Fire safe upholstered furniture
Alternative strategies to the use of chemical flame retardants
Karolina Storesund, Anne Steen-Hansen, Anna Bergstrand
SP F
ire
Rese
a
Fire safe upholstered furniture
Alternative strategies to the use of chemical flame retardants
VERSION 1 DATE 2015-12-18 KEYWORDS: Fire safety Upholstered furniture Flame retardants AUTHORS
Karolina Storesund, Anne Steen-Hansen, Anna Bergstrand
CLIENT
Swedish Contingencies Agency (Myndigheten för samhällsskydd och beredskap, MSB) CLIENT’S REF. Per Karlsson PROJECT NO. 20124 NUMBER OF PAGES/APPENDICES: 46 pages + 3 Appendices ABSTRACT
It is well known that upholstered furniture represents a fire risk due to the fact that it is composed of relatively large amounts of easily ignited and very combustible materials. The fire properties are usually improved by adding chemical flame retardants to the upholstery materials.
The goal of this investigation is to demonstrate how sufficient fire safety in upholstered furniture may be achieved, without the use of flame retardant chemicals.
A number of cover materials in different combinations with other materials including wadding, barrier materials and foam, have been tested in small scale cone calorimeter tests and in mock-up chair tests. Time to ignition, heat release and smoke production of the different combinations have been examined.
It is shown that there certainly are possibility to improve these properties by means of alternative strategies other than by adding chemical flame retardants.
PREPARED BY Karolina Storesund SIGNATURE CHECKED BY Anne Steen-Hansen SIGNATURE APPROVED BY
Paul Halle Zahl Pedersen
Document history
VERSION DATE VERSION DESCRIPTION
Contents
Preface 5
Summary
6
1
Introduction
7
1.1 Background 7 1.2 Goal 7 1.3 Methods 72
Testing of upholstered furniture components and fire
barrier materials
9
2.1 Experience from CBUF (Combustion Behaviour of Upholstered
Furniture) 9
2.2 Experience from testing of fire barrier materials in the USA 9
3
Selection of materials
11
4
Cone calorimeter tests
14
4.1 Test set-up 14
4.2 Results 17
4.2.1 Time to ignition 18
4.2.2 Heat release 19
4.2.3 Smoke production 22
4.2.4 Assessment of the results 24
5
Mock-up tests
29
5.1 Test set-up 29
5.2 Results 30
5.2.1 Heat release 34
5.2.2 Smoke production 37
5.2.3 Assessment of the results 39
6
Discussion and conclusions
43
6.1 Effects of cover material and fire barrier 43
6.1.1 Even light barriers may have positive effects 43
6.1.2 Barriers can reduce the heat release 43
6.1.3 Barriers can reduce the smoke production 44
6.1.4 Predicting larger scale fire behaviour on the basis of small scale tests 44
6.1.5 Conclusions 44
6.2 Recommendations for further work 44
6.2.1 Other ways of improving fire safety of upholstered furniture 44 6.2.2 A system for fire classification of upholstered furniture is needed 45
6.2.3 Assessment and evaluation based on testing 45
References
46
Appendix A Results cone calorimeter tests
A.1 Time to ignition
A.3 Smoke production
Appendix B Results mock-up tests
B.1 Heat release rate
B.2 Smoke production
Appendix C Pictures from mock-up chair experiments
C.1 Flame application
Preface
This report is prepared by SP Fire Research on behalf of the Swedish Contingencies Agency (MSB).
The goal of the project has been to demonstrate alternative strategies to adding chemical flame retardants in order to improve the fire safety of upholstered furniture. Different materials, commonly used in furniture on the Scandinavian market, have been investigated.
We wish to thank those who have contributed to the project with advices, knowledge as well as materials.
Trondheim 2015-12-18
Karolina Storesund Project manager
Summary
Background
It is well known that upholstered furniture represents a fire risk due to the fact that it is composed of relatively large amounts of easily ignited and very combustible materials. The goal of this investigation is to demonstrate how sufficient fire safety in upholstered furniture may be achieved, without the use of flame retardant chemicals. It is the aim that the findings may form a basis for discussions concerning fire safety in upholstered furniture and alternative strategies to the use of chemical flame retardants.
Methods and materials
In order to investigate materials that are common on the Scandinavian market we have selected the materials to be tested in cooperation with suppliers.
One important limitation of our project is that flame retardants shall not be used in any of the components. This means that the materials shall not be chemically treated or
otherwise have chemical flame retardants added to them.
To be able to analyse the fire properties of material combinations the test method ISO 5660-1 (the cone calorimeter method) was used initially for small scale testing and the measuring equipment normally used for testing according to EN 13823 (the SBI test), was used for mock-up tests. In the mock-up tests both a small gas flame and crib 5 of BS 5852 were applied as ignition sources.
Results and conclusions
In this project we have included barrier fabrics as a means to improve the fire behaviour of the upholstered material combinations. The greatest influence of barriers is to separate the flaming surface from the flammable non-flame retardant foam. So even by using very light fabrics as barriers, in addition to the cover fabric, it is possible to improve the reaction-to-fire-properties of these material combinations.
Barriers can reduce the heat release, this is shown in both our cone calorimeter tests and mock-up chair tests. The effectiveness is higher for the denser fabrics, while the
effectiveness from the light barrier fabric is more dependent upon the nature of the cover fabric. This is also true with regards to smoke production.
It is certainly possible to improve the fire safety in upholstered furniture without the use of flame retardants, but it requires thorough examination and a proper choice of material combination. From a fire safety point of view, there already exist alternative strategies and materials that have the potential to improve fire safety in upholstered furniture without the use of chemical flame retardants. We have noted that there are many interesting materials available - mainly for completely other applications than for furniture - but that could be interesting to investigate further for furniture applications. Much can also be achieved through the possibilities offered by modification of design and construction parameters of the complete furniture. We recommend that these topics are further explored.
A voluntary fire classification system for upholstered furniture may be a tool that would give the purchasers a possibility to choose products that fits the desired fire safety level, and would be useful both for private households and for commercial and public
1
Introduction
1.1
Background
It is well known that upholstered furniture represents a fire risk due to the fact that it is composed of relatively large amounts of easily ignited and very combustible materials. In a fire, upholstered furniture contributes to a rapid fire development and in modern
dwellings the time to flashover has decreased significantly over the decades. A majority of the fire fatalities in western countries are related to dwelling fires.
If stricter requirements were to be made on upholstered furniture with regards to
flammability it would most likely lead to a greater use of flame retardants. There is a risk that increased use of chemicals will cause adverse health- or environmental effects. It is therefore a need to further study how the choice of non-flame retardant material
combinations can be used as a strategy for improving the fire safety of upholstered furniture.
1.2
Goal
The goal of this investigation is to demonstrate how sufficient fire safety in upholstered furniture may be achieved, without the use of flame retardant chemicals. It is the aim that the findings may form a basis for discussions concerning fire safety in upholstered furniture and alternative strategies to the use of chemical flame retardants.
1.3
Methods
In order to investigate materials that are common on the Scandinavian market we have selected the materials to be tested in cooperation with suppliers. More information of the selection of the materials follows in section 3 below.
