ISO/TR 9705-2:2001

TECHNICAL ISO/TR
REPORT 9705-2
First edition
2001-05-01
Reaction-to-fire tests — Full-scale room
tests for surface products —
Part 2:
Technical background and guidance
Essais de réaction au feu — Essais dans une pièce en vraie grandeur pour
les matériaux de revêtement intérieur —
Partie 2: Données techniques et lignes directrices
Reference number
©
ISO 2001
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but shall not
be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In downloading this
file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat accepts no liability in this
area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation parameters
were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In the unlikely event
that a problem relating to it is found, please inform the Central Secretariat at the address given below.
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means, electronic
or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or ISO's member body
in the country of the requester.
ISO copyright office
Case postale 56 � CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.ch
Web www.iso.ch
Printed in Switzerland
ii © ISO 2001 – All rights reserved

Contents Page
Foreword.iv
Introduction.v
1 Scope .1
2 Characteristics of the ignition sources .1
2.1 Standard ignition source .1
2.2 Alternative ignition source .1
3 Sensitivity analyses.6
3.1 General.6
3.2 Specimen configurations.6
3.3 Effect of the burner size.7
3.4 Effect of the stand-off distance of the burner.7
4 Heat balance in the room .7
4.1 General.7
4.2 Heat release by combustion.7
4.3 Heat loss by convection.8
4.4 Heat loss by conduction .8
4.5 Heat loss by radiation .8
4.6 Results of heat balance calculations.9
5 Measuring techniques.9
5.1 Mass flow through the doorway and interface height .9
5.2 Measurement of toxic gases.10
5.3 Mass loss rate from the sample .10
6 Extended calculations.10
6.1 Filling time of room and hood .10
6.2 Prediction of mass flow and interface position.11
6.3 Estimate of sample mass loss.14
6.4 Fire growth models.14
7 Precision data .14
7.1 General.14
7.2 ISO round robin.15
7.3 ASTM round robin .16
8 Other test protocols using similar equipment.16
9 Specimen mounting .17
Annex A Calculation of HRR by means of different gas analysis data.18
Annex B Practical example of the measurements of toxic gases by FTIR and ion chromatography .26
Annex C Estimation of mass loss rate by means of HRR and gas analysis measurements.32
Annex D Overview of other test protocols using similar equipment .35
Bibliography.38
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies (ISO
member bodies). The work of preparing International Standards is normally carried out through ISO technical
committees. Each member body interested in a subject for which a technical committee has been established has
the right to be represented on that committee. International organizations, governmental and non-governmental, in
liaison with ISO, also take part in the work. ISO collaborates closely with the International Electrotechnical
Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 3.
The main task of technical committees is to prepare International Standards. Draft International Standards adopted
by the technical committees are circulated to the member bodies for voting. Publication as an International
Standard requires approval by at least 75 % of the member bodies casting a vote.
In exceptional circumstances, when a technical committee has collected data of a different kind from that which is
normally published as an International Standard ("state of the art", for example), it may decide by a simple majority
vote of its participating members to publish a Technical Report. A Technical Report is entirely informative in nature
and does not have to be reviewed until the data it provides are considered to be no longer valid or useful.
Attention is drawn to the possibility that some of the elements of this part of ISO/TR 9705 may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO/TR 9705-2 was prepared by Technical Committee ISO/TC 92, Fire safety, Subcommittee SC 1, Fire initiation
and growth.
ISO 9705 consists of the following parts, under the general title Reaction-to-fire tests — Full-scale room tests for
surface products:
� Part 1: Full-scale test for surface products (currently published as ISO 9705:1993,Fire tests — Full-scale room
test for surface products)
� Part 2: Technical background and guidance [Technical Report]
iv © ISO 2001 – All rights reserved

Introduction
ISO 9705:1993 specifies a test method simulating a fire that starts under well-ventilated conditions, in a corner of a
small room with a single open doorway.
The method is intended to evaluate the contribution to fire growth provided by a surface product using a specified
ignition source. The method provides data for a specified ignition source for the early stages of a fire from ignition
up to flashover. ISO 9705:1993 also describes different measurement techniques inside and outside the room. This
part of ISO 9705 gives background information and support to the potential users of the test. It gives the user of the
test technical information on the ignition source, heat fluxes in the room from the burner, heat balance in the room
during a fire, aspects of smoke production and toxic gas species production, as well as aspects of modelling the
results of these tests. It gives the user the information necessary to select the testing procedure for specific projects
or regulations.
