Mattresses:
BURNING BEHAVIOUR - FULL SCALE TEST
New Nordtest Method
SP Fire Technology SP REPORT 2005:05
SP
Swedish National
T
Abstract
A new test method for testing of fire properties of mattresses is proposed. The method is intended to evaluate the full scale burning behaviour of a complete mattress or bed system when exposed to a flaming ignition source. Fire performance criteria, for mattresses used in public occupancies, are proposed. The criteria are intended to give protection against fully developed fires in mattresses in the event of, for example, arson, in premises where people may be prevented from escape.
A test series has been performed, on 10 mattresses from the market place, to demonstrate the test method proposed. The test results show that mattresses exist on the market today that meets the proposed criteria.
Key words: mattresses - fire, full-scale fire test, heat release rate, smoke production rate, fire test
SP Sveriges Provnings- och SP Swedish National Testing and Forskningsinstitut Research Institute
SP Rapport 2005:05 SP Report 2005:05 ISBN 91-85303-36-4 ISSN 0284-5172 Borås 2005 Postal address: Box 857,
SE-501 15 BORÅS, Sweden
Telephone: +46 33 16 50 00 Telefax: +46 33 13 55 02
E-mail: info@sp.se
Content
Abstract 2 Content 3 Acknowledgements 4 Background 5 1 Project programme 62 Proposed test method 7
3 Tested products 8
4 Test results and discussion 9
5 Proposed criteria and motivation 13
6 Conclusion 14
7 References 15
Appendix A Proposed test method 16
1 SCOPE AND FIELD OF APPLICATION 17
2 REFERENCES 17 3 DEFINITIONS 18 4 SAMPLING 18 5 TEST PRINCIPLES 18 6 IGNITION SOURCE 20 7 WEIGHING PLATFORM 20
8 HOOD AND EXHAUST DUCT 20
9 INSTRUMENTATION IN THE EXHAUST DUCT 21
10 CALIBRATION 23
11 PREPARATION OF TEST SPECIMENS 25
12 TESTING 26
13 TEST REPORT 28
Acknowledgements
This project was sponsored by Nordic Innovationcenter, project no 04021.
The tests were performed at SINTEF NBL. The mattress samples were sponsored by Norwegian mattress suppliers.
Background
In public premises, such as health care facilities, locked psychiatric wards, jails etc., where people may be behind locked doors, mattresses constitute a fire hazard. People are prohibited from escaping without assistance. Upholstered furniture, mattresses and other items are typical objects of ignition in the case of arson or incendiary fires or even common accidental fires. Similarly, fires in mattresses may be a substantial fire hazard in the case of passenger ships.
There are a number of possible actions to take to minimize the risk, one of them is to choose fire resistant products in a systematic way.
Traditionally, mattresses are tested with “smoker’s materials” (i.e. cigarettes or small open flames) according to ordinary ignitability tests. To evaluate the contribution of a mattress to the heat release and smoke production in the early stages of a fire, these ordinary ignitability tests are not sufficient. In such cases, a full-scale test with an ignition source simulating arson is needed.
This report proposes a new full scale fire test method for mattresses based on the same testing technique as NT FIRE 032 [1]. This test method will enable Nordic authorities, public purchasers etc. to define strict fire criteria for special premises where arson is a risk. Nordic industry will have a common tool to define fire characteristics of their mattress products. In an informative annex to the test method, criteria on heat release rate and smoke production rate are proposed.
The proposed test method also includes an annex describing how to measure the amount of certain toxic, gaseous species in the combustion gases during the fire test using the FTIR technique. The risk of being exposed to smoke gases away from a mattress fire source is discussed in SP report 2005:29 [2].
1
Project programme
The project, which is a co-operation between SP, Fire Technology and SINTEF-NBL, was divided in two parts, the writing of the test method and the performance of a verification test series.
The test method draft was written by SP, Fire Technology. Based on the NT FIRE 032 “Upholstered furniture: Burning behaviour – full scale test” [1] and experience from the CBUF project (Combustion Behaviour of Upholstered Furniture) [3], the draft was specifically designed to test mattresses. Information from other test methods, such as “Technical Bulletin 133” [4], “Technical Bulletin 129” [5] and “Technical Bulletin 603” [6], were taken into account during this phase of the project.
Based on the draft test method a test series was performed by SINTEF-NBL on a number of mattresses from the market place. The test series was preformed to confirm the
capability of the draft test method to measure full-scale fire properties and to discriminate between products.
Finally, the proposed test method description was completed, including annexes describing optional toxic gas analysis and proposed fire characteristics criteria.
2
Proposed test method
The proposed test method is intended to evaluate the contribution of a burning mattress or bed system to fire growth and smoke production under well ventilated conditions, when subjected to a specified ignition source. The ignition source is primarily intended to simulate ignition, typical of arson or incendiary fires. Heat release and smoke production rate data, describing the burning behaviour from ignition until all burning has ceased, is obtained.
The heat release rate of the burning specimen is measured using oxygen consumption methodology [7]. The hazard of reduced visibility is estimated by measuring the
production of light obscuring smoke with a white light system. The burning behaviour is visually documented by photographic and/or video recordings.
The technique for toxic gas analysis using FTIR measurements is outlined in an annex as an option.
A mattress that meets the proposed criteria will not significantly contribute to the early stages of a developing room fire, it will produce limited amounts of heat and smoke, thereby reducing the amount of toxic gases produced.
3
Tested products
A test series was performed on 10 mattresses from the market place. The test series was performed by SINTEF-NBL according to the proposed test method, measuring heat release rate and smoke production rate. Optional gas analysis with FTIR equipment was not performed. The description (as given by the mattress sponsors) of the tested products is given in table 1 below.
Table 1: Description of tested mattresses Test
number
Product information Dimensions (length
x width x height)
1 Foam: Flame retardant high resilience cold cured PU
-foam, quality Wef 39, nominal density 40 kg/m3. 30 mm
thickness, horizontally, and 50 mm thickness vertically (sides). Spring core. Cover: 240 g/m2 Proban flame retardant cover of 100 % cotton quilted together with 200 g/m2 Dacron-fibre. Approved according to IMO FTP Code Part 9.
2000 x 900 x 190 mm
2 Foam: Flame retardant high resilience cold cured PU
-foam, quality Wef 39, nominal density 40 kg/m3 -
comprising 80 mm compact-foam glued to 50 mm profiled foam. Cover: PU coated PA-based fabric, 185 g/m2. Approved according to BS 7175 (crib 5).
2000 x 800 x 130 mm
3 Foam: Non flame retardant standard polyether foam,
quality WE 30, nominal density 27 kg/m3 approx. Cover:
100 % cotton.
2000 x 800 x 100 mm 4 Foam: Non flame retardant foam, “35 N”, thickness 130
mm, density 35 kg/m3 (no cover). Approved according to
EN 597-1.
2000 x 750 x 130 mm 5 Foam: Non flame retardant foam, “35 XS”, thickness
130 mm, density 35 kg/m3, (no cover). Approved according to EN 597-1.
2000 x 750 x 130 mm 6 Foam: Flame retardant treated foam, “35 HR”, thickness
130 mm, density 35 kg/m3, (no cover). Approved
according to EN 597-1.