One important limitation of our project is that flame retardants shall not be used in any of the components. This means that the materials shall not be chemically treated or
otherwise have chemical flame retardants added to them. One exception has been made, by including a cover fabric of Trevira TC polyester, an inherently flame retardant fibre. The aim has been to use materials that does not need to be added any treatment and use them in such a way that a satisfactory fire safety level is still achieved.
The information about the materials are based on specifications and information given by the suppliers. The chemical content of the materials has not been analysed in this project. To be able to analyse the fire properties of material combinations the test method ISO 5660-1 (the cone calorimeter method) [1] was used initially for small scale testing and the measuring equipment normally used for testing according to EN 13823 (the SBI test) [2], was used for mock-up tests. The mock-up tests were performed in order to expose a selection of the materials tested in the cone calorimeter to flaming ignition sources.
We have not tested the material combinations for smouldering fire (e.g. EN 1021-1[5]), as this was not a part of the project.
The mock-up test was an adjusted method were the exhaust hood and measuring
equipment according to EN 13823 [2] was used. Two ignition sources were used; match flame equivalent as described in European standard EN 1021-2 [3] and Crib 5 as
The tests were performed in SP Fire Research laboratories in Trondheim (cone calorimeter tests) and in Borås (mock-up tests).
The experiments and their results are described further in chapters 0 and 5. The scope of the testing has been relatively limited, and in order to include as many different material combinations as possible, the experiments with certain material combinations under specific test modes have been limited to only one. This will affect the statistical significance of the results and of our conclusions, but we are convinced that the results will still give a good indication on the relevant properties.
2
Testing of upholstered furniture
components and fire barrier materials
2.1
Experience from CBUF (Combustion Behaviour
of Upholstered Furniture)
CBUF (Combustion Behaviour of Upholstered Furniture) was a European fire research programme where the final report was published in 1995 [6]. The programme developed fire test procedures and fire models for the assessment of the burning behaviour of upholstered furniture and its components. The test methods used were based on ISO 5660 (cone calorimeter), ISO 9705 (room corner test) and NT Fire 032 (furniture calorimeter). CBUF also presented their findings concerning how different components (e.g. cover fabric, filling, interliners/barriers) and design can influence the fire behaviour.
The CBUF report offers important information about the test methods and the possibility to predict fire behaviour based on smaller scale tests. It is also an important source of information with regards to material selection and design strategies.
2.2
Experience from testing of fire barrier materials
in the USA
A report published by Underwriters Laboratories (UL) in 2013 [7] has been of great interest for this project as it shows how the cover, foam, polyester wadding and fire barrier interact. Our project has had a slightly different approach in that it attempts to demonstrate how different cover fabrics may assist to achieve an increased fire safety level. The UL study used one and the same cover fabric throughout the study. One
important difference is also the use of flame retardants. Our project has not used materials that are treated with flame retardants and hence we would not be able to use a number of the barriers tested in the UL study.
The development of fire barriers has been pushed forward largely due to the Californian regulations. Using fire barriers is a method of protecting the upholstery foam from heat or flame exposure. A recent study from Underwriter's laboratory Inc. (UL) explored
"whether commercially available flame retardant treated foam and fire barrier technology can retard and/or reduce the fire growth rate of upholstered furniture when exposed to small open flames" [7]. The approach of the study was to use commercially available cover fabrics, foams and fire barriers and both flame retardant and non-flame retardant polyurethane foam were used in the study. A number of different fire barrier materials were studied, ranging in fibre composition of the barrier and whether or not the barrier was chemically flame retardant. Experiments were conducted both with and without the use of polyester wadding, a common feature in upholstered furniture. The same polyester microsuede was used as a cover fabric throughout the study. This was chosen because it was a common cover fabric sold from the largest cover fabric supplier in the United States.
The different materials of the study were characterised with regards to a number of properties such as density and thickness, but also through chemical analyses.
material combination was tested with three single tests. Variables like time to ignition, re-ignition time, flame duration, weight loss, peak- and average heat release rates, effective heat of combustion, total generated heat were measured.
Medium scale tests – mock-up assembly
A three sided mock-up assembly (305 mm x 305 mm) bottom cushion with two side cushions (229 mm x 305 mm and 229 mm x 229 mm) was used on an expanded steel mesh frame and placed on load cell. The ignition source was a match flame equivalent (BS 5852 source 1 - gas flame). Three single tests of the same material combination were performed with the initial ignition time of 20 seconds. Combustion products were
collected by a hood and exhaust system in order to measure exhaust flow and oxygen concentration. The variables time to ignition, weight loss, peak heat release rate, total heat release and effective heat of combustion were measured.
Full-scale furniture tests
Four material combinations experiments in a single seat upholstered chair were tested. The selection was based on previous results from medium scale tests. There were three different ignition locations used in the experiments. The ignition source was a match flame equivalent (BS 5852 source 1 - gas flame) applied for 20 seconds.
Some results and findings relevant for our project
The cone calorimeter tests showed that the incorporation of a barrier material may be more effective at prolonging the time to ignition and to reduce the heat release rate than using a flame retardant foam instead of a flame retardant free foam.
Different types of physical interference may enhance the fire properties of upholstered furniture:
1. Thermal barrier that limits the heat transfer from the ignition source to the foam. 2. Physical barrier that limits the transport of volatile gases from the foam to the
surface.
3. Increased thermal mass of the material prolongs the time to heat up the material and to start the pyrolysis.
3
Selection of materials
Irrespective of any possible adverse effects of chemical flame retardants on health and environment there are other incentives to avoid using foam that is flame retardant. One common flame retardant upholstery is CMHR foam (Combustion Modified High Resilience Foam), which is a flexible polyurethane foam that has been added flame retardants to improve its flammability [9]. However, we have been informed from producers that these types of foams does not offer the required comfort in some applications.
The use of fire barriers in upholstered furniture does not appear to be a common production method in Scandinavia yet, and publications in this area are mainly found in USA. The article "A review of fire blocking technologies for soft furnishings" [8] gives examples of different types of barrier fabrics that are in use and are available on the American market. These examples include materials both with and without flame retardants. Glass fibre fabric is one example of an inherently fire resistant material. Its effectiveness lies in preventing an ignition source from reaching the flammable upholstery. Its disadvantage is a relatively poor durability and resiliency. In our experiments we have used light glass fibre scrims of 25 and 80 g/m2 respectively, but these are also manufactured in considerably denser qualities. There are also methods of improving the durability of the glass fibre, e.g. core spun yarn, where polyester fibres are protecting a core of glass filament.
Adding a barrier introduces one more material and hence increases the labour involved in production. However, this can partly be reduced by using composites or laminated fabrics.
The starting point for the selection of the materials for the tests was to include common textiles from the perspective of a supplier. We wanted to test different aspects of the textiles with regards to:
Fibre composition
Tightness of the weave
Area density
The premises for the project were to not use any chemical flame retardants. There are many types of barrier materials on the market, especially on the American market. However, many of these are based on addition of a chemical flame retardant, and was therefore not relevant to use in this project. One exception was made. In the selection of cover materials, a textile made of Trevira CS, an inherent flame retardant polyester fibre, was used. The flame retardancy is built into the fibre and the fabric is thereby not chemically treated .
