during the
required period of fire resistance.
• Voids in the connection area must be protected by appropriate measures, e.g. by filling the voids
with mineral fibre insulation (melting point ≥ 1000 °C and density ≥ 26 kg/m³).
• Effects caused by swelling and shrinkage must be allowed for when designing timber buildings.

Detailed proposals for connections between components can be taken from construction catalogues
[3], [4], [5], [6]. Examples of how to evaluate connections and element joints are described below.

8.2 Fire stops and element joints

8.2.1 Evaluation of structural measures in connections

The fire resistance of assemblies and the failure time of panels and claddings of these elements are
important when deciding the necessary performance of connections. When using EN 1995-1-2: 2006
wall and ceiling components are separated into single layers, of which the failure times and the
thermal heat transfer can be calculated. The fire resistance for the separating function of the wall and
flooring components can be determined as the sum of the failure times of the single layers. The
calculation method in EN 1995-1-2 is limited to a fire resistance of 60 minutes.

If wall- and flooring constructions cannot be calculated in accordance with EN 1995-1-2 fire tests are
necessary. That is the case, if other configurations of walls and floorings or a fire resistance above 60
minutes is required. Fire tests on wall- and floor assemblies will be planned on basis of the series of
standards EN 1363 [8] and EN 1365 [9]. The disadvantage of fire tests is that only minor changes of
the test specimen are possible. For instance the thickness of panels can be heightened.

Fire tests for connections of assemblies are not included in standards. Some details of element
connections have been fire-tested in research projects [3], [5], [10], but until now, no guidelines for
element connections have been published.

In addition to the fire test performance of separating walls and floorings, the leakage rate of and
temperature rise inside the connections must also be quantified. This criterion is necessary in order to
avoid leakage of toxic gases into neighbouring areas. For measuring the leakage rate, a smoke duct
have to be fixed in front of the element joint on the unexposed side, enabling the transmission factor
and the concentrations of CO, CO2 and O2 to be measured. The ignition of combustible parts inside the
construction and the therewith an uncontrollable fire spread into neighbouring units must be avoided
in view of fire fighting brigades.

To evaluate such connections different fire spread paths must be determined, as shown in Figure 8.4.
In addition to fire protection measures, detailing of timber connections is affected by other factors,
such requirements for thermal insulation, noise insulation or airtightness.

138
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

C2
C1

Figure 8.5. Connection of a fire separating internal wall to flooring construction

(after failure of panels for sound insulation)

The fire resistance of assemblies (paths A and B) can be calculated as specified in EN 1995-1-2, or be
taken from test reports. To evaluate the fire behaviour of the connection as a whole requires fire spread
paths C1 and C2 to be taken into account.

The following principles must be fulfilled from the connection above:

• The element joint must be sealed with sealing compounds (path C1).
• The layers of the cross-laminated timber ceiling must also be glued on the narrow sides. If not, the
ceiling must be cut above the separating wall, and a fire stop as shown in Figure 8.5, Path 4 is
required (path C2).
• For the connection between panels and the timber frame member on the top, the distance between
fasteners must not exceed 150 mm.

Section 8.2.2 describes some working examples.

8.2.2 Working examples

Knowing the weak points in timber constructions, connections could be improved by using
straightforward measures. The measures vary, depending on the different timber constructions and the
arrangement of services in the building. Fire protection measures for installation services are explained
in Section 8.3.

The working examples are removed from the construction catalogue [11]. In the figures are illustrated
connections regarding to the building code requirements. These requirements have to be fulfilled from
separating walls and floorings and their connections. Furthermore the integration of rod-shaped
assemblies such as girders in flooring constructions has to be designed in view of fire protection.

In addition to the requirements of the building codes those of insurance companies have to be
determined. In opposite to the life safety of persons the loss of property comes to the fore considering
insurance aspects. Planning principles to fulfil insurance requirements will be included in the
construction catalogue, regarding the degree of damage against the duration of fire stress.

139
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

To integrate building connections in the construction catalogue, those will be divided into two types:
1) Connections of spacial assemblies
2) Connections of spacial and rod-shaped assemblies

a) Timber-frame construction b) Timber-frame construction

Floor ⊥ External wall Floor ⎜⎜ External wall

c) Timber-frame construction d) Timber-frame construction

Floor ⊥ Internal wall Floor ⎜⎜Internal wall

Figure 8.6. Connections of spacial assemblies

a) Platform construction - external wall to floor, b) Balloon-frame construction – external wall to
floor, c) Platform construction, internal wall to floor, with fire stop made of timber-derived boards,
d) Platform construction, internal wall to floor, with fire stop made of cellulose insulation

Figure 8.6 a-d give examples of fire stops in wall and ceiling constructions. The fire stops are
coloured. Hot fire gases must not reach the voids of the floors above the insulation layer or
neighbouring areas. In connection with fire stops made of insulation or stillage boards, the passage of
fire gases will be restricted.

The fire stops must fulfil requirements, e.g. the stillage board (see Figure 8.6c) must be grooved or
sealed to adjacent boards with sealing compounds. The stillage board must be fixed with fasteners on
top and bottom with a maximum separation of 150 mm. Cellulose insulation (see Figure 8.6d) may
also be used for the fire stop, because of its excellent clamp force, the connection will be sealed.

140
 Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

a) b)

Inner gutter
∅ 125 mm
Front
plate of
the outer

Metal sheet
Support
hinge Lower plate of
outer gutter

Figure 8.7 a) Connections of spacial assemblies – wall to roof - Diagram of fire spread delaying
eave (installation method in tests) and b) photograph

An eaves structure has been developed which substantially delays fire spread into roof void spaces
[12]. With the arrangement, as shown in Figure 8.7, the spread of a flashover fire to the eaves can be
delayed and the spread of an exterior fire to the roof void spaces can even be prevented. According to
the tests, the developed eaves structure delays the spread of a flashover fire by approximately 10 to 15
minutes, which is often long enough for the fire services to prevent the fire from spreading to the roof
structures.

a) Timber-frame or Skeleton construction b) Section A-A

Floor ⊥ main girder
A

Figure 8.8. Connections of spacial and rod-shaped assemblies

Figure 8.8 shows the connection from a girder to the ceiling in a timber frame construction with a
suspended ceiling. The element joint between the batten and girder must be sealed, in order to prevent
two-sided charring of the panels before the failure time is reached.

An additional and important question is related to the service installations in timber structures. Fire
shutters and penetration sealing systems are necessary to guarantee the fire resistance of wall and
ceiling structures when installations have to pass through these structures. Most of these products and
systems for electrical, ventilation, water or heating systems are designed for, and have been tested
with, masonry and concrete structures. Certified items for use with timber structures are rare, and often
not even available. Penetration seals of pipes, cables, ducts etc. in residential, office or public
buildings are particularly important in multi-storey timber structures.

141
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

8.2.3 Ventilation voids

The main threat of a fire that spreads in the ventilation void of a façade is that the fire can spread from
the façade to other spaces, such as apartments or attic/roof void space. The latter of these two events
has a higher probability, unless it has been prevented with a special eave structure. The fire can spread
to apartments through for example gaps in the windows or if a window is open.

The following focuses on façades, where the outer surface of the façade and the ventilation void has
been made from class D building materials (=wood). The threat is external ignition. The most
important technical problem in the structural sealing of the fire spread in the ventilating void of the
façade is that the sealed structure has to retain its moisture technical functionality. To create a
functional fire sealing solution for the ventilation void, a compromise has to be made between the fire
and moisture technical requirements.

Two fire sealing solutions [10], [11] and [12] are described below: perforated steel profile (Figure
8.9a) and fire seal made from two horizontal wooden bars (Figure 8.9b). In the steel profile seal the
diameter of the holes was 18 mm and the spacing 140 mm. On one side of the fire seal there were 6
holes with an area of 15 cm² in total, which equals 5 % of the cross-sectional area of the seal.
According to fluid dynamic examination it is enough for moisture technical functioning, as long as no
significant amounts of rainwater can flow into the ventilation slot. For instance a plain tongued and
grooved wooden wall was a tight enough structure considering the requirements. The air flow
restricting type of seal was made wooden bars resulting into was 7 mm wide gap, which equals ca. 20
% of the cross-sectional area of the sealed gap. In practice the gap of the wooden bar seal cannot be
made much narrower because then not much tolerance would be left for the dimensional accuracy,
moisture related dimensional changes, litter/dust and other similar matters.

a) b)

Figure 8.9. A fire seal made from a) non-combustible and b) combustible material

142
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

8.3 Building service installations

The increase in the standard of residential buildings is largely accounted for by the increase in the use
of technological equipment over the last few years. Features such as electrical installations,
communication and TV, bus and heating systems, water and sewage or sprinklers are generally
included in the planning and construction phase of new buildings. Approximately 10 % of the costs of
multi-storey buildings are associated with these installations.
Despite an increasing number of these building services system installations and the associated
penetrations in fire separating walls and floors, the requirements for fire resistance of all fire
separating components are not reduced. Any reduction in the fire resistance of elements for fire
compartmentation due to these installations is unacceptable. Time and again, inspections of new and
existing buildings report that inappropriate or defectively sealed service installations in fire separating
wall- and floor constructions, whether due to poor design or poor workmanship, create weak spots
(Figure 8.1) and reduce resistance to the spread of fire.
The need of building services system installations and the guarantee of fire protection in separating
elements with penetrations, requires for all building materials and type of constructions (concrete,
masonry, drywall or timber) approved and certificated sealing compounds or fire shutters, which fulfil
the same requirements for fire protection capacity as the separating elements, to avoid the spread of
fire and the pass of hot gases as well the ignition on unexposed side.
Until now, tests of fire shutters and sealing compounds for penetrations in timber constructions have
been rare. Approved sealing systems for penetrations and fire shutter systems are typically available
only for concrete or drywall constructions.
The basis for practical use of approved penetration sealings and fire shutters for building services
system installations are fire tests in accordance with series of standards EN 1366 [13].

Conceptual arrangements of fire compartments for building services installation

Penetrations of building services systems in separating elements are unavoidable, and must be allowed
for the start of the planning phase. All building regulations specify that the fire rating of separating
building elements between fire compartments must not be impaired by services that pass through
them. A basic concept in planning and design for building services systems is to distribute them via a
central configuration of conduit, like in shafts or ducts, with horizontal branches for distribution to
individual areas and compartments. This avoids unnecessary penetrations and causes to a considerable
degree.
The aim of all arrangements in phase of planning should be an assignment of building services
installation on previously defined fire compartments. This can be done by the application of the
following design concepts:

143
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 8.1. Concepts and arrangements of fire protection for installation of building services / fire
compartmentation
Concepts Description
• service shaft / This concept based on continuous conduit in separate service shafts or
service ducts service ducts. The fire resistance rate of these shafts and ducts corresponds
to the fire resistance of the separation elements. These service shafts or ducts
include a multitude of cables or pipes for building services installation. The
areas of penetrations of services have to be sealed in the quality of the
separating elements.
All installations that pass through the separating elements of shafts or ducts
have to be sealed with approved systems without to impair the fire resistance
of separating elements in the area of the penetration.
• penetration seals Sealing of penetrations for building services installations in separating
elements (walls, floors) with approved sealing systems, to ensure the fire
protection quality of the separating elements.
• continuous encasing The arrangement is similar to the principle of service shafts or ducts, but
with a single encase of each cable or pipes over the entire length in
accordance with quality of the fire resistance of separating elements. This
allows the renouncement of a continuous service shaft or duct.

installation shaft with sealing in each fire separating encasing of each installation
penetration sealing element lines
Figure 8.10. Concepts for arrangement of building services installations with regards to fire
compartmentation [13], [15]

144
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Separation and decoupling of pipes, wires, ducts etc. from the structure of the building must
be considered and worked out as part of the planning stage. These solutions must also fulfil
the requirements in respect of thermal, acoustic and moisture separation and protection, while
also allowing for access for maintenance and installation of additional pipes, wires, ducts etc.
if required some time in the future.

8.3.1 Types of building services installations

Penetration seals are structural arrangements to prevent an early spread of fire and smoke in separating
elements such as walls, floors, shafts and ducts passed through of building services installations. These
systems ensure the classified fire resistance of the separating elements despite penetrations of services.
Simultaneously the penetration sealing systems should also ensure no early spread of fire in the
separating elements themselves by the penetrations of building services installations.
For the classification of building services lines passes through separating elements, between type,
number, size, substance and material of building services installation are distinguished.
These information have decisive influences on arrangements and type of penetrations seals to be taken.

Types of service
installations

Ducting Electrical Pipes

cables

Fire rated With fire Steel or cast Non ferrous

Plastic
damper iron metals

Single cables Bundles Cable trays Conduit and

trunking

Combinations of types of service installations

Figure 8.11. Types of service installations.

Every type of building services installations passing through fire separating elements has own
characteristics and reacts differently in the case of fire, so there is no single solution or product that
will protect all services. Each penetration for service installation in separating elements needs the
application of specific and individual solutions, like sealing compounds or fire shutters adapted to the
type of separating element.

145
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

8.3.2 General requirements

This section provides a general description of penetration sealing systems for building services
installations such as electrical wiring, heating and cooling pipes, water and sewage pipes, ducts and
exhaust systems suitable for use in timber buildings. Because of the diversity of applications and
ongoing test programmes it is only of a general nature, and it is essential to confirm that the system
specified or being installed is approved for use. For additional information contact the technical
services of producers to confirm correct specifications.

Services must be tested in accordance with the test methods set out in appropriate standards. The
failure criteria of penetration sealing systems and linear joints in accordance to EN 13501-2 [16] are
measured in terms of integrity (E) and insulation (I): see Chapter 3. Stability or structural adequacy is
not recorded for these systems. Practical confirmation of the performance of penetration seals is
provided by full-scale fire tests in accordance with EN 1366-3, and for linear joint seals in accordance
with 1366-4.

• An Integrity Failure (E) occurs when cracks, holes or openings appear, through which flames
or hot gases can pass, measured by:
- Ignition on cotton pad

- flame ejection on unexposed surface

- gaps equal or lager then 150 mm x 6 mm national requirements in some countries

- holes larger than ∅ 25 mm

• An Insulation Failure (I) occurs when the temperature rise on the unexposed surface of the
service installation or on the unexposed side of the separating building element, 25 mm from
the penetration or on the seal itself exceed 180°C above the initial average temperature.

According to the ETAG 026 European Technical Approval Guideline (Fire Stopping and Fire Sealing
Products), European Technical Approvals for penetration sealing systems are now possible, which
means that new products may be CE-marked.