2000 x 750 x 130 mm 7 Foam: Flame retardant foam, CMHR (Combustion
Modified High Resilience), thickness 130 mm, density 37 kg/m3, (no cover). Approved according to BS 5852 Crib
5.
2000 x 750 x 130 mm
8 Foam: Flame retardant foam, CMHR (Combustion
Modified High Resilience), thickness 130 mm, density 47 kg/m3, (no cover). Approved according to BS 5852 Crib 5.
2000 x 750 x 130 mm
9 Foam: Non flame retardant foam, “35 XS”, thickness
130 mm, density 35 kg/m3. Cover: 100 % polyester,
“Trevira CS”, 200 g/m2. Approved according to BS 5852,
part 1 (cigarette), classified M1.
2000 x 750 x 130 mm
10 Foam: Flame retardant treated foam, “35 HR”, thickness
130 mm, density 35 kg/m3. Cover: 100 % polyester,
“Trevira CS”, 200 g/m2. Approved according to BS 5852,
part 1 (cigarette), classified M1.
4
Test results and discussion
The tested mattresses exhibited, in principle, two different kinds of burning behaviour. Four mattresses (tests 3, 4, 5 and 9) ignited easily and showed progressive flaming, producing high heat release and smoke production rates. Six mattresses (tests 1, 2, 6, 7, 8 and 10) were ignited and showed flaming under the influence of the burner, but after the burner was withdrawn (120 s) no progressive flaming was shown. Hence only limited rates of heat release were measured and smoke production was low. Below, a tabulated summary of the test results is given, see table 2.
During tests 4, 5 and 9 the test was interrupted due to progressive flaming. If those products had not been extinguished, higher peak heat release and smoke production rates would have been measured. Similarly, higher amounts of total heat and smoke would have been found.
Table 2: Test results regarding heat release rate, smoke production rate and mass loss from the ten tested mattresses
Test HRR peak (kW) HRRsmooth peak (kW) THR10 (MJ) THR (MJ) SPR peak (m2/s) TSP10 (m2) TSP (m2) Initial mass (kg) Mass loss (kg) EHC (MJ/kg) 1 11 9 1,41 2,12 <0,1 8,1 45,0 12.90 -** -2 31 -28 2,90 3,06 0,4 30,0 43,1 8.42 -** -3 549 514 88,6 88,6 4,2 692 692 4.83 4,04 21,9 4* >699 699 >39,2 >39,2 >4,5 >253 >253 6.27 1,88 20,8 5* >586 566 >31,2 >31,2 >3,5 >197 >197 6.57 2,26 13,8 6 20 14 1,14 1,14 0,5 71,9 71,9 6.07 0,33 3,4 7 22 16 1,11 1,11 0,5 43,8 43,8 6.33 0,12 9,1 8 54 44 3,17 3,17 0,9 65,2 65,2 8.43 0,20 15,5 9* >565 565 >35,6 >35,6 >3,9 >276 >276 7.47 1,47 24,2 10 34 30 2,12 2,12 0,4 32,8 32,8 7.08 0,17 12,8 HRRpeak = maximum heat release rate, exclusive burner heat output (30 kW)
HRRsmooth peak= peak of 30 s sliding average of HRR
THR10 = total heat release rate (first 10 minutes)
THR = total heat release rate
SPRpeak = maximum smoke production rate TSP10 = total smoke production (first 10 minutes)
TSP = total smoke production Initial mass = mass prior to test EHC = effective heat of combustion
* The test was interrupted due to progressive fire behaviour
** Mass loss was too small to be measured accurately by the load cell system
In appendix B HRR (30 s sliding average ) and SPR graphs are given, as a function of time. Maximum HRR, THR and TSP are shown as bar graphs below for the sake of comparison between the different mattresses.
The measured heat release rate was smoothed as a 30 second sliding average. This was done to avoid influence of noise etc. on HRR peak values. Figure 1 below shows the smoothed HRR peak values from each test. Test no 4, 5 and 9 were interrupted due to progressive flaming. Those peaks would have been even higher if the fire would not have been extinguished. There is a clear distinction between those mattresses that did not show progressive flaming and those which did. The non-progressive mattresses all show peak values below 50 kW. The CBUF project identified a peak heat release rate borderline of 80 kW as a level where mattresses did propagate or not [3].
0 100 200 300 400 500 600 700 1 2 3 4 5 6 7 8 9 10 HRR smooth peak (kW) Test no >69 9 k W > 566 kW > 565 kW Proposed criteria 55 kW
Figure 1. Peak heat release rates, taken from a 30 second sliding average HRR curve. The energy released by the mattress specimens is shown in figure 2 below. Total heat release constitutes the fire load of the mattress. The energy released in the early stage of a fire has a direct implication on the hazard for people who are in the room of fire origin, especially in small and middle sized rooms. Figure 2 shows the heat released during the first 10 minutes of the mattress fire. Test no 4, 5 and 9 would have shown much higher THR10 values if they would not have been interrupted. Their THR10 is based on a 3
minutes burning period approximately (see appendix B). Test no 3 was not extinguished and burned during a period of 9 minutes approximately (see appendix B).
Similar to the peak heat release parameter there is a clear distinction between those mattresses that show progressive flaming and those which do not.
0 10 20 30 40 50 60 70 80 90 100 1 2 3 4 5 6 7 8 9 10 THR 10 (MJ) Test no >3 9,2 MJ > 31, 2 M J > 35, 6 MJ Proposed criteria 10 MJ
Figure 2. Total heat released, during the first 10 minutes of the tests (note that test 4, 5 and 9 were extinguished before 10 minutes due to progressive flaming). The total smoke produced during the first 10 minutes by each mattress is given in figure 3 below. Similar to THR, test 4, 5 and 9 would have shown much higher TSP10 if they
would not have been extinguished. The other mattresses produce limited amounts of smoke even though there is a relatively large difference within the group of mattresses that did not show progressive flaming. The amounts of smoke measured from the low producing mattresses can be compared to what is accepted from a first class wall lining material according to EN 13501-1, 50 m2. Peak smoke production rate is not proposed as
a criterion at low production rates due to too high uncertainty in measurements [8].
0 50 100 150 200 250 300 1 2 3 4 5 6 7 8 9 10 TSP 10 (m 2 ) Test no >25 3 m 2 > 1 97 m 2 >2 7 6 m 2 Proposed criteria 50 m2
Figure 3. Total smoke produced, during the first 10 minutes of the tests (note that test 4, 5 and 9 were extinguished before 10 minutes due to progressive flaming).
During another test series performed on ten health care mattresses in 2000 [9] a more varied fire performance was noted. Peak heat release rates between 45 kW to 1100 kW and smoke production rates up to 14.5 m2/s were measured. These results are summarized
in Table 3.
From a hazard point of view those mattresses performed similar to the mattresses tested within this project. There was a relatively clear distinction between mattresses that showed progressive flaming compared to those which did not, with the exception of mattress M9 and M10 which gradually decreased after the burner was redrawn.
Mattress M1, M2, M3, M7, M8 and M10 all had test durations less than 10 minutes why a calculation of TSP10 would give the same values as given for TSP.
