Smoke Characterization Project

Technical report





Prepared by:
Thomas Z. Fabian, Ph.D.
Pravinray D. Gandhi, Ph.D., P.E.
Underwriters Laboratories Inc.

FOREWORD
Residential smoke alarms provide an important notification to individuals within a
residential setting that there is a presence of smoke and/or fire. Over the last four
decades, several studies have been conducted to determine the response of
smoke alarms and to assist in establishing performance criteria for their use in
residential settings.
With the advent of new smoke particulate and the gas effluent measurement
technologies becoming commercially available, UL, with support from FPRF
initiated this research project to more fully characterize the products of flaming
and non- flaming combustion. The materials investigated included a range of
products and chemistries commonly found in today's residential settings.
The objectives of the investigation were as follows:
. Develop smoke characterization analytical test protocols using non- flaming
and flaming modes of combustion on selected materials found in residential
settings.
. Using materials from the analytical smoke program, develop smoke particle
size istribution data and smoke profiles in the UL 217/UL 268 Fire Test Room for
both nonflaming and flaming modes of combustion.
. Provide data and analysis to the fire community for several possible initiatives:
a. Develop recommendations to the current residential smoke alarm standard
(UL 217).
b. Development of new smoke sensing technology.
c. Provide data to the materials and additives industries to facilitate new
smoke suppression technologies and improved end products.
The Research Foundation expresses gratitude to the report authors Thomas
Fabian and Pravinray Gandhi of Underwriters Laboratories Inc; and to the Project
Technical Panelists and sponsors listed on the following page.
The content, opinions and conclusions contained in this report are solely those of
the authors.
Smoke Characterization Project
Technical Panel
David Albert, InnovAlarm
Thomas Cleary, National
Institute of Standards and Technology
Kenneth Dungan, PLC Foundation
Jackie Gibbs, Marietta Fire and Emergency Services
Daniel Gottuk, Hughes Associates, Inc.
Morgan Hurley, Society of Fire Protection Engineers
Arthur Lee, U.S. Consumer Product Safety Commission
James Milke, University of Maryland
Robert Polk, National Association of State Fire Marshals
Lee Richardson, NFPA
Sponsors
Underwriters Laboratories Inc.
BRK Brands/First Alert
Centers for Disease Control
GE Security
Honeywell Life Safety
Invensys Climate Controls
InnovAlarm
Kidde Safety
National Electrical Manufacturers Association
PLC Foundation
SFPE Foundation
Siemens Building Technologies
SimplexGrinnell




Smoke Characterization Project


Final Report
Project Number: 06CA08584
File Number: NC 5756


Underwriters Laboratories Inc.
333 Pfingsten Road, Northbrook, IL 60062


April 24, 2007

Prepared by

Thomas Z. Fabian, Ph.D.
Research Engineer
Fire, Signaling and Security Division


Pravinray D. Gandhi, Ph.D., P.E.
Global Director, Business Development
Fire, Signaling and Security Division

Reviewed by

Paul E. Patty, P.E.
Senior Research Engineer
UL Corporate Research

J. Thomas Chapin, Ph.D.
Director, Research & Development
UL Corporate Research

Underwriters Laboratories Inc. (UL) its trustees, employees, sponsors, and contractors, make no warranties,
express or implied, nor assume and expressly disclaim any legal liability or responsibility to any person for
any loss or damage arising out of or in connection with the interpretation, application, or use of or inability
to use, any information, data, apparatus, product, or processes disclosed in this Report. This Report cannot
be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.

Copyright 2007 Underwriters Laboratories Inc.
Smoke Characterization Project Final Report
ii
EXECUTIVE SUMMARY

INTRODUCTION
Residential smoke alarms provide an important notification to individuals within a residential
setting that there is a presence of smoke and/or fire. Over the last four decades, several studies 5
have been conducted to determine the response of smoke alarms and to assist in establishing
performance criteria for their use in residential settings. These studies have led to the
development and subsequent revisions of UL Standard 217 Single and Multiple Station Smoke
Alarms, as well as a National Fire Alarm Code (NFPA 72) that addresses smoke alarm
installation requirements. A study completed by NIST in 2004 reflected that smoke alarms were 10
working but there was a reduction in the margin between available and safe egress times from an
earlier study in 1975.

Fires in either a flaming or a smoldering phase provide several cues for smoke alarms. These
include smoke particulates, heat, and gas effluents (e.g., CO, CO
2
). Current smoke alarms 15
primarily utilize two types of detection technologies: photoelectric or ionization. The
photoelectric type has a light source and detects the scattering or obscuration caused by smoke
particulates. The ionization type detects changes in local ionization field within the detection
chamber resulting from the presence of smoke. Both types of alarms activate when a set
threshold is reached. While current technology smoke alarms were found in the NIST study to 20
operate within the established performance criteria, there was a difference in activation times for
the different technologies depending upon the combustion mode (flaming vs. non- flaming).

One of the conclusions drawn from the NIST study was that performance of smoke alarms could
be studied with greater precision, accuracy and confidence if there were better data available on 25
combustibility and smoke characteristics for a wider range of products used in todays residential
settings.

With the advent of new smoke particulate and the gas effluent measurement technologies
becoming commercially available, UL initiated this UL/FPRF research project to more fully 30
characterize the products of flaming and non- flaming combustion. The materials investigated
included a range of products and chemistries commonly found in todays residential settings. The
objectives of the investigation were as follows:

Develop smoke characterization analytical test protocols using non- flaming and flaming 35
modes of combustion on selected materials found in residential settings.
Using materials from the analytical smoke program, develop smoke particle size
distribution data and smoke profiles in the UL 217/UL 268 Fire Test Room for both non-
flaming and flaming modes of combustion.
Provide data and analysis to the fire community for several possible initiatives: 40
a. Develop recommendations to the current residential smoke alarm standard (UL 217).
b. Development of new smoke sensing technology.
c. Provide data to the materials and additives industries to facilitate new smoke
suppression technologies and improved end products.
45
Smoke Characterization Project Final Report
iii
METHODOLOGY
A survey was conducted of residential settings for products and materials commonly found in
settings there. Materials, contemporary to todays residential settings, in addition to the
prescribed UL 217 fire test materials were selected for this investigation based on product
chemistry and occurrence. 5

ASTM E1354 cone calorimeter was selected as it can simulate well- ventilated, early stage fires
under well-controlled radiant heating conditions. In these tests, material based combustion
properties were developed that included weight loss rate, heat and smoke release rates, smoke
particle size and count distribution, and effluent gas composition were characterized for a variety 10
of natural, synthetic, and multi-component materials in both the flaming and non- flaming mode.
The results from the cone calorimeter tests were used to identify materials for subsequent larger
scale investigations.

Intermediate scale calorimeters were used to develop test parameters (e.g. sample size, ignition 15
method) on the selected materials for subsequent evaluation in a UL 217/UL 268 Fire Test Room.
Evaluation of the UL 217 fire test protocols, and the developed fire scenarios in intermediate
calorimeters, also permitted characterization of heat and smoke release rates as well as smoke
and gas effluents closer to the combustion source. This enabled collection of smoke data prior to
aging that would be expected in the vicinity of smoke alarms in the UL 217/UL 268 Fire Test 20
Room. This methodology allows for the comparison of smoke particle sizes near the source of
the fire, as well as at the detector location.

Finally, the developed scenarios were evaluated along with the prescribed UL 217 fire tests in a
UL 217/UL 268 Fire Test Room. Smoke particle size and count distribution and gas effluent 25
composition were monitored along with ceiling air velocity and temperature and analog alarm
responses in the vicinity of standard UL 217 obscuration and Measuring Ionization Chamber
(MIC) equipment.

In this study smoke particle size and count distribution and effluent gas composition were 30
characterized using a particle size spectrometer and a gas-phase FTIR respectively.


KEY FINDINGS
The key findings of the research were as follows: 35

Gas Analysis and Smoke Characterization Measurement
1. Physical Smoke Particle Characterization - The particle spectrometer provides data on
smoke particle size and count distribution that is unavailable by traditional obscuration
and ionization techniques used to quantify smoke. 40
2. Relationship of Smoke Particle Characterization to Traditional Methods - Linear
relationships between the smoke particle data and the traditional techniques were
demonstrated such that:
a. Particle size and number count are linearly related to MIC signal change:
MIC ~ d
m
n
m
45
Smoke Characterization Project Final Report
iv
b. Number count is linearly related to scattering while particle size exhibits a second
order relationship:
2
i i
d n s

c. Number count is linearly related to obscuration while particle size exhibits a third
order relationship:
3
i i
d n
OD

l

3. Smoke Particle Aggregation - Tests conducted in the UL 217 Sensitivity Test smoke box 5
and the UL 217/UL 268 Fire Test Room indicate an aggregation of smaller smoke
particles to form larger particles as evidenced by the increase in smoke particle
concentrations in conjunction with increasing fractions of larger smoke particles. This
was more evident for non-flaming fires than flaming fires. While the settling of smoke
was observed in the Indiana Dunes study, this effect was measured and more pronounced 10
in this study.
4. Smoke Gas Effluent Composition - Gas effluent analysis showed the dominant gas
components were water vapor, carbon dioxide and carbon monoxide.

Influence of Material Chemistry 15
1. Combustion Behavior: Synthetic and Natural Materials - Cone calorimeter tests indicate
synthetic materials (e.g. polyethylene, polyester, nylon, polyurethane) generate higher
heat and smoke release rates than the natural materials (e.g. wood, cotton batting). This is
anticipated to be primarily due to the modes of degradation and chemical structure of
synthetic versus natural materials. 20
2. Charring Effects - Materials exhibiting charring behavior such as wood alter the size and
amount of smoke particles generated as the combustion process progresses.
3. Influence on Smoke Particle Size - In general, the synthetic materials tested generated
larger mean smoke particle sizes than natural materials in flaming mode.
25
Mode of Combustion
1. Flaming Combustion - Flaming combustion tends to create smaller mean particle sizes
than non-flaming combustion. This is primarily due to the more efficient conversion of
high molecular weight polymers to low molecular weight combustion products and
ultimately CO, CO
2
and H
2
O instead of organic by-products and soot. 30
2. Non-Flaming Combustion - Non-flaming combustion tends to generate greater volumes
of smoke particles for a given consumed mass than flaming combustion.

Small-Scale and Intermediate Scale Test
1. Cone Calorimeter Test - The cone calorimeter provided combustibility, smoke 35
characteristics and gas effluent data in flaming and non-flaming modes for a range of
materials studied. The smoke characterization data revealed the influences of material
chemistry, physical sample structure, and the mode of combustion. The data were found
to be repeatable. In the non- flaming mode, the heat and smoke release rates were lower
than the resolution of the cone calorimeter measurement system for several materials 40
investigated. However, the smoke particle spectrometer provided repeatable data on
smoke size and count distribution for both flaming and non- flaming modes.
2. Intermediate-Scale Test - The intermediate scale test provided a platform to scope
combustion scenarios, and provided data on the heat and smoke release rates as well as
Smoke Characterization Project Final Report
v
smoke size and count distribution for test samples subsequently used in the UL 217/UL
268 Fire Test Room. The tests also identified test samples with heat and smoke
characteristics that varied from UL 217 fire test samples such as Douglas fir, newspaper,
heptane/toluene mixture, and Ponderosa pine. In the non- flaming mode, the method used
for heating the test sample was observed to influence the smoke characteristics. The 5
heating by a hot plate provided larger particle size as compared to radiant heating.

UL 217/UL 268 Fire Test Room Tests
1. Smoke Particle Size and Count Distribution - The tests provided smoke particle size and
count distribution data in conjunction with traditional obscuration and Measuring 10
Ionization Chamber data. PU foams in the flaming mode produced the smallest particle
sizes of all materials tested.
2. Combustion Mode Effects - Changes in the combustion mode (flaming versus non-
flaming) resulted in different smoke particle size and count distributions that influenced
the response of photoelectric and ionization smoke alarms. The particle size distribution 15
for the non- flaming fires yielded larger mean smoke particle diameter than the flaming
mode fires. The ionization alarm responded quicker to flaming fires; the photoelectric
responded quicker to non- flaming fires.
3. Smoke Alarm Response to Flaming Fires - In all but one flaming test the ionization alarm
activated first. Both alarm types activated within the 4 minute time limit specified in UL 20
217 for the three UL 217 flaming test targets (Douglas fir, heptane/toluene mixture, and
newspaper). In one of two flaming tests involving PU foam with cotton/poly fabric the
photoelectric smoke alarm did not activate, however the ionization alarm did activate in
both tests. In a flaming PU foam with cotton/poly fabric test using a smaller sample size
neither alarm type activated. It should be noted that the maximum obscuration in these 25
PU foam tests was less than for Douglas fir, heptane/toluene mixture, and newspaper test
samples.
4. Smoke Alarm Response to Non-Flaming Fires - The photoelectric alarm activated first in
the non-flaming tests with the exception of the higher energy bread/toaster test in which
the ion alarm activated first. The UL 217 smoldering Ponderosa pine test triggered both 30
the ionization and photoelectric smoke alarms. For many of the other materials, the
ionization smoke alarm did not trigger. In each of these cases, the obscuration value was
less than the 10 %/ft limit specified in UL 217. It was also found that there was settling of
the smoke particles in the test room over time. Measurements from several non- flaming
tests showed that the obscuration values at the ceiling dropped over time, and the 35
maximum obscuration values were observed at the 2 feet measurement location below the
ceiling.
5. Smoke Stratification - Non- flaming fires result in changes in the smoke build up over
time, such that stratification of smoke below the ceiling occurs. This time-dependent
phenomenon results in less obscuration at the ceiling than below the ceiling. This caused 40
both detection technologies to drift out of alarm.

Future Considerations
Based upon the results of this Smoke Characterization Project, the following items were
identified for further consideration: 45

Smoke Characterization Project Final Report
vi
1. The addition of other test materials such as polyurethane foam in the flaming and non-
flaming combustion modes in UL 217.
2. Whether a smoke alarm, once triggered, should remain activated unless deactivated
manually.
3. Requiring the use of combination ionization and photoelectric alarms for residential use 5
in order to maximize responsiveness to a broad range of fires.
4. Characterize materials described in UL 217 using cone calorimeter, smoke particle
spectrometer and analytical testing.

10

Smoke Characterization Project Final Report
vii
KEY WORDS
Smoke, smoke alarm, smoke detector, alarm response, UL 217, optical density, smoke
composition, fire tests, smoke particle size and count distribution, gas effluent, ASTM E1354
cone calorimeter, natural products, synthetic materials, polymer combustion.
5

Smoke Characterization Project Final Report
viii
TABLE OF CONTENTS

EXECUTIVE SUMMARY ii
INTRODUCTION ii
METHODOLOGY iii 5
KEY FINDINGS iii
Small-Scale and Intermediate Scale Test iv
UL 217/UL 268 Fire Test Room Tests v
KEY WORDS vii
SMOKE CHARACTERIZATION PROJECT: FINAL REPORT 16 10
INTRODUCTION 16
OBJECTIVES 18
TECHNICAL PLAN 19
TASK 1 SELECTION OF TEST SAMPLES 20
Task Objectives 20 15
review, Selection and Procurement of Materials and Products in Residential Setting 20
Experimental 24
Results 24
TASK 2 DEVELOP SMOKE CHARACTERIZATION ANALYTICAL TEST
PROTOCOL USING FLAMING AND NON-FLAMING MODES OF 20
COMBUSTION 25
Task Objectives 25
Smoke Characterization 25
Characterization of Smoke in UL 217 Sensitivity Test 29
Small-Scale Tests 34 25
Intermediate-Scale Tests 55
Intermediate-Scale Tests 55
TASK 3 DEVELOP SMOKE PROFILES AND PARTICLE SIZE AND COUNT
DISTRIBUTIONS IN THE UL 217/UL 268 FIRE TEST ROOM 82
Introduction 82 30
Task Objectives 82
Test Samples 83
Experimental 83
Test Procedure 88
Test Results 88 35
TASK 4 CORRELATE ANALYTICAL DATA AND PERFORMANCE IN THE
FIRE TEST ROOM 129
Introduction 129
Smoke Particle Distribution Measurements 129
Influence of Materials and Combustion Mode: Cone Calorimeter 129 40
Influence of Materials and Combustion Mode: Fire Test Room 131
Influence OF Testing Method 137
TASK 5 - IDENTIFY FUTURE CONSIDERATIONS 146
SUMMARY OF FINDINGS 147
Gas Analysis and Smoke Characterization Measurement 147 45
Influence of Material Chemistry 147
Smoke Characterization Project Final Report
ix
Mode of Combustion 147
Small-Scale and Intermediate-Scale Test 148
UL 217/UL 268 Fire Test Room Tests 148
APPENDIX A: Material Chemistry 150
APPENDIX B: Test Sample Documentation and Characterization 156 5

Note: Appendices C through I are provided only in electronic format.
APPENDIX C: Small-Scale Flaming Combustion Test Results
APPENDIX D: Small-Scale Non-Flaming Combustion Test Results
APPENDIX E: Intermediate-Scale Flaming Combustion Test Results 10
APPENDIX F: Intermediate-Scale Non-Flaming Combustion Test Results
APPENDIX G: UL 217/UL 268 Fire Test Room Flaming Combustion Test Results
APPENDIX H: UL 217/UL 268 Fire Test Room Non-Flaming Combustion Test Results
APPENDIX I: UL 217/UL 268 Fire Test Room Smoke Color
15
Smoke Characterization Project Final Report
x
TABLE OF FIGURES