In case of fire the performance of penetration sealing systems for building services installations in
separating elements is decisive determined by the type of penetration seal, the technical quality of
workmanship as well as the supporting structure for the service lines in front of and behind the sealing
systems. Movements of service installation caused by mechanical loading, thermals stretches as well
as self load of the service installations must be excluded and considered in planning phase. Mechanical
stress such as entering the penetration sealing system or load application must be eliminated by design
measures.

8.3.3 Penetration sealing compounds

Special sealing compounds are available for sealing gaps and joints around penetrations of
electrical cables and pipes through separating elements such as walls and floors. Nearly all of
them are certified for drywall and solid constructions under defined boundary conditions.
However, as drywall and timber structures fail in almost the same way (attrition and collapse
of cladding / lining) when exposed to fire, many of these products can also used for timber
constructions.

146
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Table 8.2. Suitable sealing compounds for penetrations for service installations
type of media Suitable in

Combination
Single cable

construction

construction

construction
combustible

combustible
Cable trays

Pipes non-

Ventilation
Bundle of

Drywall

Timber
cables

Pipes

ducts

Solid
type of shutter
Sealing compound ++ + + + - - - ++ ++ ++
mastic
foam
silicone,
Plaster / mortar
Mineral wool
description: ++ suitable
+ suitable to only a limited extent
- unsuitable

Based on realised fire tests [17] with penetration sealing compounds in timber constructions
for practical use, the following recommendation can be given:

• filling of joints in the complete thickness of cladding / lining necessary

• distance d between cable penetrations d ≥ largest hole diameter (see Figure 8.1)
• for more than 5 cables in a bundle use of specific systems or installation in non-
combustible stone wool pipe lining insulation (see chapter 0)
• for smoke proof connections, gaps and joints use of permanent elastic sealing compounds
or mastics
• density of penetrated wooden panel ρk≥ 400 kg/m³
a) b)

specific information to sealing compounds and distance

(dsealing) of annular gap/joints are given in Table 8.3

Figure 8.12. Principle of penetration sealing of cables and cable bundles for use with timber
construction (a); distance between multi-cable penetrations (b)

Table 8.3. Recommendation of sealing compounds for cables

Type of electrical cable Sealing systems
Single electrical cable - None for joints ≤ 0,5mm
NYM ∅ ≤ 5x16mm² - gypsum putty, intumescent stripes, intumescent sealing
compound or mastics, stone wool plugs, fire-retardant
polyurethane foam - for annular gap / joints = 5 – 10 mm

Cable bundle - gypsum putty, intumescent stripes, intumescent sealing

(max 3)NYM ≤ 1(5x16mm²) compound or mastics, stone wool plugs, fire-retardant
+ polyurethane - foam for annular gap / joint = 5 – 10 mm
NYM ≤ 2(3x2,5mm²)
Cable bundle - gypsum putty, intumescent stripes, intumescent sealing

147
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

(max. 5) compound or mastics

for annular gap / joints = 5 – 10 mm
NYM ≤ 5 (3x2,5mm²) - for combustible cavity insulation use of non-combustible stone
wool pipe lining insulation around the cables

8.3.4 Penetration sealing systems

Many approved systems for the penetration of cable bundles, trunkings, pipes and ducts
through concrete, masonry and drywall constructions are available on the market. The type of
penetration sealing system used depends on the type, size, material, content and number of
pipes, cables etc., as well as on any further additional installations.

Table 8.4. Suitable penetration sealing systems for building services installations
Type of media Suitable in

Ventilation line
Single cable

combination

construction

construction

construction
combustible

combustible
Cable trays

Pipes non-
Bundle of

Drywall

Timber
cables

Pipes

Solid
Type of sealing system
Mortar ++ ++ ++ + + + ++ ++ +/- +/-
Coated mineral wool slab ++ ++ ++ + + + ++ ++ ++ ++
shutter (intumescent or
ablative coating)
Intumescent formed ++ ++ + - - - - ++ + +
components, blocks, plugs
Cable boxes ++ ++ + - - - - ++ + +
Pipe collars - - - ++ ++ + - ++ + +
Fire pillows ++ ++ + + - - - ++ + +
Sand gaskets ++ ++ +/- ++ ++ - + ++ + +
Mineral wool pipe linings + + - ++ - + +/- ++ + +
Foams/ mastics + + - + - + - ++ +/- +/-
description: ++ suitable
+ suitable to only a limited extent
- unsuitable

These approved penetration sealing systems can also be used for timber structures [19], [20] and [20]
in consideration and compliance of specific basic conditions and design regulations. So for installation
of penetration sealing systems in separating timber structures same installation conditions such as in
concrete or drywall constructions must be exist.
Therefore one concept is the lining of reveal area of the penetrations/openings with a non-combustible
encasing cladding over the entire thickness of the separating element (see Figure 8.12). Depends on
the quality of the fire separating structure the used encasing cladding in the reveal area should have at
least the following classification:
Table 8.5. Type of encasing claddings for reveal area
Fire resistance of separating structure Type of encasing cladding in reveal area
EI 30 / REI 30 K230
EI 60 / REI 60 and EI 90 / REI 90 K260
Continuous joints of the encasing claddings without step joints should be avoided (see Figure 8.12c).
In timber frame constructions a framed timber structure in the opening area is necessary for
stabilisation and fixing of the encasing cladding. The minimum thickness of the timber frame should
exceed 40 mm. For fixing of the encasing cladding, at least two rows of pin-shaped fasteners (staples,
screws, nails), not more than 100 mm apart, should be used. Additional information provided by the
manufacturer of the classified encasing cladding must be observed. The reveal timber frame is not
required for solid timber components.

148
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

a) b)

1 lining / cladding of separating element

2 encasing cladding in reveal area
(see Table 8.5)
c) 3 penetration sealing system (see Table 8.4)
4 framed timber structure minimal width 40 mm
5 framed lining of encasing cladding minimal width
wall / floor

wall / floor

50 mm
a timber frame structure
b solid timber structure

Figure 8.13. Fire save lining of reveal area for timber structure with encasing cladding

For fire separating building elements in

5
timber construction with combustible
lining a spread of fire in the area between
reveal and timber structure must be
avoided. Therefore a joint step have to
be ensure by fixing of at least 50 mm
wide additional lining (5) around the
opening, based by non-combustible
encasing cladding classified at least to
class K230. 4
Figure 8.14. Design of reveal for separating timber
elements with combustible lining

For fire separating elements whose

thickness (hs) is as less as the required 5
thickness of the penetration sealing
system / fire shutter kit (hc), the required
minimal thickness of the separating
element can achieved by using a non-
combustible lining around the opening
on the structure. The minimal width of
the used framed lining (5) should be at
least 50 mm. 4
Figure 8.15. Enlargement of the thickness of building
components

149
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In addition to the above recommendations research studies [21] shows that penetration sealing systems
for drywall constructions without additional reveal linings fulfil the fire protection requirements, when
the joints to the timber structure are sealed in accordance to the technical approval document.
A final certification of these tests must be provided by the producers in normative fire tests in
accordance to EN 1366-3.

8.3.5 Service installations in building elements

Normally, the fire resistance tests of separating building elements are measured without service
installations or installed equipment inside the construction. The installation of valves and accessories,
electrical switches, cables and pipes in fire separating building elements is permitted, provided that the
remaining cross-section of the separating element delivers the required fire resistance. It must
therefore prevent a fire caused by electrical equipment, or the spread of fire by service installations.
For this reason, only non-combustible pipes or wires that directly supply equipment in the rooms, such
as switches and sockets, are permitted.
Basically two types of design and execution for wiring of services installation exist, to affect the
classified fire resistance of the separating building element not negatively.

• installation gap / pre wall installation without fire resistance function:

With regards to fire prevention a definite solution

is probably given by the arrangement of
installation gaps in front of separating walls or
floor elements. Therefore the quality of fire
protection for separating building elements is not
affected negatively by services installations, so
for the inside lining of the installation gap no fire
protection requirements exists. Thus the
installation of switches and sockets is enable on
each position of the lining without additional
sealing measures.
Figure 8.16. Installation gap

• without installation gap / wiring of cable inside the separating element:

The cables that directly supply equipment in the room be installed inside the separating
elements, valves, such as connectors, switches, sockets, junction boxes be mounted into the
cladding/lining of the separating building elements with special encasing systems.

Installation must not reduce the fire resistance of the building elements, and so the penetration
size of the installations and equipment is limited: 200 cm² is recommended as the maximum.
Multiple and opposite installation in separating elements, such as switches and sockets should
be arranged in different compartments, with distance more than 150 mm away from
combustible components, such as studs or beams. Where cables penetrate through the cladding
or lining of the fire separating element, the remaining joint should be sealed as described in
Section 0.

150
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

a) Fire prevention insulation layer:

melting point ≥ 1000°C [22], density ≥ 26 kg/m³,
d > 30 mm, Distance to combustible structural
elements, such as studs or beams ≥ 150 mm. This
distance can be reduced for non load-bearing elements
or the addition of an encasing cladding (1) around
combustible load-bearing structures.

b) Gypsum putty:
encasing of sockets and switches with gypsum putty
for the same protection time as that of the
cladding/lining of fire separating building element, but
d ≥ 30 mm
This arrangement should only be used in combination
with full insulation of the cavity with flexible
insulating mats.
c) Gypsum box:
encasing with Type F plaster boards (EN 520
[23]) or gypsum fibre boards for the same protection
time as that of the cladding/lining of fire separating
building element, but d ≥ 15 mm. In this design type a
full cavity insulation is not required.

As an alternative to these measures, tested and approved switches and sockets based by intumescent
materials are available, which close the installation voids in the separating elements with reactive fire
retardant foam in the case of fire. For practical use, items must be installed in accordance with their
manufacturer's instructions.

These regulations do not apply for niches or slots

with a size larger than 200 cm² that reduce the cross-
sectional area of separating elements, such as floors
or walls. In such cases, special measures are
required, such as additional insulation panels or
cladding to maintain the remaining cross section as
need in order to meet the fire prevention
requirements.
Figure 8.17. Arrangements of fire protection
in remaining cross section

8.3.6 Services installation shafts and ducts

The use of vertical installation shafts or ducts in buildings assists the construction of separate fire
compartments for building services installation. The inclusion of such classified systems in timber
constructions has already been implemented and is very suitable, but is more expensive that sealing
each penetration in the separation building elements. In addition to the shaft structure itself, sealing
systems for entry and exit of service systems are required: see Sections 8.3.3 and 8.3.4.

When using services installation shafts in timber structures, a basic point is that of differential
movements and settlement of the shafts, service installations and the timber structure, particularly for
constructions using platform framing. Flexible spacers or movable connectors must be used when
connecting to walls and floors, as well as for penetrations for pipes, cables, ducts etc.

151
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

8.3.7 Air ventilation ducts

The technical equipment in buildings for ventilation, air conditioning or extraction must meet fire
protection requirements in respect of the spread of fire and smoke. The temperature of air in such
systems does not normally exceed 100 °C. General ventilation systems are made from non-
combustible materials, such as metal, to ensure on permanent airtightness at the typical operating
pressures and adequate hygiene. Two ways of fulfilling the fire prevention requirements for separating
building elements (spread of fire and smoke to other compartments) are available:

• ventilation ducts with classified fire resistances equal to those of the separating building
elements,
• ventilation ducts without classified fire resistance, but with the installation of self-actuating
fire dampers in the area of the separating building element.

Figure 8.18. Design of fire-rated ventilation ducts Figure 8.19. Fire damper

See the requirements and recommendation for design in Sections 8.3.3 and 8.3.4 for installation and
sealing of penetrations of ventilation ducts in timber structures.
The proof of usability for requirements of fire protections based on classified fire tests in accordance
with EN 1366-1 for ventilation ducts and in accordance with 1366-2 for fire dampers.

8.3.8 Penetrations of exhaust systems

Generally for penetrations of exhaust systems in fire separating building elements the same
requirements are needed as for building services installations.
Thus, two design principles can be used to ensure the fire safety,
• continuous service shafts,
• seal of penetration in every separating building element.

If exhaust systems penetrates separating elements in timber construction, sufficient clearance distances
for uninsulated exhaust pipes must be ensured or covering with insulation material must be provided
in dependency of exhaust gas temperature.
Following distances are suggested in subject to exhaust gas temperature [20]:

Table 8.6. Clearance distance in subject to exhaust temperature

Class of exhaust gas temperature [°C] Clearance distance x (Figure 8.20 )
80 – 160 0,1 m
200 – 400 0,2 m
450 - 600 0,4 m

152
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In the case that combustible materials respectively surfaces are lined directly with non
combustible encasing claddings of class K230 or covered by non combustible cladding with a
20 mm ventilation void on the backside, the distance in Table 8.6 can be reduced to the half.
For the penetration of exhaust systems through separating building elements approved
systems have to be used. These systems have to avoid the ignition of the adjacent timber
structure (see Figure 8.21). Minimum thicknesses of insulation material respectively cladding
around the exhaust systems in the area of the penetration are determined by national
standards.