TECHNICAL REPORT ISO/TR 9705-2:2001(E)
Reaction-to-fire tests — Full-scale room tests for surface
products —
Part 2:
Technical background and guidance
1 Scope
This part of ISO 9705 provides guidance on ISO 9705:1993. It describes the technical background of the test and
gives information which may be used for determining a testing procedure for a specific scenario, or how results can
be utilized in a total hazard analysis for the specified scenario.
2 Characteristics of the ignition sources
2.1 Standard ignition source
The standard ignition source consists of a sandbox burner with dimensions of 0,17 m � 0,17 m. This source is used
in reference [1] (see Bibliography). An important characteristic of the ignition source is its heat transfer towards the
material. Figures 1 and 2 show a detailed mapped overview of the total heat flux towards the specimen and the gas
temperatures. The measurements are performed in an open wall configuration, at an energy release rate level of
100 kW [2]. These values will slightly change when the burner is located in a room environment. Figures 3 and 4
give the contours of a constant heat flux of 10 kW/m at different heat outputs of the burner and also where areas
of total heat flux are higher than a given value.
2.2 Alternative ignition source
One of the alternative heat sources is a box burner, with dimensions of 0,3 m � 0,3 m. It is described in
ASTM E603-98 [3]. Figures 5 and 6 give a detailed mapping of heat fluxes and gas temperatures for a burner
energy release rate of 160 kW [2]. Other heat sources may be more appropriate (see annex B of ISO 9705:1993).
Figure 7 gives results of heat fluxes towards the specimen for a heat source level of 40 kW and 160 kW, with
different gases (natural gases and a mixture of natural gas and toluene) [4]. Figures 8 and 9 show a comparison of
different burner sizes for contours of constant heat flux of 10 kW/m , at an energy release rate of 100 kW in an
open corner and for areas exposed to a certain irradiant heat flux [4].
Finally, an example is given of the difference between the total heat flux produced by a burner in a corner and a
wall position. Table 1 gives an overview of the total heat flux towards the floor and the total heat flux to the wall at
0,9 m and 1,5 m height for energy release rates of 40 kW and 160 kW using the alternative ignition source of
ISO 9705:1993. Results show that, for the corner position, heat flux levels are higher in almost all cases.
Figure 1 — Heat flux distribution at an energy Figure 2 — Gas temperature distribution 30 mm
release rate of 100 kW for the standard ignition from the wall at an energy release rate of 100 kW
source in an open corner for the standard ignition source in an open corner
2 © ISO 2001 – All rights reserved

NOTE Contours of 10 kW/m .
Figure 3 — Contours of constant heat flux for the standard ignition source in an open corner at different
irradiant heat flux levels
Figure 4 — Areas of total heat flux levels higher than a given value for the standard ignition source at
different irradiant heat flux levels in an open corner
Figure 5 — Heat flux distribution at 160 kW for the Figure 6 — Gas temperature distribution 30 mm
alternative ignition source in an open corner from the wall at 160 kW for the alternative ignition
source in an open corner
4 © ISO 2001 – All rights reserved

Figure 7 — Heat flux distribution for the alternative ignition source in an open corner at 40 kW and 160 kW
with different types of gas
NOTE Contours of 10 kW/m .
Figure 8 — Contours of constant heat flux for the different sizes of box ignition sources in an open corner
at a 100 kW heat source level
Figure 9 — Areas of total heat flux levels higher than a given value for different box ignition sources at
100 kW in an open corner
Table 1 — Comparison between corner and centre wall position
Burner in the corner Burner at centre of back wall
Heat source Heat flux to Heat flux to Heat flux to Heat flux to Heat flux to Heat flux to
level floor wall at 0,9 m wall at 1,5 m floor wall at 0,9 m wall at 1,5 m
2 2 2 2 2 2
kW/m kW/m kW/m kW/m kW/m kW/m
40 kW 0,6 12,5 6,5 0,6 8,5 4
160 kW 5,4 56 60 4,2 62 33
3 Sensitivity analyses
3.1 General
Various sensitivity analyses have been performed over the last 25 years. All studies used the room described in
ISO 9705:1993, but differed in the type of ignition source (the standard ignition source or the alternative ignition
source of ISO 9705). These sensitivity analyses contained different specimen configurations and different ignition
positions and levels. An overview is given below of some of the findings as guidance for testing in the ISO 9705
room.