Figure 1 Schematic of the sampling method 26
Figure 2 UL 217 Smoke Box 29
Figure 3 WPS Spectrometer connected to the UL 217 Smoke Box 30 5
Figure 4 UL 217 Smoke Box mean smoke particle size diameter for non- flaming cotton
wick 31
Figure 5 UL 217 Smoke Box relative smoke particle count for non- flaming cotton wick 31
Figure 6 Relationship between smoke particle size and optical density (UL 217
Sensitivity Test) for non- flaming cotton wick 32 10
Figure 7 Relationship between the MIC signal and particle density in the UL 217
Smoke Box for non-flaming cotton wick 33
Figure 8 Cone Calorimeter sample holder 35
Figure 9 Schematic of ASTM E 1354 cone calorimeter 35
Figure 10 Schematic of the gas effluent and smoke measurement system for the cone 15
calorimeter 37
Figure 11 Effective HOC (top) and peak HRR (bottom) for flaming combustion 45
Figure 12 Smoke production for flaming combustion 46
Figure 13 Mean particle diameter for flaming combustion 47
Figure 14 Mean specific particle count for flaming combustion 48 20
Figure 15 Heat release rate per unit area and smoke particle size for flaming Douglas fir
wood 48
Figure 16 Heat release rate per unit area and smoke particle size for flaming
heptane/toluene mixture 49
Figure 17 Heat release rate per unit area and smoke particle size for flaming HDPE 49 25
Figure 18 Carbon dioxide yield for flaming combustion 50
Figure 19 Carbon monoxide yield for flaming combustion 50
Figure 20 Smoke production for non- flaming combustion 51
Figure 21 Mean particle diameter for non- flaming combustion 52
Figure 22 Mean specific particle count for non- flaming combustion 53 30
Figure 23 Carbon dioxide yield for non-flaming combustion 54
Figure 24 Carbon monoxide yield for non-flaming combustion 54
Figure 25 Schematic of NEBS calorimeter 57
Figure 26 Schematic of the IMO calorimeter 58
Figure 27 Intermediate calorimeter evolved smoke and gas sampling cone and tube 58 35
Figure 28 Intermediate calorimeter flaming mode sampling arrangement 59
Figure 29 Intermediate calorimeter non-flaming mode sampling arrangement 60
Figure 30 Photograph of test set-up for UL 217 smoldering test 62
Figure 31 Schematic of smoke sampling for smoldering Ponderosa pine test 62
Figure 32 Heat (top) and smoke (bottom) release rates for heptane/toluene mixture 64 40
Figure 33 Heat (top) and smoke (bottom) release rate for Douglas fir 65
Figure 34 Heat (top) and smoke (bottom) release rate for newspaper 66
Figure 35 Heat (top) and smoke (bottom) release for coffee maker 67
Figure 36 Heat (top) and smoke (bottom) release for nylon carpet 68
Figure 37 Heat (top) and smoke (bottom) release for cotton/poly sheet wrapped PU 45
foam 69
Smoke Characterization Project Final Report
xi
Figure 38 Smoke release rate for bread in non- flaming combustion 70
Figure 39 Smoke release rate for PU foam in non- flaming combustion 70
Figure 40 Smoke release for cotton/poly sheet wrapped PU foam in non- flaming
combustion 71
Figure 41 Smoke particle data from the UL 217 smoldering Ponderosa pine test 74 5
Figure 42 UL 217 smoldering Ponderosa pine particle size distribution 75
Figure 43 Peak HRR for flaming combustion tests 76
Figure 44 Peak SRR for flaming combustion tests 76
Figure 45 Particle size distribution for flaming combustion of natural and synthetic
materials 77 10
Figure 46 Particle size distribution for non- flaming combustion of natural and synthetic
materials 78
Figure 47 Average smoke particle diameters for flaming combustion tests 79
Figure 48 Average smoke particle density for flaming combustion tests 79
Figure 49 Mean smoke particle diameter for non-flaming tests 80 15
Figure 50 Average particle count for non- flaming combustion tests 81
Figure 51 Fire Test Room. Drawing not to scale. 87
Figure 52 Smoke OBS for heptane/toluene mixture in flaming combustion 90
Figure 53 Smoke OBS for newspaper in flaming combustion 91
Figure 54 Smoke OBS for Douglas fir in flaming combustion 91 20
Figure 55 Smoke OBS for coffee maker in flaming combustion 92
Figure 56 Smoke OBS for PU foam in flaming combustion (35 kW/m
2
radiant heating) 92
Figure 57 Smoke OBS for PU foam (100100 mm) with cotton-poly sheet in flaming
combustion 93
Figure 58 Smoke OBS for PU foam (150150 mm) with cotton-poly sheet in flaming 25
combustion 93
Figure 59 Smoke OBS for nylon carpet in flaming combustion 94
Figure 60 Photo and ionization alarm analog signals for flaming PU foam tests 96
Figure 61 Photo and ionization alarm analog signals for flaming nylon carpet tests 97
Figure 62 Photo and ionization alarm analog signals for flaming Douglas fir test 97 30
Figure 63 Comparison of smoke particle size data for selected flaming test 98
Figure 64 Mean smoke particle diameter and count for flaming Douglas fir tests 100
Figure 65 Mean smoke particle diameter and count for flaming newspaper tests 100
Figure 66 Mean smoke particle diameter and count for flaming heptane/toluene tests 100
Figure 67 Mean smoke particle diameter and count for flaming coffee maker tests 100 35
Figure 68 Mean smoke particle diameter and count for flaming PU foam (100100 mm)
tests 101
Figure 69 Mean smoke particle diameter and count for flaming PU foam (100100100
mm) tests 101
Figure 70 Mean smoke particle diameter and count for flaming PU foam (150150150 40
mm) tests 101
Figure 71 Mean smoke particle diameter and count for flaming nylon carpet tests 101
Figure 72 OBS for Ponderosa pine in non- flaming tests 105
Figure 73 OBS for bread in non-flaming tests 105
Figure 74 OBS for polyisocyanurate foam in non- flaming tests 106 45
Figure 75 OBS for PU foam in non- flaming tests 106
Smoke Characterization Project Final Report
xii
Figure 76 OBS for cotton sheet wrapped PU foam in non- flaming tests 107
Figure 77 OBS for polyester microfiber wrapped PU foam non- flaming tests 107
Figure 78 OBS for nylon carpet in non-flaming tests 108
Figure 79 OBS for polystyrene in non-flaming tests 108
Figure 80 Beam vs. MIC response: Ponderosa pine 110 5
Figure 81 Beam vs. MIC response for PU foam in non- flaming combustion 111
Figure 82 Beam vs. MIC response for cotton sheet wrapped PU foam 112
Figure 83 Beam vs MIC response for polyester microfiber wrapped PU foam 112
Figure 84 Beam vs MIC response for Polystyrene in non- flaming combustion 113
Figure 85 OBS changes in the test room for heptane/toluene mixture 114 10
Figure 86 OBS changes in the test room for bread 114
Figure 87 OBS changes in the test room for polyester microfiber wrapped PU foam 115
Figure 88 OBS changes in the test room for cotton fabric wrapped PU foam 115
Figure 89 Mean smoke particle diameter and count for Ponderosa pine in non- flaming
tests 118 15
Figure 90 Mean smoke particle diameter and count for bread in non- flaming tests 119
Figure 91 Mean smoke particle diameter and count for polyisocyanurate foam in non-
flaming tests 120
Figure 92 Mean smoke particle diameter and count for PU foam in non- flaming tests 121
Figure 93 Mean smoke particle diameter and count for PU foam in non- flaming tests 20
(Data from Test 12261 were found to be suspicious and were not plotted) 122
Figure 94 Mean smoke particle diameter and count for cotton fabric wrapped PU foam
in non- flaming tests 123
Figure 95 Mean smoke particle diameter and count for cotton-poly wrapped PU foam in
non- flaming tests 124 25
Figure 96 Mean smoke particle diameter and count for polyester microfiber wrapped
PU foam in non-flaming tests 125
Figure 97 Mean smoke particle diameter and count for nylon carpet in non- flaming
tests 126
Figure 98 Mean smoke particle diameter and count for polystyrene in non-flaming tests 127 30
Figure 99 Specific extinction area for small-scale flaming and non-flaming combustion 130
Figure 100 Mean particle diameter for small-scale flaming and non- flaming combustion 130
Figure 101 Specific particle count for small-scale flaming and non- flaming combustion 131
Figure 102 Mean particle diameters at an obscuration of 0.5 %/ft in the Fire Test Room 132
Figure 103 MIC signal versus particle size data for Fire Test Room flaming tests 133 35
Figure 104 MIC signal versus particle size data for Fire Test Room non- flaming tests 133
Figure 105 Analog ion signal versus particle size data for Fire Test Room flaming tests 134
Figure 106 Analog ion signal versus particle size data for Fire Test Room non- flaming
tests 134
Figure 107 Obscuration versus particle size data for Fire Test Room flaming tests 135 40
Figure 108 Obscuration versus particle size data for Fire Test Room non-flaming tests 135
Figure 109 Analog photo (scattering) signal versus particle size data for Fire Test Room
flaming tests 136
Figure 110 Analog photo (scattering) signal versus particle size data for Fire Test Room
non- flaming tests 136 45
Smoke Characterization Project Final Report
xiii
Figure 111 Small-scale smoke release rate versus Fire Test Room obscuration for
flaming PU foam tests 139
Figure 112 Intermediate-scale smoke release rate versus Fire Test Room obscuration
for flaming heptane/toluene mixture tests 139
Figure 113 Intermediate-scale smoke release rate versus Fire Test Room obscuration 5
for flaming nylon carpet tests 140
Figure 114 Intermediate-scale smoke release rate versus Fire Test Room obscuration
for flaming coffee maker tests 140
Figure 115 IMO and Fire Test Room smoke particle mean diameter for flaming
heptane/toluene mixture tests 141 10
Figure 116 IMO and Fire Test Room smoke particle mean diameter for flaming
Douglas fir tests 142
Figure 117 IMO and Fire Test Room smoke particle mean diameter for flaming
newspaper tests 142
Figure 118 IMO and Fire Test Room smoke particle mean diameter for flaming PU 15
foam tests 143
Figure 119 IMO and Fire Test Room smoke particle mean diameter for flaming coffee
maker tests 143
Figure 120 Intermediate-scale and Fire Test Room smoke particle mean diameter for
non- flaming Ponderosa pine tests 144 20
Figure 121 IMO and Fire Test Room smoke particle mean diameter for non-flaming
bread tests 145
Figure 122 IMO and Fire Test Room smoke particle mean diameter for flaming nylon
carpet tests 145
25
Smoke Characterization Project Final Report
xiv
TABLE OF TABLES

Table 1 Items commonly found in residential settings 20
Table 2 Project test samples 21
Table 3 Sample description and material chemistry 22 5
Table 4 Cone calorimeter test samples 34
Table 5 Test parameters for cone calorimeter flaming mode tests 39
Table 6 Test parameters for cone calorimeter non- flaming mode tests 39
Table 7 Cone calorimeter combustibility data for small-scale flaming mode tests 41
Table 8 Cone calorimeter combustibility data for small-scale non- flaming mode tests 42 10
Table 9 Smoke particle and gas effluent data for small- scale flaming mode tests 43
Table 10 Smoke particle and gas effluent data for small- scale non- flaming mode tests 44
Table 11 Intermediate calorimeter test samples 56
Table 12 Intermediate calorimeter sample exposure scenario 61
Table 13 Intermediate calorimeter combustibility results 63 15
Table 14 Intermediate calorimeter smoke particle data 72
Table 15 Maximum observed carbon monoxide and carbon dioxide concentrations 73
Table 16 Test samples for UL 217 Fire Test Room Test tests 83
Table 17 Fire Test Room Tests 84
Table 18 Data acquisition sampling intervals 88 20
Table 19 Summary of obscuration for flaming tests 89
Table 20 Flaming mode alarm response times 95
Table 21 Smoke particle data at 0.5 %/ft and 10 %/ft OBS: flaming tests 99
Table 22 Observed Fire Test Room test signals for flaming mode at 240 seconds 102
Table 23 Fire Test Room ceiling test signatures for flaming combustion tests 103 25
Table 24 Summary of smoke obscuration for non- flaming tests 104
Table 25 Non- flaming mode alarm response times 109
Table 26 Observed UL 217 room test signals at ceiling location for non- flaming mode
tests at 0.5 % /ft 116
Table 27 Observed UL 217 room test signals at ceiling location for non- flaming mode 30
tests at 10 % Obs/ft 117
Table 28 UL 217 Fire Test Room ceiling test signatures for non-flaming combustion
tests 128
Table 29 Theoretical smoke particle dependency for traditional smoke sensor
technologies 129 35
Table 30 Fire Test Room alarm trigger times 137
Table 31 Influence of scale on mean smoke diameter 138
Table 32 Influence of heating mode on smoke characteristics: non-flaming 138

40
Smoke Characterization Project Final Report
xv
NOMENCLATURE

Acronymns Description
Organizations
ASTM American Standards for Materials and Testing
FPRF Fire Protection Research Foundation
NFPA National Fire Protection Association
NIST National Institute of Standards and Technology
UL Underwriters Laboratories Inc.

Equipment
DMA Dynamic Mobility Analyzer (part of WPS spectrometer)
FTIR Fourier Transform Infrared Spectrometer
LPS Light Particle Spectrometer (part of WPS spectrometer)
MIC Measuring Ionization Chamber
TGA Thermogrametric Analyzer

Notation Description Units
Ionization chamber physical characteristics (constant) s
Attachment coefficient of air- molecule ions to the soot particles s
-1

CO Carbon monoxide ---
CO
2
Carbon dioxide
---
C
s
Smoke concentration
kg/m
3

D Ion diffusion coefficient
cm
2
/s
d
m
Mean smoke particle diameter for one WPS Spectrometer scan
10
-6
m
D
m
Average smoke particulate diameter over the duration of the test
10
-6
m
HOC Heat of combustion
kJ/g
HDPE High density polyethylene ---
HRR Heat release rate
kW or
kW/m
2

l Path length m or ft
n
m
Mean smoke particle count density for one WPS Spectrometer scan cc
-1

N
m
Average particle count density over the duration of the test cc
-1

OBS Smoke obscuration (UL 217 definition) ---
OD Optical density ---
Peak HRR Maximum heat release rate for the duration of the test
kW or
kW/m
2

Peak SRR Maximum smoke release rate for the duration of the test m
2
/s
ppm parts per million ---
PU Polyurethane ---
SRR Smoke release rate m
2
/s
T Ceiling temperature in Fire Test Room
o
C
Vel. Velocity measured in Fire Test Room m/s

Smoke Characterization Project Final Report P. 16 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
SMOKE CHARACTERIZATION PROJECT: FINAL REPORT

INTRODUCTION
Residential smoke alarms provide an important notification to individuals within a residential
setting that there is a presence of smoke and/or fire. Fires and incipient fires (non-flaming phase) 5
provide several cues for detection equipment. These include smoke particulates, heat, and gas
effluents (e.g. CO, CO
2
). Current smoke alarms primarily utilize two types of detection
technologies: photoelectric or ionization. The photoelectric type has a light source and detects
the scattering or obscuration of light caused by smoke particulates. The ionization type detects
changes in local ionization field within the detection chamber resulting from the presence of 10
burning materials. Both types of alarms activate when a set threshold is reached.

Over three decades ago following a seminal research study to develop data on smoke alarm
performance and location requirements for the alarms
1,2
known as the Indiana Dunes
investigation. The use of smoke alarms began to increase. In the Indiana Dunes study, tests were 15
conducted in actual homes with representative sizes and floor plans, utilized simulated furniture
component mock-ups, actual furnishings and household items for fire sources, and tested actual
smoke alarms sold in retail stores. That report concluded that smoke alarms of either
photoelectric or ionization type generally provided the necessary escape time for different fire
types and locations. However, materials used in this investigation were not characterized for 20
their physical and chemical properties. There were several findings worth noting: (i) smoke
particulates from flaming and non-flaming fire provide different smoke signatures; (ii) detection
technologies (ionization vs. photoelectric) respond differently to flaming and non- flaming smoke
particulates; and (iii) the location of the alarms had a significant influence on the safe egress time.
25
The Indiana Dunes investigation contributed to the ongoing development of a smoke alarm
performance standard (UL 217
3
) by Underwriters Laboratories Inc. (UL). The development of
this standard accelerated the use of smoke alarms in residential setting such that smoke alarms
are now found in more than 90 % of residential structures in the USA. In the UL certification
program smoke alarm models are evaluated for response to three flaming fire tests (wood, paper, 30
and heptane/toluene) and one smoldering smoke test (Ponderosa pine). The materials used for
these tests are intended to represent fuels commonly found in buildings in the USA, and produce
gray and black smoke during either flaming or smoldering conditions. The non- flaming test
represents the basic smoke profile that occurs during a typical slow non- flaming cushion fire.
Thus, the UL performance tests assess the ability of an alarm to respond to several different fire 35
sources. The UL standard and the Indiana Dunes test also led to the development of a new
national code (NFPA 72
4
).

Statistics
5
developed by National Fire Incident Reporting System (NFIRS) provide evidence that
smoke alarms have a significantly beneficial impact towards preventing fatalities from fires. It 40
has been estimated that installation of smoke alarms achieves a 40-50% reduction in the fire
death rate relative to number of fires. However, over a period from 1996 to 1998, data
6
show that
smoke alarms did not operate in 22% of the residential structure fires involving one and two-
family homes and apartments. In general, the fire data shows that the number of fatalities
increases when smoke alarms are either absent or fail to operate. Poor maintenance, disabling of 45
Smoke Characterization Project Final Report P. 17 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
alarms (e.g., due to nuisance alarms), and inability for the working alarms to trigger in sufficient
time (i.e., respond to smoke particulate) are some of the reasons for the inability of smoke alarms
to provide sufficient time to execute an evacuation plan.

Substantial changes have occurred in the typical household since the Indiana Dunes study. 5
Residential settings are now larger, with more synthetics, and contain a wide variety of
manufactured products that are driven by consumer demand. Synthetic materials are now the
norm with regards to textiles, thermoplastic enclosures and engineered materials. This has been
accelerated by the global petrochemical and polymer industry that has exponentially advanced
since the mid 1940s. With the advent of global manufacturing and shipping, these products are 10
now manufactured and distributed throughout the world. In contrast, materials derived from
natural processes, such as photosynthesis and metabolism, are less common on a percentage
basis.

It is thought that synthetic materials currently found in the home tend to ignite and burn faster 15
than materials used in the original study and this may be explained by analyzing the chemical
structures of the synthetic and natural materials and investigating their modes of decomposition
in a fire scenario. Accelerated decomposition is expected to result in faster growing fires and
therefore an overall reduction of safe egress time. At the same time there have also been
advances in fire retardant additives and compounding technology thereby improving material fire 20
resistance. This would result in longer period of non-flaming decomposition of materials,
especially with smaller ignition sources. These changes in materials are expected to alter the
chemistry and the nature of smoke particulates, heat and gas component signatures. It has been
suggested that non- flaming material decomposition also generate more carbon monoxide and
other gases that can lead to incapacitation before occupants can respond to the smoke alarm. 25

The influence on smoke alarm response to changes in available materials was investigated in a
recent study by NIST
7
. This work followed a design similar to that of the Indiana Dunes
investigation. Tests were conducted in actual homes with representative sizes and floor plans,
utilized actual furnishings and household items for fire sources, and tested commercially 30
available smoke alarms. However, as in the Indiana Dunes investigation, the materials of these
furnishings were not physically or chemically characterized.

NIST concluded that smoke alarms, of either photoelectric or ionization type, installed on every
building level generally provided the necessary escape time for different fire types and locations 35
though significant differences were measured between the response times of photoelectric and
ionization alarms to flaming and non- flaming fires. Adding smoke alarms in bedrooms
lengthened the escape time, especially for non-flaming fires. The main difference with the NIST
study and the previous Indiana Dunes investigation is that the calculated safe egress time was
consistently shorter and the fire growth rates were faster. In addition to developing smoke alarm 40
performance data, the NIST study also measured smoke particle size distribution and
components of gas effluents from the fire tests but did not characterize the materials.

The influence of material chemistry on smoke production is significant. Except for
noncombustible materials (for example metals, minerals, glasses, ceramics), the vast majority of 45
materials found in residential settings are carbonaceous and thus, susceptible to decomposition
Smoke Characterization Project Final Report P. 18 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
and burning. The combustion behavior of carbonaceous materials (ignition, heat release, smoke
release) with attendant softening, melting and liquefaction, and charring is dictated by chemistry.
Polymeric materials (either natural or synthetic) have chemical structures and morphology that
affect degradation, heat release and smoke production. In general, synthetic materials are
chemically less complex than natural materials as they are derived from monomers from crude 5
oil (ethylene, propylene, acetylene, styrene, vinyl chloride, acrylic acid, acrylonitrile and so on).
Natural materials have polymeric structures that are highly complex linear and crosslinked
structures (carbohydrates, proteins, glycerides, etc.) and tend to char rather than soften and
liquefy.
10
Despite significant advances in the knowledge of alarm performance with typical products found
in residential settings gained from the NIST study, it was determined that further study was
needed to develop combustibility and smoke characteristics for a wider range of synthetic
materials and natural products found in residential settings. These materials also need to be fully
characterized for their physical and chemical composition as well their combustibility behavior. 15

Thus, the current research project was initiated to fully characterize the products of combustion
for both the flaming and non-flaming modes on a variety of materials and products commonly
found in residential settings. The study would also take advantage of advances in the smoke
particle and gas effluent characterization technology that was not previously conducted. 20


OBJECTIVES
The objectives of this research investigation were as follows:
25
1. Develop smoke characterization analytical test protocols using flaming and non- flaming
modes of combustion on selected materials found in residential settings;

2. Using materials from the analytical smoke program, develop smoke particle size and count
distribution data and smoke profiles in the UL 217/UL 268 Fire Test Room for both flaming 30
and non- flaming modes of combustion.

3. Provide data and analysis to the fire community for several possible initiatives:
Develop recommendations to change the current residential smoke alarm standard (UL
217). 35
Development of new smoke sensing technology.
Provide data to the materials and additives industries to facilitate new smoke suppression
technologies and improved end products.

40
Smoke Characterization Project Final Report P. 19 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
TECHNICAL PLAN
A technical plan was developed to meet the project objectives as following:

Task 1 Selection of test samples
5
Task 2 Develop smoke characterization analytical test protocol using non- flaming and flaming
modes of combustion

Task 3 Develop smoke profiles and particle size and count distributions in the UL 217/UL 268
Fire Test Room 10

Task 4 Correlate analytical data and performance in the UL 217/UL 268 Fire Test Room

Task 5 Identify future considerations
15
Task 6 Develop Final Report

The results of this investigation (Task 6) are described herein.

20
Smoke Characterization Project Final Report P. 20 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
TASK 1 SELECTION OF TEST SAMPLES

TASK OBJECTIVES
The objectives of this task were as follows:
Survey materials and products in contemporary residential settings 5
Select materials for the research investigation
Procure samples
Document and characterize the samples

10
REVIEW, SELECTION AND PROCUREMENT OF MATERIALS AND PRODUCTS IN
RESIDENTIAL SETTING
An informal review of typical products and materials found in contemporary residential settings
was performed to assist in the selection of test samples for investigation in this study. A list of
typical items and their corresponding combustible base materials is presented in Table 1. 15

Table 1 Items commonly found in residential settings
Residential Area Common Items Common Base Materials
Bedroom and Living Room
Appliance wiring
Bed clothing
Candles
Carpeting
Drapes and blinds
Mattress

Paper products
Plastic enclosures for electrical
products
Upholstered furniture

Wallpaper

Wood furniture
Flexible PVC (plasticized)
Cotton, Polyester, Acrylic, Blends
Hydrocarbon wax, Cotton wick
Polyolefin, Nylon, Polyester
Cotton, Linen, Wood, PVC
Polyurethane foam, Cotton,
Polyester
Paper
Polyolefin, ABS, Nylon

Polyurethane foam, Polyester,
Cotton, Wood
Paper, PVC plastisol, Polyacrylates
coatings
Wood, Polyurethane, Cotton,
Polyester, Adhesives
Kitchen
Appliance enclosures
Appliance wiring
Cabinets
Counter tops
Food containers
Foods
Wallpaper
Polyolefins, ABS, Polycarbonate
Flexible PVC (plasticized)
Wood, MDF, Adhesives
Laminates, Acrylics, Wood
Polyolefins, PVDC
Fats, Oils, Carbohydrates, etc.
Paper, PVC plastisol, Polyacrylates
coatings
Storage Areas
Paints

Fuels
Packaging materials
Acrylic latex, Oil, Polyurethane,
Thinner
Hydrocarbons
Paper, Polystyrene, Starch

Smoke Characterization Project Final Report P. 21 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Representative test samples were selected based upon the prevalence of items in residential
settings, the chemistry of their base material components, and their role in residential fires.

All of the selected materials were procured from commercial sources. Where the selected
material was a composite item such as a mattress, individual components of the final item were 5
also investigated to provide a connection between the components and the end product. The
selected materials and UL 217 test samples are listed in Table 2 along with their corresponding
base material description.
Table 2 Project test samples
Residential Item Samples Material Description
Appliance wiring
Electrical wire (duplex
lamp cord)
Duplex wire (16 gauge, stranded copper), brown PVC
insulation
Appliance Coffee maker
12 cup capacity; atactic polypropylene housing, PVC
wire
Mattress Mattress Twin size, no fire barrier
Cotton batting 7 mm thick; 0.7 kg/m
2
Mattress components
(from mattress) Polyurethane foam 25 mm thick; 1.2 kg/m
2

Pillow
Queen size; white
Cover: 70% polyester/30% cotton
Fill: 100% polyester with silicone finish
Cotton sheeting White; plain weave; 102 g/m
2
(CA TB 117 sheeting)
Cotton/Poly sheeting
White; plain weave; 50:50 blend; 763 g/m
2
(CA TB 117
sheeting)
Bed/Upholstered
furniture cover
Polyester sheeting White, plain weave; 790 g/m
2
microfiber
Fabric Rayon White, Plain weave, 763 g/m
2

Nylon
Nylon 6 yarns; Polypropylene backing; 3.0 kg/m
2

finished product Carpeting
Polyester Polyester yarns; 2.7 kg/m
2
finished product
Bread Wonder

white
Cooking oil Wesson Vegetable oil (polyunsaturated oil)
Lard Natural; Saturated fat
Cooking material
and fuels
Heptane Flammable liquid (represents aliphatic chemistry)
Insulation Polyisocyanurate inch thick; 43 kg/m
3

Plastic enclosures HDPE sheet 6 mm thick; 930 kg/m
3

Cotton wick Diameter: 4.3 mm; Weight: 7.2 g/m
Douglas fir 6 6 2-1/2 inch; Weight: 450 g
Ponderosa pine 3 1 inch stick, 10 sticks weighing 160g
Newspaper Black print only, 42.6 g. of inch wide strips
UL 217 Test sample
Heptane/Toluene 30 mL Heptane and 10 mL Toluene (ACS reagent grade)
10
Table 3 describes the material chemistry of the test samples
8
. A cross-reference code assigned to
natural (N) and synthetic (S) materials is included for reference to additional technical
descriptions found in Appendix A.