Figure 8.20. Clearance distance of exhaust pipes Figure 8.21. Penetration of exhaust pipes
to combustible materials through timber structures

References:
[1] Stürmer, M.: Ein Beitrag zum Qualitätsmanagement im vorbeugenden baulichen Brandschutz.
Untersuchung von ausgewählten Brandschutzmängeln in der Ausführungsphase, Mensch und
Buch Verlag, Berlin 2006
[2] Wirtschaftsministerium Baden-Württemberg (Hrsg): Holzbausysteme im Wohnungsbau, 2.
aktualisierte Auflage, Stuttgart 1998
[3] Stein, R.; Winter, S.: Mechanismen der Brandweiterleitung bei Gebäuden in
Holzrahmenbauweise, Teilprojekt 11; HTO Abschlussbericht 2008
[4] Frangi, A.: 4.1 Bauteile in Holz – Decken, Wände und Bekleidungen mit Feuerwiderstand,
Lignum-Dokumentation Brandschutz, Lignum (Hrsg.), 2005
[5] Teibinger, M. u.a.: Deckenkonstruktionen für den mehrgeschossigen Holzbau – Schall- und
Brandschutz, Detailkatalog, Holzforschung Austria (Hrsg.), 1. Auflage 2009
[6] Muster-Richtlinie über brandschutztechnische Anforderungen an hochfeuerhemmende Bauteile
in Holzbauweise – M-HFHHolzR, Fassung Juli 2004
[7] EN 1995-1-2 (Eurocode 5), Design of Timber Structures – Part 1-2, General Rules Structural
Fire Design, 2006
[8] EN 1363-1: Fire resistance tests,
part 1: General requirements, 1999
part 2: Alternative and additional procedures, 1999
[9] EN 1365-1: Fire resistance tests for load-bearing elements,
part 1: Walls, 1999
part 2: Floors and roofs, 1999
part 3: Beams, 1999
part 4: Columns, 1999

153
Chapter 8 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

part 5: Balconies and walkways, 2004

part 6: Stairs, 2004
[10] Hosser, H.; Dehne, M.: Theoretische und experimentelle Grundlagenuntersuchungen zum
Brandschutz bei mehrgeschossigen Gebäuden in Holzbauweise, Abschlussbericht zur Stufe 2 -
Bauteilversuche. 2000
[11] Winter, S.; Stein, R., Werther, N.: Konstruktionskatalog für Gebäude in Holzbauweise
unter Berücksichtigung bauordnungsrechtlicher und versicherungstechnischer Aspekte, 2010
[12] Jukka Hietaniemi, Jukka Vaari, Tuula Hakkarainen, Jaakko Huhta, Ulla-Maija Jumppanen, Timo
Korhonen, Ilpo Kouhia, Jaakko Siiskonen and Henry Weckman: Fire safety of cavity spaces -
Characteristics of fires in building voids, their structural prevention and extinguishing (In
Finnish), VTT Tiedotteita - Research Notes 2249. Espoo 2004.
[13] EN 1366: Fire resistance tests for service installations:
part 1; ducts: 1999
part 2; fire dampers: 1999
part 3; penetration seals: 2009
part 4; Linear joint seals: 2006
part 5; Service ducts and shafts: 2003
part 6; Raised access and hollow core floors: 2004
part 7; Conveyor systems and their closures: 2004
part 8; Smoke extraction ducts: 2004
part 10; Smoke control dampers: 2004
[14] Merk, M. ; Winter, S.: Brandsicherheit mehrgeschossiger Holzbau, Teilprojekt 02; HTO
Abschlussbericht 2008
[15] Huber, C. , Hilger, M.; Modulare, vorgefertigte Installationen in mehrgeschossigen
Holzbauwerken Teilprojekt 12; HTO Abschlussbericht 2008
[16] EN 13501-2: Fire classification of construction products and building elements –Part 2:
Classification using data from fire resistance tests, excluding ventilation services, 2007
[17] Werther, N.; report of fire tests of penetration sealings for electrical cables and cable bundles in
connection with timber elements; Technische Universität München - chair of timber structures
and Building construction, 2009-10
[18] MPA Braunschweig: Prüfbericht (3049/9435)-13-TP, Prüfung einer etwa 285 mm dicken,
raumabschließenden, wärmedämmenden Deckenkonstruktion mit einer unterseitigen „K260
Brandschutzbekleidung“ aus 2 x 18 mm dicken Gipskartonfeuerschutzplatten (GKF) zum
baupraktischen Nachweis des „Kapselkriteriums“ in Verbindung mit Durchführungen von
Kabeln und Rohren bei Brandbeanspruchung von der Deckenunterseite; 25.08.2006
[19] MPA Braunschweig: Prüfbericht (3049/9435)-7-TP, Prüfung einer etwa 162 mm dicken,
raumabschließenden Holzständerwand mit beidseitiger „K260 Brandschutzbekleidung“ aus
2 x 18 mm dicken Gipskartonfeuerschutzplatten (GKF) zum baupraktischen Nachweis des
„Kapselkriteriums“ in Verbindung mit Durchführungen von Kabeln und Rohren bei
Brandbeanspruchung von der Deckenunterseite; 17.03.2006
[20] Lignum Dokumentation Brandschutz: 6.1 Haustechnik – Installationen und Abschottungen; 2009
[21] Werther, N.; Report of fire tests on penetration sealings and fire shutter kits for cable bundles
and combustible and non-combustible pipes in connection with timber constructions; Technische
Universität München - Chair of timber structures and Building construction, 2009-10
[22] DIN 4102:17: Brandverhalten von Baustoffen und Bauteilen; Schmelzpunkt von Mineralfaser-
Dämmstoffen; Begriffe, Anforderungen, Prüfung, 1990
[23] EN 520 Gypsum plasterboards; Definitions, requirements and test methods, 2004
[24] EN 15882-3: 2009; Extended application rules for penetration seals
[25] Kaiser, K.O.; Haustechnik, Brandschutztechnische Bauüberwachung; FeuerTrutz 2008

154
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

9 Novel products and their implementation

Note the highlight colours:
yellow: editing or checking needed
green: new text, not reviewed nor language checked

This chapter is aimed primarily at product developers. It describes guidelines for

introducing novel structural materials and products. The basic performance
requirements and potential solutions for adhesives, insulating materials, encasing
claddings and board materials, thin thermal barriers and fire-retardant wood products
are included. The innovation process from idea to approved product ready for the
.
market is outlined.

9.1 Performance requirements and potential solutions

9.1.1 Adhesives
Structural adhesives for use in Europe must comply with performance requirements in
accordance with EN 301 [1] and EN 15425 [2]. As far as performance at elevated
temperatures is concerned, the highest temperature in the tests is 70 °C, to be held for two
weeks under constant loading of the specimens [3]. For the time being, no relevant European
standard exists with respect to the performance of bonded connections in fire. In the U.S., a
new standard – ASTM D 7247 [4] – was published in 2006, prescribing a method of
performing oven tests with pre-heated specimens with lap-shear joints heated to a temperature
of 225 °C. No justification of the relevance of this particular temperature to a real fire scenario
has been presented: the goal seems rather to be to exclude adhesives that soften before the
wood starts to char or, in other words, to exclude adhesives with lower performance than
phenolic resorcinol formaldehyde adhesives (PRF), which are recognised for their structural
performance, long-term performance and insensitivity to elevated temperatures, the latter
guaranteeing excellent structural performance in a fire.

Since both phenolic resorcinol formaldehyde adhesives (PRF) and melamine urea
formaldehyde adhesives (MUF) satisfy the performance criteria for Type 1 adhesives
according to EN 301 [1], it was also assumed that they were equivalent in fire conditions [5].
Since EN 15425 [2] has been published, one-component polyurethane adhesives (PUR) have
also been added to this group of adhesives with high fire performance.

However, several experimental investigations have shown that some PUR adhesive bonds did
not perform as expected in fire [6, 7, 8]. Premature fall-off of the char-layer caused increased
charring rates once the preceding lamination had completely charred. Another investigation
[9] showed considerably lower finger-joint strength for finger joints bonded with PUR and
MUF compared to finger joints bonded with PRF. Other investigations showed that char layers
tend to fall off the timber member once the char line has reached the bond between the
laminations of cross-laminated timber, giving rise to increased charring rate when the
protecting char layer has fallen off.

155
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The effect of this premature failure of bonded joints depends on the structural element
considered. The effect of ablation of the char layer is illustrated in 6.6.2.4.5. It may be large
for unprotected CLT with very thin laminations (below 15 mm). For normal lamination
thicknesses of about 20 mm or more, the effect is small. The effect is greater for CLT in
horizontal position, however the consequences are small since floors are normally oversized
with respect to bending resistance. No premature ablation of the char layer has been observed
regarding glued-laminated beams, probably due to larger thickness of lamellae.

The reduced finger joint strength has very small effect on mechanical resistance in fire of
finger joints in CLT plates due to their system effect. The effect on the resistance in fire of
glued-laminated beams is not known. Since the bending resistance of I-joists is strongly
dependent on finger joint strength in the flanges, the effect of reduced finger joint strength in
fire is a potential risk.

It should be noted that there is little information on adhesive performance: in general, it cannot
be said that all adhesives in one group, e.g. MUF, PUR or others, perform in the same way.
Novel adhesives should be tested in relevant fire tests in order to see if design by calculation is
possible. As an alternative, design by testing is a way to introducing novel adhesives on the
market, although only for specific building products that have been tested.

9.1.2 Insulating materials

9.1.2.1 General
Insulation is an important component in all construction forms. The choice of materials is
generally driven by requirements for thermal performance and air tightness, ultimately aimed
at reducing the energy use of the built environment, particularly in buildings. A FireInTimber
industry questionnaire has been circulated amongst European partners to establish innovative
insulation materials considered as potential alternatives to traditional insulation materials
currently used in the industry. The main suggestions were:
• Glass wool with high maximum service temperature
• Lightweight foam from natural fibres
• Straw bale infill
• Recycled EPS
• Insulation made from recycled tyres
• Sheep wool

Whilst not generally available to, and used in, mainstream construction at present, these
materials are of interest to the timber industry to monitor and observe for the future. The drive
for a more sustainable way of building, awareness of the impact of the choice of materials, and
the importance of appropriate insulation of the building envelope will mean that alternatives
will need to be sought and can act as real differentiator between building solutions.

In addition to the materials listed below, a number of established insulation materials,

although not classed as “innovative” as they are used extensively in the construction industry,
are currently not included in EN 1995-1-2:2004. The most commonly used insulation
materials in this category would be:

• Cellulosic fibre insulation

156
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

• EPS (Expanded polystyrene)

• XPS (Extruded polystyrene)
• PUR (Polyurethane)
• PIR (Polyisocyanurate)

Both categories of materials require supportive work: the former to enable their wider uptake
in the industry, and the latter to facilitate their contribution to the performance of timber-based
solutions in standard construction.

Whilst fire performance is an important factor for insulation, it is not the only one and
certainly not the one single decisive factor for use. In order to provide a meaningful overview
of the decision-influencing performance parameters for industry, it is suggested that the
following additional information about materials should be captured and assessed:

• U-value (overall heat transfer coefficient)

• Density
• Cost
• Environmental profile

EN 1995-1-2:2004 provides a reliable, established basis for performance expectation in order

to benchmark the fire performance and suitability of materials for use in timber-based
applications. The relative fire performance levels of different materials are also of importance,
and help to gauge the suitability of materials and discuss potential application areas.

Information required for considering insulation materials in EN 1995-1-2:2004 is:

• Proof of meeting requirements in EN 13162 (Thermal insulation products for buildings
– Factory-made mineral wool (MW) products – Specification)
• Thickness of the layer (hi), in mm
• Density of the layer ( ρ i ), in kg/m
3

The following products/types of insulation are established in EN 1995-1-2:2004. The

associated design clauses are given in Annex E of EN 1995-1-2:2004.
• Rock fibre batts
• Glass fibre

9.1.2.2 Separating and protection function of insulation

An insulation material must be assessed in order to determine the required performance
information for input into the equations of Annex E of EN 1995-1-2:2004. Varying
experimental procedures are available, and more detailed references can be found in Chapter 5
’Separating timber structures’ for determining the performance levels of selected insulation
materials. Generally, testing in accordance with EN 1364-1 can be used to establish the fire
resistance of a single layer that is not influenced by adjacent materials to establish tins,0. The
test establishes the 140 K or 180 K temperature rise on the unexposed surface of the material.
The extended method, as presented in Chapter 5, also determines a basic protection value
( t prot ,0 ) using test data in accordance with EN 13501-2.

However, especially in the early stages of product development, performance levels of novel
material combinations must be readily and cost-efficiently compared using alternative,

157
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

indicative methods, as described in [10]. This is especially important for assessing innovative
material solutions, where prototype manufacture is generally able to produce only small
quantities of material. At this stage, it is considered important to test several replicates in order
to gather information about the repeatability of material performance obtained, giving useful
information on manufacturing and quality processes. Realistically achievable performance
levels and typical failure behaviours can also be examined in replicate tests.

9.1.2.3 Load-bearing function of timber members protected by insulation

For the time being, no test standards exist to determine the degree of protection of structural
timber members provided by thermal insulation. The following provides some guidance for
the assessment of novel insulation materials in design by calculation.

With respect to the load-bearing function of timber members in timber frame construction,
effective protection of the sides of the timber member facing the cavities against charring and
elevated temperatures is essential: see 6.5.3.1. The capability of the insulation to remain in
tight contact with the timber member over the greater part of the depth of the timber member
is important, and must be verified by fire testing. If, for a novel insulation material, the shape
of the residual (uncharred) cross-section is similar to the cross-section shown in Figure 6.10b,
the design model given in 6.5.3.2 can be used, although it must be checked whether the design
parameters listed in 6.5.3.5 need to be changed.

The charring depth dchar as shown in Figure 6.10b should be determined by fire testing.
Reference tests with rock fibre insulation should be performed. It should be observed that,
when ks was derived (Expression (6.13) or (6.14)), the non-linear relationship between the
charring depth dchar and time was replaced with a linear relationship giving the same charring
rate for a charring depth of 30 mm. Fire tests in bending should be performed in order to
determine the reduction of strength and stiffness properties due to elevated temperature. The
ambient bending strength of timber members tested in fire should be predicted from the results
of reference bending tests performed at ambient temperature.

The design model for assemblies with cavities filled with glass fibre insulation, shown in
Section 6.5.3.4, may be applied to other insulation materials that undergo thermal degradation
above a critical temperature. In such cases, the surface recession rate must be determined, e.g.
by performing fire tests with the insulation in tight contact with the timber member. The
effective surface recession rate with respect to charring of the timber member can be
determined using thermocouples on the timber member surface in contact with the insulation,
recording the time when a temperature of 300 °C is reached at different positions.

The design model given in Section 6.5.3.2 cannot be used when novel insulation materials are
used that perform in such a way that a gap opens between the insulation and the timber
member. When designing by calculation, the protection provided by the insulation should not
be taken into account unless a special investigation is performed.

9.1.3 Encasing claddings and board materials

Performance requirements
Encasing claddings are the outermost parts of building elements, such as walls, floors and
roofs. The main performance requirement for these claddings is to provide protection for the
wood material behind the cladding for a determined time against ignition, charring and other

158
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

damage. This implies that there is no two-sided fire within the same period of time. According
to the test standard EN 14135, the average temperature measured on the unexposed side of the
covering shall not exceed the initial temperature by more than 250 °C and the maximum
temperature measured at any point of these sides shall not exceed the initial temperature by
more than 270 °C. In addition, there shall be no ignition or charring at any point on the
supporting timber structure and no collapse of the covering during the fire exposure.
The purpose of the temperature rise limits and charring restrictions is to avoid the formation of
pyrolysis gases and to prevent the ignition of the structure for a determined time.

During the design and development of new covering materials for encasing claddings, the
following appropriate information should be considered:
− a low rate of shrinkage and heat conductivity, or a high proportion of bound water, to
make single-layer claddings possible
− a low weight, to reduce the additional load on the structure (for comparison: the density of
gypsum boards is 800 – 1200 kg/m³)
− no loss of integrity, melt, or crack formation (or only at very high temperatures)
− no contribution to fire, or only a limited contribution to fire
− based on sustainable materials

After a fire, it would be beneficial if the cladding material were able to absorb and retain water
to reduce its penetration into the building structure (for instance, flooring systems including
gypsum) and minimize the degree of damages by extinguish water.