Table 3: Test results regarding heat release rate, smoke production rate and mass loss from ten tested health care mattresses in the year 2000 [9] (HRR is given inclusive the burner heat output, 30 kW)
Mattress HRRpeak (kW) THR10 (MJ) THR (MJ) SPRpeak (m2/s) TSP (m2) Mass loss (kg) EHC (MJ/kg) Comment M1 60 6,2 6,2 0,2 15,2 -** - non-propagating fire M2 52 6,2 6,2 0,2 12,9 -** - non-propagating fire M3 72 6,8 6,8 0,2 11,0 -** - non-propagating fire M4 1120 204 210 6,0 1820 9,3 22,6 propagating fire M5 850 150 165 6,4 1470 7,3 22,6 propagating fire M6 1120 162 164 8,5 1100 7,3 22,4 propagating fire M7 45 5,0 5,0 0,7 43,0 -** - non-propagating fire M8 64 5,7 5,7 0,3 11,9 -** - non-propagating fire M9 193 46,4 62,8 2,9 714 2,7 23,2 propagating fire M10 223 20,3 20,3 14,5 1180 1,1 18,3 propagating fire* HRRpeak=maximum heat release rate
THR10=total heat release rate (first 10 minutes)
THR=total heat release rate
SPRpeak=maximum smoke production rate TSP=total smoke production
EHC=effective heat of combustion
* the whole mattress was not consumed, flame spread stopped after 6 minutes. ** Mass loss was too small to be measured accurately by the load cell system.
The test equipment used for the proposed test method has certain limitations. Heat release rates up to in the order of 1.5 MW can be measured with acceptable accuracy; above this level losses of smoke from the hood system significantly reduce the accuracy of the results. From a classification point of view, however, test results above 1.5 MW are of limited interest since this would be beyond acceptable fire properties for mattresses in the types of applications studies here. If test results are used as input to fire engineering, measured heat release rates above this level should be assessed from case to case.
5
Proposed criteria and motivation
Heat release rate is one of the most important fire parameters used for describing the fire properties of a burning object. Findings from the CBUF project [3] define four types of burning behaviour for upholstered furniture and mattresses:
• Quick development, high peak HRR • Delayed development, moderate peak HRR • Slow development, low peak HRR
• Very limited burning
A mattress which starts to burn in a small room scenario can easily develop into a hazardous situation. People in the room must escape before the fire grows beyond a certain critical size. The available time for escape is generally defined as the time from discovery of the fire until untenable conditions are reached.
A mattress that fulfils the proposed criteria should not constitute an immediate threat to people who are in the room of fire origin. The criteria do not cover the situation when a mattress is burning in a small room without ventilation (closed door, windows etc.). In such a situation even a small fire always constitutes a high risk. The criteria are designed to prevent the use of a mattress for the purpose of arson. Due to the low heat release and smoke production properties, a mattress that adheres to the predefined criteria will allow more time for escape in a fire emergency situation than a non-classified mattress. The risk for ignition of secondary objects also decreases.
To be able to control the size of the mattress fire a criterion is proposed on peak HRR. To avoid a scenario with an ongoing mattress fire a criterion on total heat released is also proposed. Some materials develops relatively large amounts of smoke even at low burning rates why also a criterion on total smoke produced is proposed.
The following performance criteria are recommended for fire classification of mattresses tested according to the proposed test method (see also figure 1, 2 and 3):
• The maximum heat release rate, HRRsmoothpeak shall not exceed 55 kW
(exclusive the burner heat output).
• The total heat released during the first 10 minutes of the test, THR10 shall not
exceed 10 MJ (exclusive the burner heat output).
• The total smoke produced during the first 10 minutes of the test, TSP10 shall not
6
Conclusion
The project has successfully developed a test method for full scale fire testing of mattresses (see Appendix A).
The chosen burner heat output represents a relatively large ignition source, in the range of what can be expected in the case of arson etc. This means that the test method is suitable to discriminate between good and worse fire behaviour from a heat and smoke production point of view. The most accurate measuring technique for heat release and smoke
production is applied in the proposed test method.
Proposed criteria have been chosen to secure safe mattresses which produce low quantities of heat and smoke and which do not add significantly to the fire load of the burning room. The criteria offer protection against fully developed fires in mattresses in the event of arson in premises where people may be prevented from escaping without external assistance.
The proposed test method has been applied to 10 commercial mattresses and proved sufficient stability and discrimination of the products tested.
The proposed test method will give Nordic manufacturers of mattresses a common test platform to define fire properties for mattresses used in public occupancies.
7
References
1. Nordtest Method NT FIRE 032 Upholstered Furniture: Burning Behaviour – Full Scale Test, edition 2, 1991-05, ISSN 0283-7188
2. Hertzberg, Tommy, Tuovinen, Heimo, Blomqvist, Per, SP report 2005:29, Measurement and simulation of fire smoke BRANDFORSK PROJEKT 702-041 031.
3. Sundström, Björn, CBUF, Fire safety of upholstered furniture – The final report of the CBUF research programme, European Commission Measurement and Testing Report EUR 16477 EN, Interscience Communication Limited.
4. Technical Bulletin 133, Flammability Test Procedure for Seating Furniture for Use in Public Occupancies, State of California Department of Consumer Affairs Bureau of Home Furnishings and Thermal Insulation, January 1991.
5. Technical Bulletin TB 129 Flammability test procedure for mattresses for use in public buildings, State of California Department of Consumer Affairs Bureau of Home Furnishings and Thermal Insulation, January 1992.
6. Technical Bulletin 603 Requirements and test procedure for resistance of a mattress/box spring set to a large open-flame, State of California Department of Consumer Affairs Bureau of Home Furnishings and Thermal Insulation, January 2004.
7. CEN prEN 14390 Fire test – Large-scale room reference test for surface products, April 2004.
8. Axelsson, Jesper et al, SP report 2001:04, Uncertainties in measuring heat and smoke release rates in the room/corner test and the SBI. Nordtest technical report 477, Nordtest project 1480-00.
9. Thureson Per, SP Report 2000:15, Brand i vårdanläggningar, SP Fire Technology, Borås 2000.
Appendix A Proposed test method
Contents
(The contents are for practical reasons omitted and should be added when the test method is put into the standardization template).
Mattresses:
BURNING BEHAVIOUR - FULL SCALE TEST
Key words: mattresses - fire, full-scale fire test, heat release rate, smoke production rate, fire test
0
INTRODUCTION
0.1
This test method is intended to describe the fire properties of a product under controlled laboratory conditions.
0.2
Users of this test method should observe the following warning:
SAFETY WARNING - IN ORDER THAT SUITABLE PRECAUTIONS MAY BE TAKEN TO SAFEGUARD HEALTH, THE ATTENTION OF ALL ENGAGED IN FIRE TESTS IS DRAWN TO THE POSSIBILITY THAT TOXIC OR HARMFUL GASES MAY BE EVOLVED DURING COMBUSTION OF TEST SPECIMENS. When handling the gas system for the ignition source, measures must be taken to avoid any risk of explosion.