The selection of materials was limited due to the number of tests that was planned within the framework of the project. But the aim was to select materials and material
combinations that could be used to improve the following parameters:
Time to ignition
Heat release rate (HRR)
Total heat release (THR)
Rate of smoke production (RSP)
the fire development. By using furniture that do not contribute significantly to the heat development and smoke production, the time available for evacuation in a fire may be prolonged.
Better reaction-to-fire properties can be achieved by choosing materials that may
act as a thermal barrier - limiting heat transfer
act as a physical barrier - limiting transport of volatile gases from the foam
have an increased thermal mass - prolonging the time to heat up and pyrolyse the material
The materials to be tested were selected to cover a range of materials common in upholstered furniture on the Scandinavian market.
The same foam was used in all the experiments. This was a non-flame retardant polyurethane foam with a nominal density of 35 kg/m3.
Two types of polyester wadding was used, one type in the cone calorimeter tests (W1), and one type in the model chair tests (W2). These were different for practical reasons and came from different suppliers, however their properties were corresponding. Wadding was not used in all tests. It was shown in the cone calorimeter tests that the wadding had little influence on the results, and for practical reasons it was decided to reduce the number of specimens with wadding in the mock-up tests.
The different materials and their identification used in the tests are presented in Table 3-1 through Table 3-4. In the presentation of the test results the test samples are identified by the composition. I.e. "C1_W1_F3" means cover 1 (plain weave of 83 % cotton, 9 % modal, 8 % polyester) on top of foam 3, with wadding 1 in the middle.
A few of the materials that were received in the project have been deselected from the study. However they have been registered, but not included in the following tables, and therefore results from testing of C7, C8 and C9 are not reported.
Table 3-1 Cover material identification and specification ID Description Composition [%] Nominal area
density [g/m2]
Measured area density [g/m2]
C1 Plain weave with pile 83 % cotton 9 % modal 8 % polyester
460 435
C2 2/2 plain weave 54 % cotton 46 % viscose
519 568
C3 Plain weave 100 % polyester,
Trevira CS
250 262
C4 2/2 plain weave with fleece backing
100 % polyester 430 515
C5 Felted plain weave 70 % wool 25 % polyester
400 386
C6 Artificial leather PVC coating on polyester cotton/jersey
610 ±50 648
Table 3-2 Wadding material identification and specification ID Description Composition [%] Nominal area
density [g/m2]
W1 Heat set polyester wadding
100 % polyester 200
W2 Polyester wadding 100 % polyester 200
Table 3-3 Barrier material identification and specification ID Description Composition [%] Nominal area
density [g/m2]
B1 Glass fibre plain weave scrim
100 % glass fibre 25
B2 Glass fibre twill scrim 100 % glass fibre 80 B3 Aramid fibre plain
weave scrim
100 % aramid fibre
36
Table 3-4 Foam material identification and specification ID Description Composition [%] Nominal
density [kg/m3] Measured density [kg/m3] F3 Polyurethane foam 100 % polyurethane 35 31
4
Cone calorimeter tests
4.1
Test set-up
The test specimens for the cone calorimeter tests were prepared with (100 x 100) mm2 foam samples. Polyester wrap, when used, was also 100 mm in width and breadth (see Figure 4-1). The cover material was cut and sewn to cover both the top surface and the sides of the filling (see Figure 4-2 and Figure 4-3).
The tests were performed using 35 kW/m2 heat flux, with an electric spark igniter as ignition source. The samples with foam and cover were tested in series of three parallel tests, their average results are presented in the report. The other specimens with
combinations including wadding and barriers were only tested once. The specimens were placed in the specimen holder as described in ISO 5660-1 using a retainer frame, but not a wire grid (see Figure 4-4).
Figure 4-2 Foam specimen with cover textile C1.
4.2
Results
Table 4-1 presents tabulated results with regards to time to ignition, total heat release rate, peak heat release rate and total smoke production from the cone calorimeter experiments for the different combinations tested. The results are evaluated over a period of 600 seconds from start of the test. In the following sections examples of results are presented and discussed.
Table 4-1 Results from the cone calorimeter with regards to time to ignition, total heat release rate, peak heat release rate and total smoke production. tign [sec] THR0-600 sec [MJ/m2] HRRmax [kW/m2] TSP0-600 sec [m2/m2] C1 14 57 252 133 C1_W1 11 53 173 73 C1_B1 14 39 201 57 C1_B2 15 42 187 52 C1_B3 15 54 220 84 C2 25 55 286 46 C2_W1 23 61 242 125 C2_B1 20 48 294 43 C2_B2 23 50 282 44 C2_B3 26 42 273 30 C3 12 48 430 692 C3_W1 9 48 410 771 C3_B1 12 46 263 309 C3_B2 14 41 182 305 C3_B3 17 45 257 295 C4 18 67 386 841 C4_W1 20 64 406 852 C4_B1 25 73 347 391 C4_B2 27 42 358 227 C4_B3 34 57 333 354 C5 17 62 428 433 C5_W1 12 66 377 564 C5_B1 18 20 391 471 C5_B2 18 32 349 136 C5_B3 18 40 365 102 C6 10 64 301 1444 C6_W1 7 65 290 1304 C6_B1 9 69 310 854 C6_B2 12 48 276 474 C6_B3 11 49 256 533 C10 51 79 361 343 F3 (Foam only) 5 37 492 223
4.2.1
Time to ignition
Figure 4-5 below shows time to ignition for the different material combinations, when exposed to a heat flux radiation of 35 kW/m2 in the cone calorimeter. More detailed column charts for the individual cover materials are presented in Appendix 0.
Figure 4-5 Time to ignition in the cone calorimeter test at heat flux 35 kW/m2 for all material combinations.
0 10 20 30 40 50 60 Ti m e [ sec ] C1_F3 C1_W1_F3 C1_B1_F3 C1_B2_F3 C1_B3_F3 C2_F3 C2_W1_F3 C2_B1_F3 C2_B2_F3 C2_B3_F3 C3_F3 C3_W1_F3 C3_B1_F3 C3_B2_F3 C3_B3_F3 C4_F3 C4_W1_F3 C4_B1_F3 C4_B2_F3 C4_B3_F3 C5_F3 C5_W1_F3 C5_B1_F3 C5_B2_F3 C5_B3_F3 C6_F3 C6_W1_F3 C6_B1_F3 C6_B2_F3 C6_B3_F3 C10_F3 F3
4.2.2
Heat release
Figure 4-6 shows the comparisons between heat release rate for the different cover materials in combination with only foam. The result for foam alone is also shown. More detailed graphs showing heat release rate results for the different combinations are presented in appendix 0. Two examples are shown in Figure 4-7 and Figure 4-8, with different shapes of the graphs, and different degree of benefit from the use of barriers.
Figure 4-6 Heat release rate. Different covers directly on foam, and foam alone (F3). 0 50 100 150 200 250 300 350 400 450 500 0 100 200 300 400 500 600 H e at R e le ase Ra te [ kW/m2] Time [sec] C1_F3 Average C2_F3 Average C3_F3 Average C4_F3 Average C5_F3 Average C6_F3 Average C10_F3 Average F3 Average
Figure 4-7 Heat release rate, material combinations with cover C2.