15



Smoke Characterization Project Final Report P. 22 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 3 Sample description and material chemistry
Sample Description
Reference
Code
Material Chemistry
Lamp wire
compounded PVC
S20
Flexible PVC is produced by the incorporation of 20-60% by weight
aromatic or aliphatic ester plasticizers in the PVC powder. This
plasticization produces compounds with exceptional flexibility,
toughness and weatherability. Typical aromatic plasticizers are based
upon terephthalic acid (di-carboxylic acid) or trimellitic acid (tri-
carboxylic acid). Alcohols used in these plasticizers usually contain
from 8 to 16 carbon atoms. Elemental composition C, H, O;
structure aromatic or aliphatic depending upon type of acid used.
Coffee maker
Polypropylene
S14
Polymers based on the polymerization of propylene (CH
2
=CHCH
3
),
or copolymers with other unsaturated monomers. PP polymers and
copolymers have a range of properties due to factors, such as cross-
link density, molecular weight, degree of branching, incorporation of
co-monomers, etc. Elemental composition essentially C, H
depending upon type and percentage of co-monomers; structure
aliphatic.
Mattress
Combination of
cotton, polyester
batting, and
polyurethane foam
N4
S10
S16
Cotton - Staple fiber consisting primarily of cellulose (88-96%) with
other natural-derived aliphatic organic compounds (C, H, O).
Cellulose is a natural carbohydrate polymer (polysaccharide)
consisting of anhydroglucose units joined by an oxygen linkage to
form essentially linear high molecular weight chains.

Polyester - A generic term for commercially available textile and
thermoplastic products based upon ester polymers with the
characteristic linkage (R-COO-R) where R or R can be various
hydrocarbon groups. Ester polymers are produced by either the
condensation reaction of dicarboxylic acids with dihydroxy alcohols
or the reaction of lactones (cyclic esters) or hydroxy-carboxylic acids.
Polyester textiles are usually composed of PET polyethylene
terephthalate. PET is formed by the reaction of terephthalic acid
(aromatic compound) and ethylene glycol (aliphatic compound).
Elemental composition C, H, O; structure aliphatic and aromatic.

For Polyurethane (S15) see Polyisocyanurate rigid foam (S16)
Mattress Cotton
batting
N4 See Cotton (N4)
Mattress
Polyurethane foam
S16 See Polyisocyanurate rigid foam (S16)
Pillow
- Cover: cotton/
polyester blend
- Fill: polyester
N4, S10
See Cotton (N4)
See Polyester (S9)
Cotton sheeting N4 See Cotton (N4)
Cotton/Polyester
sheeting
N4, S10
See Cotton (N4)
See Polyester (S9)
Polyester microfiber
sheeting
S10 See Polyester (S9)

Smoke Characterization Project Final Report P. 23 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Sample Description
Reference
Code
Material Chemistry
Rayon fabric S23
Generic name for a manufactured fiber composed of regenerated
cellulose in which >15% of hydroxyl substituents have been replaced
by chemical modification (for example by acetate groups). The fiber
ignites and burns readily. Chemical composition C, H, O; structure
aliphatic
Carpeting Nylon 6 S7
Generic name for a family of polyamide polymers characterized by
the presence of an amide group (R-CONH-R) where R and R are
various hydrocarbon groups. As with polyesters, nylons are used in
various applications, such as textiles and structural housings. The
nylon properties are dictated by the various monomers used in the
polymerization and subsequent compounded fillers that may be
incorporated into the structure in post processing steps. Nylon 6 is
formed from the homopolymerization of caprolactam. Chemical
composition C, H, O, N; structure aliphatic
Carpeting
Polyester
S10 See Polyester (S9)
Bread N1 Composed primarily of starch, sugar, fats and oils.
Cooking oil N13
Edible oils extracted from the seeds, fruit or leaves of plants.
Generally considered to be mixtures of glycerides (safflower,
sunflower, peanut, walnut, etc.).
Polyisocyanurate
rigid foam
S17
Rigid polyurethane or polyisocyanurate foams have a high cross-link
density. Crosslinking is achieved by the ratio of co-monomers and
reactive group functionality. One example of rigid foam is produced
by MDI (diphenyl methane diisocyanate), water, catalyst and blowing
agents. Water readily reacts with isocyanates to form amine groups,
which further react to form urea linkages (R-NH-CO-NH-R) in the
polymer structure. Rigid foams typically have a close-cell structure
and more resistant to degradation (liquefaction) due to the high cross-
link density. Elemental structure C. H. O. N; structure - aromatic
Plastic enclosure
HDPE sheet
S11
Polyethylene (PE) is based on the polymerization of ethylene
(CH
2
=CH
2
). PE polymers can have a range of properties due to
factors, such as cross-link density, molecular weight, degree of
branching, incorporation of co-monomers, etc. High density
polyethylene is characterized by a linear structure and high molecular
weight. Elemental composition essentially C, H depending upon
type and percentage of co-monomers; structure aliphatic.
Cotton wick N4 See Cotton (N4)
Douglas fir N15
Wood is typically composed of 40-60% cellulose and 20-40% lignin,
together with gums, resins, variable amounts of water and inorganic
matter.
Ponderosa pine N15 See Wood (N16)
Newspaper N8
A processed product of cellulosic fibers primarily made from
softwoods. Carbon black is used in the printing ink.
Heptane/Toluene
S5
S24
Heptane is a 7-carbon, hydrocarbon liquid with the formula C
7
H
16

Toluene (methyl benzene) is a 7-carbon aromatic hydrocarbon liquid
composed of a 6-membered aromatic ring (benzene C
6
H
6
)

with an
attached methyl (-CH
3
) group.
Smoke Characterization Project Final Report P. 24 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
EXPERIMENTAL
The selected plastics materials were characterized for their chemistry by FTIR, and the TGA for
their thermal decomposition profile.

FTIR - Infrared spectral response of the materials was characterized in the solid-state using a 5
Nicolet Nexus 470 FTIR with a Golden Gate KRS-5 diamond ATR accessory. Samples were
scanned from 400 to 4000 cm
-1
wavenumber at a 4 cm
-1
resolution; 32 scans were averaged per
recorded spectra.

TGA - Thermal decomposition of the materials were characterized using a TA Instruments 10
model Q500 TGA with an evolved gas analysis (EGA) furnace. Samples weighing between 10 to
50 milligrams were heated from 40 to 825 C at 20 C/min under a 90 mL/min dry air flow rate.


RESULTS 15
The material characterization results are provided along with photographs in Appendix B.
Smoke Characterization Project Final Report P. 25 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
TASK 2 DEVELOP SMOKE CHARACTERIZATION ANALYTICAL
TEST PROTOCOL USING FLAMING AND NON-FLAMING MODES OF
COMBUSTION

TASK OBJECTIVES 5
The objectives of this task were as follows:
Develop sampling method for smoke particle size and gas effluent analysis
Develop smoke particle size and count distribution data from UL 217 Sensitivity Test
(Smoke Box)
Develop combustibility, smoke particle size and gas effluent data using small and 10
intermediate scale tests
Develop flaming and non- flaming scenarios for potential use in Task 3 UL 217/UL 268
Fire Test Room tests

15
SMOKE CHARACTERIZATION
Equipment
A smoke particle analyzer and a gas FTIR analyzer were used to characterize the smoke particle
size and gas effluents.
20
Smoke Particle - Smoke particle size and count distribution was characterized using a Model
WPS 1000XP wide range particle size spectrometer from MSP Corporation (WPS spectrometer).
The WPS spectrometer combines laser light scattering, electrical mobility and condensation
particle counting technologies in a unique, single instrument with the capability of measuring the
concentration and size distribution of aerosol particles ranging from 10 nm to 10,000 nm (0.01 25
m to 10 m) in diameter. The instrument divides a 1 Liter/min sample flow between the
dynamic mobility analyzer (DMA) and the light particle spectrometer (LPS) modules to develop
the particle size distribution. The LPS module is sensitive to particle sizes greater than 200 nm
(0.2 m) whereas the DMA module is sensitive to particle sizes ranging from 10 nm to 500 nm
(0.01 m to 0.50 m). The instrumentation measurement sensitivity is limited to a particle 30
concentration not exceeding 210
7
particles/cc.

Effluent Gas Composition - Gas effluent composition was characterized using a MIDAC #I
1100 Fourier Transform Infrared (FTIR) Spectrometer equipped with a 10 meter path length
optical cell. The UL FTIR equipment has gas calibration library to calculate the concentration of 35
the key gas components detected. The instrument has a measurement range of 600 to 4000 cm
-1

wavenumber and a resolution of 0.5 cm
-1
.

Measurement Method
Smoke samples were extracted from the respective test apparatus for particle size distribution 40
and effluent gas composition analyses as depicted in Figure 1. The smoke samples were diluted
with nitrogen gas (UHP grade, 99.999%) as necessary to prevent saturation of the detection
instrument. The sample flow and the nitrogen gas flows were controlled using rotameters.

Smoke Characterization Project Final Report P. 26 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
FTIR
Smoke Particle
Size
Measurment
Extracted smoke sample
N
2
N
2


Figure 1 Schematic of the sampling method

Smoke Particle - Particle sizes were measured by the DMA module at a rate of 2 seconds per 5
size interval (bin). For the data reported herein, the DMA analyzer was set to obtain data for 24
size intervals resulting in an ensemble measurement time of 48 seconds. Particle size
measurements by the LPS module are instantaneous, however the recorded count is an average
over the 48 second ensemble measurement time. The analyzer was purged between successive
ensemble measurements resulting in subsequent measurements being collected at 67 second 10
intervals.

Effluent Gas Composition - Infrared spectra of the effluent gas were continuously collected at
15 second intervals. Each spectrum was based on the signal average of 8 individual scans at a
resolution of 0.5 cm
-1
. Prior to testing, a background reference spectrum was collected. The 15
background reference spectrum was based on the signal average of 32 individual scans at a
resolution of 0.5 cm
-1
.

Smoke Particle Analysis
In order to interpret collected smoke particle data, a correlation based on Beers Law was 20
developed for smoke obscuration and smoke particle size and count. Beers Law as applied to
smoke relates optical density per unit path length to smoke concentration as shown in Eq. 1.
s
C
OD

l
Eq. 1
Where OD is the optical density, l is path length, and C
s
is the smoke concentration at a given
time. The smoke concentration is related to the smoke number density as shown in Eq. 2.
3
i i s
d n C

Eq. 2
Where n
i
, and d
i
are the number count (density) and particle diameter for a given particle size i. 25
Thus a relationship between optical density per path length and the number count at a given time
may be established as described in Eq. 3.
3
i i
d n
OD

l
Eq. 3

Smoke Characterization Project Final Report P. 27 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The following notation is used in the remaining body of this report to distinguish the three levels
of particle data collected on the WPS spectrometer:

n
i
, d
i
individual bin size data
5
n
m
, d
m
mean ensemble data (the arithmetic mean of the 24 bins of data measured per
ensemble) such that:

24
n
n
24
1 i
i
m

Eq. 4

24
1 i
i
24
1 i
i i
m
n
d n
d Eq. 5
10
N
m
, D
m
time averaged mean ensemble data (the arithmetic mean of all measured
ensembles) such that:

scans of number
n
N
finish
0 t
m
m

Eq. 6

finish
0 t
m
finish
0 t
m m
m
n
n d
D Eq. 7
15
Effluent Gas Analysis
A simple mixing model was used to deconvolute the effects of the FTIR gas cell retention time
on the measured effluent gas concentrations. The relevant quantities are the fixed volumetric
flow rate,
in
v& =
out
v& = v& , of the effluent gas sample through a well- mixed controlled volume
V
o
(the FTIR cell) at atmospheric pressure and a temperature of 120 C. The mass flow rate for a 20
given effluent gas component i leaving the control volume at constant air density ? is:
( )
dt
] i [ d
V v ] i [
dt
] i [ d
V
dt
dV
] i [
dt
] i [ V d
m
out out , i
+ +

& & Eq. 8

The mass flow rate for the given component i entering the control volume is:
Smoke Characterization Project Final Report P. 28 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
( )
in int , i
] i [ v
dt
] i [ d
V
dt
dV
] i [
dt
] i [ V d
m & & +

Eq. 9
since d[i]/dt = 0 for the incoming gas species at [i]
in
. The mass balance for the gas is:
0 m m
out , i in , i
& & Eq. 10

Combining Eq. 8, Eq. 9, and Eq. 10 results in the deconvoluted incoming gas concentration:
out in
] i [
dt
] i [ d
] i [ + Eq. 11
such that the FTIR gas cell retention time is defined as v& / V
o
.
5
The following values were used for the calculations:
v& = measured FTIR sample flow rate
V
o
= FTIR cell volume = 2 liters

10
Smoke Characterization Project Final Report P. 29 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
CHARACTERIZATION OF SMOKE IN UL 217 SENSITIVITY TEST

Introduction
The UL 217 Sensitivity Test (Section 37) is used to determine the relative sensitivity of smoke
alarms to smoke/aerosol buildup. In this test a smoke alarm is enclosed in a sealed case with a 5
constant re-circulating airflow and subjected to a prescribed rate of smoke/aerosol buildup. The
smoke alarm must operate within specified visible smoke obscuration value between 0.5 and 4.0
%/ft, and MIC signal 93 to 37.5 pA.

Analysis of smoke generated during UL 217 Sensitivity Tests was used to (i) develop smoke 10
particle size data for the reference smoke alarm test; (ii) compare smoke particle size to
obscuration data; and (iii) develop understanding of smoke aggregation as a function of test time.

Experimental
UL 217 Sensitivity Tests were conducted in accordance with Section 37 of UL 217 Single and 15
Multiple Station Smoke Alarms using Underwriters Laboratories UL 217 Sensitivity Test case
(smoke box). Aerosol buildup, by smoke generated by a non- flaming cotton wick, followed the
relationship between the MIC (Electronikcentralen Type EC 23095) output and the percent light
transmission remains within the Beam and MIC curves illustrated in UL 217 (Figures 37.1, and
37.2). The air velocity in the test compartment was maintained at 32 +/-2 fpm (0.16 +/-0.001 20
m/s). A photograph of the UL 217 Smoke Box is shown in Figure 2; detailed descriptions of the
smoke box assembly are available in the UL 217.

Figure 2 UL 217 Smoke Box
25
Smoke particle size and count density was characterized using the WPS spectrometer. The
sampling was accomplished by inserting a 6.25 mm O.D. conductive silicone tube 90 mm into
the Smoke Box from the top. Thus, the sample point was located in the center of the flow path.
5 ft Light Path Length
Test sample
holder
MIC
Flow
Flow
Smoke Characterization Project Final Report P. 30 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The other end of the conductive tubing was connected directly to the WPS Spectrometer. The
collected smoke sample was not diluted with nitrogen as relatively low concentrations of smoke
were anticipated. The schematic of the WPS connected to the Smoke Box is presented in Figure
3.
5

Figure 3 WPS Spectrometer connected to the UL 217 Smoke Box

Prior to testing, the Smoke Box was exhausted and a background check was conducted with the
WPS spectrometer to ensure low particle count density (less than 10
3
particle/cc). The test was 10
initiated after igniting the cotton wick, placing it in the sample holder (Figure 2), and closing the
lid. The data acquisition for both the smoke box and the WPS spectrometer were then initiated
simultaneously.

A total of two tests were conducted and both were terminated after approximately 15 minutes. 15

Results
The mean smoke particle diameter (d
m
) and mean smoke particle count (n
m
) for the non-flaming
cotton wick are plotted as a function of test time in Figure 4 for both of the test runs. The results
from the two tests show repeatability of particle measurements over the duration of the tests. 20


Smoke Characterization Project Final Report P. 31 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.00E+00
5.00E+05
1.00E+06
1.50E+06
2.00E+06
2.50E+06
0 100 200 300 400 500 600 700 800 900
Time (s0
P
a
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
0
0.05
0.1
0.15
0.2
0.25
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
s
)
Test 1 -Count
Test 2 - Count
Test 1 - Mean Dia
Test 2 - Mean Dia.

Figure 4 UL 217 Smoke Box mean smoke particle size diameter for non-flaming cotton wick
Smoke particle count was separated into three relative size groups to differentiate the population
of small, medium, and large particles. The 0.03 to 0.109 m range characterizes small particles,
0.109 to 0.500 m range for medium particles, and 0.500 to 10 m range for large particles. 5
Relative particle size counts plotted in Figure 5 indicate that over time there is a gradual increase
in the number of large particles and a gradual decrease in small particles. Aggregation of smaller
particles into fewer larger particles is a potential explanation for the observed phenomenon.
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
0 100 200 300 400 500 600 700 800 900
Time (s)
P
a
r
t
i
c
l
e

C
o
u
n
t

D
e
n
s
i
t
y

(
1
/
c
c
)
0.000
0.001
0.010
0.100
1.000
C
o
u
n
t

F
r
a
c
t
i
o
n
Total Count
0.01-0.109 microns
0.109-0.5 microns
0.5 - 10 microns
Figure 5 UL 217 Smoke Box relative smoke particle count for non-flaming cotton wick 10
Smoke Characterization Project Final Report P. 32 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Particle size density,
3
i
i
d n , was calculated for each WPS spectrometer measured particle
ensemble data. This calculated data was plotted against optical density per path length calculated
from the measured smoke obscuration data and averaged over the same time period as the smoke
particle ensemble data. The results, depicted in Figure 6, show agreement with the expected
relationship described in Eq. 3. 5
0
100
200
300
400
500
600
700
800
900
1000
0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016 0.018 0.020
Optical Density per Path Length (1/ft)
S
u
m
(
n
i
d
i
3
)
Test 1
Test 2

Figure 6 Relationship between smoke particle size and optical density (UL 217 Sensitivity Test) for non-
flaming cotton wick

The MIC response is related to the physical characteristics of the ionization chamber and the 10
attachment coefficient of air- molecule ions to the soot particles such that = 2Dd
m
, where D
is the ion diffusion coefficient.
9
Thus MIC response is related to the product of particles count
and diameter as shown in Eq. 12.
MIC ~ d
m
n
m
Eq. 12
15
The MIC data were averaged over the sampling time of the particle analyzer and the number
density and diameter product was plotted on the y-axis as shown in Figure 7. The data shows the
linear relationship between the particle density and the MIC signal as expected from Eq. 12.

Smoke Characterization Project Final Report P. 33 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0E+00
5.0E+04
1.0E+05
1.5E+05
2.0E+05
2.5E+05
3.0E+05
3.5E+05
0 10 20 30 40 50 60 70
Percent
MIC Signal Change (pA)
S
u
m
(
n
i d
i )
Test 1
Test 2

Figure 7 Relationship between the MIC signal and particle density in the UL 217 Smoke Box for non-
flaming cotton wick
Smoke Characterization Project Final Report P. 34 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
SMALL-SCALE TESTS

Introduction
The ASTM E1354 cone calorimeter was selected to investigate the combustion of various
materials on a small-scale because it can simulate well- ventilated, early stage fires and allows 5
control of the heating conditions leading to thermal decomposition and ignition of the test sample.

In this portion of the investigation, solid and liquid test samples were evaluated under flaming
and non- flaming combustion conditions.
10
Test Samples
Test samples were selected from the list in Table 2 and included both natural and synthetic
materials with different chemical structures. The selected samples are presented in Table 4.
Table 4 Cone calorimeter test samples
Test Sample Comment
3:1 Heptane/Toluene mixture
UL 217 test material mixture of short straight chain and
simple aromatic hydrocarbon molecules
Douglas fir UL 217 test material
Newspaper UL 217 test material
Ponderosa pine UL 217 test material
Heptane Hydrocarbon liquid short straight chain hydrocarbon
HDPE Polyolefin plastic long straight chain hydrocarbon
Bread Potential nuisance source
Lard Used in cooking; Potential nuisance source
Cooking oil Hydrocarbon liquid intermediate length hydrocarbon
Mattress composite
Natural and synthetic materials; Commonly found in home
furnishings
Mattress PU foam
Synthetic; Flexible, open cell structure; Commonly found in
home furnishings
Cotton batting Natural material; Commonly found in home furnishings
Polyester pillow stuffing Aromatic; Commonly found in home furnishings
CA TB 117 50:50 Cotton/
Polyester blend fabric
Natural and synthetic materials blend; Commonly found in bed
clothing and apparel
Rayon fabric Synthetic; Commonly found in apparel
Nylon carpet Synthetic; Commonly found as a flooring product
PET carpet Synthetic; Commonly found as a flooring product
Polyisocyanurate insulation
foam
Synthetic; Rigid, closed cell structure; Commonly found as
insulation
PVC wire Common electrical wiring
15
Solid test specimen measuring 100 100 mm square were cut and tested in a horizontal
orientation using an edge frame sample holder with a restraining grid (HEG) such that the
intended outer surface of the material was exposed to the applied radiant heat flux. Liquid
samples were tested in 50 mL quantities using a glass Petri dish with a surface area of 0.0061 m
2
.
Examples of a solid and liquid sample are presented in Figure 8. 20

Smoke Characterization Project Final Report P. 35 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.