Different types of board materials which fulfil the classification criteria of encasing claddings
are shown in Table 9.1.

Table 9.1. Examples of possible classification of coverings.

Type Description Classification
EN 13501-2
gypsum boards single layer ≥ 10 mm gypsum fibre board on timber battens K210
45 x 10 mm, butt joints [11]
two layers 12,5 mm gypsum board type F according EN 520, butt K230
joints, displaced layers [12]

single layer 18 mm gypsum fibre board, butt joints [13]

two layers 18 mm gypsum board type F according EN 520, butt K260
joints, displaced layers [14]

two layers 18 mm gypsum fibre board, butt joints, displaced layers

[15]
wood based panels see table 4.7 K210 - K260
and solid wood depending on material, density and thickness

cement bonded single layer ≥ 12,5 mm cement bonded board on timber battens K210
board 50 x 10 mm, butt joints [16]
lightweight concrete single layer 25 mm lightweight concrete board butt joints [17] K230
board
lightweight silicate single layer 25 mm calcium silicate board, butt joints [18] K230
board
combination / single layer 15 mm gypsum board and single layer 15 mm calcium K260
composite silicate board
butt joints, displaced layers [19]
for practical application the documents of proof of usability are to consider

159
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In the test procedure, weak spots are often the joints of the cladding and the areas around
fasteners, which lead to faster temperature rise and charring at the wooden structure compared
to the plane and undisturbed cladding. Therefore, the existing application documents avoid
continuous fasteners and butt joints. New research projects investigate prefabricated glued
cladding and joint formations to minimise this influence.

Failure times of board materials can also be determined in accordance with EN 1995-1-2:2004
(see Chapter 6). In the calculation method, wall and ceiling components were separated into
single layers. For these layers, failure times and thermal heat transfer were calculated on the
basis of the duration of fire exposure in accordance with the standard temperature-time curve.
The fire resistance of wall and ceiling components can be determined as the sum of the failure
times of single layers. The European standard contains the following materials for panelling
wall and ceiling constructions:
- wood panelling
- plywood
- wood-based panels other than plywood, e.g. particleboards and oriented strand boards
- Type A, H and F gypsum plasterboards (EN 520)

Cavities can be uninsulated, or insulated by glass or rock wool insulation. Fire protection
properties of modified panelling products such as cement-bonded particle boards, fire
protection plasterboards, as well as renewable insulation materials such as wood fibre
insulation boards, are not included.

9.1.4 Thin thermal barriers

Performance requirements

The main performance requirements for thermal barriers are the ability to influence ignition
and/or charring, and the sustainability of the protection.

A thermal barrier can prevent or delay ignition and/or charring by its thermal insulation
function, or by acting as a gas barrier. A protective layer with low thermal conductivity can
insulate the underlying material from heat and thus reduce and delay its temperature rise. A
gas barrier layer reduces the evaporation of pyrolysis gases from, and access of oxygen to, the
surface of the protected material, with the result that ignition of the surface is prevented or
delayed.

In order to provide efficient protection, the thermal barrier layer must be sustainable. First of
all, it must withstand temperatures typical for fires, i.e. it must not melt, crack or otherwise
lose its integrity when exposed to high temperatures. It must adhere firmly to the material
beneath, so that it does not detach under heat exposure or in the course of time. In relation to
its intended end use application, the thermal barrier system (i.e. the barrier material and its
attachment) must have a sufficient long-term durability.

Further requirements can be specified for the appearance of the protective system. In some
applications, it is desirable to maintain visibility of the surface features of wood, which limits
the choice of thermal barriers.

160
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Potential solutions

Thin thermal barriers considered as suitable for improving the fire performance of wooden and
wood-based products include intumescent coatings and layers of incombustible materials (e.g.
metal sheets and ceramic layers). Combinations of these two methods can provide further
possibilities for some applications: for example, intumescent coating can be supplemented
with a metal sheet acting as an additional gas barrier.

In principle, ceramic coatings can be used for thermal protection. However, their suitability
for fire protection of wood must be considered case-specifically. Thermally sprayed ceramic
layers are used as thermal insulation of gas turbine components, but the method is relatively
costly and its applicability to wood products is unknown. Another possibility might be
ceramic materials with water of crystallisation, providing a heat sink due to the evaporation of
water and thus slowing down the temperature rise. Ceramic coatings can also be used for
enhancing light and moisture resistance, as shown in a study of the effects of ceramic coatings
on the properties of fire retardant treated wood [20]. The fire performance of the material was
also slightly improved.

Intumescent coatings are commonly used for protecting steel structures. Their application on
wood has been studied [21], and some commercial products intended for wood are available.
At present, however, fire protection of wooden and wood-based products with intumescent
coatings is not widely employed.

Intumescent coatings consist of a binding agent, a carbonising substance, a foam-producing

substance, a dehydrating agent, and an esterification catalyst. A wide variety of compounds
can be used for these purposes [21]. In addition to the basic constituents, other substances
such as expandable graphite flakes [22, 23] can be included.

Intumescent coatings are inert at low temperatures. When exposed to high temperatures, the
coating swells and produces a porous char layer of low thermal conductivity. The char layer
protects the underlying material by delaying its temperature rise and hindering the transport of
oxygen to and pyrolysis gases from the surface. A well-selected intumescent coating can
effectively delay the onset and propagation of charring in wooden or wood-based products
[24].

In the future, nanocomposite coatings might be applied to improve the fire performance of
wooden and wood-based products. In this approach, nanometer-scale particles are added to a
macroscopic material to produce novel properties. Intumescent polymer-clay nanocomposites
have been applied to protect polymers [25] and steel [26], but studies for wood products are
not currently available.

When selecting a thin thermal barrier for protecting wooden or wood-based products, attention
must be paid to the following issues:
− suitability for specific end-use conditions
− suitability for protecting wood
− method of attachment/application, and its reliability in the long term and in fire conditions

161
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

9.1.5 Fire retardant wood products

Fire retardant treatment of wood products may considerably improve their reaction to fire
performance, and the highest fire classifications for combustible products may be reached, see
Chapter 4.2. Treatments may be either by vacuum impregnation, by incorporation during the
manufacturing process of wood products, or by surface treatment of the final product. In all
cases, the amount of fire retardant chemicals in the final product is decisive for its reaction to
fire performance.

The durability of the reaction to fire performance also needs to be addressed, since the
chemicals may migrate in the product during service life and even leach out when used on
exteriors.

The following procedure is recommended when developing novel fire retardant wood
products:
1. Determine the amount of chemicals to be used for reaching, say, Class B-s1, d0 on a small
scale. A suitable procedure is to use three retention levels and test in, for example, the
cone calorimeter, ISO 5660. An optimum retention level can then be predicted from the
results.
2. Check the durability of the reaction to fire performance for interior applications, DRF Int 1
and/or DRF Int 2 (DRF = Durability of Reaction to Fire performance), in accordance with
prEN 15912: see Chapter 4.2. The optimum retention level should then be used.
3. If Items 1 and 2 are successful, continue by full-scale testing using the SBI test, Single
Burning Item, EN 13823.
4. If the products are to be used in exterior applications, check for DRF Ext in accordance
with prEN 15912. Please note that a protective coating is usually needed in order to
maintain the fire performance after weathering.
5. Protective coatings for exterior applications must be chosen carefully in order to be
compatible with the treated product, both initially and during exterior use, and not to
significantly decrease the reaction to fire performance.

Fire retardant treatments do not usually have much influence on the charring rate of wood
products (except those based on intumescent materials, see 9.1.4). Special attention is
strongly recommended if fire retardant wood products are being used in load-bearing
applications, since the strength of the products may be reduced during their service life due to
deterioration of the wood structure. Experience in this field has been gained mainly in North
America, and special standards for controlling this behaviour have been developed, e.g.
ASTM ….

A full-scale test method for testing the fire performance of intumescent coatings is ENV
13381-7 [27].

162
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

9.2 Procedure for introducing novel products

The procedure for introducing novel products to the market is shown in Figure 9.1.

Product idea

Potential to meet basic refine

performance no the idea
requirements?

yes

Other relevant functional

requirements met? no

yes

Product development
• prototypes
• indicative testing / assessment

improve
Classification testing perform-
ance

Required / desired
class achieved? no

yes

Approval

Product
to the market

Figure 9.1. Flow chart of the procedure for introducing novel products to the market.

163
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In the early phase of the product development process, novel product ideas must be critically
assessed in respect of the basic performance requirements related to fire protection properties
(reviewed above). The ability of the product to meet these requirements in general and in the
intended end-use application must be estimated. In addition, other relevant properties of the
product must also be estimated in order to ensure the overall functionality and usability.

The essential requirements for construction products are listed in the Construction Products
Directive (89/106/EEC):
1. Mechanical resistance and stability
2. Safety in case of fire
3. Hygiene, health and the environment
4. Safety in use
5. Protection against noise
6. Energy economy and heat retention

If it is obvious that the proposed product does not meet the basic performance requirements
related to fire protection or other relevant functional requirements, the product idea must be
refined in order to improve the inadequate property.

In the product development phase, it is often reasonable to produce several prototypes with
varying properties for comparison, and in many cases small-scale testing can be utilised. On
the basis of the test results, the potential of the product to pass classification tests can be
assessed, and in some cases the class of the product can be predicted. When the product has
reached the stage of showing a potential to meet the classification criteria of the intended
class, it can be submitted for classification testing. The test laboratory issues a classification
report on the basis of the classification test results, while approval of the product is granted by
a certification body.

Approval of a product can be attested by CE marking, which declares that the product meets
harmonised European standards or has been granted European Technical Approval. CE-
marking via harmonised standards (hEN) is intended to be the normal route for most products.
These standards are published by the European standardisation body, CEN. European
Technical Approvals (ETAs) provide a means of CE-marking of products for which there is no
hEN. An ETA is a technical assessment of a product's fitness for its intended end use. ETAs
are issued by one of the national Member Bodies of EOTA (European Organisation for
Technical Approvals). They can be issued individually for a given manufacturer's product
under rules laid down in a Common Understanding of Assessment Procedures (CUAP)
document. When there is more than one manufacturer, EOTA drafts an ETAG (European
Technical Approvals Guideline), and individual ETAs are issued accordingly.

It should be noted that even though the classification and fire test methods for construction
products are harmonised within the EU, the required classes for different applications can be
defined nationally. It is therefore essential to check the requirements for the intended
applications of the product in the market area.

In the case of wooden and wood-based products, the intended main class in the European
reaction-to-fire classification of construction products must be at least Class D, i.e. Class E
and F products have no practical end uses. In addition to the main class, possible requirements
for smoke production and flaming droplets/particles may apply.

164
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Wooden and wood-based products with improved fire performance can meet the requirements
of Class B or C. Better classes (i.e. A1 and A2) cannot be reached in practice due to the
organic substances in these products. The smoke production class of wood-based products
with fire retardant modifications is typically s1 or s2. However, in the product development
phase, it must be checked that (for example) the fire retardant used does not significantly
increase the smoke production of the product, leading to a poorer classification. The
requirements for flaming droplets/particles can be easily met, provided that the thickness of
the product is sufficient (about 12 mm or more, depending on the product, specimen structure
and installation).

References of Chapter 9

1. EN 301:1997 Adhesives, phenolic and aminoplastic for load-bearing timber structures –

Classification and specifications. European Committee for Standardization, Brussels,
1997.
2. EN 15425:2008 Adhesives - One component polyurethane for load bearing timber
structures - Classification and performance requirements. European Committee for
Standardization, Brussels, 2008.
3. EN 302-2:2004 Adhesives for load-bearing timber structures – Test methods – Part 2:
Determination of resistance to delamination. European Committee for Standardization,
Brussels, 2004.
4. ASTM D 7247 Test Method for Evaluating the Shear Strength of Adhesive Bonds in
Laminated Wood Products at Elevated Temperature. 2006.
5. EN 1995-1-2:2004 Eurocode 5: Design of timber structures – Part 1-2: General –
Structural fire design. European Standard. European Committee for Standardization,
Brussels, 2004.
6. Frangi, A. & Fontana, M. A Design Model for the Fire Resistance of Timber-Concrete
Composite Slabs. International Conference on Innovative Wooden Structures and
Bridges, Lahti, Finland, August 29–31, 2001. Conference Report edited by IABSE, ETH
Zurich.
7. Frangi, A., Fontana, M. & Mischler, A. Shear behaviour of bond lines in glued laminated
timber beams at high temperatures. Wood Science and Technology, Vol. 38, No. 2, May
2004, pp. 119–126. Springer Verlag, Berlin.
8. Frangi, A. & Fontana, M. Effect of temperature on the structural behaviour of bond lines
in glulam. 1st International Symposium on Advanced Timber and Timber-Composite
Elements for Buildings, Conference Report. COST Action E-29. Florence, Italy, October
27–29, 2004.
9. König, J., Norén, J. & Sterley, M. Effect of adhesives on finger joint performance in fire.
CIB W18, Meeting 41, St. Andrews, Canada. Lehrstuhl für Ingenieurholzbau, University
of Karlsruhe, Karlsruhe, Germany, 2008.
10. Bregulla, J. (2003). Investigation into the fire and racking behaviour of structural
sandwich panel walls – a methodology to assess structural sandwich walls in fire. Civil
Engineering. Guildford, University of Surrey: 610.
11. Danish Institute of Fire and Security Technology: test report PG 11522 fire test of single
layer 10 mm FERMACELL gypsum fibre board for encasing cladding in accordance with
EN 14135; Xella Trockenbau System GmbH.