Figure 4-8 Heat release rate, material combinations with cover C3.
From the graphs one can see the effect from adding wadding or a barrier material to the material combination. 0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C2_F3 Average C2_W1_F3 C2_B1_F3 C2_B2_F3 C2_B3_F3 0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C3_F3 Average C3_W1_F3 C3_B1_F3 C3_B2_F3 C3_B3_F3
Total heat release during 0-600 seconds for covers C1 to C6 and C10 with different component combinations are presented in Figure 4-9 below, and corresponding results for peak heat release rate are shown in Figure 4-10. Results for foam alone are also shown (F3).
Figure 4-9 Total heat release during 0-600 seconds. Covers C1 to C6 and C10 with different component combinations, and foam only (F3).
Figure 4-10 Peak heat release during 0-600 seconds. Covers C1 to C6 and C10 with different component combinations, and foam only (F3).
0 10 20 30 40 50 60 70 80 90 C1 C2 C3 C4 C5 C6 C10 F3 To tal h e at re le ase [M J/ m 2] Cover+Foam Cover+wadding+foam Cover+Barrier1+Foam Cover+Barrier 2+Foam Cover+Barrier3+Foam F3 0 50 100 150 200 250 300 350 400 450 500 C1 C2 C3 C4 C5 C6 C10 F3 Peak h e at re le ase [kW/ m 2] Cover+Foam Cover+wadding+foam Cover+Barrier1+Foam Cover+Barrier 2+Foam Cover+Barrier3+Foam F3
4.2.3
Smoke production
Figure 4-11 shows the comparisons between smoke production rate for the different cover materials in combination with only foam, and for the foam alone. Smoke production rate results for the different combinations are presented in appendix 0. From the graphs one can see the effect from adding wadding or a barrier material to the material combination.
Figure 4-11 Smoke production rate. Different covers directly on foam, and foam alone (F3).
Figure 4-12 is one example where adding one of the barrier materials has a substantial effect. The figure describes C3 (Trevira CS polyester) in different combinations.
Figure 4-12 Smoke production rate, material combinations with cover C3.
0 2 4 6 8 10 12 14 0 100 200 300 400 500 600 R SR [ (m ²/ s)/ m ²] Time [sec] C1_F3 Average C2_F3 Average C3_F3 Average C4_F3 Average C5_F3 Average C6_F3 Average C10_F3 Average F3 Average 0 2 4 6 8 10 12 14 0 100 200 300 400 500 600 R SR [ (m ²/ s)/ m ²] Time [sec] C3_F3 Average C3_W1_F3 C3_B1_F3 C3_B2_F3 C3_B3_F3
Figure 4-13 Smoke production rate, material combinations with cover C2.
Total smoke production during 0-600 seconds for covers C1 to C6 and C10 with different component combinations are presented in Figure 4-14 below.
Figure 4-14 Total smoke production during 0-600 seconds. C1 to C6 and C10 with different component combinations, and foam only (F3).
0 2 4 6 8 10 12 14 16 18 20 0 100 200 300 400 500 600 R SR [ (m ²/ s)/ m ²] Time [sec] C6_F3 Average C6_W1_F3 C6_B1_F3 C6_B2_F3 C6_B3_F3 0 200 400 600 800 1000 1200 1400 1600 C1 C2 C3 C4 C5 C6 C10 F3 To tal h e at re le ase [M J/ m 2]
Total smoke production 0-600 sec
Cover+Foam Cover+wadding+foam Cover+Barrier1+Foam Cover+Barrier 2+Foam Cover+Barrier3+Foam F3
4.2.4
Assessment of the results
In Figure 4-15 the results for peak heat release rate (HRRmax), total heat release (THR),
total smoke production (TSP) and time to ignition (tign) for each cover material applied
directly on the foam F3 are shown. All results presented are calculated over a test period of 600 s. These values are used as reference values for assessing the effects of using different barrier materials in the upholstery combinations.
Figure 4-15 Results from the different covers mounted directly on the upholstery foam F3 tested in the cone calorimeter at a heat flux density level of 35 kW/m2. All results are evaluated over a test period of 600 seconds.
To study how test results from a cover fabric change when it is combined with different underlying materials (wadding and barriers) we have used radar diagrams, as shown in Figure 4-16 and Figure 4-17. The results in these diagrams are normalized against the results from each cover mounted directly on the foam, i.e. test results from cover + foam are given the value 100 %. The cover designated C10 (leather) has not been further tested with any barriers in this project, and is therefore not shown in a radar diagram.
Results on smoke and heat release are shown in one diagram, while time to ignition is shown in a separate diagram. The results are separated because low values for heat and smoke are regarded as good, while a high value for time to ignition is desirable. For heat release rate and smoke production a value below 100 % means an improvement in fire properties compared to the combination of cover directly on upholstery foam, while a value above 100 % means a worse result. For ignitability a value above 100 % represents an improvement compared to the combination of cover directly on upholstery foam.
Figure 4-16 Comparison of results from different material combinations of covers C1, C2 and C3 compared to results with the same cover mounted directly on the upholstery foam F3. The materials were tested in the cone calorimeter at 35 kW/m2. All results are evaluated
Figure 4-17 Comparison of results from different material combinations of covers C4, C5 and C6 compared to results with the same cover mounted directly on the upholstery foam F3. The materials were tested in the cone calorimeter at 35 kW/m2. All results are evaluated over a test period of 600 seconds.
Two examples of assessments:
For C1 the diagram shows that C1_B2 gives better results on all four parameters than for C1 on foam – i.e. below 100 % for the heat release and smoke
production, and above 100 % for time to ignition.
C2 on wadding gives a better result with regards to peak heat release rate, but worse results with regards to time to ignition and total heat release. The results are significantly worse on total smoke production (outside the scale of the diagram).
These diagrams clearly show that the important reaction-to-fire properties may change in different directions when introducing changes in the material combination. One
combination may release much heat and little smoke, while another combination may produce much smoke but little heat. It will not be sufficient to only regard one single variable when assessing the overall fire safety of the upholstered furniture. An assessment of what material combination that will be the "optimal" one (if any at all) must take this into account.
Effects of wadding
Wadding is normally used between the cover fabric and the upholstery foam to enhance comfort and providing a "cushion appearance". The polyester wadding used in the cone calorimeter tests was designated W1.
Time to ignition, tign
The wadding tended to decrease time to ignition with a few seconds, except for C4 (heavy polyester fabric with a fleece back) where a small increase was observed. This effect can be seen in the diagrams to the right in Figure 4-16 and Figure 4-17. The reduction in time to ignition is probably due to a higher insulation effect, and hence a reduction in the surface thermal inertia1 compared to the cover directly applied on the foam. The effect of the wadding on time to ignition is, however, so small that it most probably is not important with regard to the fire safety of a complete furniture in a real life situation.
Peak heat release rate, HRRmax
The wadding tended to decrease the peak heat release rate. The largest relative decrease of 31% was found for cover C1 (cotton –modal –polyester blend) in combination with wadding (C1_W1). For several of the combinations only a small decrease or increase in HRRmax was observed.
Total heat release, THR0-600 sec
The wadding did not lead to any evident changes in THR, only very small positive or negative changes could be observed.