3.2 Specimen configurations
Sensitivity analyses revealed that testing with linings on both ceiling and walls resulted in a more severe condition
than tests with linings on the walls only [5]. When only the walls are covered with linings, a ceiling lined with
ceramic wool is more severe than a ceiling lined with gypsum boards and will show less discrimination between the
different materials [6].
6 © ISO 2001 – All rights reserved

In order to achieve comparable tests data between laboratories and high discrimination, it is recommended in
ISO 9705 that the walls (excluding the wall containing the doorway) and the ceiling are covered with the product.
When other specimen configurations are used, this should be clearly stated in the report.
3.3 Effect of the burner size
The effect of the burner size has been studied extensively within the Eurefic programme [7]. Results have been
shown for heat flux distribution and gas temperatures. Moreover, tests were done in a room lined with particle
board. Little effect was seen on the time to flashover at rates of heat release of 160 kW and 300 kW. At a lower
heat release of about 40 kW, the time to flashover with a large burner (0,5 m by 0,5 m) was significantly longer than
for the other burners (standard and alternative ignition source of ISO 9705). The reason for this was explained by
the smaller area which is exposed to a given heat flux level (see Figure 9), hence producing a slower flame spread.
3.4 Effect of the stand-off distance of the burner
Experiments at lower heat source levels with the alternative ignition source of ISO 9705 showed that there was a
considerable influence of the stand-off distance of the burner [8]. With the standard ignition source, the stand-off
distance seems to be less critical. The influence can in most cases be predicted by heat flux measurements at the
walls behind the burner flame.
4 Heat balance in the room
4.1 General
An energy balance calculation was carried out at the early stages on the development of the ISO 9705 room corner
test [9]. The energy balance in the room can be given as follows:
�� � � �
QQ��Q�Q�Q
cco w r b
where
�
Q is the heat released by combustion (kW);
c
�
Q is the heat loss by convection through the doorway (kW);
co
�
Q is the heat loss by conduction into the surrounding structure (kW);
w
�
Q is the heat loss by radiation throughout the doorway (kW);
r
�
Q is the heat stored in the gas volume (kW).
b
In most cases the heat stored in the gas volume is negligible. The other terms are calculated as given in the
following paragraphs. The results of a heat balance calculation are also given below.
4.2 Heat release by combustion
Heat release by combustion might be the heat release measurement or, in the case of the calibration test, this can
be calculated as
�
�
QH��∆ m
ccf
where
∆H is the heat of combustion, equal to the net calorific value of propane (46,4 MJ/kg);
c
m� is the mass loss rate of the propane (kg/s).
f
4.3 Heat loss by convection
The heat loss at the doorway can be calculated as follows:
�
Qm��� c()T�T
co o p g a
where
�
m is the mass flow rate out of the doorway (kg/s);
o
c is the specific heat of the smoke gases (kJ/kg�K);
p
T is the smoke gas temperature (K);
g
T is the ambient temperature (K).
a
4.4 Heat loss by conduction
The heat loss by conduction through the walls can be calculated as follows:
dT
��
�
Qk�� ��
w0�� x�
��
dx
where
� 2
Q is the heat conduction per unit area (W/m );
��
w
k is the thermal conductivity (W/m�K);
dT
��
is the temperature gradient at the surface (K/m).
x�0
��
��dx
The temperature gradient can be calculated by means of temperature measurements in and on the walls. The heat
loss through the walls can also be calculated using numerical heat transfer methods.
4.5 Heat loss by radiation
The heat loss by radiation out of the doorway can be calculated by adding the contribution from a number of
smaller areas from the walls and ceiling of the room:
�
QA���� �F�T
r �ii i i
i
8 © ISO 2001 – All rights reserved

where
�8 2 4
� is the Stefan Boltzmann constant (5,67�10 W/m �K );
� is the emissivity;
i
A is the temperature gradient at the surface (K/m);
i
F is the view factor;
i
T is the absolute temperature (K).
i
4.6 Results of heat balance calculations
The heat balance calculations of a room test with a propane burner as heat source are given in Table 2 at steady-
state conditions.
