Problemy Kolejnictwa – Zeszyt 171 (czerwiec 2016) 27

Experience Gained from Fire Tests According to EN 45 545-2

and DIN 5510-2 for Testing of Seats
Dieter HOHENWARTER1

Summary
Results of fire tests according to ISO 5658-2 and ISO 5660-1 are shown for different products. The results of GFK products
are extensively analysed. The result of a body shell painting is analysed according to ISO 5658-2 and ISO 5660-1, but this
painting does not have to be analysed according to DIN 5510-2. Measurements of heat release rate of tram seats show that
it is difficult to fulfill the requirements of fire behavior and at the same time to fulfill the demands for comfort and me-
chanical requirements of tram seats. Different possibilities to improve the fire behaviour of railway coaches are presented
as well as the fact that the combination of the foam and the textile covering is essential.

Keywords: fire tests of railway materials and seats, fire tests of tram seats

1. Introduction to Fire Tests according (ofi) in Vienna. Concerning the tests of ISO 5658-2
to EN 45 545-2 (ISO 5658-2:2006, and ISO 5660-1 we are an approved fire laboratory by
ISO 5660-1:2015) CERTIFER (Railway Certification Agency) allowed to
perform tests according to EN 45 545-2.
To fulfill the requirements of EN 45 545-2 it is nec-
In the past, materials were investigated ac- essary to use flame retardant materials as an additive
cording to the old Austrian Standard VORNORM for the different plastic materials. Also it is common
ÖNORM B 3800-2: 1988 regarding the designa- to use intumescent coating of materials to fulfill the
tion low combustibility (combustibility class B1) criteria of the different standards [2].
and moderately combustible (combustibility class
B2) including the characteristics such as smoke 1.1. Fire Test according to ISO 5658-2:2006
production and droplet formation at the TGM fire
testing laboratory. The testing according to ISO 5658-2 specifies
After the catastrophic fire of the Kaprun glacier a method of testing for measuring the lateral spread
funicular 2 in 2000, the work environment and the of flame along the surface of a specimen of a prod-
fire safety investigation of constituents of coaches uct orientated in the vertical position. It provides
was substantially expanded. It liaises with the fact data suitable for comparing the performance of es-
that all interior fittings of coaches of the ÖBB were sentially flat materials, composites or assembly that
fire-safety tested approximately in accordance with are used primarily as the exposed surface of walls
DIN 5510-2. Shortly after the incident of Kaprun, in buildings or transport vehicles, such as ships and
all components were tested and found to corre- trains.
spond to the interior fittings of coaches ÖBB the A test specimen (800 mm × 155 mm) is placed
fire safety requirements approximately according to in a vertical position adjacent to a gas-fired radiant
DIN 5510-2 [1]. panel (Fig. 1) where it is exposed to a defined field of
The fire testing of materials according to EN 45 radiant heat flux for a time of 10 minutes. Following
545-2:2016 is done according to ISO 5658-2:2006 and ignition, any flame front that develops is noted and
ISO 5660-1:2015 and is carried out at the fire testing the results are expressed in terms of flame spread
laboratory. The toxicity tests and smoke tests accord- distance and at least the critical heat flux at extin-
ing to ISO 5659-2:2013 are done by another institute guishment (CFE).

1
Prof. Dipl. Ing. Dr.; Federal Institute of Technology (TGM), Plastics Technology and Environmental Engineering, Austria; e-mail:
[email protected].
28 Hohenwarter D.

Fig. 1.

The correlation between flame spread distance and

CFE is shown in the diagram (Fig. 2).
Fig. 3.

Fig. 2.

Fig. 4. The calculation of MARHE from the measured heat

release rate (HRR measured in kW/m²)
1.2. Fire Test according to ISO 5660-1:2015

The test according to ISO 5658-2 specifies

a  method for assessing the heat release rate of
2. Examples of Fire Tests according to
a  specimen exposed in the horizontal orientation ISO 5658-2 and ISO 5660-1
to controlled levels of irradiance with an external
igniter. The test method is based on the observa- In the following a  number of examples of
tion that, generally, the net heat of combustion is tests according to the mentioned standards are
proportional to the amount of oxygen required for presented. Fire tests of materials according to
combustion. EN 45 545-2 are carried out according to ISO 5658-2
Approximately 13.1 MJ of heat are released per and ISO 5660-1.
kilogram of oxygen consumed. During the test speci-
mens (Fig. 3) are burned under ambient air conditions 2.1. GFK Material 4 mm with glas fibre
(radiation in the range of 0 to 100 kW/m²) and at the + polyester resin
same time oxygen concentration and exhaust gas flow
rates are measured. Using these measurements the The testing of this GFK Material shows a signifi-
Average Rate of Heat Emission (ARHE) and then the cant difference of the heat release rate measured ac-
Maximum Average Rate Heat Emission MARHE are cording to ISO 5660-1 between the outside and the
calculated. The principle of the MARHE Calculation rear side of the material. The outside shows a signifi-
is shown in the diagram (Fig. 4). cant influence of the painting and gel coat while the
Experience Gained from Fire Tests According to EN 45 545-2 and DIN 5510-2 for Testing of Seats 29