Total smoke production, TSP0-600 sec
The tests with wadding lead to an increase in total smoke production for all covers except for C6 (artificial leather). The largest relative increase in TSP of 172% was measured for the combination C2 (cotton/viscose blend) and wadding (C2_W1).
For the majority of covers the wadding did not seem to have any considerable effect on any of the measured properties. It is therefore assumed that the polyester wadding will have no significant effect on the test results from fire tests in larger scale.
Effects of fire barriers
Three different barriers were tested in the project. In the cone calorimeter tests the barriers were tested without any layer of wadding.
Time to ignition, tign
There was some variation in time to ignition between the different material combinations. For the covers C1, C3, C4 and C5 time to ignition was increased independently of which of the barriers that was added. For the covers C2
(cotton/viscose blend) and C6 (artificial leather) an increase or decrease in time to ignition seemed to be dependent on type of barrier. The heavier glass fibre barrier B2 gave longer times to ignition than the lighter version B1.
Peak heat release rate, HRRmax
The barriers tended to decrease the peak heat release rate, except for the cover C2 where HRRmax seemed to be unchanged in combination with any of the barriers.
The largest relative decrease of 58% was found for cover C3 (polyester Trevira CS) in combination with the 80 g/m2 glass fibre barrier B2 (C3_B2). For several of the combinations only a small decrease (or increase in two cases) in HRRmax
was observed.
Total heat release, THR0-600 sec
The barriers B2 (glass fibre) and B3(aramid) decreased the total heat release in combination with all cover materials. The largest decrease of 48 % was measured for barrier B2 in combination with cover C5 (wool/polyester blend). The lightest glass fibre barrier B1 gave results that seemed to be dependent on the cover, either a clear reduction in THR or a nearly unchanged value.
Total smoke production, TSP0-600 sec
Total smoke production was reduced for all combinations of cover and barriers, except for the combination of cover C5 and barrier B1 where a small increase in TSP was measured. The largest relative reduction in TSP of 76 % was measured for the combination C5_B3.
As an overall assessment the barrier B1 gave nearly unchanged results or improved results dependent on the type of cover fabric. B1 is the lightest of the two glass fibre cloths, with an area density of 25 g/m2. The heavier glass fibre fabric B2, with an area density of 80 g/m2, improved the test results in all tests, except for the cover C2 where all results are close to the results from C2 (cotton/viscose blend) applied directly on foam F3. The aramid barrier B3 improved all the test results in all tests, especially concerning the total smoke production. None of the barriers lead to worse test results compared to tests without the barrier.
5
Mock-up tests
5.1
Test set-up
A corner mock-up assembly was used for the mock-up test (see Figure 5-1), based on the setup described in [7]. The dimensions used were:
Back cushion: (305 x 229 x 75) mm
Side cushion: (229 x 229 x 75) mm
Bottom cushion: (305 x 305 x 75) mm
The ignition source (both the gas flame and the crib) was applied in the corner of the mock-up. The gas flame was applied for 15 seconds, as described in EN 1021-2 [3].
Figure 5-1 Test set-up, mock-up assembly with cover textile C1 during match flame equivalent exposure.
The different test performed using this set-up is presented in Table 5-1 below.
Table 5-1 Material combinations and scope of test Material combination Ignition source No. of single
tests C1 + W2 + F3 Flame *) 2 Crib 5 1 C1 + B1 + F3 Crib 5 1 C2 + W2 + F3 Flame *) 2 Crib 5 1 C2 + B1 + F3 Crib 5 1 C2 + B2 + F3 Crib 5 1 C3 + W2 + F3 Flame *) 2 Crib 5 1 C3 + B1 + F3 Crib 5 1 C3 + B2 + F3 Crib 5 1 C6 + W2 + F3 Flame *) 2 Crib 5 1 C6 + B1 + F3 Crib 5 1 C10 + W2 + F3 Flame *) 2 Crib 5 1 C10 + B1 + F3 Crib 5 1 C10 + B3 + F3 Crib 5 1 *) 15 seconds application
5.2
Results
Table 5-2 and Table 5-3 presents tabulated results with regards to total heat release rate, peak heat release rate and total smoke production from the mock-up chair experiments, for the different combinations tested. In the following sections examples of results are presented and discussed.
Table 5-2 Results from the mock-up chair tests with regards to total heat release rate, peak heat release rate and total smoke production. Ignition source: gas flame
THR0-1200 sec [MJ] HRRmax [kW] TSP0-1200 sec [m2] C1_W2 21.54 60.68 54.89 C2_W2 0.03 0.59 0.01 C3_W2 1.45 0.14 0 C6_W2 0.08 1.12 0.08 C10_W2 0.04 0.81 0.02
Table 5-3 Results from the mock-up chair tests with regards to total heat release rate, peak heat release rate and total smoke production. Ignition source: crib 5
THR0-1200 sec [MJ] HRRmax [kW] TSP0-1200 sec [m2] C1_W2 21.36 63.97 47.84 C1_B1 19.14 38.82 29.11 C2_W2 21.60 35.92 34.95 C2_B1 20.09 33.07 18.79 C2_B2 20.21 40.89 20.15 C3_W2 18.63 75.42 178.93 C3_B1 16.40 71.29 119.24 C3_B2 0.57 3.57 5.25 C6_W2 22.87 80.19 562.80 C6_B1 0.97 5.53 21.32 C10_W2 25.03 47.78 86.57 C10_B1 23.05 53.63 47.31 C10_B3 22.90 52.81 59.24
Pictures from the different test modes are presented in appendix 0. Examples are shown in the following sections.
Figure 5-2 shows the combination with the worst behaviour after the gas flame exposure. The other cover materials tested left very little damage to the test specimen (see also appendix 0).
t=0 t=1 min
t=2 min t=5min
Figure 5-2 C1_W2_F3, cover C1: Plain weave with pile (83 % cotton, 9 % modal, 8 % polyester). Flame application.
The Trevira CS in combination with polyester wadding showed relatively poor resistance to the crib exposure, as compared to the other cover materials, see Figure 5-3. However, as can be seen from the experiments, this behaviour can be improved by adding a barrier material (see also 0).
t=0 t=1 min
t=2 min t=5min
Figure 5-3 C3_W2_F3, cover C3: Plain weave (100 % Trevira CS). Crib 5 exposure.
5.2.1
Heat release
Figure 5-4 shows the heat release rate for the different cover materials used together with polyester fibre wadding, tested in the mock-up tests and exposed to crib 5. Figure 5-5 and Figure 5-6 are examples of the possible effects of a barrier material on the heat release rate. The results from the flame application are presented in Table 5-2.
Figure 5-4 Heat release rate. Different covers, with foam and wadding. Crib 5 exposure. 0 10 20 30 40 50 60 70 80 90 0 200 400 600 800 1000 1200 H e at R e le ase R ate [ kW] Time [sec] C1_W2_crib 5 C2_W2_crib 5 C3_W2_crib 5 C6_W2_crib 5 C10_W2_crib 5
Figure 5-5 Heat release rate, material combinations with cover C3. Both ignition sources flame and crib 5.