Table 2 — Results of heat balance calculations
Heat release by Heat loss by Heat loss by Heat loss by Total heat Difference
combustion convection conduction radiation loss
kW kW kW kW kW %
125 105 19 6 129 3
250 208 32 12 252 1
5 Measuring techniques
5.1 Mass flow through the doorway and interface height
One of the methods referred to in ISO 9705 to calculate the mass flow out of the door is by means of bi-directional
probes and suction pyrometers in the door opening. In many cases this is an extensive and expensive method. In
the next clause some calculation methods will be given for determination of the interface height and the mass flow
through the door opening. A possible set-up of instrumentation for such calculations is given in Figure 10 [6]. It
should be noted that in some cases small pressures are to be measured which can influence the accuracy of the
measurement.
Key
1 Door TC tree
2 �p transducer
3 Corner TC tree
4 Corner burner
Figure 10 — Experimental set-up for different calculation methods of the interface height and mass flow
rates at the door opening of the ISO 9705 room
5.2 Measurement of toxic gases
Additional to the measurement techniques given ISO 9705, techniques such as FTIR and ion chromatography have
recently been applied successfully in full-scale tests. A practical example how this can be performed is given in
annex B. The reader is also referred to the documents developed within ISO/TC 92/SC 3 for a complete overview
of the measurement of toxic gases in combustion gases produced in fire tests.
5.3 Mass loss rate from the sample
Direct mass loss measurements of the linings can be performed by means of putting the complete room on load
cells or by putting the structure on which the linings are fixed on load cells. Due to the high tare value obtained by
the weight of the room, it should be noted that only limited accuracy can be obtained. For items positioned in the
room, a weighing platform as used in furniture calorimeters can be used and has been successfully applied.
6 Extended calculations
6.1 Filling time of room and hood
At the beginning of a test there is some delay time in order to fill the part above the soffit level of the door in the
room. Filling of the hood in the beginning of the test is almost negligible since the smoke gases will enter
immediately into the duct. Some filling of the hood might occur later on in the test if the extraction rate is close to
the limit of the system. This is close to flashover conditions if the maximum exhaust flow rate is used. Delay time
correction can be easily incorporated into the time shifting of the data.
10 © ISO 2001 – All rights reserved

Although mixing of the gases is enhanced by the baffle plates into the hood, corrections can be made to take into
account mixing of the gases. However, this will only be necessary if one wants to perform calculations which are
better than the actual accuracy as given in ISO 9705. The following formulae developed by Kokkala can be used
for correction of mixing [2]:
Ct()��C [exp(t/t )�exp(t/��)]/(1��)
mr,max d m
where
C is the measured concentration;
m
C is the maximum concentration;
r,max
�
m
�´ is the dimensionless time constant =
t
d
�
� is the time constant of mixing = volume of “mixing chamber” (V)/volume flow rate (V );
m
t is the duration time of phenomena (s).
d
6.2 Prediction of mass flow and interface position
6.2.1 General equations
General equations for the vent flow rates as a function of temperature profiles are described in this subclause. The
equations are based on the orifice concept.
Flows in and out of the compartment are driven by pressure differences across the vent. Inside the compartment,
velocities are negligible except locally in flames, plumes and wall jets. Thus, (static) pressure varies vertically only
due to gravity. The velocity at height z is given according to Bernouilli's equation [10] as
pz� �p �z
� �
i �
vz ��C 2 (1)
� �
� z
� �
d
where
–1
v is the velocity (m�s );
C is the orifice coefficient;
p is the pressure inside the compartment (Pa);
i
p is the pressure outside the compartment (Pa);
�
z is the height above floor level (m);
�3
� is the density of gases in the doorway (kg�m ).
d
The height z at which there is no pressure difference (and no flow) between the compartment and the
n
environment, is called the neutral plane. There is a maximum of one neutral plane for the case of a room connected
to the outside (or a large reservoir). Hydrostatic pressure outside the compartment can be written as a function of
height:
pz��pz z�z � g (2)
� � � � � �
��nn
where
�3
� is the density of ambient air (kg�m );
�
�2
g is the acceleration due to gravity (m�s ).
Hydrostatic pressure differences are very small (typically a few pascals) compared to the magnitude of the absolute
pressure itself, which is of the order of 10 Pa. Therefore, p may be written as
�
� T
352,8
ref ref
��� (3)
�
TT
��
where
�3
� is the density of ambient air at temperature T and atmospheric pressure (kg�m );
ref
ref
T is the reference temperature (K);
ref
T is the temperature of ambient air (K).