1 SCOPE AND FIELD OF APPLICATION
1.1
This method is intended to evaluate the contribution to fire growth and smoke production under well ventilated conditions provided by a mattress, bed system etc. when subjected to a specified ignition source. Data describing the burning behaviour from ignition until all burning has ceased is obtained.
The heat release rate of the burning specimen is measured by an oxygen consumption technique. The hazard of reduced visibility is estimated by measuring the production of light obscuring smoke. The burning behaviour is visually documented by photographic and/or video recordings.
The amount of certain toxic gas species in the combustion gases can be optionally analyzed.
2 REFERENCES
1. ISO 13943, Fire safety - Vocabulary.
2. ISO 9705 – Fire tests – Full scale room test for surface products.
3. CIE. “The basis of physical photometry”. 2nd Rev. Ed. of Publication CIE No.
18.2.1983. UDC 662.61:645.4
4. Parker, W.J., “Calculations of the heat release rate by oxygen consumption for various applications”, Journal of Fire Sciences, Vol. 2, September/October 1984.
5. Technical Bulletin 133 – Flammability Test Procedure for seating Furniture for use in Public occupancies, January 1991.
6. NT FIRE 047, Combustible products: Smoke gas concentration, continuous FTIR analysis
3 DEFINITIONS
For the purposes of this test method, the definitions given in ISO 13943 apply together with the following:
Product: The mattress, bed system etc. about which information is required. Specimen: A representative item of the product which is to be tested.
4 SAMPLING
A representative sample of the product shall be tested.
5 TEST PRINCIPLES
5.1
A mattress is placed on a weighing platform. The platform is located under a hood that extracts all the combustion gases. There should be virtually no obstructions to the air supply to the test set-up. Probes for sampling of gas and for measurement of volume flow rate are placed in the exhaust duct leading from the hood. A photocell lamp system for measurement of light obscuration is installed across the exhaust duct. A general test set-up is shown in Fig. 1.
The procedure can be extended to cover measurement of certain gaseous combustion products under highly ventilated conditions as input data for studies of toxicological hazard.
5.2
The specimen is ignited with a propane gas burner from above. During the test,
concentrations of oxygen, carbon dioxide and carbon monoxide, light obscuration by the smoke and volume flow rate are measured in the exhaust duct. Mass loss rate of the burning sample is measured by means of a weight measuring device (optional). From these measurements the heat release rate, the production rate of gas species and smoke production rate (light obstructing) are calculated.
These values together with visual recordings constitute the results from the test. Concentrations of certain toxic gases could optionally be measured in the exhaust duct using FTIR (Fourier Transform InfraRed) technique. Guidelines are given in Annex F.
6 IGNITION SOURCE
6.1
The ignition source is a square gas burner as given in annex E [5]. All equipment such as tubes, couplings, flow meters, etc., shall be approved for propane. The installations shall be performed according to existing national regulations.
The burner shall be run with natural grade propane (at least 95% purity) at a heat output of 30 kW. The gas flow to the burner shall be measured with an accuracy of at least 3%. The heat output from the burner shall be controlled within 5% of the prescribed value. The burner is placed centrally above the mattress as given in 12.2.3.
7 WEIGHING PLATFORM
7.1
A weighing platform can be used to continuously measure the mass loss of the burning sample (optional). The weighing platform shall consist of a slab placed on top of a weight measuring device.
7.2
The slab shall have the dimensions 1.2 m x 2.4 m and be of a noncombustible material, e.g. calcium silicate boards. The boundary shall have a frame of 10 cm height in order to prevent melting or falling material from the tested sample to fall off the slab.
7.3
The weight measuring device, e.g. load cells, shall measure the specimen mass with an accuracy of at least ±150 g up to at least 90 kg of specimen mass. It shall be installed in such a way that the heat from the burning sample and any eccentricity of the load is not affecting the accuracy. All parts of the weight measuring device should be below the top level of the slab. The distance from the upper surface of the slab to floor level shall not exceed 0.5 m.
8 HOOD AND EXHAUST DUCT
The system for collecting the combustion products shall have a capacity and be designed in such a way that all of the combustion products leaving the burning specimen are collected. There shall not be any leakage of flames or smoke. The system shall not disturb the fire enduced flow. The exhaust capacity shall be at least 3.5 m3 s-1 at normal pressure
and a temperature of 25 °C. Exhaust system design based on natural convection is not permitted. A design example of a hood and an exhaust duct is given in Annex A.
9 INSTRUMENTATION IN THE EXHAUST
DUCT
The following specifications are minimum requirements. Additional information and technical solutions can be found in Annex B and Annex F.
9.1 Volume flow rate
9.1.1 Specification
The volume flow rate in the exhaust duct shall be measured with an accuracy of at least ±5 %.
9.2 Gas analysis
9.2.1 Sampling system
The gas samples shall be taken in the exhaust duct at a position where the combustion products are uniformly mixed. The sampling line tubes shall be made of a material not influencing the concentration of the gas species to be analysed, see Annex B and Annex
F.
9.2.2 Oxygen analyser
The O2 analyser shall be of the paramagnetic type or equivalent in performance and
capable of measuring a range of at least 0-21 Volume % oxygen (VolumeO2/Volumeair).
The uncertainty of measurement shall be ≤ 0.1 Volume % O2 or better. The stability of
the analyser shall be within 0.01 Volume % O2 over a period of 30 minutes. The output
from the analyser and also the data acquisition system shall have a resolution of 0.01 Volume % O2 or better. A procedure to check the stability of the oxygen analyser is given
in Annex B.3.
9.2.3 Carbon dioxide analyser
The CO2 analyser shall be of the IR type or equivalent in performance and capable of
measuring a range of at least 0 Volume % to 10 Volume% carbon dioxide. The
uncertainty of measurement shall be ≤ 0,1 Volume % CO2 up to 5 Volume % CO2 and ≤
0,2 Volume % CO2 from 5 to 10 Volume % CO2. The linearity of the analyser shall be
1 % of full scale or better. The output from the analyser and also the data acquisition system shall have a resolution of 0,01 Vol % CO2 or better.
9.2.4 FTIR analysis
For special purposes other gases may be analysed. A FTIR instrument equipped with a heated gas cell could be used for optional continuous analysis of certain toxic gases. Guidelines are given in Annex F.
The optical density of the smoke is determined by measuring light obscuration with an incandescent lamp photometer system. The system shall be constructed so that soot deposits during a test do not reduce the light transmission by more than 5 %. The light beam shall cross the exhaust duct along its diameter at a position where the smoke is homogenous. More information on the system and its calibration is given in Annex B.4.
10 CALIBRATION
This paragraph describes specific system calibrations before testing or after changes of the system.
10.1
A calibration test according to paragraphs 10.2, 10.3, 10.4 and 10.5 (if mass loss is measured) shall be performed prior to each test or continuous test series. Paragraphs 10.6 and 10.7 refer to a basic calibration to be performed on a new system, when modifications are introduced or at any other occasion when required. Equations for calculations are given in Annex C.