165
Chapter 9 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

12. MPA Braunschweig: test report 3049/9435-12, fire tests of 2 x 12,5 mm fire protection
plaster board for encasing cladding in accordance with EN 14135; Deutsche Gesellschaft
für Holzforschung.
13. MFPA Leipzig: classification report KB III / B-07-059, K2 30 encasing cladding of
FERMACELL Gipsfaserplatten; Xella Trockenbau System GmbH.
14. MPA Braunschweig: test report 3049/9435-16, fire tests of 2 x 18 mm fire protection
plaser board for encasing cladding in accordance with EN 14135; Deutsche Gesellschaft
für Holzforschung.
15. MFPA Leipzig: classification report KB III / B-07-060, K2 60 encasing cladding of
FERMACELL Gipsfaserplatten; Xella Trockenbau System GmbH.
16. MFPA Leipzig: test report PB III/B-06-193; fire tests of 12,5 and 16 mm Fermacell
Powerpanel H2O boards for encasing cladding in accordance with EN 14135; Xella
Trockenbau System GmbH.
17. MFPA Leipzig: pre-fire tests of 25 mm lightweight concrete board under consideration of
14135; TU München.
18. MFPA Leipzig: test report PB III/B-08-282; fire tests on 25 mm ACICO FES boards for
encasing cladding in accordance with EN 14135; Aerated Concrete Industries
CO.K.S.C.C.
19. MFPA Leipzig: classification report KB III / B-07-054, K2 60 encasing cladding of
15 mm PROMAXON and 15 mm PROMATECT boards; PROMAT GmbH.
20. Harada, T., Nakashima, Y. & Anazawa, Y. The effect of ceramic coating of fire-retardant
wood on combustibility and weatherability. Journal of Wood Science, 2007. Vol. 53, pp.
249–254.
21. Wladyka-Przybulak, M. & Kozlowski, R. The thermal characteristics of different
intumescent coatings. Fire and Materials, 1999. Vol. 23, pp. 33–43.
22. Qureshi, S. P. & Krassowski, D. W. Intumescent resin system for improving fire
resistance of composites. In: Proceedings of the 29th International SAMPE Technical
Conference. 28 October – 1 November 1997. The Society for the Advancement of
Material and Process Engineering, 1997. Pp. 625–634.
23. Lyon, R. E., Speitel, L., Walters, R. N. & Crowley, S. Fire-resistant elastomers. Fire and
Materials, 2003. Vol. 27, pp. 195–208.
24. Hakkarainen, T. & Mikkola, E. Thin thermal barriers for wood based products to improve
fire resistance. VTT Research Report VTT-R-07061-09.
25. Bourbigot, S. & Duquesne, S. Intumescence and nanocomposites: A novel route for
flame-retarding polymeric materials. In: Morgan, A. B. & Wilkie, C. A. (eds.). Flame
retardant polymer nanocomposites. Hoboken, New Jersey: John Wiley & Sons, 2007. Pp.
131–162.
26. Wang, Z. Y., Han, E. H. & Ke, W. Fire-resistant effect of nanoclay on intumescent
nanocomposite coatings. Journal of Applied Polymer Science, 2007. Vol. 103, pp. 1681–
1689.
27. ENV 13381-7:2002. Test methods for determining the contribution to the fire resistance
of structural members - Part 7: Applied protection to timber members.

166
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

10 Active fire protection

Active fire protection is used to achieve a more flexible fire safety design of buildings and an
acceptable level of fire safety in large and/or complex buildings.
This chapter introduces common active fire protection systems, including fire detection and
alarm systems, fire suppression and smoke control systems.
Sprinkler installation provides special benefits for an increased use of wood in buildings
especially on visible surfaces.

The main reason for using active fire protection is to gain time for safe evacuation. The use
of active fire protection measures is also the only way of controlling or suppressing a fire in
order to minimise fire damage to contents and buildings. It is often advantageous to include
expert fire safety input in the design of a building at a very early stage in order to ensure that
the building will be acceptably safe in the event of fire, and also be cost-effective to design,
build, operate and maintain.

In contrast to passive fire protection, active fire protection systems become operational only
when a fire occurs. They will include a means of automatic fire detection and devices that
control the growth, suppression or extinguishment of the fire. They may also include
provisions to assist the work of the fire and rescue services to rescue occupants and minimise
property damage and/or fight the fire. Fires in a particular building are rare events, and both
active and passive fire protection measures must be serviced and maintained in order to
ensure they will function in the event of a fire. Large fires are usually due to inbuilt fire
precautions being disabled or compromised – for example, doors propped open or a delivery
of equipment in combustible packaging temporarily being stored in an inappropriate location
such as an atrium or stairwell.

Active fire protection measures include:

• Provision of alarm systems – audible (voice alarm or siren/bell), visual (strobe etc),
tactile (vibrating pagers etc)
• Provision of automatic fire detection – smoke, heat, flame, combustion gas etc. to
trigger alarm or other active fire protection systems
• Indication of the location of fire on visual display, mimic diagram etc
• Provision of smoke/heat venting & exhaust control systems. Note that these systems
extend escape time or aid firefighting: they are not designed to control or suppress
the fire
• Closing of smoke dampers in ducts, closing of doors to provide fire compartments,
refuges etc – note it is not unusual for a fire to self-extinguish in a well sealed fire
compartment due to lack of oxygen (needs above approximately 15 % for flaming
combustion).
• Release of access control systems on doors, gates, turnstiles etc to facilitate people
movement and escape from the building
• Provision of automatic fire suppression systems – water (sprinklers, water mist,
drenchers, foam), chemical agent (gas – halon replacement), inerting gas (nitrogen,
CO2 etc to reduce oxygen level to below 15 %).

The distinction between active and passive fire protection measures is not always clear-cut, as
some measures are a combination of both, such as smoke control doors which remain open
under normal circumstances but which are automatically closed on receiving a signal from a
detecting device, or protective coatings which intumesce and become effective only when a

167
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

fire occurs. The doors and the coatings in the two cases remain part of the passive defence
system, and it is only the detection and operating mechanism that is an active measure.

10.1 Fire detection and fire alarm systems

All active fire protection systems remain ‘inactive’ until they are triggered. The most
common active fire protection system installed in buildings is the fire alarm. In general, all
occupied buildings need to be fitted with some means of raising an alarm in the event of fire
in order to alert the occupants and enable them to make a timely escape.

Simple fire alarm systems comprise a number of interconnected electrically operated bells or
sounders which are audible throughout the premises. If there are high ambient noise levels or
occupants with hearing impairment, flashing strobes may also be linked into the fire alarm
system. The fire alarm system in most buildings is triggered by manual operation of a break
glass alarm or manual call point. Manual call points are typically located adjacent to exit
doors.

The zoning of the alarm will be configured to comply with the evacuation strategy for the
building – for example, total evacuation, phased evacuation or progressive horizontal
evacuation. In some buildings, the initial alarm may be delivered only at a manned security
desk in order to permit the incident to be investigated before evacuating a large number of
people from a place of entertainment, for example.

Figure 10.1 Fire alarms

Automatic fire detection linked to the fire alarm system will normally be required under fire
safety law only when it is needed to safeguard the lives of people who may not be aware of a
developing fire. Examples include:
• Premises where people sleep, such as hotels and hospitals
• Alerting people who are isolated or remote, and who could become trapped by a fire
because they are unaware of its development, such as lone workers
• Premises where there are unoccupied or rarely visited rooms, and where a fire would
compromise the escape route
• as a compensating feature, e.g. for inadequate structural fire protection, in dead-ends
or where there are extended travel distances
• where other active systems such as smoke control, ventilation or automatic fire
suppression systems are controlled by the automatic fire-detection system and form
part of the fire life safety system.

Automatic fire detectors are capable of responding to a range of combustion products

including:
• smoke,
• heat (fixed temperature or rate of rise of temperature),

168
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

• flame (infra-red, ultra-violet)

• combustion gas (carbon monoxide)

Smoke detectors are the most sensitive and most common form of automatic fire detection.
There are several types of smoke detector, which may respond differently to different types of
smoke. Point detectors are the most common type. Optical-type point smoke detectors
respond most quickly to smoke which comprises large particles, such as from smouldering
fires. Ionisation chamber smoke detectors are particularly sensitive to smoke from fast
flaming fires. Other examples of smoke detector may sense the obscuration of an infrared
beam up to 100 m in length or use small-diameter plastic tubing to draw air to a remote
sensitive detection chamber. Smoke detectors are essentially particle detectors, and may
respond to aerosols, dusts, exhaust fumes etc., which means that false alarms may be a
problem with smoke detectors. Work is still progressing with the development of multi-
criteria or multi-sensor fire detectors in order to provide more reliable automatic detection of
fires.

Heat detectors tend to be fitted in locations where smoke detectors would be unsuitable, such
as in kitchens, boiler rooms or laundries. Point heat detectors have a similar appearance to
smoke detectors, and may monitor for temperature thresholds or rate of rise of temperature.
Linear heat detectors are available which have a similar appearance to electric cables, and
may be non-resettable, where thermoplastic insulation between conductors melts, or
resettable, for example using fibre optic cable.

Flame detectors respond very quickly providing that they have direct line of sight visibility of
the flame, and tend to be used in more specialist applications including large high spaces or
even supervising outdoor storage areas. They may monitor for UV or IR radiation or a
combination of the two. False alarms may be created by arc welding or other sources of IR,
including solar gain.

Combustion gas detectors (carbon monoxide in particular) tend to be used in combination

with smoke or heat detectors to provide reliable fire detection with fewer false alarms, or fire
detection in confined spaces where the fire is likely to be oxygen-starved and so produce CO.

Figure 10.2 Fire detectors

Fire detectors and alarm sounders are usually linked to and controlled by control and
indicating equipment – often referred to as the fire alarm panel. This, coupled with a high
integrity backup or emergency power supply, makes up the basic automatic fire alarm system.

It is important that the fire detection and alarm system is correctly specified, designed,
installed and maintained. Maintenance typically includes weekly testing of the alarm
sounders on the system and regular testing of the automatic detectors. To this end, the fire
detection and fire alarm systems area is supported by a large number of European and
international standards, EN 54 and ISO 7240.

169
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

10.2 Fixed fire suppression systems

Fixed fire suppression systems control, suppress or extinguish a fire by cooling and/or wetting
unburnt material, lowering the oxygen level or chemical reaction, and thus inhibit the
combustion process. Types of fixed fire suppression systems include:

• automatic water based systems – sprinklers, water mist, foam, deluge

• gaseous fire suppression systems – chemical agents and inert gases
• powder

Typically, fixed fire suppression systems will be designed and installed to control or suppress
fire growth, as opposed to completely extinguishing the fire – it is assumed that manual
intervention will complete the suppression. This section covers the most commonly
encountered types of fixed fire suppression systems.

10.2.1 Sprinkler systems

Sprinkler systems were developed by the insurance industry for property and asset protection,
and have been in existence for more than a century. It is relatively recently that their
applicability to the protection of life has begun to be recognised, and standards developed in
support of sprinklers for life safety. Historically, sprinkler installations have been fitted to
cover an entire building or complex. However, their applicability to deal with localised high-
risk areas is now accepted.

The installation of a sprinkler system may result in a reduction in insurance premiums, the
amount of portable fire-fighting equipment necessary, and more flexibility in the design and
use of the building. A sprinkler system is usually part of a package of fire precautions in a
building.

A sprinkler system consists essentially of a reliable water supply feeding an array of

individual sprinkler heads mounted at the correct spacing on an appropriately sized network
of hydraulic pipes. Water may be supplied from one or more tanks by gravity or pumps, or
taken directly from the water main, if this can provide sufficient pressure and flow at all
material times.

Each sprinkler head is an individual heat detector and, the heads do not operate in unison. A
typical operating temperature for a sprinkler head is 68 oC. However, there is thermal inertia
in the mass of the sprinkler head and associated thermal fuse, so there is an inherent delay
once fire gases have reached this temperature in the proximity of the head before the head
will activate and discharge water. This delay is characterised by the Response Time Index
(RTI).

There are several variations on the basic sprinkler design, intended to deal with unusual
ambient conditions or particular risks. Many codes and standards are available to cover all
aspects of the design, specification, installation and maintenance of sprinkler systems.

Sprinkler systems have a long and successful history. Insurers claim that sprinklers have a
99 % success rate in controlling or suppressing fires when they have been correctly specified,
installed and maintained. The design and calculations take into account the size and
construction of the building, the category of goods stored in it and the characterisation of
occupancy. Water damage from sprinkler systems is minimal. Accidental discharge of water
from sprinkler heads is an extremely rare event, as is water leaking from sprinkler pipework.

170
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

If a sprinkler system forms an integral part of the fire strategy for a building, it is necessary to
ensure that management procedures are in place to cater for those periods when the sprinkler
system is not functional. Such procedures may include the following:
• Limit any planned shutdown to lower-risk periods, when building occupant numbers
are low (e.g. at night, at weekends or during public holidays)
• Isolate the area without the benefit of working sprinklers from the rest of the premises
by fire-resisting material
• Avoid high-risk processes such as ‘hot-work’
• Training and deployment of additional staff as fire patrols
• Suspension of phased evacuation. Evacuation of occupants should be immediate and
total. (Exercise caution, as stairway widths may have been designed for phased
evacuation only)
• Inform the local fire and rescue service

Automatic sprinklers are the most commonly installed systems and are often used to reduce
risks that arise from the building’s use – as in manufacturing – to protect vulnerable
individuals or high-value contents, or to compensate for the materials used in the building’s
construction. They are also often used in modern buildings – such as airports or storage
facilities – to allow extensive open spaces to be created without structural compartment walls.
They may also be used to protect premises that are geographically isolated. It is important
that the reason for installation is understood and that, in those cases where the system has
been provided as part of a fire engineering design for the building, the responsible person is
made aware of and understands the interaction of the system and other building design and
service features.

Figure 10.3 Typical sprinkler heads Figure 10.4 Sprinkler in action

10.2.2 Water mist systems

Water mist systems were originally developed to suppress fires in ship’s engine rooms. They
discharge water as much finer droplets than do conventional sprinkler systems, and typically
have a relatively small supply of water. Unlike conventional sprinkler systems, water mist
systems tend not to use interchangeable components and therefore the design and installation
of the complete system must carefully controlled by a single supplier.

Water mist systems from different manufactures may operate at different pressures, they may
use pumps or pressurized cylinders to provide the driving force to discharge water through
heads which have activated, and they may produce different droplet sizes. Therefore
performance-based burning commodity tests are used to evaluate the main water mist system
design parameters, such as nozzle type, nozzle spacing, nozzle installation height, minimum
working pressure etc. In addition, individual components are subjected to component tests,

171
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

many of which have been adopted from similar tests for conventional sprinkler system
components.

Land-based water mist systems are proving to be particularly cost-effective in retro-fit

applications such as residential units, or to protect specific risks such as computer cabinets.
They discharge water through specially developed nozzles to produce fine droplets – much
smaller than traditional sprinkler systems. Triggering is typically similar to that used with
conventional sprinkler systems, and typically uses an individual glass bulb or fusible link to
activate individual heads. The droplets are sufficiently small to be entrained by convective
currents produced by the fire and so can be drawn into and quickly cool the combustion zone
or seat of the fire. They work best when they are used to control a growing fire in a relatively
small compartment or room (as opposed to say a hall or place of assembly or warehouse),
where the relatively fine water mist provides fire suppression by a combination of wetting,
cooling and oxygen displacement.