Figure 5-6 Heat release rate, material combinations with cover C6. Both ignition
0 10 20 30 40 50 60 70 80 0 200 400 600 800 1000 1200 H e at R e le ase R ate [ kW] Time [sec] C3_W2_crib 5 C3_W2_flame C3_B1_crib 5 C3_B2_crib 5 0 10 20 30 40 50 60 70 80 0 200 400 600 800 1000 1200 H e at R e le ase R ate [ kW] Time [sec] C6_W2_crib 5 C6_W2_flame C6_B1_crib 5
Total heat release during 0-1200 seconds for covers C1, C2, C3, C6 and C10 with different component combinations when exposed to crib 5 are presented in Figure 5-7 below. Corresponding peak heat release rate is presented in Figure 5-8. The results from the flame application are presented in Table 5-2.
Figure 5-7 Total heat release during 0-1200 seconds. C1, C2, C3, C6 and C10 with different component combinations. Crib 5 exposure.
Figure 5-8 Peak heat release rate during 0-1200 seconds. C1, C2, C3, C6 and C10 with different component combinations. Crib 5 exposure.
0,00 5,00 10,00 15,00 20,00 25,00 30,00 C1 C2 C3 C6 C10 To tal h e at re le ase [M J] Cover+wadding+foam Cover+Barrier1+Foam Cover+Barrier 2+Foam Cover+Barrier3+Foam 0 10 20 30 40 50 60 70 80 90 C1 C2 C3 C6 C10 H e at re le ase rate [kW] Cover+wadding+foam Cover+Barrier1+Foam Cover+Barrier 2+Foam Cover+Barrier3+Foam
5.2.2
Smoke production
Figure 5-9 below shows the smoke production rate for the gas flame mock-up tests. Only C1 has a substantial smoke production, the other curves are marked with red. C1 was the only cover material that showed sustained burning after removal of the burner tube.
Figure 5-9 Smoke production rate. Different covers, with foam and wadding. Flame exposure. 0 0,05 0,1 0,15 0,2 0,25 0,3 0 200 400 600 800 1000 1200 Sm o ke p ro d u ction [m 2/s] Time [sec] C1_W2_flame C2_W2_flame C3_W2_flame C6_W2_flame C10_W2_flame
Figure 5-10 below shows the smoke production rate for the crib 5 mock-up tests.
Figure 5-10 Smoke production rate. Different covers, with foam and wadding. Crib 5 exposure.
Figure 5-11 Total smoke production during 0-1200 seconds. C1, C2, C3, C6 and C10 with different component combinations. Crib 5 exposure.
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2 0 200 400 600 800 1000 1200 Sm o ke p ro d u ction [m 2/s] Time [sec] C1_W2_crib 5 C2_W2_crib 5 C3_W2_crib 5 C6_W2_crib 5 C10_W2_crib 5 0 100 200 300 400 500 600 C1 C2 C3 C6 C10 Sm o kw p ro d u ction [m 2] Cover+wadding+foam Cover+Barrier1+Foam Cover+Barrier 2+Foam Cover+Barrier3+Foam
5.2.3
Assessment of the results
As only one of the tested combinations (C1 on wadding) ignited when tested with the small flame ignition source, these tests are not assessed in detail in this section. In Figure 5-12 the results for peak heat release rate (HRRmax), total heat release (THR)
and total smoke production (TSP) for the upholstery combinations tested with ignition source crib 5 in the mock-up are shown. All combinations are tested on the polyurethane foam F3, and the results are calculated over a test period of 1200 s. The average of these values are used as reference values for assessing the effects of different barrier materials in the upholstery combinations.
Figure 5-12 Results from the different upholstery combinations tested in the mock-up with exposure to crib 5. All results are evaluated over a test period of 1200 seconds. The value of C6_W2 on total smoke
production was 563 m2 and the column extends beyond the scale of the diagram above.
As for the assessment of cone calorimeter test results we have also here used radar diagrams, as shown in Figure 5-13 and Figure 5-14. The results in the diagram in Figure 5-13 are normalized against the averaged results from all the tests with crib 5 in the mock-up. The results in the diagram in Figure 5-14 are normalized against the results from each cover mounted on wadding W2 and the foam, i.e. test results from cover + wadding W2 + foam F3 are given the value 100 %.
A value below 1 means an improvement compared to the average value for the assessed fire property (i.e. less heat or less smoke than the average).
Figure 5-13 Comparison of results from for the different combinations of covers, wadding and barriers mounted on the upholstery foam F3 tested with ignition source crib 5 in the mock-up. All results are evaluated over a test period of 1200 seconds, and normalized against the average test results for each variable. Values below 1 are assessed as better than the average. C6_W2 had a TSP value of 5.9 times the average TSP.
From Figure 5-12 and Figure 5-13 we can see that the best results on both smoke production and heat release are achieved by cover C3 (polyester, Trevira CS) in
combination with the 80 g/m2 glass fibre barrier B2 and by cover C6 (artificial leather) in combination with the 25 g/m2 glass fibre barrier B1.
For assessing the effect of barriers on each cover fabric, a radar diagram for each cover is presented in Figure 5-14. In this figure the axes for the two cases where only one barrier was tested (C1 and C6) are changed compared to the diagrams where test results from more than one barrier are presented.
Figure 5-14 Comparison of results from for the different combinations of covers, wadding and barriers mounted on the upholstery foam F3 tested with ignition source crib 5 in the mock-up. All results are evaluated over a test period of 1200 seconds, and normalized against the test results for each cover in combination with wadding W2 and foam F3.
Effects of fire barriers
Material combinations with three different barriers were tested in the mock-up. Barrier B3 was only tested in combination with cover of leather, C10.
Peak heat release rate, HRRmax
The barriers either decreased the peak heat release rate significantly, or in some cases the peak heat release rate value was approximately unchanged. The largest effect was observed in the test of barrier B2 with cover C3 and cover C6 with cover B1.
Total heat release, THR0-1200 sec
The effects on THR0-1200 sec were the same as for the peak heat release rate.
Total smoke production, TSP0-1200 sec
Total smoke production was reduced for all combinations of cover and barriers. The largest relative reductions in TSP of 96-97 % were measured for the combinations C3_B2 and C6_B1.
These results show that it is possible to improve important reaction-to-fire properties by adding a fire barrier fabric to the upholstery combination. Even the light-weight glass fibre fabric had dramatic effect in the mock-up scale test with the cover of artificial leather, C6. None of the barriers lead to worse test results compared to tests without the barrier.
6
Discussion and conclusions
In this study we have concentrated on the initial fire development in an upholstered piece of furniture. The aim of the study is to evaluate if – and to which degree – reaction-to-fire properties of upholstered furniture can be improved without addition of any chemical fire retardants to the materials. For evaluation of material fire properties we have used the small scale cone calorimeter test. For assessment of the fire behaviour in real scale, we have tested the material combinations in a semi-scale mock-up, exposing the test objects to the recognised ignition sources crib 5 according to the British standard BS 5852, and the small flame according to European standard EN 1021-2.
The focus has been on properties that are of importance with respect to detection of fire, human reaction and evacuation. Results on ignitability, heat release and smoke
production from the tests are considered useful for evaluation of the fire hazard of different material combinations in upholstered furniture. Because there is a limited number of tests in this project, and thus a limited data set, proper statistical analysis of the data is not possible. The assessments below are therefore subjective and based on
observations of the test results.