Figure 8 Cone Calorimeter sample holder

Experimental
Cone Calorimeter - Cone calorimeter tests were conducted in accordance with test method 5
ASTM E1354 Standard Test Method for Heat and Visible Smoke Release Rates for Materials
and Products Using an Oxygen Consumption Calorimeter. The apparatus consists of a conical
shaped electrical heater capable of heating a test sample with radiant heat flux of up to 100
kW/m
2
, a load cell, a laser smoke obscuration system, and gas analysis equipment. A schematic
of the Cone Calorimeter is shown in Figure 9. 10

Figure 9 Schematic of ASTM E 1354 cone calorimeter

Flaming mode tests were performed at 35 kW/m
2
radiant heat flux setting on the conical heater
and using an electric spark igniter to ignite the thermal decomposition gases. Non- flaming mode 15
tests were conducted at a radiant heat flux of 15 kW/m
2
but the combustion products were not
Smoke Characterization Project Final Report P. 36 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
ignited using the electric spark igniter. Since heptane is a flammable liquid, it was tested without
the application of external radiant heating, but a spark was used to ignite the vapors.

For the flaming mode. data was collected until flaming or other signs of combustion ceased. For
the non-flaming mode, the test duration was ten minutes in order to collect sufficient data for this 5
investigation. Observations regarding ignition time and physical changes to the sample (i.e.
melting, swelling, or cracking) were also noted.

The heat and smoke release rates, effective heat of combustion, and specific extinction area were
calculated using the procedures described in ASTM E1354 and are summarized in the following 10
equations.

Heat release relations:
HRR =
area Sample
heat Measured
[=] kW/m
2

Eq. 14

Total Heat =
kJ / MJ 1000
dt HRR
completion
ignition

[=] MJ/m
2

Eq. 15
15
Effective Heat of Combustion =
kJ/MJ 1000 loss weight Total
area Sample Heat Total

[=] kJ/g
Eq. 16

Smoke release relations:
SRR = Volumetric flow rate
length path Sample
density Optical
[=] m
2
/s
SRR = Extinction Coefficient () Mass flow rate
Eq. 17

Total Smoke =

completion
ignition
dt SRR [=] m
2
Eq. 18

Specific extinction area =
loss weight Total
Smoke Total
[=] m
2
/g
Eq. 19
20
Combining Eq. 17 through Eq. 19, it may be observed that the Smoke Yield is proportional to the
Extinction Coefficient () and Specific Extinction Area () as:
Smoke Yield =

[=] dimensionless Eq. 20
Babrauskas and Mulholland
10,11
have been found that the Extinction Coefficient is relatively
constant at 8,500 m
2
/kg for well- ventilated combustion of a wide variety of fuels.
25
Smoke Characterization Project Final Report P. 37 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Smoke Particle and Gas Effluent Sampling - A custom gas effluent and smoke sampling
system for the Cone Calorimeter was designed and constructed to condition the evolved smoke
for analyses in the WPS spectrometer and the gas FTIR spectrometer. A schematic of the
sampling system is shown in Figure 10. The sampling port was located 0.6 m away from the
cone hood in the exhaust duct and the sample line was divided to the two spectrometers. Smoke 5
and gas samples lines were diluted with nitrogen gas (UHP grade, 99.999%) to prevent saturation
of the respective detection instrument. The dilution ratio for the FTIR spectrometer was 2 and the
dilution ratio for the WPS spectrometer ranged from 8 to 21. The actual dilution flow rates were
documented for each test and used in the calculation of the smoke particle counts and gas
effluent concentration. 10

Sample lines to the spectrometers were 3 m long with a 3.2 mm I.D. The sample line to the FTIR
was maintained at 120 C to prevent condensation of generated water vapor in the effluent gas
stream.
15
Because the sampling port was facing downstream, it is anticipated that the data obtained will be
biased towards the smaller particles. In addition, some particulates are anticipated to be lost due
to adhesion to the sampling tube. The sampling tubes were cleaned prior to each test.

FTIR
N2
Smoke Particle
Size
Measurement
Sample Holder
Exhaust Duct
2 ft
N2
20
Figure 10 Schematic of the gas effluent and smoke measurement system for the cone calorimeter

Prior to each test, the FTIR gas spectrometer and the WPS spectrometer were purged with
ambient air. Both the analyzers were checked to ensure that the background signal was
insignificant prior to initiating a test. 25

Smoke Particle Characterization - Smoke particle size and count was characterized using the
WPS spectrometer previously described in the Smoke Characterization section.

Effluent Gas Composition Characterization - Gas effluent composition was characterized 30
using the FTIR spectrometer and deconvoluted as previously described in the Smoke
Characterization section (Eq. 8 through Eq. 11).

In order to determine the mass of the generated effluent gases, the deconvoluted FTIR
concentrations [i]
in
must be corrected for temperature differences between the FTIR cell and the 35
Smoke Characterization Project Final Report P. 38 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
cone calorimeter sampling port, the cone calorimeter mass flow rate, and respective gas
molecular weight:
( ) dt
MW
MW
Rate Flow Cone
T
T
] i [ Mass
air
gas
air
cone
FTIR
in gas

,
_

,
_

[=] g Eq. 21
such that the density of air is 353.22/T
cone
.

The following values were used for the calculations: 5
T
FTIR
= FTIR cell temperature = 393 K
T
cone
= Cone effluent gas temperature measured at photocell
MW
air
= Molecular weight of air = 28.97 g/mol

Exposure Scenario - The exposure scenario used to conduct the flaming and non-flaming tests 10
are summarized in Table 5 and Table 6 respectively.


Smoke Characterization Project Final Report P. 39 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 5 Test parameters for cone calorimeter flaming mode tests
Initial Weight (g) Dilution Rate
Test Sample
Heat
Flux
(kW/m
2
)
Sample
Area
(m
2
)
Test 1 Test 2 FTIR WPS
UL 217 Heptane/Toluene mixture 0 0.0061 32.8 -- 2 16
Heptane 0 0.0061 32.7 33.3 2 16
UL 217 Douglas fir 35 0.0088 98.8 94.3 2 16
UL 217 Newspaper 35 0.0088 7.0 7.0 2 16
UL 217 Ponderosa pine 35 0.0088 91.9 93.4 2 16
HDPE 35 0.0088 61.8 61.9 2 13
Bread 35 0.0088 22.8 22.1 2 21
Cooking oil 35 0.0061 40.0 40.2 2 16
Mattress composite 35 0.0088 9.0 9.1 2 16
Mattress PU foam 35 0.0088 7.2 7.2 2 16
Cotton batting 35 0.0088 5.9 6.0 2 16
Polyester pillow stuffing 35 0.0088 4.0 4.0 2 16
CA TB 117 50:50 Cotton/
Polyester blend fabric
35 0.0088 10.1 10.2 2 16
Rayon fabric 35 0.0088 9.9 9.8 2 8.5
Nylon carpet 35 0.0088 29.2 30.0 2 18
PET carpet 35 0.0088 29.5 29.0 2 16
Polyisocyanurate insulation foam 35 0.0088 6.0 5.6 2 16
PVC wire 35 0.0088 78.5 78.5 2 16

Table 6 Test parameters for cone calorimeter non-flaming mode tests
Initial Weight (g) Dilution Rate
Test Sample
Heat
Flux
(kW/m
2
)
Sample
Area
(m
2
)
Test 1 Test 2 FTIR WPS
UL 217 Douglas fir 15 0.0088 100.9 99.0 2 21
UL 217 Newspaper 15 0.0088 7.0 7.0 2 16
UL 217 Ponderosa pine 15 0.0088 91.1 90.9 2 16
HDPE 15 0.0088 60.6 61.6 2 21
Bread 15 0.0088 20.7 24.0 2 16
Lard 15 0.0061 63.5 -- 2 16
Cooking oil 15 0.0061 40.0 40.0 2 16
Mattress composite 15 0.0088 9.3 9.3 2 16
Mattress PU foam 15 0.0088 7.2 7.3 2 16
Cotton batting 15 0.0088 7.0 7.8 2 16
Polyester pillow stuffing 15 0.0088 4.0 4.1 2 16
CA TB 117 50:50 Cotton/
Polyester blend fabric
15 0.0088 9.9 10.0 2 16
Rayon fabric 15 0.0088 9.9 10.0 2 16
Nylon carpet 15 0.0088 30.0 28.9 2 21
PET carpet 15 0.0088 29.5 27.6 2 16
Polyisocyanurate insulation foam 15 0.0088 5.8 5.7 2 16
PVC wire 15 0.0088 78.5 78.5 2 16

Smoke Characterization Project Final Report P. 40 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Test Results
The cone calorimeter combustibility results from the tests included ignition time, sample weight,
heat and smoke release rates, effective heat of combustion, and specific extinction area.

Sample ignition occurred in all flaming mode tests. Sample ignition was not observed in any of 5
the non-flaming tests, however thermal degradation was observed in some of the tests.
Combustibility data for flaming and non- flaming tests are summarized in Table 7 and Table 8
respectively.

The smoke particle size distribution data measured on the WPS spectrometer were analyzed to 10
calculate the mean particle diameter D
m
and count N
m
for each test as described by Eq. 6 and Eq.
7. Mean particle count was further corrected to compensate for weight loss differences between
the evaluated materials as described in Eq. 22.
Specific N
m
= N
m
/ weight loss [=] cm
-3
g
-1
Eq. 22
Similarly the gas concentrations were also normalized by weight loss to determine the yield.
15
Mean smoke particle size, specific mean particle counts, maximum specific carbon monoxide
and carbon dioxide concentrations, and carbon monoxide and carbon dioxide yields for flaming
and non- flaming tests are summarized in Table 9 and Table 10 respectively.

Individual results for flaming and non- flaming combustion tests are plotted in Appendix C and D 20
respectively.



25
Smoke Characterization Project Final Report P. 41 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 7 Cone calorimeter combustibility data for small-scale flaming mode tests
Sample
Description
Ignition
Time
(s)
Total
Weight
Loss
(g)
Weight
Loss
Fraction
(%)
Effective
HOC
(kJ/g)
Peak
HRR
(kW/m)
Peak
SRR
(m/s)
Specific
Ext.
Area
(m/g )
UL 217 Heptane/
Toluene mix
42 32.80 100.0 40.7 715 0.066 0.492
6 32.70 100.0 43.0 543 0.010 0.117
Heptane
10 33.25 100.0 44.1 577 0.010 0.111
87 85.76 86.8 12.5 155 0.010 0.048 UL 217 Douglas
fir
86 84.13 89.2 11.4 133 0.008 0.016
15 7.00 100.0 15.1 89 0.010 0.010 UL 217
Newspaper
7 7.00 100.0 13.8 109 0.004 0.007
58 77.50 84.3 11.3 142 0.005 0.004 UL 217 Pond.
pine
90 76.05 81.4 12.2 154 0.011 0.010
144 29.97 48.5 30.0 467 0.051 0.285
HDPE
140 47.88 77.4 22.2 629 0.060 0.215
17 20.11 88.5 6.8 83 0.021 0.117
Bread
63 19.65 89.1 6.3 67 0.016 0.084
130 39.97 100.0 32.7 549 0.069 0.743
Cooking oil
138 40.15 100.0 33.5 584 0.069 0.736
16 8.99 100.0 20.6 193 0.021 0.142 Mattress
composite 14 9.08 100.0 21.2 196 0.020 0.158
3 7.22 100.0 23.7 250 0.014 0.077 Mattress PU
foam
6 7.22 100.0 23.3 240 0.014 0.083
13 5.13 86.9 14.2 164 0.040 0.239
Cotton batting
12 5.29 88.2 15.4 175 0.040 0.242
73 4.04 100.0 15.9 176 0.050 0.323 Polyester pillow
stuffing 144 4.00 100.0 16.5 204 0.057 0.414
24 9.89 97.5 15.1 338 0.066 0.271 Cotton/Polyester
blend fabric
37 10.16 100.0 16.9 318 0.072 0.295
68 9.85 100.0 14.1 222 0.010 0.052
Rayon fabric
38 9.77 100.0 16.0 213 0.008 0.078
105 21.27 72.9 29.1 410 0.084 0.467
Nylon carpet
125 21.40 71.3 31.9 453 0.094 0.458
114 19.11 64.9 18.3 259 0.080 0.545
PET carpet
94 18.32 63.2 19.4 260 0.076 0.521
9 2.66 44.6 7.9 67 0.005 0.117 Polyisocyanurate
foam
16 2.84 51.1 9.1 94 0.008 0.078
43 26.47 33.7 16.2 197 0.100 0.739
PVC wire
39 27.30 34.8 14.9 182 0.094 0.733

Smoke Characterization Project Final Report P. 42 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 8 Cone calorimeter combustibility data for small-scale non-flaming mode tests
Sample Description
Total
Weight
Loss
(g)
Weight
Loss
Fraction
(%)
Peak
HRR
(kW/m)
Peak
SRR
(m/s)
Total
Smoke
(m)
Specific
Ext.
Area
(m/g)
4.22 4.2 trace
[1]
trace trace ---
UL 217 Douglas fir
4.32 4.4 trace trace trace ---
6.71 95.9 22 0.012 2.1 0.315
UL 217 Newspaper
5.78 82.6 14 0.012 2.2 0.371
9.04 9.9 trace trace trace ---
UL 217 Ponderosa pine
9.49 10.4 trace trace trace ---
3.29 5.4 trace trace trace ---
HDPE
0.33 0.5 trace trace trace ---
11.79 57.0 trace 0.008 2.1 0.176
Bread
18.13 75.7 trace 0.009 4.4 0.244
Lard 0.24 0.4 trace trace trace ---
0.51 1.3 trace trace trace ---
Cooking Oil
0.61 1.5 trace trace trace ---
4.89 52.5 trace 0.014 4.2 0.849
Mattress composite
5.00 53.8 trace 0.016 3.3 0.668
3.43 47.4 trace 0.009 2.7 0.786
Mattress PU Foam
4.56 62.6 trace 0.009 4.8 1.042
2.34 33.4 trace 0.004 1.4 0.604
Cotton Batting
3.25 41.6 trace 0.005 2.3 0.714
0.41 10.4 trace trace trace --- Polyester pillow
stuffing 0.42 10.2 trace trace trace ---
5.35 54.1 trace 0.007 2.8 0.530 Cotton/Polyester blend
fabric 5.28 53.0 trace 0.007 3.0 0.560
9.90 100.0 19 0.012 2.7 0.273
Rayon fabric
9.99 100.0 19 0.014 3.0 0.297
1.22 4.1 trace trace trace ---
Nylon Carpet
1.20 4.2 trace trace trace ---
PET Carpet 1.26 4.3 trace trace trace ---
1.44 24.9 trace trace trace ---
Polyisocyanurate foam
1.62 28.4 trace trace trace ---
18.34 23.2 trace 0.005 2.3 0.127
PVC wire
12.21 15.6 trace 0.006 2.2 0.177
Note to Table 8:
[1]
A value of trace indicates that the measured values were less than the resolution of the instrument.

5
Smoke Characterization Project Final Report P. 43 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 9 Smoke particle and gas effluent data for small-scale flaming mode tests
Smoke Particles Effluent CO Effluent CO
2

Sample Description
D
m

(m)
Specific N
m

(1/cc/g)
Max
(ppm)
Yield
(g/g)
Max
(ppm)
Yield
(g/g)
UL 217 Heptane/Toluene mix 0.264 9.60E+04 318 0.069 69 2.143
0.199 1.10E+05 63 0.020 20 2.471
Heptane
0.195 1.28E+05 68 0.022 22 2.413
0.073 4.36E+04 297 0.087 87 0.998
UL 217 Douglas fir
0.040 9.09E+04 291 0.093 93 0.928
0.041 9.63E+05 434 0.259 259 1.194
UL 217 Newspaper
0.046 1.25E+06 429 0.264 264 1.203
0.037 5.14E+04 386 0.092 92 1.468
UL 217 Ponderosa pine
0.034 8.02E+04 344 0.071 71 1.147
0.167 8.48E+04 229 0.039 39 1.199
HDPE
0.158 3.40E+04 369 0.043 43 1.439
0.059 4.96E+05 161 0.099 99 0.488
Bread
0.071 6.31E+05 190 0.113 113 0.474
0.226 4.20E+04 341 0.097 97 2.162
Cooking oil
0.293 1.40E+05 372 0.101 101 2.276
0.045 2.04E+06 158 0.140 140 0.881
Mattress composite
0.048 6.13E+05 190 0.146 146 1.812
0.050 2.13E+06 64 0.029 29 1.060
Mattress PU foam
0.048 1.83E+06 79 0.044 44 1.455
0.095 9.92E+05 326 0.310 310 1.360
Cotton batting
0.092 8.03E+05 301 0.278 278 1.179
0.091 1.29E+06 229 0.187 187 1.362
Polyester pillow stuffing
0.093 1.01E+06 242 0.137 137 1.516
0.083 2.62E+05 414 0.217 217 1.593
Cotton/Polyester blend fabric
0.085 5.68E+05 393 0.227 227 1.426
0.054 1.69E+05 226 0.113 113 1.559
Rayon fabric
0.067 1.44E+05 164 0.092 92 1.034
0.134 3.11E+05 347 0.066 66 1.725
Nylon carpet
0.112 5.28E+05 431 0.069 69 1.800
PET carpet 0.128 1.91E+05 385 0.141 141 1.211
0.070 2.42E+05 133 0.041 41 0.204
Polyisocyanurate foam
0.063 3.11E+06 104 0.164 164 0.562
0.135 2.90E+06 88 0.132 132 0.430
PVC wire
0.138 3.15E+05 492 0.115 115 0.859
Smoke Characterization Project Final Report P. 44 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 10 Smoke particle and gas effluent data for small-scale non-flaming mode tests
Smoke Particles Effluent CO Effluent CO
2

Sample Description
D
m

(m)
Specific N
m

(1/cc/g)
Max
(ppm)
Yield
(g/g)
Max
(ppm)
Yield
(g/g)
0.136 1.05E+05 10 0.017 17 0.000
UL 217 Douglas fir
0.141 1.05E+05 12 0.023 23 0.000
0.101 4.41E+05 319 0.673 673 0.549
UL 217 Newspaper
0.103 4.91E+05 275 0.901 901 0.687
0.132 7.28E+04 59 0.129 129 0.141
UL 217 Ponderosa pine
0.156 8.08E+04 63 0.129 129 0.054
0.076 1.64E+05 10 0.019 19 0.246
HDPE
0.076 1.65E+06 12 0.218 218 0.019
0.095 2.15E+05 84 0.043 43 0.164
Bread
0.104 2.28E+05 94 0.106 106 0.210
Lard 0.075 5.13E+06 3 0.085 --
[1]
--
[1]

0.079 1.94E+06 2 0.093 93 0.612
Cooking Oil
0.077 1.89E+06 2 0.055 55 1.299
0.061 5.66E+05 194 0.255 255 0.112
Mattress composite
0.072 5.32E+05 203 0.266 266 0.273
0.085 1.86E+06 14 0.044 44 0.699
Mattress PU Foam
0.076 2.89E+06 14 0.047 47 0.152
0.086 7.09E+05 42 0.262 262 0.745
Cotton Batting
0.105 5.94E+05 107 0.318 318 0.298
0.041 1.33E+06 2 0.033 --
[1]
--
[1]

Polyester pillow stuffing
0.047 6.95E+05 2 0.036 --
[1]
--
[1]

0.136 1.18E+05 138 0.388 388 0.391
Cotton/Polyester blend fabric
0.116 3.01E+05 60 0.311 311 0.884
0.088 2.64E+05 502 0.738 738 0.340
Rayon fabric
0.093 2.21E+05 503 0.686 686 0.311
0.072 1.86E+06 12 0.095 95 0.138
Nylon Carpet
0.079 1.66E+06 13 0.104 104 0.002
0.133 5.71E+05 25 0.215 215 0.243
PET Carpet
0.120 3.41E+04 28 0.011 11 0.009
0.082 7.71E+05 7 0.065 65 1.230
Polyisocyanurate foam
0.073 1.01E+06 6 0.063 63 0.179
0.132 3.70E+04 16 0.008 8 0.145
PVC Wire
0.100 3.19E+05 103 0.085 85 0.258
Note to Table 10:
[1]
Observed carbon dioxide levels are suspect.

Discussion of small-scale flaming combustion results 5
Comparison of heat release rates and an effective inherent heat of combustion in the flaming
mode (note that heptane and the heptane-toluene mixture were ignited without any incident heat
flux), plotted in Figure 11, indicate that natural cellulosic materials generally have the lowest
heat release whereas hydrocarbon and synthetic materials have the highest heat release. The heat
releases exhibited by the natural cellulosic materials and synthetic materials prescribed by UL 10
Smoke Characterization Project Final Report P. 45 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
217 are in the same range as the other evaluated materials. Materials with higher effective heat of
combustion exhibit greater peak heat release rates.