Currently, the NFPA 750 standard is available for design and installations, and further
standardisation work is ongoing. Users of water mist fire suppression systems are very
dependent on literature, information and data produced by the manufacturers. Consequently
the design and installation of a successful water mist system must be based on a risk
assessment, taking into account the probable type and location of fire, the fuel and the
immediate environment.

10.2.3 Gaseous fire suppression systems

Gaseous fire suppression systems operate by one or more of the following processes:
• Interruption, at a molecular level, of the chemical combustion process and so
extinguishing the flame
• reducing the oxygen concentration level below one which will support flaming
combustion (i.e. <15 % oxygen) by totally flooding a space with an inert gas
• providing cooling

A properly designed and installed fixed gaseous suppression system will leave little residue.
Such systems are typically used where the risk being protected has a high intrinsic value or is
sensitive to the application of water. They are therefore designed for specific ‘risks’ and used
in areas such as electronic data processing rooms, telephone installations and other electrical
facilities. Gaseous fire suppression systems are not normally used to provide a safe
environment and to permit escape in the event of a fire.

There are two main types of fixed gaseous system:

• chemical gases act on fires by absorbing heat, lowering the flame temperature
and chemically interfering with the combustion process to such an extent that the
process cannot continue.
• inert gases deplete the oxygen concentration to below 15 %, at which point
flaming combustion is inhibited.

10.3 Smoke control

Smoke control is the name given to a range of physical and mechanical techniques used to
dilute or remove the influence of smoke in selected locations in a building in the event of a
fire.

Active or passive smoke control, for example using fans or opening vents, may be provided
for the safety of the occupants and to assist firefighting and post-fire smoke clearance. These
systems are designed to restrict the spread of fire and smoke, usually by naturally or

172
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

mechanically venting the heat and smoke through the roof or via other routes to the outside.
Inlets at low level to provide make-up air are essential for the operation of such smoke control
systems and no openings provided for this purpose should be obstructed during the life of the
building

Smoke control may also be combined with other active fire protection measures, such as
sprinklers or automatic door closers or fire shutter release mechanisms, and it is relatively
common to provide ‘indefinite’ escape time in shopping malls by using a combination of fire
compartmentation and/or fire suppression with smoke control techniques.

Special down-stands may be installed to create a reservoir which will contain the buoyant
smoke and hot gases at ceiling/roof level, while vents allow the smoke to escape.
Alternatively, concealed smoke curtains may be fitted which drop down upon receiving a
signal from the fire detection and alarm system to form smoke reservoirs. It is important that
any smoke can flow easily into the reservoirs and escape through the vents, and that nothing
which could cause an obstruction, such as large displays, is fixed near the vents or reservoirs.
Smoke control techniques may also be used to protect structure from heat where necessary
(e.g. glass facades). Typical large-scale applications of smoke control are warehouses,
exhibition halls, terminals, shopping centres and atria.

Smaller-scale applications include common access areas of residential buildings, corridors,

lobbies, stairwells, and fire-fighting shafts.

Figure 10.5. Example of smoke exhaust from an open sided compartment

10.3.1 Smoke control techniques

Pressurisation: Air is blown into an escape route (usually a stairway) to hold back smoke
trying to pass through leakage paths in what would otherwise be a smoke containment design.

Depressurisation: Smoky gases are removed from the smoke-affected space in a way that
maintains the desired pressure differences and/or air speeds across leakage openings between
that space and adjacent spaces.

173
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

10.4 Fire safety design with sprinklers

In addition to saving lives, sprinklers may allow for an alternative design of buildings.
Requirements on passive fire protection to provide means of safe egress may be at least partly
reduced. This will facilitate a more flexible use of alternative building products. Wooden
facades may, for example, be used in sprinkled buildings, which is logical since the risk for
flames out of a window from a fully developed fire is eliminated.

Figure 10.6. Principle for fire safety design by sprinklers:

Increased fire safety by installation of sprinklers may lead to relaxations in the passive fire
means, and still fulfil the same or higher safety level.

The implementation of alternative fire safety design with sprinklers may vary between
countries. Some examples are given in table 10.1.

Table 10.1. Examples of reduced passive fire protection in sprinkled buildings in some
countries
Country Reductions in requirements for passive fire protection in sprinkled buildings
(Requirements for non-sprinkled buildings in parentheses)
Number of Wooden Wooden linings, Fire resistance Miscellaneous;
storeys in facades, number of storeys requirements, comments
timber number of Flats Escape minutes
frame storeys routes
Estonia 4 (2) 4 (3) 60 (60) Longer distance in escape
routes
Finland 4 (2) * * * *
Germany 5 (5) 8 (3) ** 3 (3) REI 60 *** Draft proposal for timber
construction guideline
Sweden 8 (2) 8 (2) Longer distance in escape
routes
Switzerland
UK
Eurocode
1-2

Australia With fire shield 60 (>60) To be checked

Canada 4 (3) 4 (3) 45 (60) -“-
New 3 (1) 15 (30) -“-
Zealand 30 (45)
USA 5 (4); 4 (3) -“-

174
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Comments to table 10.6:

In Estonia two of following requirements could be different when using sprinklers:
1. Area of each storey of the building
2. Area of fire compartment
3. Length of escape route
4. Fire resistance of load-bearing structures
5. Fire resistance of separating structures between fire compartments
6. Reaction to fire requirements of structures
7. Fire spread limitation requirements to neighbouring buildings

In Finland, derogations may be allowed concerning:

– gross floor area of the building and the area of its fire compartment
– length of passageways to exit
– classification of fire load
– constructions (slower rise in temperature and cooling of load-bearing constructions)
– reaction to fire requirements of surfaces (internal & external)
– distance between buildings

Germany
** Combustible surfaces limited to 50%
*** ≤ 400 m²

A Nordic design guideline specifies methods for verification of design options (trade-offs)
when sprinklers are installed. Both residential and conventional sprinklers are included. (ref)
Ongoing work, text to be updated.

175
Chapter 10 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Fire safety design with sprinklers – a case study

The effect of residential sprinklers on a fire in an ordinary living room has been demonstrated
in a Swedish project, see picture below. It is evident that the fire damage is very limited if
sprinklers are installed. No water damage has been experienced, since the water supply from
sprinklers is far less than from a fire brigade arriving at a much later stage.

Figure 10.7. After a fire in a living room without sprinkler (left) and with sprinkler (right).

Examples of the implementation of fire safety design with sprinklers are two modern timber
buildings, one in Malmö and one in Sundsvall, both with exterior wood claddings (Figures
10.9 and 10.10). Another example is a three-storey building for guest researchers at
Stockholm University in solid timber structure, called Science City (Vetenskapsstaden), in
which interior wood linings have been used to an extent greater than is allowed without
sprinklers.

Figure 10.8. Four-storey timber building in Figure 10.9. Six-storey timber building in Sundsvall,
Malmö, Sweden (Trähus 2001 at the Bo01 Sweden (Norra hamnen – North harbour).
building fair).
Exterior wood claddings have been permitted, since both buildings are sprinkled.

176
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

11 Performance-based design
Note the highlight colour:
green: new text, not language checked

This chapter describes the basic principles for performance-based design, requirements
and verification. Fire risk assessment principles are described in terms of objectives,
fire safety engineering design, design fires, calculation/simulation methods and
statistics. A case study on probabilistic approach is also included.

A performance-based approach to fire safety design relies on the use of fire engineering
principles, calculations and/or appropriate modelling tools to satisfy building regulations.
Instead of prescribing exactly which protective measures are required, it is the required
performance of the overall system that is presented against a specified set of design objectives.
Fire and evacuation analysis is used together with experimental evidence to assess the
effectiveness of the protective measures proposed in the fire safety design of a building.

11.1 Performance-based requirements

The main principle in applying performance-based requirements is that the building should be
designed and executed on the basis of design fire scenarios, which must cover the conditions
which are likely to occur in the building. The following objectives must be shown to be
fulfilled:
• Load-bearing capacity of constructions must remain in fire for a minimum specific period
of time
• The generation and spread of fire and smoke must be limited
• Spread of fire to neighbouring areas or buildings must be limited
• Occupants must be able to leave the building or be rescued
• The safety of rescue teams must be taken into consideration

This leads to the need for defining criteria to satisfy life safety objectives (safety of occupants
and rescue teams) and criteria for loss prevention.

National building regulations may define performance criteria to be applied in structural fire
safety engineering (FSE) design, such as in the following way:
• A building of more than two storeys must not collapse during the fire or cooling phase
or
• A building of not more than two storeys must not collapse during the period of time
required for securing evacuation, rescue operations and controlling the fire.

There are basically two ways to attest that a design solution fulfils the performance criteria:
1) based on absolute values (see 11.1.1):
• determination of the risk of non-performance of the solution
• comparison of this risk with a limiting risk value agreed to represent a tolerable risk
2) based on relative values (see 11.1.2):
• assessment of the risks related to the FSE design solution
• assessment of the risks that would result from application of the prescriptive rules of
fire regulations (deemed-to-satisfy solution)

177
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

• comparison of these risks.

11.1.1 Verification of safety through comparison with a tolerable risk

level
The risks involved in a potential structural failure differ widely, depending on the damage that
might be caused. For example, the collapse of a basically unmanned single-storey warehouse
is completely different from the collapse of a multi-storey building having a large number of
occupants. Fire safety design aims at providing solutions with risk levels that our society can
tolerate. As the risk is composed of the probability pf of the failure - or more generally, the
unwanted event - and the associated consequence C, high-consequence events must have
considerably lower probability of occurrence than low-consequence events. However, it is
important to realise and acknowledge the fact that there is always some - though very small -
probability that the unwanted and unexpected event will happen. Consequently, a tolerable
risk level greater than zero exists.

Eurocodes are based on reliability theory and Eurocode 1990 [1] actually gives quantitative,
numerically expressed levels of when the probability of structural failure for a given level of
consequence can be deemed to be tolerably low. It distinguishes between three consequence
classes (CC1, CC2 and CC3), and associates three reliability classes (RC1, RC2 and RC3) to
these consequence classes. The reliability in each reliability class is quantified by the
reliability index β, which for normally distributed variables is related to the failure probability
pf as

β = φ −1 (1 − p f ) (11.1)
p f = 1 − φ (β )

where φ is the cumulative distribution function of the standardised Normal distribution.

It should be noted that these values depend on the risk-relevancy of the building, not on the
potential cause - excessive snow or wind load, earthquake, fire or some other cause - of the
unwanted event. Hence they give a plausible starting point for assessment of the tolerable risk
in case of fire.

11.1.2 Verification of safety through comparison with a deemed-

to-satisfy solution
Verification of safety through comparison with a deemed-to-satisfy (DTS) solution is a
straightforward method that has the great benefit that the designer is not required to justify the
safety of the solution. The drawback of this approach is that one has to carry out two analyses,
one for the Fire Safety Engineering design and one for the design that complies with the
classes and numerical requirements given in the fire regulations.
In brief, using fire safety assessment of structures as an example, the process is as follows:
1) Establish the design fires using the fire-load entities (see 12.2.3).
2) Carry out the Fire Safety Engineering (FSE) analysis of the design that you plan to
have accepted, including the assessment of structural adequacy.
3) Assess the likelihood of structural failure in the FSE design.

178
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

4) Carry out the Fire Safety Engineering (FSE) analysis, including assessment of
structural adequacy for a similar building (or part of it) with design solutions taken
from the fire regulations (the DTS design solution).
5) Assess the likelihood of structural failure in the DTS design, using the same failure
criterion as used for the FSE solution.
6) Compare the results and, if the FSE solution gives at least as low a failure likelihood as
the DTS solution, the FSE solution is acceptable.

11.2 Fire risk assessment

The concept of fire risk can be defined as the probability of a fire causing a loss of life (or
injury) and/or damage to property. Fire-risk assessment is the process of estimating the fire
safety level of the building.

11.2.1 The objective of risk analysis

The use of fire risk analysis aims at avoiding fatalities, ensuring the safety of persons and
reducing financial losses caused by fires. Fire risk analysis methods provide tools to assess
the different alternatives for fire safety measures in buildings, e.g. fire precaution measures
which aim to prevent ignition, to detect and extinguish fires effectively and to minimise losses
occurring. Using fire risk analysis, the fire risk and the costs of the different fire safety
measures can be compared in quantitative terms and, in addition, a solution achieving the
optimal fire safety level can be found.

11.2.2 Fire safety engineering design

In performance-based design, the assessed level of fire risk defines the acceptability of the
design. It is therefore essential that all features affecting the total fire risk are included in the
analysis. All factors must be quantified, such as ignition and fire development, performance
of structural elements (including detailed solutions), performance of the building occupants,
level and reliability of the fire safety systems (incorporating both the active and passive
measures), intervention of the fire department, and damage caused by fire.

The Fire Safety Engineering (FSE) design process is shown schematically in Figure 12.1, and
the basic factors of FSE analysis in Figure 12.2. In assessing whether the results are safe (the
analysis part of the diagram) the following abbreviations have been used:
• ASET = available safe egress time
• RSET = required safe egress time
• F-N curves are frequency-number plots, showing the cumulative frequencies (F) of events
causing N or more losses (fatalities, injuries, etc.).

179
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

DESIGN PROCESS: ISO TC 92/SC 4/WG 10 N55Rev4

AGREEMENT BUILDING (OR PART OF IT)
ON THE
BASIS OF FIRE SCENARIOS
THE WORK - what burns, where & when - how many people are
- operation/operation failures threatened
of safety systems - what is the percentage of children,
elderly people, disabled

DESIGN FIRES ACCEPTANCE CRITERIA

- fire growth rate - maintaining tolerabre conditions
- max. fire size (HRR, MW) for safe evacuation (heat & smoke)
- production of smoke & - ensuring operable conditions for
hazardous gases the fire department
- duration - other, if needed

TREATMENT OF UNCERTAINTIES: SENSITIVITY ANALYSIS

FIRE SAFETY
ENGINEERING ANALYSIS OF A FIRE SCENARIO
ANALYSIS - heat, smoke (fire simulation)
incl. influence of safety measures
- evacuation (evac. calculations)

IS THE RESULT SAFE?