�
�2
With acceleration of gravity g=9,81m�s , equation (2) then becomes
3 461
p ()z=p z + z �z (4)
� � � �
nn
�
T
�
Inside the compartment, temperature is not constant with height. Thus, pressure as a function of height follows from
zz�
n
3 461
()z=p z + d�z (5)
p � �
n
i
�
()z�
T
i
Combining equations (4) and (5) leads to the following expression for the pressure difference:
zz�
n
��11
∆pz( )= 3 461 ��d z (6)
���
��()zT�
T
i �
The mass flow rate out of the compartment follows from integration of equation (1):
Hz��H z
dn dn
(7)
�=z��()��v()z d�z � 2 ()z� ∆p()z� d�z
mCW CoW
ood d d
d
��
where
W is the door width (m);
d
H is the door height (m).
d
As the outflowing gases mainly consist of nitrogen, the density is not too different from that for air at the same
temperature and pressure. Substitution of equation (6) and an expression analogous to equation (3) for � into
d
equation (7) yields
12 © ISO 2001 – All rights reserved

1/2
Hz� zz'�
dn��n
��
11 1
��
=z1563 ��d" dz (8)
�
mCW ��
ood
��
��
()zz� (")
TTT
di���
��
Similarly, the inflow rate is equal to
1/2
zz��
z��
nn
��
11 1
��
=z1563 ��d" dz (9)
mC� W
iid ��
��
��
()zz� (")
TTT
di�
��
��
��
It should be noted that a distinction is made between the orifice coefficient for inflow C and that for outflow C .This
i o
allows implementation of the recommendations in reference [11].
6.2.2 z from temperature profiles and one�p measurement
n
Algorithms developed at NIST to reduce room fire data include a procedure to obtain z and mass flow rates through
n
the vent [12]. These algorithms are referred to as RAPID. Equation (6) shows that�p can be calculated as a function of
height on the basis of the temperature profile measured inside the room if z is known. The NIST RAPID procedure [12]
n
requires measurement of�p at one reference height z in addition to the temperature profile inside the room. z can
ref n
then be found by evaluating equation (6) at z :
ref
zz�
ref
n
��
∆p=( ) 3 461 ��dz (10)
z
ref
��
�
()z�
��TT
i �
The best reference height is at the soffit as the pressure difference is usually the largest at this height. Once z is
n
known, mass flow rates can be obtained according to equations (8) and (9). This also requires the temperature
profile in the doorway.
6.2.3 z via temperature profiles only
n
The RAPID procedure outlined in 6.2.2 has some practical difficulties.�p(z ) is in the order of a few pascals and is
ref
very difficult to measure. Moreover, pressure data at such a low level are very noisy mainly due to turbulence. Another
important drawback of the procedure is that it does not necessarily conserve mass. Therefore, a procedure is outlined
here, based on temperature profiles only [6]. The requirement for conservation of mass replaces equation (10) as the
equation for obtaining z . The mass balance equation has the following form:
n
dm
r
=+ + � (11)
mm�� m� m�
ib v o
dt
where
m is the mass accumulated inside the compartment (kg);
r
m� is the ignition source mass flow rate (kg/s).
i
The rate of change of mass inside the room can be calculated from the temperature profile measured inside the
room via
H
d d�z
dm
r
=W352,8L (12)
�
ddtt ()z�
T
i
where
W is the room width (m);
L is the room length (m);
H is the room height (m).
The burner gas flow rate m� is measured. m� consists of water vapour and pyrolysis gases emerging from the walls.
b v
Both m� and m� are usually very small compared to the other terms in equation (12) and can be neglected. m� and
b v o
m� are functions of z as indicated in equations (8) and (9). Therefore, equation (11) is a non-linear equation in z which
n n
i
canbesolvediteratively.
6.3 Estimate of sample mass loss
When no mass loss measurements are made during tests it is possible to estimate the mass loss rate as a function
of time using one of two methods. The first method is to divide the measured heat release by an effective heat of
combustion of the product, which might be determined, for example, in the cone calorimeter. The second method is
to estimate the mass loss rate by means of the gas analysis measurements. A procedure for this method is outlined
in annex C [13].