10.2
The calibration of the instrumentation in the exhaust duct shall be performed by the burning of propane gas and comparing the heat release rates calculated from the metered gas input and the measured oxygen consumption. A suitable burner is the ISO 9705 sand burner. The burner flame shall not be premixed with air. The propane shall be of at least 95 % purity. The gas flow to the burner shall be measured with an accuracy of at least ±3 %. The burner shall be positioned directly under the hood. The heat output shall be 100 kW.
Measurements are taken at least every 3 s and shall be started 1 minute prior to ignition of the burner. The difference between the time average value of the heat release rate
(calculated from the measured oxygen consumption) and the heat release rate calculated from the metered gas input must not exceed 5 %. The measurement shall be recorded over a period of one minute, and should only be made when steady state conditions have been reached
10.3
10.3.1 Daily calibration of oxygen analyzer
The oxygen analyzer shall be adjusted for zero and span, each day on which tests are performed. The span width shall be within 0.04 % of the width defined by the calibration gases used. The analyzer output for dried ambient air shall be (20.95 ± 0.01) %. Further information is given in Annex B.
10.3.2 Daily calibration of carbon dioxide analyser
The carbon dioxide analyser shall be adjusted for zero and span, each day on which tests are performed. The span width shall be within 0.1 % of the width defined by the
calibration gases used. The analyzer output for carbon dioxide-free nitrogen shall be (0.00 ± 0.02) %.
10.4
For the daily calibration of the smoke measuring system adjust the signal from the light receiver to 100 % with the smoke measuring system and the exhaust duct system equipment operating. Further information is given in Annex B.
10.5
Calibration of the weight measuring device shall be performed by loading the weighing platform with known masses corresponding to the measuring range of interest, to ensure that the requirements of accuracy in paragraph 7.3 are fulfilled.
10.6
In order to minimize the errors due to time delays in the gas analysis system the delay times shall be established by the following test sequence:
Time Burner heat output
Min kW
0 to 2 0
2 to 7 100
7 to 9 0
The stepwise changes of the burner heat output shall be completed within the same time period as the time period between two scans of the data acquisition system. Measure the time delay from the moment the burner output is changed from 0 kW to 100 kW to when the analysers start to respond. The time delay shall be maximum 20 seconds. Note that the time delay for the oxygen analyser in most cases differs from that of the CO and CO2
analyser. The time delays for the different gas analysers shall be introduced in the heat release rate calculations as a time shift between volume flow and gas concentrations.
10.7
The precision of the system at various volume flow rates is checked by increasing the volume flow in the exhaust duct in four steps, starting from 1.5 m3 s-1 (at 0.1 MPa and 25 °C) up to maximum. The heat output from the burner shall be 100 kW. The drift, in heat release rate, comparing time average values over one minute, shall not be more than 10 % of the actual heat output from the burner.
11 PREPARATION OF TEST SPECIMENS
11.1
Full size items of mattresses shall be tested on the mock-up bed frame (see Annex D) as they are delivered except for conditioning according to section 11.2. Note that for self supporting mattresses, bed systems etc. the mock-up bed frame may not be used.
11.2
Prior to testing, the specimen shall be conditioned for two weeks or to equilibrium in an atmosphere of (50 ± 5) % relative humidity at a temperature of (23 ± 2) °C. Equilibrium shall be deemed to be reached when the item has achieved constant mass. Constant mass is considered to be reached when two successive weighing operations, carried out at an interval of 24 hours, do not differ by more than 0.1% of the mass of the specimen or 0.1g, whichever is the greater. All packing and preserve material (i.e. plastic wraps) shall be removed prior to conditioning.
12 TESTING
12.1 Initial conditions
12.1.1
The ambient temperature shall be 20 ± 5 °C.
12.1.2
The environment around the specimen shall be a draught free area with no more than two enclosing walls. An enclosing wall is defined as a wall closer than 2 m from the outer edge of the smoke collection hood. The horizontal wind draught measured at a distance of 0.5 m from the boundary of the weighing platform in level with the slab shall not exceed 0.5 m s-1.
12.1.3
The specimen shall prior to the test be weighed with an accuracy of ± 100 g. The mock-up bed shall be centered on the weighing platform before the test. The test set-mock-up shall be centrally placed under the hood of the calorimeter. The time from the end of the
conditioning to the start of the test shall be noted
12.1.4
The specimen shall be photographed or video recorded prior to testing with a sign identifying:
Testing laboratory Date
Sample identification
12.1.5
Means for extinguishing a fully developed fire shall be available.
12.2 Procedure
12.2.1
All recording and measuring devices shall be started and data taken at least two minutes prior to the test. Measurements are taken at least every 3 s.
12.2.2
The volume flow rate shall be set to approximately 1.5 m3 s-1. This flow rate shall be
maintained unless it is necessary to raise it in order not to loose any combustion products.
The burner ring shall be positioned level in horizontal plane. The burner shall be positioned immediately over the centre of the mattress, 25 mm above the mattress surface.
12.2.4
After the baseline recording, the burner, which has previously been put in position, shall be lit. The propane gas flow shall be controlled to 0.65 g s-1 (19.5 L min-1 at 273 K and
100 kPa), corresponding to 30 kW. The burner shall be applied to the test object for 120 ± 2 s. After the ignition period the burner shall be moved out of the testing area without disturbing the mass loss measurement.
12.2.5
At the time of the ignition of the gas burner a clock shall be started. A photographic and video recording shall be performed during the test. In all visual records, preferably, the time from ignition of the burner shall appear.
12.2.6
During the test it is essential that the following events are observed if they occur and the time for their occurrence:
- ignition of the product - position of flame front
- flame front reaching the extremities of the specimen - melting and dripping
- occurrence of pool fire under the product
- excessive smoking in some local area (suggesting smouldering combustion) - signs of combustion below the cover
- smoke not captured by the hood - any changes to the duct flow rate
- early termination of test due to excessive burning - any other event of interest
12.2.7
The test shall continue until 2 minutes after all visible flaming has ceased or for maximum 60 minutes. After the test, photographs and notes shall be taken showing the extent of damage on the tested object.
Note: Due to safety of staff and equipment, tests may have to be interrupted due to excessive flaming and smoke production. This shall be clearly stated in the test report.
13 TEST REPORT
The test duration used for calculations shall be from ignition of the burner until end of test as defined in 12.2.7.
The test report shall contain the following information: a) Name and address of the testing laboratory
b) Date and identification number of the report c) Name and address of the client
d) Purpose of the test e) Method of sampling
f) Name of manufacturer or supplier of the product
g) Name or other identification marks and description of the product
h) Density or weight per square unit and thickness of the main components in the product i) Description of the specimens
j) Conditioning of the specimens k) Date of test
l) Test method
m) Test results (see also Annex C)
graphs of heat release rate (30 s sliding average, HRRsmooth)
graphs of production rate of light obstructing smoke (SPR) description of the fire development (notes and photographs) table of numerical results containing:
peak heat release rate (30 s sliding average, HRRsmooth)
total heat released until end of test (THR) total heat released at 10 minutes from ignition peak production rate of light obstructing smoke total smoke produced until end of test (TSP) total smoke produced at 10 minutes from ignition When appropriate also:
graphs of mass burning rate
graphs of production rate of carbon dioxide graphs of mass flow in the exhaust duct
Effective heat of combustion determined from the quotient between the measured total heat released and the total mass loss
The production rates and amounts of gas species as analysed according to Annex F n) When appropriate, designation of the product according to criteria expressed in official
standards or regulations
o) Deviations from the test method, if any
ANNEX A
DESIGN OF EXHAUST SYSTEM
A.1
Hood and exhaust duct, recommended design
A.1.1
The combustion gases from the burning specimen are collected by a hood. Below a system is described which has been tested in practice and proved to fulfill the requirement specifications given in the method.