0
5
10
15
20
25
30
35
40
45
50
C
o
o
k
in
g

o
il
H
e
p
ta
n
e
H
e
p
ta
n
e
/
T
o
lu
e
n
e

m
ix
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
.
p
in
e
H
D
P
E
B
re
a
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

fo
a
m
C
o
tto
n

b
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

c
a
r
p
e
t
P
E
T

c
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
ra
te

fo
a
m
P
V
C

w
ir
e
E
f
f
e
c
t
i
v
e

H
O
C

(
k
J
/
g
)
UL 217 materials

0
100
200
300
400
500
600
700
800
C
o
o
k
in
g

o
il
H
e
p
ta
n
e
H
e
p
ta
n
e
/
T
o
lu
e
n
e

m
ix
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
. p
in
e
H
D
P
E
B
re
a
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttre
s
s

P
U

fo
a
m
C
o
tto
n

b
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

c
a
r
p
e
t
P
E
T

c
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
ra
te

fo
a
m
P
V
C

w
ir
e
P
e
a
k

H
R
R

(
k
W
/
m

)
UL 217 materials
5
Figure 11 Effective HOC (top) and peak HRR (bottom) for flaming combustion

Similarly, smoke production during flaming combustion is greater for synthetic materials than
that for natural cellulosic products, plotted in Figure 12. Material chemistry plays a significant
role in the amount of smoke produced such that: 10
Smoke Characterization Project Final Report P. 46 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
1. Introduction of aromatic groups to simple straight chain hydrocarbons increases smoke
production (heptane-toluene mixture versus heptane alone).
2. Materials with aromatic molecular groups exhibited the highest smoke production
polyester products (carpet, pillow stuffing, sheet), PVC wire, and heptane-toluene
mixture. 5
3. Unsaturated cooking oil very likely decomposes to soot.
4. Substitution of nitrogen and chlorine atoms into the base polymer molecule as well as
aromatic additives (nylon carpet, PVC) also increases smoke production.

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
C
o
o
k
in
g

o
il
H
e
p
ta
n
e
H
e
p
ta
n
e
/
T
o
lu
e
n
e

m
ix
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
.
p
in
e
H
D
P
E
B
r
e
a
d
M
a
ttre
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

fo
a
m
C
o
tto
n

b
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

c
a
r
p
e
t
P
E
T

c
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
ra
te

fo
a
m
P
V
C

w
ire
S
p
e
c
i
f
i
c

E
x
t
i
n
c
t
i
o
n

A
r
e
a

(
m

/
g
)
UL 217 materials
10
Figure 12 Smoke production for flaming combustion

The mean particle sizes and specific counts for the evaluated materials are plotted in Figure 13
and Figure 14. Smokes generated by materials such as heptane, toluene, cooking oil, and HDPE
have the largest mean sizes whereas the natural cellulosic materials and PU foam based materials 15
have the smallest. The natural cellulosic materials and synthetic materials used in UL 217 are in
the same range as the other evaluated materials. It was observed that materials generating larger
smoke particles, e.g. cooking oil, heptane/toluene mixture, also have larger specific extinction
areas, Figure 12. The cooking oil contains unsaturated, long-chain hydrocarbon components that
resemble the behavior of the heptane-toluene mixture. 20

It may be observed that the mean smoke particle sizes generated by the different samples trends
with the energy required to vaporize the respective material for subsequent combustion such that
materials requiring the least amount of energy generate the largest mean particle sizes. The liquid
samples (heptane, heptane-toluene mixture, cooking oil) that generate the largest mean particle 25
sizes require the least amount of energy for vaporization as they do not need to be first liquefied
like solid samples. HDPE, a long chain analog of heptane that is a solid at room temperature, is
easily liquefied prior to vaporization and has the next largest particles, followed by the PVC wire
which incorporates an easily liquefiable plasticizer in the PVC compound. The smallest particles
Smoke Characterization Project Final Report P. 47 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
are from the cross- linked materials (PU and polyisocyanurate foams) and the two wood samples
which form a cross- linked char structure during combustion.

0.00
0.05
0.10
0.15
0.20
0.25
0.30
C
o
o
k
in
g

o
il
H
e
p
ta
n
e
H
e
p
ta
n
e
/
T
o
lu
e
n
e

m
ix
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
.
p
in
e
H
D
P
E
B
r
e
a
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

fo
a
m
C
o
tto
n

b
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

c
a
r
p
e
t
P
E
T

c
a
r
p
e
t
P
o
ly
is
o
c
y
a
n
u
r
a
te

fo
a
m
P
V
C

w
ire
M
e
a
n

P
a
r
t
i
c
l
e

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
UL 217 materials

Figure 13 Mean particle diameter for flaming combustion 5

Specific smoke particle counts indicate that the materials with the highest surface area to sample
volume ratios (the two foam materials, newspaper, cotton batting, and polyester fill) generate
more particles per consumed mass than the other evaluated materials. It is also worth noting that
the two most prolific particle producers, the two foam materials, contain nitrogen atoms in the 10
polymer backbone. The higher particle production from PVC versus HDPE is in part due to the
high percentage of easily liquefiable aromatic plasticizers in the PVC wire insulation compound.

Smoke Characterization Project Final Report P. 48 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0E+00
5.0E+05
1.0E+06
1.5E+06
2.0E+06
2.5E+06
3.0E+06
3.5E+06
C
o
o
k
in
g

o
il
H
e
p
ta
n
e
H
e
p
ta
n
e
/
T
o
lu
e
n
e

m
ix
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
.
p
in
e
H
D
P
E
B
re
a
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

fo
a
m
C
o
tto
n

b
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

c
a
rp
e
t
P
E
T

c
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
r
a
te

fo
a
m
P
V
C

w
ire
M
e
a
n

S
p
e
c
i
f
i
c

P
a
r
t
i
c
l
e

C
o
u
n
t

(
1
/
c
c
/
g
)
UL 217 materials

Figure 14 Mean specific particle count for flaming combustion

The smoke particle characteristics also depend upon the specific combustion reaction mechanism
as a function of time. For example the particle size and count change significantly for Douglas fir 5
wood during the combustion process. After initial ignition of this material a char layer develops
that reduces the heat release rate per unit area. The smoke particle size also changes and the
smoke particle size reduces. The particle size then increases in conjunction with the heat release
rate per unit area as depicted in Figure 15.
10
0
20
40
60
80
100
120
140
160
180
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Time (s)
H
R
R
(k
W
/m
)
0
10
20
30
40
50
60
70
80
90
100
W
e
ig
h
t F
ra
c
tio
n
(%
)
HRR
Weight %

0
0.05
0.1
0.15
0.2
0.25
0.3
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Time (s)
M
e
a
n
D
ia
m
e
te
r (m
ic
ro
n
)
0
10
20
30
40
50
60
70
80
90
100
W
e
ig
h
t F
ra
c
tio
n
(%
)
Mean Diameter
Weight Fraction

Figure 15 Heat release rate per unit area and smoke particle size for flaming Douglas fir wood

In contrast to such charring materials, liquid samples such as the heptane/toluene mixture and
liquefied materials such as the HDPE after 200 s exposure result in consistent particle sizes 15
throughout the test, Figure 16 and Figure 17 respectively.

Smoke Characterization Project Final Report P. 49 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
100
200
300
400
500
600
700
800
0 100 200 300 400 500 600 700
Time (s)
H
R
R
(k
W
/m
)
0
10
20
30
40
50
60
70
80
90
100
W
e
ig
h
t F
ra
c
tio
n
(%
)
HRR
Weight %

0
0.05
0.1
0.15
0.2
0.25
0.3
0 100 200 300 400 500 600 700
Time (s)
M
e
a
n
D
ia
m
e
te
r (m
ic
ro
n
)
0
10
20
30
40
50
60
70
80
90
100
W
e
ig
h
t F
ra
c
tio
n
(%
)
Mean Diameter
Weight Fraction

Figure 16 Heat release rate per unit area and smoke particle size for flaming heptane/toluene mixture

0
50
100
150
200
250
300
350
400
450
500
0 200 400 600 800 1000 1200
Time (s)
H
R
R
(k
W
/m
)
0
10
20
30
40
50
60
70
80
90
100
W
e
ig
h
t F
ra
c
tio
n
(%
)
HRR
Weight %

0
0.05
0.1
0.15
0.2
0.25
0.3
0 200 400 600 800 1000 1200
Time (s)
M
e
a
n
D
ia
m
e
te
r (m
ic
ro
n
)
0
10
20
30
40
50
60
70
80
90
100
W
e
ig
h
t F
ra
c
tio
n
(%
)
Mean Diameter
Weight Fraction

Figure 17 Heat release rate per unit area and smoke particle size for flaming HDPE 5

Effluent gas analysis indicates water and carbon dioxide are the predominant species, and carbon
monoxide to a lesser extent. This is consistent with the chemical reaction for hydrocarbon
combustion. Average carbon dioxide and carbon monoxide yields for the different materials are
plotted in Figure 18 and Figure 19 respectively. In general carbon dioxide yield ranged between 10
1 to 1.5 g/g for the various materials; liquid materials exhibited the highest CO2 yields ranging
between 2 to 2.5 g/g. Carbon monoxide yield was less than 0.16 g/g with the exception of the
higher unmodified cellulose content materials (newspaper, cotton batting, and cotton/poly sheet)
which ranged between 0.2 to 0.3 g/g.
15
Smoke Characterization Project Final Report P. 50 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
C
o
o
k
in
g

o
il
H
e
p
ta
n
e
H
e
p
ta
n
e
/
T
o
lu
e
n
e

m
ix
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
e
r
o
s
a

p
in
e
H
D
P
E
B
r
e
a
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

fo
a
m
C
o
tto
n

b
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

c
a
r
p
e
t
P
E
T

c
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
r
a
te

fo
a
m
P
V
C

w
ir
e
C
a
r
b
o
n

D
i
o
x
i
d
e

Y
i
e
l
d

(
g
/
g
)
UL 217 materials

Figure 18 Carbon dioxide yield for flaming combustion

0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
C
o
o
k
in
g

o
il
H
e
p
ta
n
e
H
e
p
ta
n
e
/
T
o
lu
e
n
e

m
ix
D
o
u
g
la
s
fir
N
e
w
s
p
a
p
e
r
P
o
n
d
e
r
o
s
a

p
in
e
H
D
P
E
B
re
a
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

fo
a
m
C
o
tto
n

b
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

c
a
r
p
e
t
P
E
T

c
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
r
a
te

fo
a
m
P
V
C

w
ir
e
C
a
r
b
o
n

M
o
n
o
x
i
d
e

Y
i
e
l
d

(
g
/
g
)
UL 217 materials

Figure 19 Carbon monoxide yield for flaming combustion 5


Smoke Characterization Project Final Report P. 51 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Discussion of small-scale non-flaming combustion results
Heat release rate per unit area for non- flaming combustion of most materials were below the
cone calorimeter resolution limit (less than 6 kW/m
2
). The three materials that generated
measurable amounts of heat had peak heat release rate per unit area of less than 20 kW/m
2
,
which is an order of magnitude less than observed for flaming combustion. 5

Similar to the heat release rate measurements on the non- flaming combustion tests, smoke
release rates for some of the materials evaluated under non- flaming combustion were also below
the cone calorimeter resolution limit (less than 0.004 m
2
/s). These materials are attributed as
having a smoke extinction area of zero for smoke production plotted in Figure 20. It may be 10
noted that the materials with measurable smoke release rates are the same materials identified as
having either a high surface area to volume ratio or loaded with easily liberated aromatic
plasticizers (PVC wire). In comparison to flaming combustion, most of the materials generate
more smoke per unit of consumed mass under non-flaming conditions. The most significant
effect of the combustion mode on smoke production is for the polyurethane and polyisocyanurate 15
foams, possibly due to the high surface area to volume ratio resulting from their unique physical
structure.

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
C
o
o
k
in
g

O
il
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
e
r
o
s
a

p
in
e
H
D
P
E
B
re
a
d
L
a
r
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

F
o
a
m
C
o
tto
n

B
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

C
a
rp
e
t
P
E
T
C
a
r
p
e
t
P
o
ly
is
o
c
y
a
n
u
ra
te

fo
a
m
P
V
C

W
ire
S
p
e
c
i
f
i
c

E
x
t
i
n
c
t
i
o
n

A
r
e
a

(
m

/
g
)
UL 217 materials

Figure 20 Smoke production for non-flaming combustion 20

The mean particle sizes and mean specific particle size counts for the evaluated materials are
plotted in Figure 21 and Figure 22 respectively. Smoke particles generated by the polyester
materials, Douglas fir, and Ponderosa pine are amongst the largest observed whereas the PU and
polyisocyanurate foams are amongst the smallest. Specific mean smoke particle counts indicate 25
that Douglas fir and Ponderosa pine are amongst the least prolific particle producers on a per
consumed mass basis whereas the lard, cooking oil, PU foam and nylon carpet are amongst the
next most prolific materials.
Smoke Characterization Project Final Report P. 52 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
C
o
o
k
in
g

O
il
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
e
ro
s
a

p
in
e
H
D
P
E
B
re
a
d
L
a
rd
M
a
ttre
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

F
o
a
m
C
o
tto
n

B
a
ttin
g
P
o
ly
e
s
te
r
p
illo
w
s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

C
a
rp
e
t
P
E
T

C
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
r
a
te

fo
a
m
P
V
C
W
ir
e
M
e
a
n

P
a
r
t
i
c
l
e

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
UL 217 materials

Figure 21 Mean particle diameter for non-flaming combustion
Larger mean particle size observed for cooking oil versus lard may be explained by its higher
unsaturated fat content. The carbon-carbon double bonds in unsaturated fats (referred to as
unsaturated bonds by chemists) can undergo an endothermic chemical reaction during thermal 5
degradation to form a cross- linked polymer network of saturated fats. This polymerization
reaction would retard particle formation. Smaller particle formation from higher molecular
weight materials is also observed for HDPE, despite being a saturated hydrocarbon. It was also
observed that for some materials (cooking oil, HDPE, PE/pillow stuffing and nylon carpet) the
mean particle size was smaller in the non-flaming mode than in the flaming mode. 10

Smoke Characterization Project Final Report P. 53 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0E+00
1.0E+06
2.0E+06
3.0E+06
4.0E+06
5.0E+06
6.0E+06
C
o
o
k
in
g

O
il
D
o
u
g
la
s

fir
N
e
w
s
p
a
p
e
r
P
o
n
d
e
ro
s
a

p
in
e
H
D
P
E
B
re
a
d
L
a
r
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttre
s
s

P
U

F
o
a
m
C
o
tto
n

B
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r b
le
n
d
fa
b
ric
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

C
a
r
p
e
t
P
E
T

C
a
r
p
e
t
P
o
ly
is
o
c
y
a
n
u
r
a
te

fo
a
m
P
V
C

W
ire
M
e
a
n

S
p
e
c
i
f
i
c

P
a
r
t
i
c
l
e

C
o
u
n
t

(
1
/
c
c
/
g
)
UL 217 materials

Figure 22 Mean specific particle count for non-flaming combustion
Comparison of the mean smoke particle sizes and mean specific particle counts measured for
non- flaming combustion to those measured for flaming combustion indicate that particle sizes
are generally larger for non- flaming combustion. This is particularly true for the two wood 5
species where the particle sizes are approximately three times larger. The specific particle counts
were up to an order of magnitude lower for non- flaming combustion. It may be noted that under
non- flaming combustion HDPE generated more, but smaller smoke particles than PVC wire
whereas under flaming combustion the HDPE generated less, but larger smoke particles.
10
Effluent gas analysis indicates water, carbon dioxide, and carbon monoxide are the predominant
species. This is consistent with the chemical reaction for incomplete hydrocarbon combustion.
Average carbon dioxide and carbon monoxide yields for the different materials are plotted in
Figure 23 and Figure 24 respectively. Carbon dioxide yield was less than 1 g/g for all of the
various materials; the only liquid material evaluated under non- flaming conditions, cooking oil, 15
exhibited the highest CO
2
yield. Carbon monoxide yield was less than 0.15 g/g with the
exception of the higher unmodified cellulose content materials (newspaper, cotton batting,
cotton/poly sheet, cotton batting topped PU foam mattress composite), Rayon (which is acetate
modified cellulose), and PET carpet.
20
Smoke Characterization Project Final Report P. 54 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
C
o
o
k
in
g

O
il
U
L

2
1
7

D
o
u
g
la
s

fir
U
L
2
1
7
N
e
w
s
p
a
p
e
r
U
L

2
1
7

P
o
n
d
e
ro
s
a

p
in
e H
D
P
E
B
re
a
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttre
s
s

P
U

F
o
a
m
C
o
tto
n

B
a
ttin
g
C
o
tto
n
/P
o
ly
e
s
te
r
b
le
n
d

fa
b
ric
R
a
y
o
n

fa
b
ric
N
y
lo
n

C
a
rp
e
t
P
E
T

C
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
ra
te

fo
a
m
P
V
C
W
ir
e
C
a
r
b
o
n

D
i
o
x
i
d
e

Y
i
e
l
d

(
g
/
g
)
UL 217 materials

Figure 23 Carbon dioxide yield for non-flaming combustion

0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
C
o
o
k
in
g

O
il
D
o
u
g
la
s

f
ir
N
e
w
s
p
a
p
e
r
P
o
n
d
e
r
o
s
a

p
in
e
H
D
P
E
B
re
a
d
L
a
r
d
M
a
ttr
e
s
s

c
o
m
p
o
s
ite
M
a
ttr
e
s
s

P
U

F
o
a
m
C
o
tto
n

B
a
ttin
g
P
o
ly
e
s
te
r

p
illo
w

s
tu
ffin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

fa
b
r
ic
R
a
y
o
n

fa
b
r
ic
N
y
lo
n

C
a
rp
e
t
P
E
T

C
a
rp
e
t
P
o
ly
is
o
c
y
a
n
u
r
a
te

fo
a
m
P
V
C

W
ir
e
C
a
r
b
o
n

M
o
n
o
x
i
d
e

Y
i
e
l
d

(
g
/
g
)
UL 217 materials

Figure 24 Carbon monoxide yield for non-flaming combustion 5
It is also worth noting that the textile and newspaper materials that exhibit the highest carbon
monoxide release rates are commonly found in residential settings.
Smoke Characterization Project Final Report P. 55 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
INTERMEDIATE-SCALE TESTS

Introduction
Potential flaming and non-flaming scenarios for subsequent evaluation to UL 217 Fire Test
Room alarm response parameters in Task 3 were developed using intermediate-scale tests. 5
Evaluation of the UL 217 fire test protocols and the developed fire scenarios in intermediate
calorimeters also permitted characterization of heat and smoke release rates as well as smoke and
gas effluents closer to the combustion source. This enabled characterization of the smoke
particles prior to transport and aging that would be expected in the vicinity of smoke alarms in
the Fire Test Room. Two sizes of intermediate calorimeters were used depending upon the 10
sample size. These are identified as the NEBS calorimeter and the IMO calorimeter.

Smoke characteristics of smoldering Ponderosa pine were measured in ULs Fire Test Room
because the hot plate and controller could not be readily re- located to either of the two
calorimeter areas. Thus heat and smoke release rates were not measured. 15

Evolved heat and smoke were measured by the same principles as used in the ASTM E1354 cone
calorimeter; smoke particle size and gas-phase effluent components were measured using the
same WPS spectrometer and gas FTIR analyzer equipment previously described.
20
Initial testing using the NEBS calorimeter showed that the calorimeter could not be configured to
resolve combustibility data for fires less than 10 kW. Thus, a smaller calorimeter, IMO
calorimeter, was employed. Data for the UL 217 test samples were repeated in this calorimeter
and additional tests on other materials and scenarios were performed.
25
Test Samples
Test samples were selected from the materials listed in Table 2. The selected samples, other than
the UL 217 test samples, were selected on the basis of their chemistry (synthetic, natural), and
their performance in the Cone Calorimeter tests. The selected materials are presented in Table 11.
30

Smoke Characterization Project Final Report P. 56 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 11 Intermediate calorimeter test samples
Test Sample Comment Test Area(s)
3:1 Heptane/Toluene UL 217 test material NEBS, IMO
Heptane
Provides chemistry difference from heptane/toluene
mixture. Relatively large particle size in small-scale tests.
NEBS
Douglas fir UL 217 test material NEBS, IMO
Newspaper UL 217 test material NEBS, IMO
Ponderosa pine UL 217 test material
Fire Test Room
Pillow Composite material; Co-combustion expected NEBS
Mattress Composite material; Co-combustion expected NEBS
Cotton batting
Mattress component. Particle distribution was in the middle
of the range for other materials in small-scale tests.
NEBS
PU foam
Mattress component. Relatively high particle count and
small size in small-scale tests.
NEBS, IMO
Cigarette Potential nuisance source NEBS
Coffee maker
Composite; Co-combustion expected; Synthetic base
material had high heat release and relatively large particle
size in small-scale tests
NEBS, IMO
Bread Potential nuisance source NEBS, IMO
Nylon carpet Relatively high particle count and size in small-scale tests IMO

Experimental
NEBS Calorimeter - The NEBS product calorimeter test room is 15.2 m 4.9 m 4.9 m
(lwh) with a square shaped collection hood located centrally in the room 2.2 m above the floor. 5
The dimensions of the extended hood are 3.9 m on the side and a height of 1.5 m. Collected
combustion products are exhausted by way of a 0.6 0.6 m plenum into a 0.45 m diameter
exhaust duct for the heat and smoke measurements. An exhaust flow rate of 8 m/s (bi-directional
probe measured) was used for the tests. A schematic of the NEBS Calorimeter hood arrangement
is shown in Figure 25. 10

Smoke Characterization Project Final Report P. 57 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Floor
12.75 ft
6 in
17.875 in
8 in
24 in
24 in
52 in
25 in
5 ft
12.25 ft
8 ft

Figure 25 Schematic of NEBS calorimeter

For flaming mode, data was collected until either the heat release rate exceeded 100 kW or
flaming and/or other signs of combustion ceased. For non- flaming mode, the test duration ranged 5
between 10 and 12 minutes.

IMO Calorimeter - The IMO calorimeter consists of a rectangular collection hood measuring
1.3 1.3 m. The hood is connected with a 0.18 m exhaust duct. An instrumented section is
located in the exhaust duct connected to enable the measurements of heat and smoke release 10
rates.