6.1
Effects of cover material and fire barrier
We can see from the cone calorimeter tests that, as expected, the choice of cover material is of importance for the time to ignition. Incorporating a barrier between the cover and the foam increases the time to ignition in most of the cases. Some cover materials will benefit from even a light fire barrier while others will need more optimisation of the barrier in order to be effective.
6.1.1
Even light barriers may have positive effects
The barrier fabrics we have chosen are light and transparent, they will therefore not act as a thermal barrier that would limit the heat transfer, nor would they limit the transport of volatile gases from the upholstery foam. The greatest influence of the barriers is to separate the flaming surface from the flammable non-flame retardant foam. So even by using very light fabrics as barriers, in addition to the cover fabric, it is possible to improve the reaction-to-fire properties of these material combinations.
Even though light barriers may have positive effect on the fire properties testing is needed to assess whether the barrier will work with a particular cover fabric. Our results have shown that a denser barrier tends to be more independent of the choice of cover fabric than a lighter barrier would be.
6.1.2
Barriers can reduce the heat release
The barrier fabrics had positive effect by decreasing total heat release from the test specimen in the cone calorimeter, as well as they showed a tendency to reduce the peak heat release rate. Again, the effectiveness is higher for the denser fabrics, while the effectiveness from the light barrier fabric is more dependent upon the nature of the cover fabric. When the material combinations were tested as a mock-up chair and exposed to a crib 5 open flame ignition source, the peak heat release as well as the total heat release were reduced when adding a barrier between the cover fabric and the foam.
6.1.3
Barriers can reduce the smoke production
Generally (both in cone calorimeter and mock-up tests), total smoke production was reduced by adding a barrier. The largest effect came from a denser barrier fabric. However, as this may not always be the case, an assessment of a cover fabric together with a barrier fabric should also include smoke production evaluation.
6.1.4
Predicting larger scale fire behaviour on the basis of
small scale tests
Our experiments show, not surprisingly, that it may be challenging to predict the reaction-to-fire properties of different material combinations based on small scale testing.
However, we do see a correlation in the interpretation of the experiments. This has also been explored in the CBUF report from 1995.
6.1.5
Conclusions
It has been shown that fire safety of upholstered furniture can be improved by clever choice of materials. Applying a fabric that can act as a fire barrier between the cover and the upholstery foam can reduce release of heat and smoke from the furniture to different degrees depending of the barrier fabric and the material combination in the upholstered furniture. Although our tests have not been performed strictly according to EN 1021-2 and BS 5852 respectively, it is our opinion that our experiments show that it is possible to pass tests according to these standards without the use of chemical flame retardants, but by using alternative strategies.
The differences in the effectiveness of cover-barrier combinations, as discussed in the sections above, show that it is certainly possible to improve the fire safety in upholstered furniture without the use of flame retardants. However, it requires thorough examination and a proper choice of material combination.
6.2
Recommendations for further work
6.2.1
Other ways of improving fire safety of upholstered
furniture
Further possibilities to improve the fire safety in upholstered furniture are modification of design and construction parameters of the complete furniture, which has not been studied in this project. Several publications offer advice on the subject, e.g. Fire Behavior of
Upholstered Furniture and Mattresses [10], Fire Safety of Upholstered Furniture – the final report on the CBUF research programme [6], and the article "A review of fire blocking technologies for soft furnishings" [8]. For example, chairs with large gaps in the
design, e.g. between seat and back, mean less fuel and decreased risk of fire development. The use of armrests means more fuel and a radiative feedback to a fire in the furniture. Upholstery with little space between the furniture and the ground can speed up a fire development when pools of melted polymer are created under it.
Using fire barriers is one method that has a potential and that should be further explored. We have noted that there are many interesting materials available - mainly for completely other applications than for furniture - but that could be interesting to investigate further for furniture applications. They should be investigated both with regards to relevant requirements on quality, comfort and cost, but also with regards to relevant requirements on fire safety so that the range of available fire barriers can be increased.
6.2.2
A system for fire classification of upholstered furniture
is needed
As already mentioned, even light barrier materials show positive effects, at least for smaller size ignition sources such as gas flame (EN 1021-2) and crib 5 (BS 5852). Not all barriers need to be resistant to significantly larger ignitions sources. However, it may be a large step forward to improve the fire safety of upholstered furniture from today's
Scandinavian standard to a higher level by clever choice of materials and design. All products may not satisfy some of the more severe fire tests that are in use in some countries, where the furniture is exposed to relatively large ignition sources, but that should not be seen as a problem.
Introducing a voluntary fire classification system for upholstered furniture could be a part of the solution. Such a system may be a tool that gives the purchasers a possibility to choose products that fits the desired fire safety level, and will be useful both for private households and for commercial and public customers.
6.2.3
Assessment and evaluation based on testing
The complexity in the way different cover materials behave and interact with barrier materials and other components requires that the actual material combination is tested and evaluated in real scale. A system for assessments that simplifies the evaluation, e.g. based on testing with reference materials, rules for extended application of test results etc. should be developed. This should make it possible to assess a large range of fabrics that are nearly, but not quite, identical with regards to properties like area density, colour etc. It will also allow for product development and smaller changes in design and material composition while limiting the required amount of testing.
References
[1] “ISO 5660-1:2015 Reaction-to-fire tests - Heat release, smoke production and mass loss rate - Part 1: Heat release rate (cone calorimeter method) and smoke production rate (dynamic measurement).” ISO Copyright office, published in Switzerland, 2015. [2] “EN 13823: 2010 Reaction to fire test for building products - Building products
excluding flooring exposed to the thermal attack by a single burning item.” CEN, 2010.
[3] “EN 1021-2:2014 Furniture - Assessment of the ignitability of upholstered furniture - Part 2: Ignition source match flame equivalent.” CEN-CENELEC, Brussels, 2014. [4] “BS 5852:2006 Methods of test for assessment of the ignitability of upholstered
seating by smouldering and flaming ignition sources.” BSI, 2006.
[5] “EN 1021-1:2014 Furniture - Assessment of the ignitability of upholstered furniture - Part 1: Ignition source smouldering cigarette.” CEN-CENELEC, Brussels, 2014. [6] B. Sundström, Ed., Fire Safety of Upholstered Furniture: the final report on the
CBUF research programme. London: Interscience Communications Ltd, 1995.
[7] T. Fabian, “Upholstered furniture flammability,” Underwriters Laboratories Inc., Jul. 2013.
[8] “Glossary of flexible polyurethane foam technoloy.” [Online]. Available: http://www.pfa.org/jifsg/jifsgs15.html. [Accessed: 13-Dec-2015].
[9] S. Nazaré and R. D. Davis, “A review of fire blocking technologies for soft furnishings,” Fire Sci. Rev., vol. 1, no. 1, pp. 1–23, 2012.