6.4 Fire growth models
The test results of a room corner test may be predicted by means of a simulation model which calculates wall fire
growth in a small room. An extensive overview of modelling full-scale test results is given in ISO/TR 11696.
Fundamental solutions for fire growth need to address various phenomena such as heat transfer, fluid dynamics
and combustion. Most of the developed models have introduced some simplifications for those problems.
They can be divided into a number of categories, as follows.
� Models applying straightforward empirical or statistical methods and using small scale data obtained directly
from one or more test methods such as the cone calorimeter (ISO 5660) and the LIFT apparatus. Although
they use a considerable number of simplifications, their predictions have been successful. Most of them are
limited to one specific scenario, but extensions to other scenarios are possible by using other empirical
parameters.
� Models applying semi-material characteristics. These semi-material characteristics are calculated from the
small scale data obtained in, for example, the cone calorimeter (ISO 5660) and the LIFT apparatus and can be
considered as a derivative or mean value of a fundamental material characteristic. Examples of such
characteristics are mean k� c, ignition temperature, etc. Most of these models also show satisfactory results
and are applicable for more than one scenario.
� Models applying fundamental material characteristics. Most of these models use characteristics which are less
easy to determine with standard reaction to fire apparatuses, but some progress has been made in recent
years. They have, however, been limited to certain products. In most cases these are sub-models describing
one type of flame spread (e.g. horizontal flame spread) and they must be incorporated in a zone on field
model.
For a description of the different developed models see ISO/TR 11696.
7 Precision data
7.1 General
Two round robins on the room corner test have been conducted in recent years. First, an initial round robin with five
laboratories was carried out prior to the circulation of ISO/DIS 9705 [14]. This first round robin used the standard
14 © ISO 2001 – All rights reserved

procedure as described in the standard. Later, a wider round robin was carried out as a joint activity between
ISO and ASTM but using the ASTM procedure as testing protocol, i.e. the alternative ignition source with only the
walls covered [15]. The results of the round robins are given in 7.2 and 7.3.
7.2 ISO round robin
This round robin was performed at five laboratories in Denmark, Finland, Norway, Sweden and the United
Kingdom. The results are given in Tables 3 to 6 and indicate that the reproducibility of the method is similar to other
fire test methods, such as fire resistance tests using large-scale furnaces. The 95 % confidence interval of the
mean of the time to flashover was found to be � 37 s and � 18 s for ordinary plywood and melamine-faced particle-
board, respectively. A similar range was found also for the rates of smoke and CO production. The reproducibility of
the tests on the fire-retarded plywood was about the same as the untreated plywood, although in only one of the
tests flashover conditions were reached.
The results of the tests on the fire retarded expanded polystyrene varied considerably, mainly due to the different
gluing methods.
Table 3 — Results for birch plywood
Time to reach
Rate of heat release Rate of smoke Rate of CO Heat flux
Laboratory
=1MW production production =20kW/m
=40m /s =15g/s
VTT 137 s 129 s 121 s 153 s
SINTEF NBL 152 s 129 s 120 s
RW 122 s 121 s 112 s 138 s
FRS 209 s 198 s
a a
Average 137 s 126 s 118 s 146 s
a
Confidence interval ��37 s ��11 s � 12 s
(95 %)
a
FRS results are not included, because the time to critical smoke value indicated that the test was an outlier according
to Dixon’s outlier test [11] (see also Appendix 3 of [11]).
Table 4 — Results for melamine-faced particle-board
Time to reach
Rate of heat release Rate of smoke Rate of CO Heat flux
Laboratory
=1MW production production =20kW/m
=40m /s =15g/s
VTT 182 s 158 s 172 s 190 s
SP 204 s 174 s 198 s
RW 202 s 165 s 174 s 208 s
FRS 206 s 186 s
Average 199 s 171 s 181 s 199 s
� 18 s � 19 s � 36 s
Confidence interval (95 %)
Table 5 — Results for fire-retarded plywood
Time to reach
Rate of heat release Rate of heat Rate of CO Heat flux
Laboratory
=1MW release production =10kW/m
= 700 kW =10g/s
VTT — 629 s 627 s 621 s
RW 1148 s 624 s 630 s 624 s
RW — 645 s 657 s 630 s
FRS — 637 s
Average — 634 s 638 s 625 s
Confidence interval (95 %) � 15 s � 41 s �5s
NOTE The critical values are different in Tables 5 and 6.