A.1.2
The hood is located centrally above the weighing platform. The bottom dimensions of the hood are 3 m x 3 m and the height 1.0 m, see Fig. A:1. On four sides, steel sheets are extended 1.0 m downwards. The effective height of the hood will thus be 2 m, and the distance between the lower edge of the hood and the weighing platform should be 1.5-2 m. The hood feeds into a plenum having a cross section area of 0.9 m x 0.9 m and a height of 0.9 m. In the plenum chamber two plates approximately 0.5 m x 0.9 m are located to increase mixing of the combustion gases.
A.1.3
An exhaust duct is connected with the plenum chamber. The rectilinear part of the exhaust duct must have such a length that a fully-developed flow profile is established at the point of measurement. For a 400 mm diameter exhaust duct the rectilinear part must be at least 4.8 m. The exhaust duct is connected to an evacuation system. To facilitate flow measurement, guide vanes are located at both ends of the exhaust duct.
A.1.4
The capacity of the evacuation system shall be designed to exhaust minimally all combustion gases leaving the specimen. This requires an exhaust capacity of at least 4 kg s-1 (about 12000 m3 h-1 at standard atmospheric conditions) corresponding to a
driving underpressure of about 2 kPa at the end of the duct. It shall be possible to control the exhaust flow from about 0.5 kg s-1 up to maximum flow as stated above during the
test process. If the air flow is not decreased during the initial part of the test, the measurement precision will be too low.
A.1.5
An alternative exhaust system design may be used if it has been shown to produce equivalent results. Equivalency may be demonstrated by meeting the requirements under paragraph 9.
Key 7 Exhaust duct Ø 400 mm
1 Guide vanes 8 Opening 3000 mm x 3000 mm
2 Pitot tube 9 Frame of steel profile 50 mm x 100 x 3.2 mm 3 Guide vanes 10 Steel plates 1000 x 3000 mm
4 To exhaust gas cleaning 11 Steel plates 2 mm x 500 mm x 900 mm 5 Lamp, photocell system 12 Hood of 2 mm thick steel plates 6 Gas analysis 13 Four steel plates 395 mm x 400 mm
ANNEX B
INSTRUMENTATION IN EXHAUST DUCT
B.0
Suitable locations for the probes described below are shown in Fig. A:1.
B.1
Volume flow
B.1.1
The flow may be measured by a bidirectional probe located at the centre line of the duct. The probe shown in Fig. B:1 consists of a stainless steel cylinder, 32 mm long and with an inner diameter of 14 mm. The cylinder has a massive wall in the centre, dividing it into two chambers. The pressure difference between the two chambers is measured by a pressure transducer.
B.1.2
The pressure transducer shall have an accuracy of at least ±5 Pa and be of the capacitance type. A suitable range of measurement is 0-2000 Pa.
B.1.3
Gas temperature in the immediate vicinity of the probe is measured by a thermocouple with a diameter of maximum 0.25 mm. The thermocouple shall not be allowed to disturb the flow pattern around the bidirectional probe
B.2
Sampling system
B.2.1
The sampling probe shall be located where the exhaust duct flow is well mixed, see Fig. A:1. The probe shall have a cylindrical form so that disturbance of flow is minimized. The gas samples shall be taken along the whole diameter of the exhaust duct. The intake of the sampling probe is turned downstream in order to avoid soot clogging in the probe. A suitable sampling probe is shown in Fig. B:2.
B.2.2
The sampling system, see Fig. B:3, shall be manufactured from non corrosive material, e.g. PTFE. The combustion gases shall be filtered with inert filters to protect the gas analysis equipment from soot, etc. The filtering procedure should be carried out in more than one step. Appropriate measures shall be taken (e.g. cooling) so that the gas mixture is dry before it is taken to each analyzer.
B.2.3
The combustion gas is transported by a pump which does not emit oil, grease or similar products, which can contaminate the gas mixture. A membrane pump is suitable.
B.2.4
The sampling line may end in an open container with atmospheric pressure. The volume of the container shall not be so large that concentration gradients or time lags are generated. Transport time in the sampling line shall be ≤5 s.
B.2.5
A suitable pump shall have the capacity of 10-50 I min-1, as each gas analysis instrument
consumes about 1 I min-1. The pump shall generate a pressure differential of at least 10
kPa to reduce the risk of smoke clogging of the filters.
Key
1 To ∆P instrument
2 Variable length support tubes 3 Weld
Note Taken from McCaffrey and Heskestad
50mm
10mm 50mm
Exhaust Duct
ø 10mm ø 2mm ø 3mm
Holes on down stream side of flow
Key 8, 9 Cooler system
1 Glass filters 150 µm to 200 µm 10 Membrane filter 3 µm 2 Thermocouple 11 Surplus gas
3 Sampling line, diameter 10 mm 12 Water drainage
4 Lamp 13 Filter for water absorption 5 Photocell 14 Membrane pump
6 Pitot tube 15 Paramagnetic (O2)
7 Exhaust duct 16 Infrared spectrophotometer (CO, CO2)
B.3
Procedure to check the stability of the oxygen
analyzer
The oxygen analyser shall comply with the requirements specified in 9.2.2.
The stability of the oxygen analyser output using the data acquisition system shall be checked after set up, maintenance, repair or replacement of the oxygen analyser or other major components of the gas analysis system and at least every six months.
The procedure for checking the stability of the oxygen analyser output shall be as follows. a) Feed the oxygen analyser with oxygen-free nitrogen gas, until the analyser reaches equilibrium.
b) After at least 60 minutes in oxygen-free conditions, adjust the volume flow in the exhaust duct to (1,5 ± 0,5) m3 s-1 and switch to air from the exhaust duct with the same
flow rate, pressure and drying procedure as for sample gases. When the analyser reaches equilibrium, adjust the analyser output to (20,95 ± 0,01) %.
c) Within 1 minute, start recording the oxygen analyser output at intervals 3s for a period of 30 minutes.
d) Determine the drift by use of the least squares fitting procedure to fit a straight line through the data points. The absolute value of the difference between the readings at 0 minutes and at 30 minutes of this linear trend line represents the drift.
e) Determine the noise by computing the root-mean-square (rms) deviation around the linear trend-line.
The sum of drift and noise (both taken as positive values) shall not be more than 0.01 % (
VO2 /Vair ).