heat release rate of the rear side is influenced by the

top coat. In the following the results of 3 measure-
ments per sample Fig. 5, 6 are shown.

Fig. 8. GFK grey inside CFE = 22 kW/m²

Fig. 5. GFK outside (green painted front side) Gel coat MAHRE
62 kW/m²
2.2. Material named GFK Pultrudat thickness
2.9 mm
The testing of this GFK material was made for re-
search purposes.
 Laminate,
 Reinforcement glass fiber roving – E-glass,
 Glass fiber mats,
 Matrix unsaturated polyester resin with Lp for
high surface quality halogenfree,
 Surface polyester veil on visual outside of the pro-
file
 Fiber weight content 48% tolerance ± 5%,
Fig. 6. GFK rear side (grey color)Top coat MAHRE 78 kW/m²  Profiles were painted.

The Figure 9 shows the GFK material before and

The spread of flame testing according to ISO 5658-2 after the Cone test with the different layers of the glass
of this GFK material shows no significant difference fibre roving.
between both of the different sides of the material
(Fig. 7, 8).

Fig. 7. GFK green outside CFE = 22 kW/m² Fig. 9.

30 Hohenwarter D.

This material was radiated with different radiation The material ignites approximately at a temperatu-
intensities to study the heat release rate of this GFK re of 340°C.
material. It is clear that another heat radiation intensi-
ty yields to a different time when the material ignites.
The time to ignite the material depends on the heat
radiation intensity which is analyzed later.
The results in Figure 10 show that, in general,
shortly after the ignition at higher radiation intensi-
ties, a large amount of heat is released. Upon irradia-
tion with 75 kW/m² the highest amount of heat (HRR
peak) is released and on exposure to 100 kW/m²
a smaller amount of heat (HRR peak) than at 75 kW/m²
is measured, as can be seen in the following table. This
can be explained as follows: Upon irradiation with
100 kW/m² the heating leads to (short) strong fumes
(outgassing) almost simultaneously with the ignition
of the material and thus the heat release is reduced.
Upon irradiation with 75 kW/m², however, a  rather Fig. 11.
uniform smoke formation takes place as may be seen
from the figure of the smoke production rate depen-
dent on the heat radiation intensity.

Fig. 12.

In Table 1 the results of measurements with a Cone

Fig. 10. Heat release rate (HRR) dependent on the radiation calorimeter of a  GFK product at different radiation
intensity intensities are presented.

Table 1
Likewise, the respectively extinction-coefficient
HRR peak

EHC [MJ/
Radiation

and the temperature of the surface were measured

ignite [s]
[kW/m²]

[kW/m²]

[kW/m²]
intensity

MARHE

Time [s]
Burning
Time to
[m²/s]
SPR
kg]

during the tests with different heat radiation intensi-

ties. The extinction coefficient is proportional to the
smoke production.
What is noticeable about this GFK – product is 25 67 133 20.0 0.0083 143 395
that, strong fumes (outgassing) form up at the begin- 35 81 132 19.6 0.0091 82 363
ning even before ignition. The outgassing is caused
by low molecular weight molecule fractions which 50 104 181 19,1 0.0119 48 316
evaporate before. This brief strong degassing takes 75 145 263 19.4 0.0179 23 335
place, especially in the middle of the plate so that the
specimen is deformed (just before the inflammation) 100 137 227 18.7 0.0225 18 284
in a small area and a noticeable sound is heard as well.
Figures 11, 12 show the measurement of the Tempera- Where:
ture °C: blue line, Extinction coefficient 1/m (Smoke HRR – Heat release rate (kW/m²),
production) black line, Heat release rate HRR kW/m² HRR peak – Peak heat release rate (kW/m²),
red line dependent on time and is shown for 2 differ- EHC – Effective heat of combustion per mass (MJ/kg),
ent radiation intensities. SPR – Smoke production rate (m²/s).
Experience Gained from Fire Tests According to EN 45 545-2 and DIN 5510-2 for Testing of Seats 31