[10] J. Krasny, W. Parker, and V. Babrauskas, Fire behavior of upholstered furniture and
Appendix A
Appendix A Results cone calorimeter tests
A.1 Time to ignition
Figure A- 1 Time to ignition, material combinations with cover C1, plain weave (cotton, modal, polyester)
Figure A- 2 Time to ignition material combinations with cover C2, basket weave (cotton, viscose). 0 5 10 15 20 25 30 35 Ti m e [ sec ] C1_F3 C1_W1_F3 C1_B1_F3 C1_B2_F3 C1_B3_F3 0 5 10 15 20 25 30 35 Ti m e [ sec ] C2_F3 C2_W1_F3 C2_B1_F3 C2_B2_F3 C2_B3_F3
Appendix A
Figure A- 3 Time to ignition material combinations with cover C3, plain weave (Trevira CS).
Figure A- 4 Time to ignition material combinations with cover C4, basket weave and fleece (polyester).
Figure A- 5 Time to ignition material combinations with cover C5, felted plain weave (wool). 0 5 10 15 20 25 30 35 Ti m e [ sec ] C3_F3 C3_W1_F3 C3_B1_F3 C3_B2_F3 C3_B3_F3 0 5 10 15 20 25 30 35 Ti m e [ sec ] C4_F3 C4_W1_F3 C4_B1_F3 C4_B2_F3 C4_B3_F3 0 5 10 15 20 25 30 35 Ti m e [ sec ] C5_F3 C5_W1_F3 C5_B1_F3 C5_B2_F3 C5_B3_F3
Appendix A
Figure A- 6 Time to ignition material combinations with cover C6, artificial leather.
Figure A- 7 C10 leather, with foam F3.
0 5 10 15 20 25 30 35 Ti m e [ sec ] C6_F3 C6_W1_F3 C6_B1_F3 C6_B2_F3 C6_B3_F3 0 5 10 15 20 25 30 35 Ti m e [ sec ] C10_F3 0 5 10 15 20 25 30 35 Ti m e [ sec ] F3
Appendix A
A.2 Heat release rate
Figure A- 9 Heat release rate, material combinations with cover C1.
Figure A- 10 Heat release rate material combinations with cover C2.
0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C1_F3 Average C1_W1_F3 C1_B1_F3 C1_B2_F3 C1_B3_F3 0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C2_F3 Average C2_W1_F3 C2_B1_F3 C2_B2_F3 C2_B3_F3
Appendix A
Figure A- 11 Heat release rate, material combinations with cover C3.
Figure A- 12 Heat release rate, material combinations with cover C4
0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C3_F3 Average C3_W1_F3 C3_B1_F3 C3_B2_F3 C3_B3_F3 0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C4_F3 Average C4_W1_F3 C4_B1_F3 C4_B2_F3 C4_B3_F3
Appendix A
Figure A- 13 Heat release rate, material combinations with cover C5.
Figure A- 14 Heat release rate, material combinations with cover C6.
0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C5_F3 Average C5_W1_F3 C5_B1_F3 C5_B2_F3 C5_B3_F3 0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C6_F3 Average C6_W1_F3 C6_B1_F3 C6_B2_F3 C6_B3_F3
Appendix A
Figure A- 15 Heat release rate, material combinations with cover C10.
0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 H e at R e le ase R ate [ kW/ m 2] Time [sec] C10_F3 Average
Appendix A
A.3 Smoke production
Figure A- 16 Smoke production rate, material combinations with cover C1.
Figure A- 17 Smoke production rate, material combinations with cover C2.
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 0 100 200 300 400 500 600 Sm o ke p ro d u ction r ate [(m ²/ s)/ m ²] Time [sec] C1_F3 Average C1_W1_F3 C1_B1_F3 C1_B2_F3 C1_B3_F3 0 0,5 1 1,5 2 2,5 3 0 100 200 300 400 500 600 Sm o ke p ro d u ction r ate [(m ²/ s)/ m ²] Time [sec] C2_F3 Average C2_W1_F3 C2_B1_F3 C2_B2_F3 C2_B3_F3
Appendix A
Figure A- 18 Smoke production rate, material combinations with cover C3.
0 2 4 6 8 10 12 14 0 100 200 300 400 500 600 R SR [ (m ²/ s)/ m ²] Time [sec] C3_F3 Average C3_W1_F3 C3_B1_F3 C3_B2_F3 C3_B3_F3 0 1 2 3 4 5 6 7 8 9 10 0 100 200 300 400 500 600 R SR [ (m ²/ s)/ m ²] Time [sec] C4_F3 Average C4_W1_F3 C4_B1_F3 C4_B2_F3 C4_B3_F3
Appendix A
Figure A- 20 Smoke production rate, material combinations with cover C5.
Figure A- 21 Smoke production rate, material combinations with cover C6.
0 1 2 3 4 5 6 7 0 100 200 300 400 500 600 R SR [ (m ²/ s)/ m ²] Time [sec] C5_F3 Average C5_W1_F3 C5_B1_F3 C5_B2_F3 C5_B3_F3 0 2 4 6 8 10 12 14 16 18 20 0 100 200 300 400 500 600 R SR [ (m ²/ s)/ m ²] Time [sec] C6_F3 Average C6_W1_F3 C6_B1_F3 C6_B2_F3 C6_B3_F3
Appendix A
Figure A- 22 Smoke production rate, cover C10 on foam F3.
0 0,5 1 1,5 2 2,5 3 0 100 200 300 400 500 600 R SR [ (m ²/ s)/ m ²] Time [sec]
C10_F3 Average
C10_F3 AverageAppendix B
Appendix B Results mock-up tests
B.1 Heat release rate
Figure B- 1 Heat release rate, material combinations with cover C1.
Figure B- 2 Heat release rate, material combinations with cover C2.
0 10 20 30 40 50 60 70 80 0 200 400 600 800 1000 1200 H e at R e le ase R ate [ kW] Time [sec] C1_W2_crib 5 C1_W2_flame C1_W2_flame C1_B1_crib5 0 10 20 30 40 50 60 70 80 0 200 400 600 800 1000 1200 H e at R e le ase R ate [ kW] Time [sec] C2_W2_crib 5 C2_W2_flame C2_B1_crib 5 C2_B2_crib 5
Appendix B
C2 did not ignite at exposure to flame in one of two test, therefore there is only one graph for C2_W2_flame.
Figure B- 3 Heat release rate, material combinations with cover C3.
C3 did not ignite at exposure to flame in one of two test, therefore there is only one graph for C3_W2_flame. 0 10 20 30 40 50 60 70 80 0 200 400 600 800 1000 1200 H e at R e le ase R ate [ kW] Time [sec] C3_W2_crib 5 C3_W2_flame C3_B1_crib 5 C3_B2_crib 5
Appendix B
Figure B- 4 Heat release rate, material combinations with cover C6.
C6 did not ignite at exposure to flame in one of two test, therefore there is only one graph for C6_W2_flame.
Figure B- 5 Heat release rate, material combinations with cover C10.
0 10 20 30 40 50 60 70 80 0 200 400 600 800 1000 1200 H e at R e le ase R ate [ kW] Time [sec] C6_W2_crib 5 C6_W2_flame C6_B1_crib 5 0 10 20 30 40 50 60 70 80 0 200 400 600 800 1000 1200 H e at R e le ase R ate [ kW] Time [sec] C10_W2_crib 5 C10_W2_flame C10_B1_crib 5 C10_B3_crib 5