ANOTHER INCLUDING SENSITIVITY ANALYSIS RE-
SCENARIO ANALYSE
- heat, smoke (fire simulation)
- evacuation (evac. calculations)

deterministic probabilistic risk-based

- for example: - for example: - for example:
ASET > RSET + safety margin P[ASET < RSET] < ptarget acceptable F-N -curve

YES NO CHANGE
THE DESIGN

ALL SCENARIOS AND ALTERNATIVES ANALYSED

CHECKING

SUMMARY, INTERPRETATION, FINAL REPORTING

- including statement of service and maintenance
needs as well as statement when potential changes
in the building or part of it necessitate partial or
complete re-design

Figure 11.1. The FSE design process scheme

180
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

SCENARIO

VENTILATION:
DESIGN FIRE
how much O2 fire gets

FIRE SIMULATION

PASSIVE FIRE SAFETY ACTIVE FIRE SAFETY MEASURES

CHANGES IN
CHANGES IN
MEASURES - first-aid extinguishing
CONDITIONS:
SENSITIVITY ANALYSIS

CONDITIONS
- load-bearing (R) - fire detection
heat
- heay
- compartments (EI) -- smoke - smoke control
smoke
- reaction-to-fire -- toxicity
toxicity - suppression systems
- safety distances - fire department

ASSESSMENTOF
ASSESSMENT
CONSEQUENCES:
CONSEQUENCES
- evacuation
- hazards from heat and
smoke (incl. fire dept.)
- heat transfer and
structural response

• The load-bearing capacity of the construction can be assumed for a specific period of time;
• The generation and spread of fire and smoke within the works are limited;
• The spread of fire to neighbouring construction works is limited;
• Occupants are able to leave the building in case of fire or can rescue by other means;
• The safety of rescue teams is taken into consideration.

Figure 11.2. Fire safety engineering analysis scheme.

As shown in the diagrams above, it is the definition of fire scenarios related to potential
hazards that is the starting point of the analysis. This is to be decided in co-operation with the
approving authorities, as is the whole process.

11.2.3 Design fires

A design fire is an engineering description of the development and spread of fire for use in the
fire scenario of concern. Design fire curves are usually described in terms of heat release rate
or temperature versus time. A design fire may also be a detailed description of an ignition
source.

181
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

There are basically two ways to assess the heat release rate (HRR) in fires:
1. Analysis and synthesis of experimental data.
2. Modelling and fire simulation.
Whenever reliable and relevant data is available, it should be used. However, the larger the
initial fire is, the less relevant data exists. In this case, HRR can be estimated using modelling
and fire simulation.

For defining design fires, an approach in which the combustibles are described as fire-load
entities can be used, and is described below. In this approach, the fire compartment analysed
is filled with a suitable number of fire-load entities with proper dimensions and placing [2].
The development of the fire is described as follows:
• One of the fire-load entities is selected by the user as the initial fire.
o Typically several initial fire positions must be used, even in a single analysis, in
order to achieve a sufficiently comprehensive picture of the fire incident.
• All other fire-load entities are secondary igniting objects that will ignite when their surface
temperature reaches the ignition temperature specified by the user. After ignition, the
ignited fire-load entity starts to release heat following the same time evolution as the initial
fire.

It should be noted that ventilation has a significant influence on fire development, as

combustion cannot continue unless there is a supply of fresh air and exhaust of smoke gases.
Ventilation typically occurs via:
• voids and cracks in the building envelope;
• normal heating, ventilating and air conditioning systems;
• windows that break and fall out due to heat from the fire;
• smoke exhaust systems.

The influence of fire suppression is partially incorporated into the fire-load entity descriptions
and partially assessed by the user:
• The influence of a normal sprinkler system on the heat release development is
incorporated in the fire-load entity description.
• The influence of first-aid extinguishing and the fire brigade is assessed by the user.

The fire-load entity approach is analogous to the procedure used in normal-temperature

structural design, in which live loads are categorised depending on the use of the building, i.e.
different live loads for offices, warehouses, etc.

Nominal temperature-time curves are used as exposure conditions when they describe the fire
conditions sufficiently accurately, or are included as comparative or reference hazard
conditions. The standard temperature-time curve, the external fire curve and the hydrocarbon
curve are defined in EN 1991-1-2 [3].

11.2.4 Calculation/simulation methods

Fire safety engineering utilises a variety of different types of calculation and simulation
methods, briefly summarised below.

182
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

11.2.4.1 Fire models

Simplified fire models according to EN 1991-1-2 [3]

- Simple fire models are based on specific physical parameters, with a limited field of
application.
- Uniform temperature distribution as a function of time is assumed for compartment fires.
Non-uniform temperature distribution as a function of time is assumed in the case of
localised fires.
- When simple fire models are used, the coefficient of heat transfer by convection should be
taken as α c = 35 [W/m2K].
Compartment fires
Gas temperatures should be determined on the basis of physical parameters considering at
least the fire load density and the ventilation conditions. For external members, the radiative
heat flux component should be calculated as the sum of the contributions of the fire
compartment and of the flames emerging from the openings.
The thermal properties of the compartment linings are required for the calculation procedure.
The critical parameters are density, specific heat and thermal conductivity.

Note: Fire load densities given in Annex E of EN 1991-1-2 may be too high in the cases of
small rooms of apartment buildings. Annex E is not applied by all countries.

Localised fires
Where flashover is unlikely to occur, thermal actions of a localised fire should be taken into
account.
Note: The method for calculation of thermal actions from localised fires given in Annex C of
EN 1991-1-2 describes the conditions of the tests which were made to build up this method.
Application to real structures and real fire conditions may need more advanced and accurate
methods.

Advanced fire models according to EN 1991-1-2

Advanced fire models should take into account gas properties, mass exchange and energy
exchange.

One of the following models should be used:

– One-zone models assuming a uniform, time-dependent temperature distribution in the
compartment.
– Two-zone models assuming an upper layer and a lower layer, with time-dependent
thicknesses and time-dependent uniform temperatures.
– Computational fluid dynamics (CFD) models giving the temperature evolution in the
compartment in a completely time-dependent and space-dependent manner.
The coefficient of heat transfer by convection should be taken as α c = 35 [W/m2K], unless
more detailed information is available.
In order to calculate more accurately the temperature distribution along a member in the
localised fire case, a combination of results obtained with a two-zone model and a localised
fire approach may be considered.

183
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

An example of computational fluid dynamics (CFD) simulation models of fire-driven fluid

flows is Fire Dynamics Simulator (FDS) [4]. The software solves numerically a form of the
Navier-Stokes equations appropriate for low-speed, thermally-driven flow, with an emphasis
on smoke and heat transport from fires. FDS has been demonstrated to predict the thermal
conditions resulting from a compartment fire. The model requires that the room or building of
interest is divided into small three-dimensional rectangular control volumes or computational
cells. The density, velocity, temperature, pressure and species concentration of the gas are
computed in each cell. Based on the laws of conservation of mass, momentum, species and
energy, the model tracks the generation and movement of fire gases. FDS utilises material
properties of the furnishings, walls, floors, and ceilings to compute fire growth and spread.

11.2.4.2 Methods for analysing structural performance in fires

Analysis methods for structural performance under standard fire exposure are described in
Chapters 5-8. This section briefly summarises methods and concepts for analysing non-
standard fire exposure conditions.

Annex F of EN 1991-1-2 [3] defines the Equivalent time of fire exposure concept, but which is
not applicable to timber structures because material-specific data has not been available. In
this method, the equivalent time of standard fire exposure is dependent on
- the design fire load density
- a conversion factor which is related to the thermal properties of the enclosure
- a ventilation factor, taking into account vertical and horizontal openings
- a correction factor (function of the material composing structural cross-sections).

The principles of the fire safety engineering approach for the structural performance of timber
elements are shown in Figure 11.3 [2].
VTT

Assessment of the Structural performance

acceptability of the
design solution

dchar( r, t)

Tgas( r, t)
OK
HRR developments
in each scenario

To define charred depth the

Fire dependence between
scenarios charring rate and
temperature is needed

Building

Figure 11.3. Design concept for structural performance of timber.

184
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

A cumulative radiation energy method [5] has been demonstrated in New Zealand to predict
the failure times of wooden doors under different fire exposures. The method assumes that the
predominant mode of heat transfer in a fire compartment is by radiation rather than by
convection. The radiant heat transfer is a function of the absolute temperature T (K) raised to
the fourth power. The severity of a fire can therefore be quantified by calculating the
cumulative radiation energy over the exposure time.

As part of the work of this FireInTimber project, it has been proposed [2] that the heat flux (to
which the building element is exposed) should be calculated according to the following
expression (including convection) for the standard fire curve:

((
HF std = α c ⋅ (T std − T0 ) + Φ εσ ⋅ T std + 273 D C ) − (T
4
0 + 273 D C )) 4

⎛ t ⎞ (11-2)
Tstd = 20 D C + 345 D C ⋅ log 10 ⎜ 8 ⋅ + 1⎟
⎝ min ⎠

where α c is convective heat transfer coefficient, Φ is the configuration factor, ε is the

emissivity of the surface and σ is the Stefan-Boltzmann constant. For non-standard fire
exposure (parametric or other), here named FSE curve,

(( ) (
HFFSE = α c ⋅ (TFSE − T0 ) + Φεσ ⋅ TFSE + 273 D C − T0 + 273 D C
4
)) 4
(11-3)

where TFSE is a case-specific temperature representing the chosen design fire for the case of
concern. The standard temperature-time fire curve and an example of a design fire curve
(FSE) are shown in Figure 11.4, and the corresponding heat flux curves in Figure 11.5.
Integration of these heat flux curves over time results in cumulative values of energy received
by the building element as a function of time (see Figure 11.6).

Time (min)
Figure 11.4. Standard fire curve and an example of a design fire curve (FSE).

185
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Time (min)

Figure 11.5. Heat fluxes of standard fire curve and FSE curve of Figure 11.4.

standard fire

Time (min)
Figure 11.6. Integrated heat fluxes of standard fire and FSE heat fluxes of Figure 11.5.

186
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In the example case shown in Figure 11.6, the total energy received by the structural element
is just above 100 MJ/m2, which corresponds to the total energy in 30 minutes of the standard
fire curve.

The mean value of charring rate for a given period of time is dependent on the total energy
received during that time, which means that the resulting char depth during a defined period of
time can be calculated if the charring rate is known. The following expression applies for
standard fire exposure

⎛ t std ⎞
d std (t std ) = F ⎜ ∫ HFstd dt ⎟ (11-4)
⎜ ⎟
⎝0 ⎠

and for the FSE exposure

⎛∞ ⎞
d FSE = F ⎜⎜ ∫ HFFSE dt ⎟⎟ (11-5)
⎝0 ⎠

where F is a function describing the dependence between char depth of a specific timber
product and the integrated heat flux for a given time (in the simplest case, F can be a
constant).

Figure 11.7. Temperature-time curves used in an Austrian study on charring rates [15], [16].

Charring rates are well known for standard fire exposure. In cases where the two curves do
not essentially differ from each other, mean charring rate estimates for the FSE curves can be
made based on the standard fire value. For other cases, experimental data close to the FSE
curve of concern are needed. Figure 11.7 shows parametric temperature-time curves which

187
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

were used in an Austrian study [15], [16] to determine charring rates for different fire
exposures. This study resulted to the following conclusions: Char depth in the test with
parametric curve PAR 2 after 90 minutes correspond to the char depth of standard fire curve
after 60 minutes. Char depths after 90 minutes for the parametric curve PAR 3 and the
standard curve were quite similar.

The principles of the reduced cross-section method of EN 1995-1-2 [6] could be used for
calculating the mechanical performance of timber structures. The char depth is calculated as
described above, and the material close to the char line is assumed to have zero strength and
stiffness. The strength and stiffness properties of the remaining cross-section are assumed to
be unchanged. This method could be applicable particularly for structures with large
dimensions and for target time periods of 30 – 60 minutes. It should be noted that, if the FSE
exposure time is essentially longer than the corresponding standard fire exposure time, the
assumption of unchanged strength and stiffness properties of the remaining cross-section may
not be valid for the FSE scenario.

11.2.4.3 Methods of risk analysis

In semi-quantitative methods, the hazards are ranked according to a simple numerical score
based on such factors as the floor area of the building, the level of fire safety measures etc.
Both frequency and consequences are considered, and the alternatives are compared on the
basis of the resulting scores [7]. The methods are usually quick to perform and can be used to
rank design alternatives, but they do not give absolute or even relative risk values.

The quantitative methods are the most advanced form of risk assessment. The best source of
information regarding these questions is the fire statistics. When suitable statistics are
available, the point estimates or the probability distributions can be derived.

It may often be convenient to split the analysis into parts and approach it by using either
deterministic or probabilistic procedures. Basically, in deterministic analysis, the probability
of occurrence is neglected and the approach is to describe the hazards in terms of the
consequences. The objective is to show that, on the basis of the initial assumptions, the
defined outcome will not occur. Deterministic procedures quantify the fire growth and its
spread, as well as smoke movement and the consequences to the building or the occupants,
which are based on well-known physical relationships, scientific theories and empirical
methods.

In the probabilistic approach, both the probability of occurrence and the consequences are
based on observations. The objective is to show that the probability of an occurrence of an
unwanted event is acceptably low. The outcome is presented in terms of life or property risk
levels for the whole building.

High quality, and thus credible, probabilistic risk assessment relies on system identification
and the three main components, namely exposure (fire hazard), vulnerability (damage
potential) and robustness (limit of spread of the fire). Many methods have been developed
over the last few years for the assessment of fire risk. One probabilistic risk assessment tool
that is often used is the event-tree method. If the number of scenarios and branches of the
event tree is increased, it leads to a large number of conditional probabilities, which are very
difficult to handle by the approach. In this context, the Monte-Carlo simulation is more
flexible. The main difference between this method and the event-tree method is that the

188
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

probability distributions that are used are continuous. More complex probabilistic risk
assessment tools are based on the application of Bayesian probabilistic nets [8, 9], which are a
very strong tool in quantified risk assessment. Bayesian probabilistic nets facilitate the joint
consideration of frequentistic information, e.g. on failure rate data together with partly or fully
subjective probabilities. They permit the utilisation of structural, technical, human and
organisational components.

The risk analysis techniques may be used to compute probabilities of individual predefined
events or a general relation between the probability and the hazard severity. One possibility is
to present the results as F-N curves [7, 10]. In F-N curves, the probability of an event is
plotted against the severity measure of the event, such as number of fatalities. The role of the
fire statistics as a source of information for risk analysis has been studied by Tillander [11].

A special fire risk analysis method [17] has been developed for historical and cultural heritage
buildings. In this method the assessment of fire performance is based on one hand on reaction
to fire behaviour and on the other hand on fire resistance performance including insulation and
integrity properties. The method enables identification of those measures with considerable
influence upon fire risk independently from the construction material. Thus this method is well
applicable for both new and existing timber buildings.

11.2.5 Statistics
Fire statistics are a necessary source in the development of methods of risk assessment. Using
such statistics, probabilities of unwanted events and their consequences can be estimated.