The time until the ignition of the material depends other radiation intensities a more uniform smoke pro-
on the radiation intensity which is shown in the table duction can be detected during a longer time range.
above. When radiated with high intensity, the mate-
rial ignites very quickly, while the exposure to low-
intensity takes longer to achieve ignition. In the Fi-
gure 13 this relationship is shown graphically. From
the literature [3] the following relationship results
2
 Tignition  Tambient 
tignition  C k   c  
 Heat Flux
 

Where:
C ‒ constant,
k ‒ thermal conductivity, Fig. 15.
ρ ‒ density,
c ‒ specific heat, With the results of Table 1 the smoke production
Tignition ‒ temperature at ignition, rate (SPR) increases approximately proportional to
Tambient ‒ ambient temperature, the intensity of irradiation which Figure 16 shows.
Heat Flux – radiation intensity.

The transformation of the equation above, for the

relationship between the radiation intensity and the
square root of the inverse tignition (time to ignition) re-
sults in a linear relationship. In the diagram (Fig. 14)
on the right side this relationship is shown and it
shows up as the theoretically expected linear depen-
dence.
Fig. 16.

2.3. Heat release rate of PMMA dependent on

the radiation intensity
Tests of a  PMMA material were carried out for
own research purposes. Figure 17 shows the heat
release rate for different radiation intensities in the
range of 5 kW/m² to 100 kW/m². These different ra-
diation intensities which ignite the material occur at
Fig. 13. different time spans.

Fig. 14.

The time to ignite is dependent on radiation inten-

sity and the radiation intensity depends nearly linear
on the parameter 1/√tignition.
Figure 15 shows that during a short time relatively
strong fume (extinction-coefficient) is evident espe-
cially when radiated with 100 kW/m², whereas at the Fig. 17.
32 Hohenwarter D.

From the literature [3] the following relationship quirement R7. The fulfillment of the requirements of
is used to show the relation between the time to ignite EN 45545-2 must be demonstrated by the manufac-
(tignition) and the heat flux which correspond to the heat turer of such coatings.
radiation intensity One of these test results are shown below: The
test of this Sandwich door leaf was commissioned by
2
 Tignition  Tambient  Knorr-Bremse GmbH, IFE Kematen Division Door
tignition  C k   c   Systems (Fig. 20). The sandwich door consists of the
 Heat Flux
  following components:
 Coated, faced aluminium honeycomb product
Where: 19.8 kg/m²,
C ‒ constant,  Final coating: 2K HS Clear coat 50 μm,
k ‒ thermal conductivity,  Second coating: Basecoat 10–20 μm,
ρ ‒ density,  First coating: Water based 2-K Epoxy Primer 40 μm,
c ‒ specific heat,  Facing: Aluminium coil coated 40 μm,
Tignition ‒ temperature at ignition,  Honeycomb: 0,029 kg/dm³, thickness 30 mm, cell
Tambient ‒ ambient temperature, diameter 19 mm, wall thickness 0.071 mm,
Heat Flux – radiation intensity.  Adhesive: Epoxy adhesive 0.2 mm,
 Facing: Aluminium 1mm.
As expected from the equation there is (nearly)
a linear correlation between the time to ignite and the
square of the radiation intensity (Fig. 18, 19).

Fig. 18.

Fig. 19.

2.3. Sandwich door leaf for railway application

According to DIN 5510-2: 2009 metals and al-
loys with organic coatings with a  nominal thick-
ness  <0.3  mm are considered as components with
proven requirements and therefore they do not need
to be tested (DIN 5510-2, Chapter 5.4).
According to EN 45545-2: 2016, the walls of exter-
nal body shells (including painting / coating systems, Fig. 20. Test material for the cone test: a) with honeycomb
films and windows) must be checked according to re- structure: b) and for spread of flame test closed structure
Experience Gained from Fire Tests According to EN 45 545-2 and DIN 5510-2 for Testing of Seats 33

For the cone test the honeycomb was tested with

a frame shown in the Fig. 21, 22.

Fig. 21. Heat release rate ISO 5660-1:2002; MARHE: < 50 kW/m²

Many thanks to Knorr-Bremse GmbH Division IFE

Automatic Door Systems for the permission to publish
the above mentioned results.

Fig. 22. Lateral spread of flame ISO 5658-2; CFE >20 kW/m²
34 Hohenwarter D.