A schematic of the IMO calorimeter is depicted in Figure 26.
Smoke Characterization Project Final Report P. 58 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
2
2

i
n
.
INSTRUMENTED
DUCT SECTION
EXHAUST
NON
COMBUSTIBLE
SKIRT
T
o

t
h
e

s
m
o
k
e

p
a
r
t
i
c
l
e

s
a
m
p
l
i
n
g
7
1

i
n
.
Test Sample

Figure 26 Schematic of the IMO calorimeter

Smoke Particle and Gas Effluent Sampling - A custom gas effluent and smoke sampling
system for the intermediate calorimeter was designed and constructed to condition the evolved 5
smoke for analyses in the WPS spectrometer and the gas FTIR spectrometer. The evolved smoke
and gas was sampled using 6.4 mm O.D. steel sampling tube mounted facing downstream along
the centerline of a 0.18 m diameter steel collection cone, Figure 27. The sample flow was
divided into two separate sample streams for dilution with nitrogen and subsequent smoke
particle size and gas component characterization. Smoke and gas samples lines were diluted with 10
nitrogen gas (UHP grade, 99.999%) to prevent saturation of the respective detection instrument.
The dilution ratio for the FTIR spectrometer ranged from 1.5 to 2 and the dilution ratio for the
WPS spectrometer ranged from 6 to 16. The actual dilution flow rates were documented for each
test and used in the calculation of the smoke particle counts and gas effluent concentration.
15
2.25 in
7 in
4.5 in
3.5 in
0.25 in
2 in

Figure 27 Intermediate calorimeter evolved smoke and gas sampling cone and tube
Smoke Characterization Project Final Report P. 59 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Sample lines to the spectrometers were 3 m long with a 3.2 mm I.D. The sample line to the FTIR
was maintained at 120 C to prevent condensation of generated water vapor in the effluent gas
stream.

Because the sampling port was facing downstream, it is anticipated that the data obtained will be 5
biased towards the smaller particles. In addition, some particulates are anticipated to be lost due
to adhesion to the sampling tube. The sampling tubes were cleaned prior to each test.

For tests conducted in the flaming mode the sampling cone and tube arrangement was located at
the interface between the plenum and the exhaust duct as depicted in Figure 28. For tests 10
conducted in the non-flaming mode the sampling cone and tube arrangement was located 0.27 m
above the load cell as depicted in Figure 29.

FTIR
N2
Smoke
Particle Size
Measurement
Exhaust Duct
Load Cell
10 ft
N2

Figure 28 Intermediate calorimeter flaming mode sampling arrangement 15

Smoke Characterization Project Final Report P. 60 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
FTIR
N2
Smoke Particle
Size
Measurement
N2
11 in.
Load Cell

Figure 29 Intermediate calorimeter non-flaming mode sampling arrangement

Smoke Particle Characterization - Smoke particle size and count was characterized using the
WPS spectrometer previously described in the Smoke Characterization section 5

Effluent Gas Composition Characterization - Gas effluent composition was characterized
using the FTIR spectrometer and deconvoluted as previously described in the Smoke
Characterization section (Eq. 8 through Eq. 11).
10
Smoke Characterization Project Final Report P. 61 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Ignition Scenario - Samples were evaluated for heat and smoke release, particle size and gas
effluent concentration under flaming and/or non- flaming exposure conditions as summarized in
Table 12.
Table 12 Intermediate calorimeter sample exposure scenario
Test Sample Size/Quantity Mode
Heat/Ignition
Source
Test Area(s)
Test
Duration
UL 217 Heptane/
Toluene mixture
45 mL Flaming UL 217 assembly
NEBS
IMO
250 s
200 s
UL 217 Douglas fir 1 crib Flaming UL 217 assembly
NEBS
IMO
365 s
340 s
UL 217 Newspaper 42.5 g Flaming UL 217 assembly
NEBS
IMO
190 s
270 s
Heptane 500 mL Flaming Open-Flame NEBS 500 s
Pillow 1 unit Flaming TB 604 burner NEBS 400 s
Mattress 1 unit Flaming CPSC 1633 burner NEBS 205 s
Cotton batting 300 300 6 mm Flaming TB 604 burner NEBS 535 s
PU Foam
300 300 25 mm
thick
Flaming TB 604 burner NEBS 500 s
PU Foam wrapped
in cotton/poly
sheet
100 100 100
mm
Flaming TB 604 burner IMO 480 s
Coffee maker 12 cup, no carafe Flaming TB 604 burner
NEBS
IMO
1600 s
950 s
Nylon carpet 100 100 mm Flaming
Cone heater at 35
kW/m
2

IMO 360 s
Ponderosa pine
8 sticks, 75 long
25 20 mm
Non
Flaming
UL 217 -
Temperature
controlled hot
plate
Fire Test
Room
3400 s
Bread 4 slices
Non-
Flaming
Toaster
NEBS
IMO
1035 s
600 s
Cigarettes 2
Non-
Flaming
Lighter NEBS 320 s
Mattress Quarter section
Non-
Flaming
3 Cigarettes NEBS 1940 s
Cotton batting 100 100 6 mm
Non-
Flaming
Hot Plate NEBS 450 s
PU foam 100 100 25 mm
Non-
Flaming
Hot Plate NEBS 710 s
PU foam
3- 50 100 25
mm thick
Non-
Flaming
Cone heater at 15
kW/m
2

IMO 600 s
PU foam with
cotton/poly sheet
100 100 25 mm
thick foam, 1 sheet
cotton-poly sheet
Non-
Flaming
One smoldering
cigarette
IMO 620 s
5
Smoke Characterization Project Final Report P. 62 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
UL 217 Smoldering Ponderosa Pine Test
The test sample for this test was eight Ponderosa pine sticks placed on a temperature controlled
hotplate. Each stick measured 75 25 19 mm with the 19 75 mm inch face in contact with
the hotplate. The space between sticks was 15 mm. The temperature of the hotplate was
controlled in accordance with Section 45 Smoldering Smoke Test of UL 217. A photograph of the 5
test set-up is shown in Figure 30.


Figure 30 Photograph of test set-up for UL 217 smoldering test
10
The smoke sampling collector is shown in Figure 27. The bottom of the smoke sampling
collector was held 11.5 inches above the hotplate to catch the decomposition products from the
test sample. The opening of sampling tube was pointing to the downstream flow to prevent
clogging. A schematic of the smoke sampling is depicted in Figure 31.
15
N2
Smoke
Particle Size
Measurement
11.5 in.

Figure 31 Schematic of smoke sampling for smoldering Ponderosa pine test

The test was conducted in accordance with protocol specified in the UL 217. The dilution for the
WPS spectrometer was documented. The gas sampling was initiated simultaneously with the hot 20
plate. The test was terminated at 60 minutes.
Smoke Characterization Project Final Report P. 63 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Intermediate Calorimeter Test Results
The data from the combustibility tests were analyzed to calculate the heat and smoke release
rates, specific extinction area, smoke particle size and count distribution, and gas effluent
composition for flaming and non-flaming modes of combustion. Heat and smoke release rates
were calculated using the procedures described in ASTM E1354. 5

The combustibility results for the tests performed in the NEBS calorimeter are presented in Table
13.
Table 13 Intermediate calorimeter combustibility results
Test Sample (Heat source) Area
Test
Series
Mode
Peak
HRR
(kW)
Peak
SRR
(m
2
/s)
Total
Smoke
(m
2
)
NEBS Test 1 Flaming 19 0.24 16
IMO Test 1 Flaming 14 0.34 30
3:1 Heptane/Toluene mixture
(UL 217)
IMO Test 2 Flaming 12 0.34 29
NEBS Test 1 Flaming < 10 0.08 2
IMO Test 1 Flaming 12 0.26 11
UL 217 Douglas fir
(UL 217)
IMO Test 2 Flaming 10 0.24 11
NEBS Test 1 Flaming < 10 0.53 12
IMO Test 1 Flaming 6 0.99 25
UL 217 Newspaper
(UL 217)
IMO Test 2 Flaming 6 1.04 39
Heptane (lighter) NEBS Test 1 Flaming 51 0.09 25
Pillow (TB 604 burner) NEBS Test 1 Flaming 62 1.10 141
Mattress (TB 604 burner) NEBS Test 1 Flaming 108 1.15 60
Cotton batting (TB 604 burner) NEBS Test 1 Flaming < 10 0.01 0.5
PU foam (TB 604 burner) NEBS Test 1 Flaming < 10 -- 0.3
IMO Test 1 Flaming 4 0.04 4.8 PU foam in cotton/poly sheet
(TB 604 burner) IMO Test 2 Flaming 5 0.08 6.0
NEBS Test 1 Flaming 87 1.27 461
IMO Test 1 Flaming 113 6.23 1346
Coffee maker
(TB 604 burner)
IMO Test 2 Flaming 113 4.79 1033
IMO Test 1 Flaming 4 0.15 20 Nylon carpet (cone heater at 35
kW/m
2
) IMO Test 2 Flaming 4 0.14 17
NEBS Test 1 Non-Flaming
[1]
< 10 0.28 32
IMO Test 1 Non-Flaming DNI 0.72 74 Bread (electric toaster)
IMO Test 2 Non-Flaming DNI 0.32 45
3 Smoldering cigarettes NEBS Test 1 Non-Flaming DNI -- --
Quarter mattress (3 smoldering
cigarettes)
NEBS Test 1 Non-Flaming DNI -- --
Cotton batting (hot plate) NEBS Test 1 Non-Flaming DNI 0.01 0.6
PU foam (hot plate) NEBS Test 1 Non-Flaming DNI 0.04 5.0
IMO Test 1 Non-Flaming DNI 6.1 6.1 PU foam (cone heater at 15
kW/m
2
) IMO Test 2 Non-Flaming DNI 5.8 5.8
PU foam with Poly-cotton sheet
(smoldering cigarette)
IMO Test 1 Non-Flaming DNI 0.00 0.1
Notes to Table 13: 10
[1]
Bread ignited 8:36 minutes into the test
DNI = Sample did not ignite
Smoke Characterization Project Final Report P. 64 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The heat and smoke release rates for the flaming IMO calorimeter tests are presented Figure 32
through Figure 37.

0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 60 120 180 240 300 360
Time (s)
H
e
a
t

R
e
l
e
a
s
e

R
a
t
e

(
k
W
)
Test 1
Test 2

5

0.0
0.1
0.1
0.2
0.2
0.3
0.3
0.4
0.4
0.5
0.5
0.0 60.0 120.0 180.0 240.0 300.0 360.0
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1
Test 2

Figure 32 Heat (top) and smoke (bottom) release rates for heptane/toluene mixture
Smoke Characterization Project Final Report P. 65 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 60 120 180 240 300 360 420
Time (s)
H
e
a
t

R
e
l
e
a
s
e

R
a
t
e

(
k
W
)
Test 1
Test 2




5
0.0
0.1
0.1
0.2
0.2
0.3
0.3
0.4
0.4
0.5
0.5
0 60 120 180 240 300 360 420
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1
Test 2

Figure 33 Heat (top) and smoke (bottom) release rate for Douglas fir
Smoke Characterization Project Final Report P. 66 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0 60.0 120.0 180.0 240.0 300.0 360.0
Time (s)
H
e
a
t

R
e
l
e
a
s
e

R
a
t
e

(
k
W
)
Test 1
Test 2




5
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 60 120 180 240 300 360
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1
Test 2

Figure 34 Heat (top) and smoke (bottom) release rate for newspaper
Smoke Characterization Project Final Report P. 67 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
Time (s)
H
e
a
t

R
e
l
e
a
s
e

R
a
t
e

(
k
W
)
Test 1
Test 2




5
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1
Test 2

Figure 35 Heat (top) and smoke (bottom) release for coffee maker
Smoke Characterization Project Final Report P. 68 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
0 60 120 180 240 300 360 420 480 540 600
Time (s)
H
e
a
t

R
e
l
e
a
s
e

R
a
t
e

(
k
W
)
Test 1
Test 2




5
0.0
0.1
0.1
0.2
0.2
0.3
0.3
0.4
0.4
0.5
0.5
0 60 120 180 240 300 360 420 480 540 600
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1
Test 2

Figure 36 Heat (top) and smoke (bottom) release for nylon carpet

Smoke Characterization Project Final Report P. 69 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
0 60 120 180 240 300 360 420 480 540 600
Time (s)
H
e
a
t

R
e
l
e
a
s
e

R
a
t
e

(
k
W
)
Test 1
Test 2




5
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0 60 120 180 240 300 360 420 480 540 600
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1
Test 2

Figure 37 Heat (top) and smoke (bottom) release for cotton/poly sheet wrapped PU foam

Smoke Characterization Project Final Report P. 70 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The smoke release data for the non- flaming tests conducted in the IMO calorimeter are presented
in Figure 38 through Figure 40.


0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0 60 120 180 240 300 360 420 480 540 600 660 720
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1
Test 2
5
Figure 38 Smoke release rate for bread in non-flaming combustion


0.000
0.005
0.010
0.015
0.020
0.025
0.030
0.035
0.040
0.045
0.050
0 60 120 180 240 300 360 420 480 540 600
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1
Test 2

Figure 39 Smoke release rate for PU foam in non-flaming combustion 10
Smoke Characterization Project Final Report P. 71 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.000
0.005
0.010
0.015
0.020
0.025
0.030
0.035
0.040
0.045
0.050
0 60 120 180 240 300 360 420 480 540 600
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
Test 1

Figure 40 Smoke release for cotton/poly sheet wrapped PU foam in non-flaming combustion

It was observed that only a trace amount of smoke was observed for the PU foam wrapped in the
cotton/poly sheet. 5

Smoke Characterization Project Final Report P. 72 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The smoke particle size distribution data measured on the WPS spectrometer were analyzed to
calculate the mean particle diameter D
m
and count N
m
for each test as described by Eq. 6 and Eq.
7. Mean smoke particle diameter and count from the intermediate calorimeter tests are
summarized in Table 14.
Table 14 Intermediate calorimeter smoke particle data 5
Sample Calorimeter
Test
Series
Mode
D
m

(m)
N
m

(cc
-1
)
NEBS Test 1 Flaming 0.276 1.20E+06
IMO Test 1 Flaming 0.268 1.72E+05
3:1 Heptane/Toluene mixture
(UL 217)
IMO Test 2 Flaming 0.271 1.83E+05
NEBS Test 1 Flaming 0.066 6.94E+06
IMO Test 1 Flaming 0.072 1.35E+06 Douglas fir (UL 217)
IMO Test 2 Flaming 0.061 7.87E+05
NEBS Test 1 Flaming 0.086 6.22E+06
IMO Test 1 Flaming 0.073 2.98E+05 Newspaper (UL 217)
IMO Test 2 Flaming 0.115 7.56E+04
Heptane (lighter) NEBS Test 1 Flaming 0.233 1.03E+06
Pillow (TB 604 burner) NEBS Test 1 Flaming 0.221 1.83E+06
Mattress (TB 604 burner) NEBS Test 1 Flaming 0.126 6.40e+06
Cotton batting (TB 604 burner) NEBS Test 1 Flaming 0.053 1.90E+05
PU foam (TB 604 burner) NEBS Test 1 Flaming 0.038 1.95E+06
IMO Test 1 Flaming 0.054 1.73E+06 PU foam in cotton/poly sheet (TB
604 burner) IMO Test 2 Flaming 0.058 1.27E+06
NEBS Test 1 Flaming 0.183 1.92E+06
IMO Test 1 Flaming 0.101 2.76E+06 Coffee maker (TB 604 burner)
IMO Test 2 Flaming 0.097 5.99E+06
IMO Test 1 Flaming 0.123 1.27E+06 Nylon carpet (cone Heater at 35
kW/m
2
)
IMO Test 2 Flaming 0.176 7.87E+05
NEBS Test 1 Non-Flaming 0.110 1.53E+07
IMO Test 1 Non-Flaming 0.146 3.17E+06 Bread (Electric Toaster)
IMO Test 2 Non-Flaming 0.123 2.70E+06
2 Smoldering cigarettes NEBS Test 1 Non-Flaming 0.119 5.44E+05
Quarter mattress (3 smoldering
cigarettes)
NEBS Test 1 Non-Flaming 0.175 2.11E+05
Cotton batting (Hot plate) NEBS Test 1 Non-Flaming 0.106 3.98E+06
PU foam (Hot plate) NEBS Test 1 Non-Flaming 0.118 7.50E+06
IMO Test 1 Non-Flaming 0.081 7.69E+05 PU foam (Cone heater at 15
kW/m
2
) IMO Test 2 Non-Flaming 0.085 9.98E+05
PU foam with cotton/poly sheet
(Smoldering cigarette)
IMO Test 1 Non-Flaming 0.186 3.37E+05

The results show that while mean particle diameters are similar in the two calorimeter test series,
the particle density was observed to be generally lower in the IMO calorimeter. This is expected
to be due to differences in air entrained prior to smoke extraction for the two test set-ups.
10
Gas effluent data were obtained only for the IMO test series. The data for the maximum
concentration of carbon monoxide and carbon dioxide are presented in Table 15.
Smoke Characterization Project Final Report P. 73 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 15 Maximum observed carbon monoxide and carbon dioxide concentrations
Test Sample Test Series Mode
Max CO
(ppm)
Max CO
2

(ppm)
Test 1 Flaming 78 994
Douglas fir (UL 217)
Test 2 Flaming 69 317
Test 1 Flaming 13 121
Heptane + Toluene (UL 217)
Test 2 Flaming 55 1000
Test 1 Flaming 145 179
Newspaper (UL 217)
Test 2 Flaming 79 25
Test 1 Flaming 160 2552
Nylon carpet (Cone heater at 35 kW/m
2
)
Test 2 Flaming 170 2767
Test 1 Flaming 43 717 PU foam in cotton/poly sheet (TB 604
burner)
Test 2 Flaming 18 349
Test 1 Flaming 686 9610
Coffee maker (TB 604 burner)
Test 2 Flaming 612 10546
Test 1 Non-Flaming 203 162
Bread (Electric Toaster)
Test 2 Non-Flaming 50 27
Test 1 Non-Flaming 3 17
PU foam (Cone heater at 15 kW/m
2
)
Test 2 Non-Flaming 9 34
PU foam in cotton/poly sheet
(Smoldering cigarette)
Test 1 Non-Flaming 310 629

The charts depicting the heat and smoke release rates, smoke particle size and count data, and
gas effluent for each of the flaming and non- flaming tests are presented in Appendix E and F
respectively. 5

UL 217 Smoldering Ponderosa pine Test Results
The smoke particle data were analyzed to calculate the mean diameter and count for each scan.
The data are plotted in Figure 41. The increase in smoke particle size after approximately 2,700
seconds (45 minutes) may have occurred due to the lowering of the smoke layer below the 10
sampling point.

Smoke Characterization Project Final Report P. 74 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0 500 1000 1500 2000 2500 3000 3500 4000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
S
m
o
k
e

P
a
r
t
i
c
l
e

C
o
u
n
t

(
1
/
c
c
)

Figure 41 Smoke particle data from the UL 217 smoldering Ponderosa pine test

The count distribution of the three relative particle sizes is shown in Figure 42. It was observed
that after approximately 3,000 seconds (50 minutes) into the test, the number of particles in the 5
0.109 to 0.500 micron range increase rapidly. This increase may be related to the settling of the
smoke observed during the test and/or aggregation of smoke particles as observed in the UL 217
smoke box test. The mean smoke particle diameter for the time period prior to this change (up to
than 2,000 s) was 0.142 microns versus 0.204 microns for the entire test.
10
Smoke Characterization Project Final Report P. 75 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
0 500 1000 1500 2000 2500 3000 3500 4000
Time (s)
P
a
r
t
i
c
l
e

C
o
u
n
t

D
e
n
s
i
t
y

(
1
/
c
c
)
0.03 - 0.109 micron
0.109 - 0.5 micron
0.5 - 10 micron

Figure 42 UL 217 smoldering Ponderosa pine particle size distribution

Discussion of Intermediate Scale Test Results
The data were further analyzed to develop a comparison of the samples tested with the UL 217 5
materials with respect to their smoke characteristics.

Combustibility Results
Heat and smoke release data for the flaming tests are presented in Figure 43 and Figure 44. In
order to compare heat and smoke release measurements for the coffee maker test during the same 10
experiment time frames to the other tests, maximum plotted values for the coffee maker are
through the first six minutes.

It was observed that the nylon carpet and PU foam yield smaller peak heat release rates than the
Douglas fir, heptane/toluene mixture and the newspaper test samples. The peak heat release rate 15
from the coffee maker for the duration of the test was approximately 100 kW, which was
significantly higher than the other investigated scenarios.

Smoke Characterization Project Final Report P. 76 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
2
4
6
8
10
12
14
Douglas Fir Heptane + Toluene Newspaper Nylon carpet PU Foam Coffee Maker
P
e
a
k

H
R
R

(
k
W
)

Figure 43 Peak HRR for flaming combustion tests


5
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Douglas Fir Heptane + Toluene Newspaper Nylon carpet PU Foam Coffee Maker
P
e
a
k

S
R
R


(
m
2
/
s
)

Figure 44 Peak SRR for flaming combustion tests

Smoke Characterization Project Final Report P. 77 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Influence of Material Chemistry on Smoke Characteristics
The intermediate scale tests demonstrated the influence of material chemistry on smoke
characteristics. For example, the mean smoke particle diameters were larger when aromatic
hydrocarbon molecules (toluene) were mixed with the straight chain hydrocarbon molecules
(heptane). Natural materials such as wood, newspaper, cotton batting had relatively smaller 5
average particle diameter as compared to synthetic materials (coffee maker, nylon carpet). An
exception was the PU foam that had a smaller average particle diameter in the flaming mode.
This may be due to the unique chemistry and physical cell structure of polyurethane foam. These
results are similar to those obtained in the cone calorimeter tests.
10
The influence of material chemistry on the particle size distribution is depicted in Figure 45
(vertical axis are identically scaled for the four plots).