Table 6 — Results for fire-retarded polystyrene
Time to reach
Rate of heat release Rate of smoke production Rate of CO Heat flux
Laboratory
2 2
=1MW =40m /s production =10kW/m
=15g/s
VTT — 120 s ——
SP 67 s 59 s 79 s 55 s
SINTEF NBL — 120 s ——
Average — 100 s ——
NOTE The results of SP are different because of difference in glueing.
7.3 ASTM round robin
During and after the publication of ISO 9705:1993, a second major round robin was conducted as a joint activity
between ISO and ASTM. This study involved 12 laboratories throughout the world and seven lining products. The
scenario for this round robin differed substantially from the European round robin. In the ASTM round robin, only
the walls were covered with the testing material and the alternative ignition source was used with a different heat
source programme than in the European round robin (40 kW to 160 kW). The measurements of heat release rate,
room and doorway temperatures and floor heat flux showed the best results. Smoke measurements had more
variations. As with all fire tests, the performance of the material, such as melting, delamination, etc., tended to
influence the spread of the results. Overall repeatability levels according to ISO 5725 varied between 27 % and
33 % and overall reproducibility are varying between 29 % and 41 %. In terms of overall material performance, the
round robin was successful. All materials that did not go to flashover performed the same in all tests at all
laboratories. The same is valid for the materials which went to flashover. Therefore, attainment of flashover or
reaching an HRR of 1 MW could be used as a criterion when the test is to be used for regulatory purposes.
8 Other test protocols using similar equipment
Several test protocols similar to ISO 9705:1993 or using the same test equipment, have been introduced in either
national standards or in test programmes and for products other than wall linings. An overview of the different
procedures is given in annex D.
16 © ISO 2001 – All rights reserved

9 Specimen mounting
In the test set-up of ISO 9705:1993, it is advised that both wall and ceiling linings be covered with material. Tests
where only the walls are covered can be performed, but it should be noted that this is not the standard specimen
configuration.
ISO 9705 requires that the mounting which is used in practice should be followed as much as possible. This is
given in detail in clause 11 of ISO 9705:1993. When the material is mounted with an air gap, this air gap may be
achieved by a steel framework on which the material is mounted.
The mounting method most frequently used in practice should be used in the text procedure. When this is not the
case, it should be clearly stated and reasons for alternative mounting should be explained.
Annex A
Calculation of HRR by means of different gas analysis data
A.1 General equations
A.1.1 Only O is measured
In this case all water vapour (by a cooling unit and a moisture sorbent) and CO (by a chemical sorbent) must be
removed from the sample stream before O is measured. This leads to the assumption that the sample gas only
consists of O and N . This is approximately true provided CO production is negligible. The mole fraction of O in
2 2 2
the oxygen analyser prior to a test can be written as
o
�
m
O
o M
O
A
= (A.1)
X
O
2 o o
� �
m m
O N
+
MM
ON
2 2
where
o
A
is the measured mole fraction of O ;
X 2
O
o
is the mass flow rate of O in the incoming air (kg/s);
�
m 2
O
is the molecular mass of O (approx. 32 kg/kmol);
M 2
O
is the molecular mass of N (approx. 28 kg/kmol);
M 2
N
o
is the mass flow rate of N in the incoming air (kg/s).
�
m 2
N
Equation (A.1) depends on the assumption that prior to the test the ratio of the mass flow of O to the mass flow of N
2 2
is the same in the oxygen analyser as it is in the incoming air. It is then assumed that the composition of the incoming
air does not change during a test.
Likewise, the mole fraction of O in the analyser during the test is given by
�
m
O
M
O
A
= (A.2)
X
O
m� m�
O N
+
MM
O N
2 2
where it is again assumed that the ratio of the mass flow of N and O are the same as they are in the exhaust duct.
2 2
18 © ISO 2001 – All rights reserved

o
The nomenclature is similar to that of equation (A.1) with the exception that superscript is omitted as everything
relates to the exhaust gases rather than to the incoming air. By rearranging equations (A.1) and (A.2) and using the
oxygen principle, the rate of heat release is given by:
o
��
A A
�
XX
M
OO O
��22 2��
oo
q=� E 1�� (A.3)
m� XX
a
O CO
��A � H 2�
1� M
X a
O
��
��
where
�
q is the rate of heat release (kW);
E is the heat released per O consumed (13,1 MJ/kg of O );
2 2
M is the molecular mass of the incoming air (kg/kmol);
a
m� is the mass flow rate of the incoming air (kg/s);
a
o
is themolefraction of H O in the incoming air;
X
HO
o
is themolefraction of CO in the incoming air.