3. Combustibility of tram seats according In the following an example of a plywood backrest

EN 45 545-2 is presented (Fig. 23). It was measured with a radia-
tion intensity of 35 kW/m²:
Seats of trams includes most times a hard cover seat In the following an example of a plywood seat of
shell. That is the reason that tram seats are tested ac- an old tram backrest is presented (Fig. 24). It was
cording requirement R6 of EN 45 545-2:2016 (Table 2). measured with a radiation intensity of 35 kW/m²:

Table 2

Fig. 23. Heat release rate ISO 5660-1:2015; Radiation 35 kW/m², MARHE=115 kW/m²

Fig. 24. Heat release rate ISO 5660-1:2015; Radiation 35 kW/m², MARHE=206 kW/m²
Experience Gained from Fire Tests According to EN 45 545-2 and DIN 5510-2 for Testing of Seats 35

Modern examples of tram seats show that it is dif-

ficult to reach MAHRE < 90 kW/m². This is necessary
because trams sometimes drive through tunnels.
That means, with other words that the fire behav-
ior requirements for trams seats are too high and it
is difficult to fulfill the requirements with seat mate-
rials which are comfortable and fulfill also the me-
chanical requirements.

4. Combustibility of seats according to

DIN 5510-2
The fire performance of seats for railway cars is tested
using original complete seats according to DIN 5510-2.

(Draft) E DIN 5510-2:1996

Upholstered seats must be self-extinguishing, even
if the seat covering is slit open (vandalised). A 100 g
paper cushion is positioned on the seat base (Fig. 25). Fig. 26. Seat without flame retardant covering of the foam; Fire
extinguisher was used after 13 minutes

Fig. 25.

Ventilation system: The hood can be operated so

that the smoke produced during the experiment can
be just sucked off.

DIN 5510-2: 2009

Ventilation system: The volume flow should be in the
range of 0.5 to 0.7 m³/s. The ventilation system is calibrat- Fig. 27. Seat with flame retardant covering of the foam
ed with n-Heptan. The smoke production rate should be
calculated from the measured extinction coefficient. 4.2. Comparison (according to DIN 5510-2)
of one covering material with different
4.1. ÖBB seats without and with flame foams (ÖBB)
retardant covering of the foam
These tests to analyse which foam is compatible
The fire testing of the seats from the Austrian Rail- with the standard textile covering material used by the
way coaches was carried out for the ÖBB in the year ÖBB, were made in the year 2005. In order to analyze
2008 and shows that in this case a flame retardant cov- the behaviour of the foam the textile covering material
ering was necessary to protect the foam (Fig. 26, 27). was cut (vandalised). Textile covering material: Pro-
36 Hohenwarter D.

ducer Kneitz, velour chess 85% wool, 15% polyamide The sequence of the products mentioned in the
(Velourstoff Schach 85% Wolle, 15% Polyamid). table above does not correspond with the sequence of
The foams were tested (Table 3): the pictures of the seats in the following two picture
lines (Fig. 28, 29).
Table 3
The result of this research is: one combination of
Producer Name of the foam fabric upholstery (textile covering) with foam leads
Purtec Purtec 500 to a  full-scale fire while a  different combination of
foam and textile covering extinguishes after 4 about
Metzler Metzoprotect Qualität FRM-U FT3307 minutes.
Weserland Weserland Qualität FSF 710 RA Many thanks to ÖBB-Technische Services, Flotte-
Weserland Weserland PU Muster W 5662/1 nengineering Nahverkehrs-Reisezugwagen (Manfred
Schorm), St. Pölten for the permission to publish the
Eurofoam LS. NR. 04004452 Qualität KF 55 B1 weiß above mentioned results.

Fig. 28: a) Fire extinguisher used after10 Minutes; b) Self-extinguishing after 4:25 Minutes; c) Self-extinguishing after 6:48 Minutes

Fig. 29: a) Self-extinguishing after 4:00 minutes; b) Self-extinguishing after 4:57 minutes
Experience Gained from Fire Tests According to EN 45 545-2 and DIN 5510-2 for Testing of Seats 37