B.4
Optical density
B.4.1
White light system - Incandescent lamp photometer
The lamp shall be of the incandescent filament type and shall operate at a colour
temperature of (2 900 ± 100) K. The lamp shall be supplied with stabilised direct current, stable within ± 0,2 % (including temperature, short-term and long-term stability). The lens system shall align the light to a parallel beam with a diameter D of at least 20 mm. The aperture shall be placed at the focus of the lens L2 as shown in Figure B.4 and it shall
have a diameter, d, chosen with regard to the focal length, f, of L2 so that d/f ≤0,04. Other
solutions to avoid light scattering are allowed. The detector shall have a spectrally distributed responsivity in agreement with the CIE, V (λ)-function (the CIE [3] photopic curves) to an accuracy of at least ± 5 %. The detector output shall over an output range of at least 2 decades be linear within 5% of measured transmission value or 1 % absolute transmission.
B.4.2 Calibration
The white light system calibration shall be performed before a test is conducted after set up, maintenance, repair or replacement of the smoke measurement system holder or other major components of the exhaust system and at least every six months. The calibration consists of two parts: an output stability check and an optical filter check.
B.4.2.1 Stability check
Perform the following steps with the measuring equipment operating
a) Set the volume flow of the exhaust to: V298 = 1,5 ± 0,5 m3 s-1.
b) Start the time measurement and record the signal from the light receiver for a period of 30 min.
c) Determine the drift by use of a least squares fitting procedure to fit a straight line through the data points. The absolute value of the difference between reading at 0 min and at 30 min of this linear trend line represents the drift.
d) Determine the noise by computing the root-mean-square (rms) deviation around the linear trend line.
Criterion: Both noise and drift shall be less than 0,5 % of the start value.
B.4.2.2 Optical filter check
The light system shall be calibrated with at least five neutral density filters in the optical density range of 0,05 to 2,0. The optical density calculated with the measured light receiver signal shall be within ± 5% or ± 0,01 of the theoretical value of the filters.
SMOKE PARTICLES Lamp L1 L2 Aperature Detector Wall of exhaust duct
ANNEX C
CALCULATION EQUATIONS
C.1
Volume flow in the exhaust duct
For the instrumentation described under paragraph B.1 the volume flow V298 (m3 s-1) in
the exhaust duct, related to atmospheric pressure and an ambient temperature of 25 °C is given by the relation
2 / 1 298 298
(
2
/
)
1
)
/
(
Ak
tk
ppT
o oT
sV
ρ
ρ
×
∆
×
=
&
= 22 4, (Akt /kp)(∆p T/ s)1/ 2 whereTs is the gas temperature in the exhaust duct, expressed in Kelvin (K);
To = 273,15 K;
∆p is the pressure difference measured by the bi-directional probe, expressed in pascal (Pa);
ρ298 is the air density at 25 °C and atmospheric pressure, expressed in kilograms
per cubic metre (kg m-3);
ρo is the air density at 0 °C and 0,1 MPa, expressed in kilograms per cubic
metre (kg m-3);
A is the cross-sectional area of exhaust duct, expressed in square metres (m2);
kt is the ratio of the average mass flow per unit area to mass flow per unit area
in the centre of the exhaust duct;
kp is the Reynolds number correction for the bi-directional probe, taken as
constant and equal to 1,08.
The equation assumes that density changes in the combustion gases (related to air) are caused solely by the temperature increase. Corrections due to a changed chemical
composition or humidity content may be neglected except in studies of an extinguishment process with water. The calibration constant kt is determined by measuring the
temperature and flow profile inside the exhaust duct along a cross sectional diameter. Several series of measurements are to be made with representative mass flows and with both warm and cold gas flows. The kt factor is to be determined with a maximum error of ±3 %. Further guidance on how to determine the kt factor is given in ISO 3966:1977.
C.2
Heat release rate, calibration and test process
C.2.1
During the calibration process, heat release rate from the ignition source,
&q
b, expressed in kilowatts, shall be calculated from the consumption of propane gas using the equation& & ,
qb=mb∆hc eff
where &
mb is the mass flow rate of propane to the burner, expressed in grams per second
(g s-1);
∆hc,eff is the effective lower heat combustion of propane, expressed in kilojoules
per gram (kJ g-1). ∆h
c,eff shall be set equal to 46,4 kJ g-1 (this assumes a
combustion efficiency of 100%).
C.2.2
The heat release rate (HRR) from a tested product, expressed in kilowatts, is calculated according to the equation
b H C a O q E E x V E
q& & &
8 3 2 1 298 1 1 ) 1 ( ⎟⎟⎠− ⎞ ⎜⎜ ⎝ ⎛ + − =
α
φ
φ
with Φ, the oxygen depletion factor, given by
φ = − − − − − x x x x x x x O CO O CO O CO O 2 2 2 2 2 2 2 0 0 0 1 1 1 ( ) ( ) ( ) and xOa
2 , the ambient mole fraction of oxygen, given by
xOa2 =xO02(1−xH Oa2 ) and xHa2O given by
.
exp[
23
.
2
{
3816
}]
100
2p
Ts
H
x
a O H=
−
whereE is the heat release per volume of oxygen consumed, expressed in kilojoules per cubic meter (kJ m-3),
E1 = 17,2 x 103 kJ m-3 (25 °C) for combustion of tested product;
&
V298 is the volume flow rate of gas in the exhaust duct at atmospheric pressure
and 25°C calculated as specified in (C.1), expressed in cubic metres per second (m3 s-1);
α is the expansion factor due to chemical reaction of the air that is depleted of its oxygen (α = 1,105 for combustion of tested product);
xOa
2 is the ambient mole fraction of oxygen including water vapour;
Note xOa
2 can also be measured prior to the test by measuring the oxygen content of
the ambient air without trapping of water.
xO 2
0 is the initial value of oxygen analyser reading, expressed as a mole fraction;
xO
2 is the oxygen analyser reading during test, expressed as a mole fraction;
xCO0 2 is the carbon dioxide analyser reading during test, expressed as a mole
fraction; xCO
2 is the carbon dioxide analyser reading during test, expressed as a mole
fraction; xH Oa
2 is the ambient mole fraction of water vapour.
H is the relative humidity (%) in the laboratory before the start of the test.
p is the ambient pressure (Pa) before the start of the test.
Ts is the temperature in the measurement section of the duct (K).
C.2.3
The relations above are based on certain approximations leading to the following limitations.
- No reference is made to the amount CO generated. Normally, the error is negligible. As concentration of CO is measured, corrections can be calculated in those cases where the influence of incomplete combustion may have to be quantified.
- The influence of water vapour on measurements of flow and gas analysis is only partially taken into consideration. A better correction for this error can only be obtained by a continuous measurement of partial pressure of water vapour.
- The factor E = 17.2 MJ m-3 (or 13.1 kJ g-1) is an average value for a large number of
products and gives an acceptable accuracy in most cases. It should be used unless a more accurate value is known, in which case the used E-value should be reported. The accumulated corrections of the error sources enumerated above should normally be less than 10 %. More information is given in ISO 9705 part 2.
C.2.4
Calculation of HRRsmooth is performed as follows.
HRRsmooth (t) is the 11 records seconds average, (30 s) of HRR(t).
11 ) 15 ( ) 12 ( ... ) 12 ( ) 15 ( ) (t HRR t s HRR t s HRR t s HRR t s HRRsmooth = − + − + + + + +
During the first and the last 15 s of a test the calculation of HRRsmooth according to the
equation above does not apply as the required 11 records are not available. For those cases the following apply.