X 2
CO
o
is not measured. However, it is assumed that ambient air is drawn in for
If no CO analyser is present,
2 X
CO
o
is then approximately constant and negligible (approx. 330 ppm).
combustion.
X
CO
It is furthermore assumed that even if no water vapour analyser is present, moisture content of the incoming air is
known. If the air is at temperature T , pressure p and has a relative humidity of RH %, the mole fraction of water
a a
vapour in the incoming air follows from
p()
RH T
a
o s
X= � (A.4)
HO
100 p
a
with
RH is the relative humidity (%);
p (T ) is the saturation pressure of water vapour at T (Pa);
a
s a
T is the air temperature (K);
a
p is the air pressure (Pa).
a
The saturation pressure p as a function of� is shown in Figure A.1. A curve fit in the range of 0 °C< T <0 °Cisalso
s a a
given. The fit has the functional form of a solution to the Clausius-Clapeyron equation suggested by Antoine back in
1888.
In (p )= C � C /(C + T); C =23,2; C = 3 816; C = – 46
s 0 1 2 0 1 2
Figure A.1 — Saturation pressure as a function of temperature
o
can also be determined experimentally by temporarily by-passing the water vapour trap prior to a test.
X
HO
Oxygen concentration measured with and without the trap can then be substituted into the following equation:
�� o ��
AA
=X1� (A.5)
XX
��
OHOO
��222 ��
�� ��
without trap with trap
o
Once X is known, the molecular mass of the incoming air follows from:
HO
oo
=X(1� )+M X (A.6)
MM
adry HO
HO H O
where
M is the molecular mass of dry air (approx. 29 kg/kmol);
dry
M is the molecular mass of H O (approx. 18 kg/kmol).
H O
Unfortunately, in an open system, it is the flow rate in the exhaust duct that is measured, and not the incoming
airflow rate. In order to find a relation between m� and m� , the oxygen depletion factor � is defined as
a e
o
��–
mm
O
O 2
� = (A.7)
o
m�
O
The oxygen depletion factor � is the fraction of the incoming air which is fully depleted of its oxygen. Via equations
(A.1) and (A.2), � canberewritten as
20 © ISO 2001 – All rights reserved

o
A
A
�
XX
OO
� = (A.8)
o
A A
1�
XX
��
OO
Due to the combustion chemistry, the number of moles in the fraction of the air fully depleted of its oxygen is
replaced by an equal or larger number of moles of combustion products (including N ) in the exhaust flow. Defining
the expansion factor � as the ratio of the two above-mentioned molar quantities, an expression linking m� to m� is
a e
given by
�� �
mm m
ea a
=+(1���) � (A.9)
MM M
ea a
With M � M (in absence of anything better) this can be simplified to
e a
�
m
e
= (A.10)
m�
a
1(+��1� )
As the composition of the fuel is usually not known, some average value has to be used for �. Complete
combustion of pure carbon in dry air results in � = 1. If the fuel is pure hydrogen, � is equal to 1,21. A
recommended average value for � is 1,105, which is correct for methane. The final equation for this case now
follows from equations (A.3), (A.8) and (A.10):
M o
�
O
2 oo A
q=� E 1––X X X (A.11)
�
m
e�� O
HO CO
1(+�� – 1)
M
a
Equation (A.11) is expected to be accurate to within � 10 % provided combustion is complete, i.e., all of the carbon
is converted to CO . The errors will be larger if CO or soot production is considerable or if a significant amount of
the combustion products is other than CO and H O (e.g. HCl).
2 2
A.1.2 O and CO are measured
2 2
This case is similar to that covered in the previous subclause. It is now assumed that only water vapour is trapped
before the sample reaches the gas analysers. Again, the equations are derived on the basis of conservation of
o
nitrogen. The relation between , is now given by
� � �
m m m
O O N
2 2 2
o
A
X
M
o OO
�
m= (A.12)
m�
N
O oo
AA
M
1–– N
XX
OCO
A
X
M
...