5. Conclusion Literature
1. Hohenwarter D.: Brandschutztechnische Prüfungen
5.1. Fire tests of materials according to für die Reisezugwagen der ÖBB, ZEVrail, Glasers
ISO 5658-2 and ISO 5660-1 Annalen 129, Tagungsband Moderne Schienen-
fahrzeuge 2005, p. 136‒145 (with Edith Grüner).
The test of a special GFK material shows a signifi- 2. Hohenwarter D.: Brandschutz für brennbare Ma-
cant difference of the heat release rate measured ac- terialien, Fire Safety Engineering (FSE), Brand-
cording to ISO 5660-1 between the painted side and schutz-Fachtagung, Fachhochschule St. Pölten,
the rear side of the material. Februar 2010, p. 220‒233 (with Helmut Richter).
Another GFK material was radiated with different 3. Quintiere J.G.: Principles of Fire Behavior, Delmar
radiation intensities which lead to different ignition Cengage Learning, 1998.
times. This correlation fulfills the equation of the the-
ory of fire behaviour for the ignition of thick materi- Standards
als. The tests for PMMA show also a good correlation 1. VORNORM ÖNORM B 3800-2: Behavior of
between measurements and (fire) theory. building materials and components in fire; build-
According to DIN 5510-2: 2009 metals and alloys ing materials: requirements and tests; 1988.
with organic coatings with a  nominal thickness of 2. EN 45 545-2: Railway applications – Fire protec-
<0.3 mm are considered as components with proven tion on railway vehicles, Part 2: Requirements for
requirements and therefore do not need to be tested. fire behavior of material and components, 2016.
But according to EN 45545-2: 2013 the walls of exter- 3. ISO 5658-2: Reaction to fire tests – Spread of flame
nal body shells (including painting/coating systems, – part 2: Lateral spread on building and transport
films and windows) must be checked according to re- products in vertical configuration, 2006.
quirement R7. As an example the results of one paint- 4. ISO 5660-1: Reaction-to-fire tests – Heat release,
ing are shown. smoke production and mass loss rate – Part 1: Heat
Measurements of heat release rate of tram seats release rate (cone calorimeter method) and smoke
shows that it is difficult to meet the requirements of production rate (dynamic measurement), 2015.
fire behaviour and at the same time to fulfill the de- 5. ISO 5659-2: Plastics – smoke generation – Part 2:
mands for comfort and mechanical requirements. Determination of optical density by a single-cham-
ber test, 2013.
6. DIN 5510-2: Preventive fire protection in railway
5.2. Fire behaviour of seats vehicles – Part 2: Fire behavior and fire side effects
of materials and parts – Classification, require-
Fire behaviour of seats depends on: ments and test methods, 2009.
 Density of the foam: Density 95 kg/m³ is better
than usual 85 kg/m³, however foams with 75 kg/m³
can also fulfill the requirements.
 Flame retardant fibres „glued” on the foam can
improve the fire behaviour of the seat.
 Combination of the textile covering and the
foam:
 One textile  /  foam combination leads to full
fire, while with another combination stops the
fire after 4 Minutes.
38 Hohenwarter D.

Doświadczenia zdobyte w testach palnościowych siedzeń przeprowadzonych zgodnie

z normami EN 45 545-2 i DIN 5510-2
Streszczenie
W artykule przedstawiono wyniki badań palności różnych produktów wykonanych zgodnie z  normami
ISO 5658-2 i ISO 5660-1. Szczegółowo przeanalizowano wyniki badań GFK (kompozyty poliestrowo-szklane).
Stosując normy ISO 5658-2 i ISO 5660-1, poddano analizie zgodność z normami malowanie korpusów siedzeń,
chociaż ta zgodność nie musi być analizowana według wymogów niemieckiej normy DIN 5510-2. Zaprezento-
wano różne możliwości zmniejszenia palności wagonów kolejowych, stwierdzając, że połączenie pianki z po-
kryciem tekstylnym ma zasadnicze znaczenie.

Słowa kluczowe: testy palnościowe materiałów i siedzeń używanych na kolei, testy palnościowe siedzeń tram-
wajowych

Опыт полученный во время испытания горючести сидений в соответствии

с нормами EN 45 545-2 и DIN 5510-2
Резюме
Результаты испытаний горючести проведенных в соответствии с нормами ISO 5658-2 и ISO 5660-1
были представлены для разных изделий. Были глубоко проанализированы результаты для изделий из
стеклопластика. Анализу в соответствии с нормами ISO 5658-2 и ISO 5660-1 подвергало также покры-
тие корпусов сидений, хотя их соответствие не должно быть анализировано в соответствии с требова-
ниями немецкой нормы DIN 5510-2. Измерения скорости выделения теплоты трамвайных сидений по-
казывают, что тяжело выполнить одновременно правильные пожарные требования, как и потребность
в комфорте и механические требования для этого типа сидений. Были показаны разные возможности
уменьшить горючесть железнодорожных вагонов, а также факт, что соединение пенопласта с текстиль-
ным покрытием имеет главное значение в этой области.

Ключевые слова: испытания горючести материалов и сидений используемых в железнодорожным

транспорте, испытания горючести трамвайных сидений