Beginning of test:
For t = 0 s: HRRsmooth = 0 kW
For t = 3 s: HRRsmooth = HRR average over the period (0s…6s)
For t = 6 s: HRRsmooth = HRR average over the period (0s…12s)
For t = 9 s: HRRsmooth = HRR average over the period (0s…18s)
For t = 12 s: HRRsmooth = HRR average over the period (0s…24s)
For t ≥ 15 s: HRRsmooth is calculated according to the equation above
End of test:
HRRsmooth is calculated according to the equation above until the data point which in time
is 15 s from end of test.
Note that the calculation of HRRsmooth given above is made on the assumption that a 3 s
sampling interval is used.
C.2.5
The total heat release from a tested product, THR, is calculated as:
s t HRR t THR t s 3 ) ( 1000 1 ) ( 0 • =
∑
whereTHR(t) = total heat release during 0 s – t s (MJ); HRR(t) = heat release rate (kW).
The time t is selected according to requirements. The time t may maximum be until end of test.
C.3
Combustion gases
By measuring the partial pressure of a specified gas it is possible to calculate the instantaneous rate of gas production and the total amount of gas production
& &
Vgas=V298xi
Vgas=
∫
0tV&gasdtwhere
&
Vgas is the rate of gas production, expressed in cubic meters per second at 0,1 MPa and 25 °C (m3 s-1)
Vgas is the total amount of gas production expressed in cubic meters at 0,1 MPa
and 25 °C (m3)
&
V298 is the rate of volume flow in exhaust duct, expressed in cubic meters per
second at 0,1 MPa and 25 °C (m3 s-1)
xi is the mole fraction of specified gas in the analyzer t is the time from ignition, expressed in seconds (s).
C.4
Light obscuration
The optical density is represented by the extinction coefficient, k, expressed in reciprocal meters (m-1), and shall be calculated from the following:
k L I I o = ⎡ ⎣⎢ ⎤ ⎦⎥ 1 ln where
lo is the light intensity for a beam of parallel light rays measured in a smoke
free environment with a detector having the same spectral sensitivity as the human eye;
I is the light intensity for a parallel light beam having traversed a certain length of smoky environment;
L is the length of beam through smoky environment, expressed in meters (m). The instantaneous smoke production rate, SPR, expressed in square meters per second (m2 s-1) shall then be calculated from:
s
V
k
SPR
=
&
&
Vs is the volume flow in the exhaust duct at actual duct gas temperature,
expressed in cubic meters per second (m3 s-1);
t is the time from ignition, expressed in seconds
C.4.1
The total amount of smoke, TSP, expressed in square meters (m2) is calculated as follows
s t SPR t TSP t s 3 ) ( ) ( 0 • =
∑
whereTSP(t) = total smoke production during 0 s – t s (m2);
SPR(t) = smoke production rate (m2/s).
The time t is selected according to requirements. The time t may be the time at the end of the test.
ANNEX D
DESIGN OF MOCK-UP BED FRAME
D.1
The mock-up bed shall consist of a metal frame on which the mattress is placed.
D.2
The frame shall have a total length of 2.0 m, a width of 0.9 m and an over-all height of 0.4 m, see Fig. D:1 a-d. The mock-up bed frame should be made of L-shaped or square steel bars. The horizontal surface (on which the mattress is placed) shall be made from 15 bars, evenly spaced with a c-c distance of 125 mm approximately.
25
~100
25
25
2000
50
2000
Side View
Dimensions in mm
a
b
c
d
25
ANNEX E
SQUARE GAS BURNER IGNITION SOURCE
E.1
The 250 x 250 mm square burner shall be constructed of 12 mm OD stainless steel tubing with a 1 mm wall thickness (corresponding to ½ inch OD, 0.035 inch wall thickness approximately). The front side shall have 14 holes pointing straight out and spaced 13 mm apart and 9 holes pointing straight down and spaced 13 mm apart. The right and left sides shall have 6 holes pointing straight out and spaced 13 mm apart, and 4 holes pointing at 45 degree angle inward (downward direction) and spaced 50 mm apart. All holes shall be of 1 mm diameter. The 1070 mm straight arm of the burner shall be welded onto the rear of the front side in a 30 degree angle. The burner shall be mounted on an adjustable height pole and be balanced by a counter weight or other appropriate mechanism.
Note: When the flow of propane to the burner is stopped the burner shall be removed from the mattress and away from high temperature gases. Care must be taken to allow free flow of propane through the burner holes. Periodic cleaning of soot deposits and blowing pressurized air through the tube is recommended.
ANNEX F
GAS ANALYSIS – FTIR
F.1
Principle of analysis method
By using on-line FTIR (Fourier Transform InfraRed) technique, it is possible to simultaneously measure the time resolved concentration of several gases in a fire experiment.
FTIR is based on infrared absorption. Polyatomic and heteronuclear diatomic compounds have absorption in the infrared region. Specific to FTIR is conversion of regular
irradiance from a broad band infrared source into interfered irradiance by an interferometer, and conversion of the recorded interferogram into a conventional
wavelength spectrum. The main advantage of the FTIR technique is that information from all spectral elements is measured simultaneously; another advantage is that the
measurement is made with a high optical throughput.
The practical measurement procedure is to continuously extract smoke gases to the FTIR from the furniture calorimeter gas collecting system trough a heated sampling system. The gas sample is in the FTIR drawn trough a heated IR absorption cell, in which the specific absorption patterns of IR-active gases are recorded by a detector and are presented as an absorption spectrum. The frequency on the collection of absorption spectra determines the time resolution of the measurement. The concentration data of individual gas components is evaluated from the absorption spectra by various mathematical methods.
F.2
General on the measurement of toxic gases
A number of toxic gases can be generated from the combustion of the type of products (mattresses) referred to in this standard. Products containing nitrogen (e.g. polyurethane) have the potential to generate, most important, hydrogen cyanide (HCN), nitrogen oxide (NO) and nitrogen dioxide (NO2). All of these gases can be measured by FTIR technique.
It is, however, necessary to acknowledge that the production of these gases is to a great extent determined by the present combustion conditions. The production of HCN is increased with vitiation, whereas the production of NOx (NO and NO2) is increased with
ventilation. Generally, the results from a fire test are only representative for these specific test conditions only. Other combustion conditions may change the production behaviour Chlorine may be a component of the tested product (as a flame retardant or integrated in the material, e.g. PVC, chloroprene). Hydrogen chloride (HCl) is produced from
materials containing chlorine. The production of HCl is generally less sensitive to the present combustion condition. Special attention has to be given to the sampling procedure for HCl (as well as for other halogen acids), as it dissolves easily in any condensed water and may be trapped in cold spots in the sampling system.
There may further be additional toxic gases produced appropriate for measurement by FTIR technique. Apart from carbon dioxide (CO2) and carbon monoxide (CO) that could
be measured by NDIR technique, can e.g. hydrogen bromide (HBr) and sulfur dioxide (SO2) be suitable for measurement using FTIR. Additionally, could smaller organic