To be able to use risk analysis methods, statistics are needed in order to provide quantitative
results. Ignition probabilities and consequences are the key information needed from
statistics. Unfortunately, proper statistical data is available from only a few countries.

Between 1986 and 1995 about 40 000 fires in the Canton of Berne in Switzerland were
statistically analysed with special regard to the type of construction (combustible/non
combustible), the age of the building, the type of occupancy and their influence on fatalities
and financial damage in case of fire [12]. In addition, the behaviour of modern state-of-the-art
timber buildings in comparison with traditional buildings was studied by analysing only
buildings built between 1990 and 1999 [13]. The collected data showed a clear influence of
the type of occupancy and type of construction on fire damage. Buildings of non-combustible
construction suffered lower fire damage than did buildings of combustible construction.
However, the analysis showed that fire damage increases with increasing building age. Fire
damage to new buildings was less than that to old buildings. It is interesting to note that, for
modern timber buildings, there is no difference in fire damage compared between them and
non-combustible construction. The collected data showed that the type of construction and
structural fire safety measures have little influence on fire fatalities, which depend primarily
on human behaviour, especially the careless use of fire and smoking products. This was
shown by a detailed analysis of fire fatalities in the years 1990 to 1999 in the Canton of Zurich
[13]. The same is confirmed by the UK fire statistics [14]: The leading cause of fatal accident
fires was careless handling of fire or hot substances, mostly the careless disposal of cigarettes.
It is also clear that the likelihood of structures of a building being the first item to ignite is
extremely rare, also for timber structures.

189
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

The effects of the materials used for load-bearing structures on economic fire losses have been
analysed by Tillander [11] using Finnish fire statistics over a five year period. The study
showed that, generally, the losses in wood-framed buildings were greater than those incurred
in buildings using other load-bearing materials, while their value at risk was significantly
lower. However, the analysis of data available showed that the material of load-bearing
structures was not the main factor when the fire risk of a building is assessed: a more
important factor seemed to be the type of building. A perhaps even more important element is
compartmentation, which unfortunately could not be examined more closely because the
necessary detailed information was not available.

The main findings of the Swiss and Finnish studies are clearly in line with each other, and
demonstrate the need to base fire risk analysis on properly analysed statistical values.

11.3 Example on probabilistic approach to assess fire safety

This case study summary is based on results of application of advanced, probabilistic
calculation methods to assess of the fire endurance of timber structures [18], [19].

11.3.1 Description of the building, structures and fire scenario

The building considered in this study is a grocery shop located in the South-West region of
Finland. It consists basically of a single space with the following dimensions: length 50 m,
width 34 m and height 5,5 m. The building is a timber beam-column structure constructed of
glued laminated wood (see Figure 11.8 and Figure 11.9). Deterministic design calculations
using 50 % of the snow load characteristic relevant to that part of Finland, i.e., 1,4 kN/m2,
show that an unprotected beam would survive standard fire exposure for 30 minutes.

Figure 11.8. Overview of the hall and timber structures.

190
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 11.9. The glulam beam spanning between the columns. Dimensions are given in
meters; the width is 0,165 m.

11.3.2 Analysis of the beam failure probability

In Figure 11.10 a) examples of probabilistic fire simulation curves for heat release rate (HRR)
are shown for the studied case. Based on these HRR curves the heat exposure to the structures
can be calculated and the resulting char depths as a function of time are shown in Figure 11.10
b). The distribution of possible char depths at 90 minutes are given in Figure 11.11.

a) b)

Figure 11.10. Example of probabilistic fire simulation results for heat release rate (a) and
resulting char depths (b).

Figure 11.11. Probability distribution of char depth at 90 minutes.

191
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

In the unlikely case that the fire will not extinguished, the beams will fail after 75 minutes as
shown in Figure 11.12, depending on the fire scenario and snow load. These failure time
estimates are much longer than the estimates of critical time for conditions because of smoke
(production and movement) and heat exposure (Figure 11.13 a) as well as the estimated arrival
time of fire brigade (Figure 11.13 b).

Figure 11.12. Failure probability distribution in the case the fire is not extinguished.

a) b)
100 % 1.0
Cumulative frequency
Cumulative frequency

80 % 0.8
60 % 0.6 distance to FB
40 % 0,55 km
0.4
20 %
0.2
0%
0.0
0 15 30 45
0 5 10 15
critical time: smoke and heat (min) FB arrival time (min)
Figure 11.13. a) Cumulative distribution of critical time for smoke (visibility) and heat
exposure. b) Cumulative distribution for the arrival time for fire brigade.

11.3.3 Risk analysis and conclusions

Results of performed fire risk analysis are shown in Figure 11.14. Regarding the fire safety
measures, improving the reliability of the fire detectors is the most efficient measure. Timber
structures are not critical concerning fire safety in the building studied; if one would want to
improve fire safety, investment to e.g. faster detection and alarming would be much more
efficient than investment to structural fire protection. This case study demonstrates the
possibilities which fire safety engineering / fire risk analysis can offer for showing that the
building (as a whole) can meet the required safety level without prescribing which materials
can be used or which cannot be used.

192
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

Figure 11.14. Beam failure probabilities for the building life time (50 years) for different fire
safety provision cases.

References

1. EN 1990. Eurocode. Basis of Structural Design. CEN 2002.

2. Hietaniemi, J. & Mikkola, E. Design fires for fire safety engineering. VTT Working
Papers. Draft report, to be published.
3. EN 1991-1-2. Actions on Structures - Part 1-2: General Actions – Actions on structures
exposed to fire. CEN 2002.
4. FDS, Fire Dynamics Simulator. http://fire.nist.gov/fds/.
5. Barnett, C.R. A new T-equivalent method for fire-rated wall constructions using
cumulative radiation energy. Journal of Fire Protection Engineering, 2007. Vol. 17, pp.
113-127.
6. EN 1995-1-2:2004 Eurocode 5: Design of timber structures – Part 1-2: General –
Structural fire design. European Committee for Standardization, Brussels, 2004
7. Frantzich, H. Risk Analysis and Fire Safety Engineering. Fire Safety Journal, 1998. Vol.
31, No. 4, pp. 313-329.
8. Maag, T. Risikobasierte Beurteilung der Personensicherheit von Wohnbauten im
Brandfall unter Verwendung von Bayes’schen Netzen, [A risk based approach to life
safety in dwellings during a fire using Bayesian networks], PhD Thesis Nr. 15366, ETH
Zurich, 2004.
9. Fontana, M., Maag, T. Fire risk assessment based on Bayesian networks, Proceedings of
the 5th International Conference on Performance-based Codes and Fire Safety Design
Methods, Luxembourg, 6–8 October 2004, Society of Fire Protection Engineers: Bethesda.
10. Korhonen, T., Hostikka, S. & Keski-Rahkonen, O. A proposal for the goals and new
techniques of modelling pedestrian evacuation in fires. In: Gottuk, D. & Lattimer, B.
(eds.). Proceedings of the Eighth International Symposium on Fire Safety Science.
Beijing, China, 18-23 Sept. 2005. International Association for Fire Safety Science, 2005.
pp. 557-569.
11. Tillander, K. Utilisation of statistics to assess fire risks in buildings. Espoo: VTT
Building and Transport, 2004. VTT Publications 537. 224 p. + app. 37 p.
12. Fontana, M., Favre, JP., Fetz, C. A survey of 40 000 building fires in Switzerland. Fire
Saf. J. 1999; 32: 137–158.

193
Chapter 11 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

13. Fontana, M. Fire safety concepts of buildings, Proceedings, 1th International Symposium
on Advanced Timber and Timber-Composite Elements for Buildings, COST Action E-29,
October 27-29, 2004, Florence, Italy, ISBN 88-901660-1-0, pp. 277-290.12.
14. Fire statistics, United Kingdom, 2006. Department for Communities and Local
Government, 2008, ISBN 978409800460.
15. Fornather, J., Bergmeister, K., Luggin, W., Giertlová, Z.: Versuchsbericht –
Kleinbrandversuchsreihe 1 Teil 1 (KBV 1/1) [Versuchsbericht]. Universität für
Bodenkultur, Institut für konstruktiven Ingenieurbau. Wien, 2000.
16. Fornather, J., Bergmeister, K.: Versuchsbericht – Kleinbrandversuchsreihe 1 Teil 2 (KBV
1/2) [Versuchsbericht]. Universität für Bodenkultur, Institut für konstruktiven
Ingenieurbau. Wien, 2001.
17. Vandevelde, P., Streuve E.: Fire-Tech, Fire Risk Evaluation to European Cultural
Heritage, Decision Supporting Procedure, Users Guide. Laboratorium voor Anwending der
Brandstoffen en Warmtoeverdracht, Department of Flow, Heat and Combustion
Mechanics, Gent, 2005, ISBN 908098521 X.
18. Hietaniemi, J., Toratti, T., Schnabl, S. & Turk, G. Application of reliability analysis and
fire simulation to probabilistic assessment of fire endurance of wooden structures. VTT
Working Papers 54. Espoo 2006 (http://www.vtt.fi/publications/index.jsp).
19. Hietaniemi, J. Probabilistic simulation of fire endurance of a wooden beam. Structural
Safety 29 (2007), 322–336.

194
Chapter 12 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

12 Execution and control

New chapter, to be further developed.

This chapter describes the need for execution and control of workmanship to ensure that the
planned fire safety precautions are built in.
It also emphasises the need for fire safety at building sites, when all fire safety measures are not
yet in place.
• Timber frame constructions require proper execution and monitoring during construction
period
• Self-auditing at all levels is important
• Control plans are needed for both design and execution

12.1 Control of workmanship

Timber frame constructions consisting of a combination of several different building

materials, allowing multiple functions such as fire and sound properties, are dependent on an
adequate performance at construction sites or in the factory. This is especially true for mineral
wool insulation that must be mounted carefully to be in direct contact with wooden beams and
girders. Empty voids may lead to rapid carbonization and decreased fire resistance. Careful
installation of mineral wool is particularly important in the attic floors, which often lack fire
protection on top. Long enough screws for mounting claddings are also very essential for the
fire resistance. If the screws are too short, the cladding might fall off at an early stage and
wooden beams and girders will then be exposed to fire, leading to rapid charring and
premature loss of resistance. Chapter 6 gives recommendations on the appropriate screw
lengths for mounting claddings.
Mounting of fire stops inside structures and tightening of penetrations and installations on the
construction site are essential to certify the fire performance of the structures.
Recommendations are given in chapter 8. The performance of such details can be checked
only during the construction period, but will affect the building functions both in normal use
(sound insulation) and in accident situations (fire).
The most important control in all phases and at all levels in a building project is the individual
conducting of self-monitoring. Checks carried out by others are often only random checks.
Fire engineering design involves architects, designers, consultants, fire, plumbing and
electricity, and possibly also producers and suppliers. It is therefore important that the
responsibilities of each actor are well defined and limited. A principal fire responsible should
be appointed at major construction projects. The responsibility includes the execution of the
construction site. Errors often occur in the interface between different actors.
Fire protection documentation should always be established, and normally carried out by a fire
consultant.
Control plans / checklists should also be established in both design and execution. These
should specify in detail the control areas and control responsibilities. Critical areas such as
interfaces between the various control functions, should be given special attention.

195
Chapter 12 - FireInTimber Guideline - Second draft for comments December 2009 – Not to be referenced

12.2 Fire safety at building sites

To be further developed.

References
12.1 Fire prevention on construction sites: Joint code of practise on the protection from fire of
constructions sites and buildings undergoing renovation. Fire Protection Association,
UK, 2009.
12.2 Fire safety at building sites. Swedish guideline, to be published 2010.

196
 Second draft for comments December 2009 – Not to be referenced

Description of terms
Ceiling Non-load bearing element of a building construction designed to provide
horizontal fire separation.
Cladding Sheeting attached to a construction element for different purposes. For example
linings.
Connection Fixing detail between two components which is designed to transfer loads.

External wall Wall forming the external envelope of a building which may be exposed
separately to an internal or an external fire.
Fire design Design of a structure to fulfil the required performance in case of fire.
Fire model Representation of a system or process related to fire development, including fire
dynamics and fire impacts.
Fire protection Protection attached to a structural building member such that the temperature on
the surface of a member or of a building construction is reduced throughout
exposure to fire.
Floor Horizontal element of building construction which is load bearing.
Glass wool Glass wool is a mineral wool manufactured predominantly from sand and
molten recycled glass. If it is not declared as heat resistant, glass wool is more
sensitive when exposed to fire in a fire resistance test (see also 6.3.3).
Gypsum fibre boards Gypsum fibre boards are composed of set gypsum plaster reinforced with
dispersed fibres, which may be inorganic and/or organic, to form flat rectangular
boards. They may contain additives and/or fillers to impart additional properties.

Gypsum plaster board Gypsum plaster boards are composed of a plaster core encased in, and firmly
bonded to strong durable paper liner to form a flat rectangular board. They may
contain additives to impart additional properties.

Insulation The ability of a member or of a separating element of building construction

when exposed to fire on one side, to restrict the temperature rise of the
unexposed face.
Integrity The ability of a member or of a separating element of a building construction,
when exposed to fire on one side, to prevent the passage trough it of flames and
hot gases and to prevent the occurrence of flames on the unexposed side.

Internal wall Vertical element of a building construction which can be designed for fire
resistance on one or both sides.
Joint A space between two adjacent components irrespective of whether it is filled or
not.
Lining Sheet material fixed to framework as in a partition or a ceiling. For example
insulating boards or plasterboards.
Mineral wool Generic term for insulation wool manufactured from molten rock, slag or glass
(and sand). Different production methods and raw materials can result in
insulation products which show different behaviour when exposed to
fire resistance tests (see also 6.3.3).
Separating element En element that is intended for use in maintaining separation between two
adjacent areas of a building in the event of a fire.

Stone wool Stone wool (rock wool) is a mineral wool manufactured predominantly from
molten naturally occurring igneous rocks (see also 6.3.3).
Wall Load bearing walls designed to support an applied load, non-load bearing walls
designed not to be subjected to any load other than its self weight.

197
List of terms in different languages
Please propose/delete terms and add translations

English German French Italian

access balcony
attic
board
brace
buckling mode
ceiling
cladding
column
connection
cooling
evacuation
external wall
fire design
fire model
fire protection
flashing
floor
glass wool
ground plate
gypsum plaster board
heat flow
insulation
integrity
internal wall
joint
lining
load bearing structure
nogging
mineral wool
panel
risk assessment
separating element
smouldering
steel strip
stone wool
storey
stud
surface layer
suspended ceiling
timber frame structure
timber structure
top plate
tongue and groove
wall
window

199