Newspaper Douglas Fir
PU Foam Nylon Carpet
Figure 45 Particle size distribution for flaming combustion of natural and synthetic materials
15
For the Douglas fir it was observed that there is significant reduction in the largest particle
(0.500 to 10 microns) due to charring (also observed in small-scale tests). The change in the
particle size distribution exhibited by newspaper using the UL 217 newspaper fire test protocol
can be explained by formation of more large particles prior to flame-through when smoldering
predominates and then smaller particles during the open flame portion of the test after flame- 20
through occurs. This phenomenon is also in agreement with the flaming and non- flaming results
observed in small-scale tests. Particle sizes are relatively stable for the PU foam and nylon carpet
samples.

The particle size distribution trends for non- flaming tests on Ponderosa pine, PU foam, and PU 25
foam wrapped in a cotton-poly sheet are shown in Figure 46.

1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
0 50 100 150 200 250 300
Time (s)
P
a
r
tic
le
C
o
u
n
t D
e
n
s
ity
(1
/c
c
)
0.03 - 0.109 micron
0.109 - 0.5 micron
0.5 - 10 micron
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
0 50 100 150 200 250 300 350 400
Time (s)
P
a
r
tic
le
C
o
u
n
t D
e
n
s
ity
(1
/c
c
)
0.03 - 0.109 micron
0.109 - 0.5 micron
0.5 - 10 micron
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
0 50 100 150 200 250 300 350 400 450 500
Time (s)
P
a
r
tic
le
C
o
u
n
t D
e
n
s
ity
(1
/c
c
)
0.03 - 0.109 micron
0.109 - 0.5 micron
0.5 - 10 micron
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
0 50 100 150 200 250 300 350 400 450
Time (s)
P
a
r
tic
le
C
o
u
n
t D
e
n
s
ity
(1
/c
c
)
0.03 - 0.109 micron
0.109 - 0.5 micron
0.5 - 10 micron
Smoke Characterization Project Final Report P. 78 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.

1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
0 100 200 300 400 500 600 700
Time (s)
P
a
r
tic
le
C
o
u
n
t D
e
n
s
ity
(1
/c
c
)
0.03 - 0.109 micron
0.109 - 0.5 micron
0.5 - 10 micron

PU Foam with radiant heating
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
1.E+08
0 100 200 300 400 500 600 700
Time (s)
P
a
r
tic
le
C
o
u
n
t D
e
n
s
ity
(1
/c
c
)
0.03 - 0.109 micron
0.109 - 0.5 micron
0.5 - 10 micron

PU Foam wrapped cotton/poly fabric
0.00E+00
2.00E+04
4.00E+04
6.00E+04
8.00E+04
1.00E+05
1.20E+05
1.40E+05
1.60E+05
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Time (s)
P
a
r
t
i
c
l
e
d
e
n
s
i
t
y
(
1
/
c
c
)
0.03 - 0.109 micron
0.109 -0.5 micron
0.5 - 10 micron

Ponderosa Pine

Figure 46 Particle size distribution for non-flaming combustion of natural and synthetic materials
The distribution of small and large particles for the PU foam is relatively constant throughout the
test. In contrast the PU foam wrapped with the cotton-poly sheet has a relatively higher count of
the particles in the 0.109 to 0.500 micron range and a lower count of the smaller particles. For 5
Ponderosa pine, there are very few particles in the range 0.500 to 10 microns as compared to
either of the two PU foam tests.

Comparison of Particle Size and Count
The average particle sizes (D
m
) for the test were calculated for each test sample using data from 10
both the NEBS and IMO calorimeter.

A bar chart is presented in Figure 47 displaying the comparison between the evaluated samples.

Smoke Characterization Project Final Report P. 79 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.000
0.050
0.100
0.150
0.200
0.250
0.300
Heptane +
Toluene
Doug Fire
Crib
Newspaper Heptane Pillow PU Foam Mattress Nylon carpet Coffee Maker
Test Sample
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
s
)
UL 217 Materials

Figure 47 Average smoke particle diameters for flaming combustion tests
The average particle densities from the flaming tests performed in the IMO calorimeter are
presented in Figure 48. The three nonUL 217 materials generated larger particle densities of
smoke. 5

0.00E+00
5.00E+05
1.00E+06
1.50E+06
2.00E+06
2.50E+06
3.00E+06
3.50E+06
4.00E+06
4.50E+06
5.00E+06
Douglas Fir Crib Heptane + Toluene Newspaper Nylon carpet PU foam Coffee maker
Test Sample
P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
UL 217 Materials


Figure 48 Average smoke particle density for flaming combustion tests
10
Smoke Characterization Project Final Report P. 80 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The data shows that for flaming mode, the average particle sizes from UL 217 materials are in
the same range as particle sizes observed for several products typically found in residential
occupancy areas.

The mean particle size for non- flaming tests are presented in Figure 49. 5
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
Ponderosa
Pine (Hot plate)
Mattress (3
Cigarettes)
PU Foam (Hot
plate)
PU Foam
(Radiant
heating)
PU foam with
Poly-cotton
sheet
(Cigarette)
Cotton Batting
(Hot plate)
Bread (Toaster) Cigarettes
Test Sample
M
e
a
n

d
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)

Figure 49 Mean smoke particle diameter for non-flaming tests

The average smoke particle diameter was highest for PU foam covered with poly-cotton blend
sheet, and was almost 72 % higher than the average particle size generated by Ponderosa pine. 10
Average particle diameters from other materials were in the same range as Ponderosa pine. It
may also be observed that the particle count from the PU foam covered with poly-cotton sheet
was significantly lower than other materials. This is anticipated to be due to cover sheet
obstructing the smoke flow away from the underlying polyurethane foam.
15
In these tests involving smoldering cigarette as a heat source, there was not a sustained
involvement of the target material once the cigarette extinguished or the target material around
the cigarette hot tip had gasified. Thus, this heat source scenario was not pursued.

The average particle densities for non- flaming tests are presented in Figure 50. 20

Smoke Characterization Project Final Report P. 81 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.E+00
1.E+06
2.E+06
3.E+06
4.E+06
5.E+06
6.E+06
7.E+06
8.E+06
Ponderosa Pine (Hot
plate)
PU Foam (Radiant
Heating)
PU Foam (Hot plate) PU foam with Poly-
cotton sheet
(Cigarette)
Cotton Batting (Hot
plate)
Bread (Toaster)
Test Sample
P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)

Figure 50 Average particle count for non-flaming combustion tests

A significant difference in the PU foam particle density was observed with the two heating
methods (radiant versus hot plate). Furthermore, wrapping the PU foam with poly-cotton fabric 5
decreased the particle density count. It was also observed that bread in a toaster generated
significant particle density of smoke.

Smoke Characterization Project Final Report P. 82 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
TASK 3 DEVELOP SMOKE PROFILES AND PARTICLE SIZE AND
COUNT DISTRIBUTIONS IN THE UL 217/UL 268 FIRE TEST ROOM

INTRODUCTION
Activation response of smoke alarms to different smoke scenarios is evaluated in UL 217 5
through a series of four flaming and non- flaming fire tests:
1. Paper Fire (Section 44 Fire Tests Test A)
2. Wood Fire (Section 44 Fire Tests Test B)
3. Flammable Liquid Fire (Section 44 Fire Tests Test C)
4. Wood Non-flaming Fire (Section 45 Smoldering Smoke Test) 10
The first three fire tests are open flame tests in which the alarm unit must activate within a
specified maximum time limit of 240 seconds; while the fourth test is a non- flaming fire test in
which the unit must activate within a specified obscuration range (0.5 to 10.0 percent per foot).

In this task the atmosphere in the vicinity of the alarm units during the course of the UL 217 fire 15
and non- flaming smoke tests was characterized for MIC and obscuration signals, smoke particle
size and distribution, effluent gas composition, ceiling air flow velocity, and ceiling temperature.
Atmospheres generated by flaming and non-flaming combustion of other materials were also
evaluated at the same prescribed 5.4 m sampling distance.
20

TASK OBJECTIVES
The objectives of this task were to characterize the following for UL 217 Section 44 fire test
samples and the additional test samples and fire scenarios developed in Task 2:
(i) smoke particle size and count distribution 25
(ii) gas effluent composition
(iii) analog addressable smoke alarm signals
(iv) standard light obscuration beam and MIC signals
(v) standard photoelectric and ionization alarm signals
(vi) ceiling air velocity 30
(vii) ceiling air temperature

Smoke Characterization Project Final Report P. 83 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
TEST SAMPLES
In addition to the standard UL 217 test samples, other samples were selected from Task 2 that
had unique combustibility or smoke characteristics as presented in Table 16.

Table 16 Test samples for UL 217 Fire Test Room Test tests 5
Test Sample Comments
Flaming Tests
Heptane/Toluene mixture Standard UL 217 sample
Douglas fir Standard UL 217 sample
Shredded newspaper Standard UL 217 sample
Coffee maker
Higher energy fire. Relatively more and larger particles in
intermediate scale tests
Mattress PU foam insulation
Common in residential settings. Relatively more and smaller
particles in small and intermediate scale tests
Mattress PU foam with CA TB 117
50:50 cotton/poly sheet
Common in residential settings. Relatively more and larger
particles than Ponderosa pine in intermediate scale test
Nylon carpet
Common in residential settings. Relatively more particles in
0.109-0.500 micron range in small and intermediate scale
tests
Non-Flaming Tests
Ponderosa pine Standard UL 217 sample
Mattress PU foam with CA TB 117
cotton sheet
Larger average particle diameter than Ponderosa pine in
intermediate scale test
Mattress PU foam with polyester
microfiber sheet
A more common current fabric in furnishings. Not tested in the
small-scale and intermediate scale tests.
Polyisocynanurate foam Relatively more and smaller particles in small-scale tests
Nylon carpet Relatively more and smaller particles in small-scale tests
Polystyrene pellets
Anticipate more, dark colored smoke than for UL 217
Ponderosa pine
Bread
Common nuisance alarm. Relatively larger particles and count
in intermediate scale tests


EXPERIMENTAL
All combustion tests were conducted in Underwriters Laboratories Fire Test Room. Tests were
conducted at the respective UL 217 prescribed height of 0.91 m (for flaming tests) and 0.2 m (for 10
non- flaming tests) above the floor. Test samples were preconditioned in accordance with UL 217
at a temperature of 23 t2 C (73.4 t3 F) and a relative humidity of 50 t5 % for at least 48 hours
prior to testing. The evaluated test materials and ignition scenarios are listed in Table 17.
Smoke Characterization Project Final Report P. 84 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 17 Fire Test Room Tests
Mode Target Sample Description Heat/Ignition Source Test No.
UL 217 Heptane/Toluene mixture (3:1) UL 217 prescribed ignition
12112, 12131,
12181, 12182,
01221
UL 217 Douglas fir UL 217 prescribed ignition
12123, 12124,
12127, 12146,
12183
UL 217 Shredded newspaper UL 217 prescribed ignition
12113, 12122,
12125, 12141,
12144, 12145
Coffee maker 12 cup, no carafe
CA TB 604 burner flame (50 mm
height) applied under filter
holder for 35 s
12134, 12186
Mattress PU foam 100 100 100 mm (w l
h) sample
ASTM E1354 cone heater at 35
kW/m
2

12154
Mattress PU foam wrapped in CA TB 117 50:50
cotton/poly sheet 100 100 100 mm foam
CA TB 604 burner flame (35 mm
height) applied to base for 20 s
12135
Mattress PU foam wrapped in CA TB 117 50:50
cotton/poly sheet 150 150 150 mm foam
CA TB 604 burner flame (35 mm
height) applied to base for 20 s
12142, 12156,
12191
F
L
A
M
I
N
G

Nylon carpet 100 100 mm sample
ASTM E1354 cone heater at 35
kW/m
2

12151, 12152,
12153
UL 217 Ponderosa pine
UL 217 prescribed hot plate and
temperature profile
12126, 12132,
12143, 12184,
12185
Bread 4 slices
Commercial toaster 3 cycles on
dark setting
12133, 12155,
01244
Polyisocyanurate insulation 150 150 200
mm pieces
UL 217 Ponderosa pine hot plate
and temperature profile
12271
Mattress PU foam 150 150 50 mm foam
UL 217 Ponderosa pine hot plate
and temperature profile
12192, 12193
Mattress PU foam 100 125 100 mm foam
with a 25 150 150 mm piece on two
opposing sides
UL 217 Ponderosa pine hot plate
and temperature profile
12202, 12261
Mattress PU foam wrapped in CA TB 117 cotton
sheet 100 150 200 mm foam
UL 217 Ponderosa pine hot plate
and temperature profile
01232
Mattress PU foam wrapped in CA TB 117 cotton
sheet 125 125 300 mm foam
UL 217 Ponderosa pine hot plate
and temperature profile
01241
Mattress PU foam wrapped in polyester
microfiber sheet 125 125 300 mm foam
UL 217 Ponderosa pine hot plate
and temperature profile
01233, 01245
Nylon carpet 150 150 mm sample
UL 217 Ponderosa pine hot plate
and temperature profile
12262
N
O
N
-
F
L
A
M
I
N
G

Polystyrene pellets 69.8 g
UL 217 Ponderosa pine hot plate
and temperature profile
12272

Test Facility - The Fire Test Room consists of 11.0 6.7 3.1 m (lwh) room with a smooth
ceiling with no physical obstructions. The test room is constructed to maintain a temperature of
23 t3 C and a humidity of 50 t10 % while ensuring minimal air movement during the test. The 5
room is provided with exhaust system to clear the room of smoke after each test.

Smoke Characterization Project Final Report P. 85 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Measurements and Instrumentation - The test room was equipped with the following devices
for evaluation of air quality:
Measuring Ionization Chamber (MIC) ceiling and two side walls equidistant from the test
target
Obscuration ceiling and two side walls equidistant from the test target 5
Analog addressable smoke alarms one ionization and one photoelectric unit on the ceiling
and wall
Smoke alarms one ionization and one photoelectric unit on the ceiling
Air flow velocity ceiling
Temperature ceiling 10
Sampling port for smoke particle characterization ceiling between commercial alarms
Sampling port for room gas composition characterization ceiling between commercial
alarms
Light obscuration tree located in the vicinity of the MIC. Added for the last series of tests.
15
Measuring Ionization Chamber (MIC) - An Electronikcentralen Type EC 23095 MIC

was
used to measure the relative buildup of particles of combustion during the test. The MIC utilizes
the ionization principle with air drawn through the chamber at a rate of 30 3 Lpm by a regulated
vacuum pump. The ceiling mounted monitoring head was located 6 m from the fire source and
0.1 m below the ceiling, along the centerline of the test room; side-wall mounted monitoring 20
heads were located 0.4 m below the ceiling, 6 m from the fire source and 0.1 m from the
respective wall. The MIC was not utilized during flaming mode tests.

Obscuration - A white light obscuration system consisting of a lamp and photocell assembly
spaced 1.52 m apart was used to measure the relative buildup of particles of combustion during 25
the test. The ceiling mounted obscuration system was located 5.4 m from the fire source along
the centerline of the room and 0.1 m below the ceiling; the side wall mounted systems were
located 0.4 m below the ceiling, 5.4 m from the fire source and 0.18 m from the respective wall.
Complete descriptions of the lamp and photocell assemblies are available in the UL 217.
30
Analog Addressable Smoke Alarms Commercially available residential ionization and
photoelectric type smoke alarm units were mounted on the ceiling and walls 5.4 m from the fire
source. The alarms were equipped to provide an analog output (electrical measurement) of the
alarm sensitivity during the course of the test trials.
35
Smoke Alarms - Residential ionization and photoelectric type smoke alarms were mounted on
the ceiling 5.4 m from the fire source. The automated data acquisition equipment recorded the
alarm trigger time.

Smoke Particle Characterization - Smoke for particle characterization was sampled along the 40
centerline of the room 5.4 m from the fire source and 0.01 m below the ceiling. Smoke particle
size and count were characterized using WPS Spectrometer previously described in the Smoke
Characterization section of Task 2. The sample line to the spectrometer was 10.5 m long with a
3.2 mm I.D.
45

Smoke Characterization Project Final Report P. 86 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.

Effluent Gas Composition Characterization - Gas effluent for composition characterization
was sampled along the centerline of the room 5.4 m from the fire source and 0.01 m below the
ceiling. Gas effluent composition was characterized using the MIDAC #I 1100 Fourier
Transform Infrared (FTIR) Spectrometer and deconvoluted as previously described in the Smoke 5
Characterization section of Task 2 (Eq. 8 through Eq. 11). The sample line to the spectrometer
was 8.5 m long with a 3.2 mm I.D. The utilized sample line was not heated because water vapor
condensation was not expected within the sample line as the ceiling temperatures were not
anticipated to be significantly higher than ambient conditions.
10
Air Velocity - Two-component air velocities was measured 5.4 m from the fire source and 0.1 m
below the ceiling using a CATI sonic anemometer (Applied Technologies Inc.) supplied by
NIST. The anemometer was arranged such that the two measured air velocity components are in
the radial direction away from the combustion source and in the transverse direction.
15
This device uses piezoelectric crystals to form ultrasonic transducers that can send and receive
ultrasonic pulses. The forward and backward travel time of these pulses are used to compute the
component velocity between two opposing transducers. The anemometer records the mean
velocity over a 150 mm sonic path length (which equals the distance separating opposing
transducers) at a frequency of up to 10 Hz. The measurement resolution is 10 mm/s with a stated 20
uncertainty of 10 mm/s.

Temperature - Air temperature was measured on the airflow velocity support structure 5.4 m
from the fire source and 0.15 m below the ceiling using a 0.0625 mm diameter Inconel sheathed
Type K thermocouple. 25

Light Obscuration Tree - The light obscuration tree was used in the final smoldering fire tests
to determine the obscuration in the room at three different heights during these tests. Each of the
light obscuration instruments consisted of a 12 volt DC, 20 watt, Halogen lamp (Model MR 16)
and a Huygen photocell (Weston Model 856-9901033-BB). The lamp and photocell were spaced 30
300 mm apart. The three light and photocell assemblies were mounted on an adjustable pole such
that they were located 600, 900, and 1500 mm below the ceiling.

Smoke Color - The filter paper used with the gas FTIR instrument were observed after each test
for the color of the smoke deposited during the test. 35
Smoke Characterization Project Final Report P. 87 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
A schematic of the test room with the sampling instrumentation is shown in Figure 51.

3.4 m
2.13m
11m
6.7m
Fire Source
0.91m above
the floor
2
5.4m
2
Photolamp
Photocell
PLAN
8
3
4
5
6
7
1 - MIC (Measuring Ionization Chamber)
2 - Photocell Assembly (5ft from lamp to photocell. Centerline 4in below
ceiling)
3 - Photoelectric Smoke Detector
4 - Ionization Smoke Detector
5 - Analog Addressable Ionization Smoke Alarm
6 - Analog Addressable Photoelectric Smoke Alarm
7 - Smoke Particle Size and Gas FTIR Sampling port (3-3/8in below
ceiling)
8 - Sonic anemometer, Thermocouple
9 - Obscuration Tree
1
9
2.3m
3.35m
3.57m
2.3m
5.78m

Figure 51 Fire Test Room. Drawing not to scale.
5
Smoke Characterization Project Final Report P. 88 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
TEST PROCEDURE
The flaming tests for UL 217 test samples were conducted using the procedures described in the
UL 217. For samples ignited with TB 604 ignition source, the test samples were ignited as
described in Table 17. For samples heated with the ASTM E1354 conical heater, the samples
were ignited with the aid of an electric spark. The data acquisition systems for all the instruments 5
were manually initiated upon ignition of the sample. The sampling intervals for the data
acquisition systems used are provided in Table 18.

Table 18 Data acquisition sampling intervals
Data Acquisition
Sampling
Interval (s)
Test room Beam, MIC, and smoke alarm triggers 1
Analog smoke alarms 8
Gas effluent FTIR 15
WPS spectrometer 67
[1]

Note to Table 18: 10
[1]
The first data was sampled at 48 s, followed by 67 s intervals between
subsequent measurements

For non- flaming tests, the temperature controlled hot plate described in UL 217 was used for all
the samples except for bread, where a four slice electric toaster was used. 15

TEST RESULTS
The results from these tests included:
Obscuration over the test duration
Smoke alarm trigger time 20
Smoke particle size and count distribution data
MIC and Beam signals
Gas effluent component data
Ceiling air velocity and temperature
Smoke color 25

Individual results for flaming and non- flaming combustion tests are plotted in Appendix G and H
respectively. Post-test photographs of the FTIR particulate filters for smoke particulate color
comparison are presented in Appendix I.
30
Flaming Test Results
In Table 19, is presented the obscuration measured in the room. The obscuration (OBS) was
calculated from the ceiling light beam signal data as follows:
1
1
1
]
1

,
_

d
1
Tc
Ts
1 100 OBS [=] %/ft
Eq. 23
Where Ts is the light beam signal during the test
Tc is the clear light beam signal 35
Smoke Characterization Project Final Report P. 89 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
d is path length = 5 ft
The table shows the obscuration calculated at (i) UL 217 specified time for the alarm to operate
(e.g., 240 seconds for the Douglas fir); (ii) maximum obscuration; and (iii) the time to attain
maximum obscuration.
5
Table 19 Summary of obscuration for flaming tests
UL 217 Time Max. OBS
Target Sample Description Test No.
Flame
Through
Time (s)
Time (s)
OBS
(%/ft)
Time (s) (%/ft)
12112 -- 240 13.0 143 14.6
12131 -- 240 11.9 138 12.8
12181 -- 240 11.9 153 13.2
12182 -- 240 12.9 133 13.9
UL 217 Heptane/Toluene
mixture
01221 -- 240 13.5 135 14.9
12123 189.7 240 5.0 217 20.2
12124
[1]
142.4 240 2.3 161 14.1
12127
[1]
127.6 240 1.3 189 13.2
12146 166.3 240 5.0 150 13.1
UL 217 Douglas fir
12183
[1]
102.6 240 0.6 125 9.4
12113
[1]
36.1 240 1.4 56 14.8
12122 100.3 240 6.5 125 33.3
12125 141.0 240 20.1 165 28.4
12141 60.2 240 3.4 91 21.7
12144 118.4 240 9.9 144 29.0
UL 217 Shredded newspaper
12145 83.1 240 2.8 110 23.7
12134 -- 240 0.8 605 47.4 Coffee maker 12 cup, no
carafe 12186 -- 240 0.7 510 44.2
Mattress PU foam 100 100
mm sample
12154 -- 240 [2] 64 5.5
Mattress PU foam wrapped in
CA TB 117 50:50 cotton/
poly sheet 100 100 100
mm foam
12135 -- 240 0.4 600 0.6
12142 -- 240 3.9 234 3.9
Mattress PU foam wrapped in
CA TB 117 50:50 cotton/
poly sheet 150 150 150
mm foam
12156 -- 240 3.0 167 4.7
12151 -- 240 5.1 279 6.1
12152 -- 240 4.8 343 6.2
Nylon carpet 100 100 mm
sample
12153 -- 240 4.0 323 6.8
Notes to Table 19:
[1]
Flame through time is shorter than allowed in UL 217.
[2]
Test duration was less than 240 s.
10
The OBS data for the flaming tests are shown in Figure 52 through Figure 59. There was more
variation in the newspaper tests than the others. It is believed that this was due to the influence of
the packing of the shredded material.

Smoke Characterization Project Final Report P. 90 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Repeat tests were not performed for the 44- in sample of PU foam wrapped in poly-cotton fabric
as this sample target arrangement resulted in a very low level of obscuration in the room. Testing
was repeated for this sample arrangement using a larger PU foam sample (66- in.). Also, repeat
tests for the PU foam exposed to radiant heating were not conducted as this test resulted in a
short duration fire of less than 240 s. In this test, there was rapid burn time resulting in a 5
relatively sharp smoke obscuration peak similar to that observed for the newspaper tests. It was
observed that most of the smoke remained on the ceiling. Good visibility was present throughout
the rest of the room.

It was observed that there is a good repeatability between tests, except for the shredded 10
newspaper tests. There was substantial variation observed in the shredded newspaper test with
respect to the progression of the flame out of the test specimen holder. This also resulted in
relatively larger variation in maximum OBS values.
0
5
10
15
20
25
30
35
40
45
50
0 60 120 180 240 300 360
Time (s)
O
B
S

(
%
/
f
t
)
12112
12131
12181
12182
01221

Figure 52 Smoke OBS for heptane/toluene mixture in flaming combustion 15
Smoke Characterization Project Final Report P. 91 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
5
10
15
20
25
30
35
40
45
50
0 60 120 180 240 300 360
Time (s)
O
B
S

(
%
/
f
t
)
12113
12122
12125
12141
12144
12145

Figure 53 Smoke OBS for newspaper in flaming combustion


5
0
5
10
15
20
25
30
35
40
45
50
0 60 120 180 240 300 360
Time (s)
O
B
S

(
%
/
f
t
)
12123
12124
12127
12146
12183

Figure 54 Smoke OBS for Douglas fir in flaming combustion
Smoke Characterization Project Final Report P. 92 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
5
10
15
20
25
30
35
40
45
50
0 60 120 180 240 300 360 420 480 540 600 660
Time (s)
O
B
S

(
%
/
f
t
)
12134
12186

Figure 55 Smoke OBS for coffee maker in flaming combustion


5
0
2
4
6
8
10
12
14
0 30 60 90 120 150 180 210 240
Time (s)
O
B
S

(
%
/
f
t
)
12154

Figure 56 Smoke OBS for PU foam in flaming combustion (35 kW/m
2
radiant heating)


Smoke Characterization Project Final Report P. 93 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 60 120 180 240 300 360 420 480 540 600
Time (s)
O
B
S

(
%
/
f
t
)
12135

Figure 57 Smoke OBS for PU foam (100100 mm) with cotton-poly sheet in flaming combustion


5

0
5
10
15
20
25
30
35
40
45
50
0 100 200 300 400 500 600 700
Time (s)
O
B
S

(
%
/
f
t
)
PU foam 6x6x6 flaming #2-
20061214.xls
PU foam flaming 6x6x6 #6-
20061215.xls

Figure 58 Smoke OBS for PU foam (150150 mm) with cotton-poly sheet in flaming combustion

Smoke Characterization Project Final Report P. 94 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
2
4
6
8
10
12
14
0 60 120 180 240 300 360 420 480 540 600
Time (s)
O
B
S

(
%
/
f
t
)
12151
12152
12152

Figure 59 Smoke OBS for nylon carpet in flaming combustion

Smoke Characterization Project Final Report P. 95 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The alarm trigger times for the flaming tests are presented in Table 20. The MIC was not used
for tests on the prescribed UL 217 materials.

Table 20 Flaming mode alarm response times
Analog Signal
Value
Analog Signal
Value
Target Sample Description Test No.
Ion
Alarm
Trigger
Time (s)
MIC
(pA)
Photo
(mV)
Photo
Alarm
Trigger
Time (s)
MIC
(pA)
Photo
(mV)
12123 NAP -- -- NAP -- --
12124 NAP -- -- NAP -- --
12127 164 -- 84.3 157 -- 72.1
12146 145 -- 60.5 185 -- 54.7
UL 217 Douglas fir
12183 117 -- 69.2 173 -- 88.9
12113 NAP -- -- NAP -- --
12122 NAP -- -- NAP -- --
12125 176 -- 57.1 179 -- 87.8
12141 87 -- 36.5 134 -- 80.4
12144 143 -- 21.4 160 -- 94.7
UL 217 Shredded newspaper
12145 126 -- 85.6 126 -- 85.6
12112 NAP -- -- NAP -- --
12131 -- -- -- 66 -- 69.0
12181 36 -- 89.5 70 -- 68.0
12182 34 -- 89.0 71 -- 65.8
UL 217 3:1
Heptane/Toluene mixture
01221 34 -- 88.4 72 -- 68.2
12134 210 61.5 96.0 438 36.1 85.4 Coffee maker 12 cup, no
carafe
12186 151 69.8 95.2 334 33.2 84.0
Mattress PU foam 100
100 mm sample
12154 68 84.8 77.6 NA -- --
Mattress PU foam wrapped
in CA TB 117 50:50
cotton/poly sheet 100
100 100 mm foam
12135
[1]
DNT -- -- DNT -- --
12142
[2]
112 72.9 93.0 DNT -- --
Mattress PU foam wrapped
in CA TB 117 50:50
cotton/poly sheet 150
150 150 mm foam
12156
[3]
96 74.2 94.1 171 35.6 79.7
12151 173 67.7 92.0 221 40.7 76.8
12152 162 72.3 90.8 DNT -- --
Nylon carpet 100 100
mm sample
12153 137 79.0 90.0 323 37.7 70.2
Notes to Table 20: 5
NAP = Alarm not present
NA = Alarm data not recorded
DNT = Smoke alarm did not trigger
[1]
Maximum measured OBS value was 0.59 %/ft
[2]
Maximum observed OBS value was 3.9 %/ft; 10
[3]
Maximum observed OBS value was 4.7 %/ft

Smoke Characterization Project Final Report P. 96 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
It was observed that for flaming fires, the ionization smoke alarm typically triggered prior to the
photoelectric smoke alarm. The difference in ionization and photoelectric smoke alarm trigger
times was the highest for the coffee maker where the ionization smoke alarm on average
triggered almost 2-1/2 minutes faster than the photoelectric one. It may be noted that the coffee
maker had the highest heat release rate in the intermediate scale test of the selected test samples. 5
During the first test for the PU foam (66- in.) the photoelectric smoke alarm did not trigger
while in the second one, it did trigger. This may be attributed to the higher smoke obscuration
created in the second test. The reason for the photoelectric alarm not to trigger for the second
nylon carpet test is not clear, as the OBS values for all the three tests were in the range of 6.1 to
6.8 %. Visual inspection of soot deposits on the filter paper for the PU foam and nylon carpet 10
revealed dark gray to black in color.

The analog smoke alarm signals for these tests were examined to determine the difference in the
ionization and photo alarm signals. Flaming PU foam test results are presented in Figure 60. It
was observed that the photo signal for the first test is smaller than the second one, though both of 15
these signals are relatively weak as compared to the ionization signals. This may be related to
low smoke obscuration in the room for these tests.


0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50 60 70 80 90
Time (s)
D
e
t
e
c
t
o
r

S
i
g
n
a
l
PU foam (12142)-Ion
PU foam (12142)-Photo
PU foam (12156)-Ion
PU foam (12156)-Photo
20
Figure 60 Photo and ionization alarm analog signals for flaming PU foam tests

The analog smoke alarm signals for the nylon carpet were also examined as shown in Figure 61.
The photoelectric signals for both these tests (12151, 12152) are relatively low as compared to
the ionization smoke alarm signals. 25
Smoke Characterization Project Final Report P. 97 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50 60 70 80
Time (s)
D
e
t
e
c
t
o
r

S
i
g
n
a
l
Nylon Carpet (12151)-Ion
Nylon Carpet (12151)-Photo
Nylon Carpet (12152)-Ion
Nylon Carpet (12152)-Photo

Figure 61 Photo and ionization alarm analog signals for flaming nylon carpet tests

These signals may be compared with results from the Douglas fir test (12123) as depicted in
Figure 62 where both the ionization and photoelectric reach saturation level between 3 and 4 5
minutes.

0
10
20
30
40
50
60
70
80
90
100
0 50 100 150 200 250 300 350 400 450
Time (s)
D
e
t
e
c
t
o
r

S
i
g
n
a
l
Doug Fir Brand-Ion
Doug Fir Brand-Photo

Figure 62 Photo and ionization alarm analog signals for flaming Douglas fir test
10
Smoke Characterization Project Final Report P. 98 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The role of smoke particle size in these tests was investigated. Because the optical density per
path length was shown to be correlated to
3
i
i
d n (see Eq. 3), this factor was compared for the
some of the flaming tests including those that did not activate the photoelectric alarm. The UL
217 Douglas fir flaming test and the 3:1 heptane/toluene mixture test were also included for
comparative purposes. The data are presented in Figure 63. 5

0.0E+00
5.0E+03
1.0E+04
1.5E+04
2.0E+04
2.5E+04
0 100 200 300 400 500 600 700
Time (s)

n
i
*
d
i 3
Heptane/Toluene
Douglas fir
Nylon
PU Foam with Cotton-Poly sheet

Figure 63 Comparison of smoke particle size data for selected flaming test

It was observed that this factor is significantly higher for heptane/toluene mixture and Douglas 10
fir than the other tests in which the photoelectric alarm did not trigger.

Smoke mean diameters and number counts at OBS values of 0.5 and 10 %/ft are summarized in
Table 21. The results show that the mean particle sizes increase with time. The increase in
particle count is anticipated, as there is more accumulated smoke particles in the room as the 15
smoke obscuration increases. The increase in the mean diameter during the test is smallest for the
newspaper test. This may be due to fast moving nature of this particular fire test (note the shorter
time difference between 0.5 and 10 %/ft OBS).
Smoke Characterization Project Final Report P. 99 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 21 Smoke particle data at 0.5 %/ft and 10 %/ft OBS: flaming tests
0.5 %/ft OBS 10.0 %/ft OBS
Target Sample Description
Test
No.
Time
(s)
d
m

(m)
n
m

(cc
-1
)
Time
(s)
d
m

(m)
n
m

(cc
-1
)
12123 135 0.14 3.17E+05 150 0.19 5.15E+05
12124 125 0.11 3.93E+05 151 0.17 1.12E+06
12127 117 0.08 1.16E+05 143 0.14 6.00E+05
12146 126 0.09 4.27E+05 146 0.16 9.85E+05
UL 217 Douglas fir
12183 102 0.23 5.06E+03 NA NA NA
12113 50 0.06 2.37E+04 53 0.06 5.55E+04
12122 121 0.23 2.60E+05 122 0.22 2.85E+05
12125 104 0.33 7.57E+03 116 0.35 6.71E+04
12141 82 0.19 9.87E+04 85 0.20 1.07E+05
12144 104 0.05 6.28E+03 125 0.09 4.12E+04
UL 217 Shredded newspaper
12145 108 0.15 6.33E+03 109 0.15 6.33E+03
12112 29 0.21 7.01E+03 75 0.32 1.59E+05
12131 25 0.19 3.94E+04 112 0.30 4.34E+05
12181 30 0.21 5.36E+03 112 0.30 4.94E+05
12182 29 0.22 1.70E+04 97 0.31 5.58E+05
UL 217 3:1 Heptane/Toluene
mixture
01221 28 0.19 5.62E+03 96 0.27 2.25E+05
12134 154 0.11 4.53E+05 506 0.17 7.83E+05 Coffee maker 12 cup, no
carafe 12186 122 0.23 1.92E+05 437 0.18 1.06E+06
Mattress PU foam 100 100
mm sample
12154 55 0.08 4.52E+04 NA NA NA
Mattress PU foam wrapped in
CA TB 117 50:50 cotton/
poly sheet 100 100 100
mm foam
12135 327 0.08 8.68E+05 NA NA NA
12142 93 0.09 3.60E+05 NA NA NA
Mattress PU foam wrapped in
CA TB 117 50:50 cotton/
poly sheet 150 150 150
mm foam
12156 84 0.09 2.80E+05 NA NA NA
12151 120 0.10 3.01E+05 NA NA NA
12152 110 0.10 2.73E+05 NA NA NA
Nylon carpet 100 100 mm
sample
12153 122 0.11 2.80E+05 NA NA NA
Note to Table 21:
NA = Did not attain 10 %/ft OBS

The particle size and count data trends for the flaming tests are shown in Figure 64 through 5
Figure 71.


Smoke Characterization Project Final Report P. 100 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0. 1
0.15
0. 2
0.25
0. 3
0.35
0. 4
0.45
0. 5
0 100 200 300 400 500 600
Time (s)
M
e
a
n
D
ia
m
e
t
e
r
(
m
ic
r
o
n
)
12123
12124
12127
12146
12183

0.00E+00
5.00E+05
1.00E+06
1.50E+06
2.00E+06
2.50E+06
3.00E+06
3.50E+06
4.00E+06
4.50E+06
5.00E+06
0 100 200 300 400 500 600 700
Time (s)
S
m
o
k
e
P
a
r
tic
le
D
e
n
s
ity
(1
/c
c
)
12123
12124
12127
12146
12183

Figure 64 Mean smoke particle diameter and count for flaming Douglas fir tests

0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 50 100 150 200 250 300 350 400
Time (s)
M
e
a
n
P
a
rtic
le
D
ia
m
e
te
r (m
ic
ro
n
)
12113
12122
12125
12144
12145

0.00E+00
1.00E+06
2.00E+06
3.00E+06
4.00E+06
5.00E+06
6.00E+06
7.00E+06
8.00E+06
9.00E+06
1.00E+07
0 50 100 150 200 250 300 350 400
Time (s)
S
m
o
k
e
P
a
rtic
le
D
e
n
s
ity
(1
/c
c
)
12113
12145
12144
12125
12122

Figure 65 Mean smoke particle diameter and count for flaming newspaper tests

0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 50 100 150 200 250 300 350 400 450
Time (s)
M
e
a
n
P
a
rtic
le
D
ia
m
e
te
r (m
ic
ro
n
)
12112
12131
12181
12182
01221

0.00E+00
1.00E+06
2.00E+06
3.00E+06
4.00E+06
5.00E+06
6.00E+06
7.00E+06
8.00E+06
9.00E+06
1.00E+07
0 50 100 150 200 250 300 350 400 450
Time (s)
S
m
o
k
e
P
a
r
tic
le
D
e
n
s
ity
(1
/c
c
)
12112
12131
12181
12182
01221

Figure 66 Mean smoke particle diameter and count for flaming heptane/toluene tests 5

0
0.05
0. 1
0.15
0. 2
0.25
0. 3
0.35
0. 4
0.45
0. 5
0 100 200 300 400 500 600
Time (s)
M
e
a
n
P
a
r
t
ic
le
D
ia
m
e
t
e
r
(
m
ic
r
o
n
)
12134
12186

0.00E+00
1.00E+06
2.00E+06
3.00E+06
4.00E+06
5.00E+06
6.00E+06
7.00E+06
8.00E+06
9.00E+06
1.00E+07
0 100 200 300 400 500 600
Time (s)
M
e
a
n
P
a
r
t
ic
le
D
ia
m
e
t
e
r
(
m
ic
r
o
n
)
12134
12186

Figure 67 Mean smoke particle diameter and count for flaming coffee maker tests

Smoke Characterization Project Final Report P. 101 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 100 200 300 400 500 600
Time (s)
M
e
a
n
D
ia
m
e
t
e
r
(
m
ic
r
o
n
)
12154

1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
0 100 200 300 400 500 600
Time (s)
P
a
r
tic
le
C
o
u
n
t D
e
n
s
ity
(1
/c
c
)
12154

Figure 68 Mean smoke particle diameter and count for flaming PU foam (100100 mm) tests
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
M
e
a
n
D
ia
m
e
te
r (m
ic
ro
n
)
12135

0.00E+00
1.00E+06
2.00E+06
3.00E+06
4.00E+06
5.00E+06
6.00E+06
7.00E+06
8.00E+06
9.00E+06
1.00E+07
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
S
m
o
k
e
P
a
r
tic
le
D
e
n
s
ity
(1
/c
c
)
12135

Figure 69 Mean smoke particle diameter and count for flaming PU foam (100100100 mm) tests
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 100 200 300 400 500 600 700
Time (s)
M
e
a
n
D
ia
m
e
t
e
r
(
m
ic
r
o
n
)
12142
12156

0.00E+00
1.00E+06
2.00E+06
3.00E+06
4.00E+06
5.00E+06
6.00E+06
7.00E+06
8.00E+06
9.00E+06
1.00E+07
0 100 200 300 400 500 600 700
Time (s)
S
m
o
k
e
P
a
r
tic
le
D
e
n
s
ity
(1
/c
c
)
12142
12156

Figure 70 Mean smoke particle diameter and count for flaming PU foam (150150150 mm) tests

0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 100 200 300 400 500 600
Time (s)
M
e
a
n
D
ia
m
e
t
e
r
(
m
ic
r
o
n
)
12151
12152
12153

0.00E+00
1.00E+06
2.00E+06
3.00E+06
4.00E+06
5.00E+06
6.00E+06
7.00E+06
8.00E+06
9.00E+06
1.00E+07
0 100 200 300 400 500 600
Time (s)
S
m
o
k
e
P
a
r
tic
le
D
e
n
s
ity
(1
/c
c
)
12151
12152
12153

Figure 71 Mean smoke particle diameter and count for flaming nylon carpet tests 5

Smoke Characterization Project Final Report P. 102 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
A summary of test signals for the flaming tests at 240 s are presented in Table 22.

Table 22 Observed Fire Test Room test signals for flaming mode at 240 seconds
Target Sample Description
Test
No.
OBS
(%/ft)
d
m

(m)
n
m

(cc
-1
)
CO
(ppm)
CO
2

(ppm)
T
(C)
Vel.
(m/s)
12123 5.0 0.23 1.73E+06 708 1120 25.7 0.18
12124 2.3 0.10 4.57E+06 401 1662 27.3 0.16
12127 1.3 0.09 3.66E+06 413 1733 27.7 0.14
12146 5.0 0.15 4.00E+06 468 1312 25.5 0.14
UL 217 Douglas fir
12183 0.6 0.08 4.42E+06 189 1891 28.1 0.16
12113 1.4 0.09 1.57E+06 403 1951 25.3 0.05
12122 6.5 0.07 2.02E+06 304 1643 25.0 0.08
12125 20.1 0.11 1.86E+06 661 1426 26.0 0.01
12141 3.4 0.08 1.80E+06 254 1548 26.1 0.09
12144 9.9 0.07 1.76E+06 311 1781 26.5 0.06
UL 217 Shredded newspaper
12145 2.8 0.06 2.11E+06 249 1740 27.1 0.07
12112 13.0 0.34 2.27E+05 195 2165 25.1 -0.01
12131 11.9 0.34 4.03E+05 183 2125 26.5 -0.02
12181 11.9 0.34 3.37E+05 178 1973 25.7 -0.05
12182 12.9 0.33 4.84E+05 188 1950 25.5 -0.01
UL 217 3:1 Heptane/Toluene
mixture
01221 13.5 0.34 2.48E+05 188 2143 21.4 -0.02
12134 0.8 0.09 1.52E+06 223 1218 27.0 0.13 Coffee maker 12 cup, no
carafe 12186 0.7 0.10 1.94E+06 159 969 25.8 0.15
Mattress PU foam 100 100
mm sample
12154 NA NA NA NA NA NA NA
Mattress PU foam wrapped in
CA TB 117 50:50 cotton/
poly sheet 100 100 100
mm foam
12135 0.4 0.06 8.47E+05 26 1059 25.3 0.12
12142 3.9 0.22 6.41E+05 80 2846 30.5 0.18
Mattress PU foam wrapped in
CA TB 117 50:50 cotton/
poly sheet 150 150 150
mm foam
12156 3.0 0.24 5.85E+05 78 2623 31.7 0.16
12151 5.1 0.26 3.35E+05 64 2387 28.4 0.12
12152 4.8 0.26 3.89E+05 52 952 27.6 0.16
Nylon carpet 100 100 mm
sample
12153 4.0 0.25 4.05E+05 40 893 27.4 0.11
Notes to Table 22:
NA = Not attained 5
[1]
Bad data



10
Smoke Characterization Project Final Report P. 103 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The ceiling test signatures are summarized in Table 23.

Table 23 Fire Test Room ceiling test signatures for flaming combustion tests
Alarm Trigger
Time (s)
Ceiling Analog
Ionization Alarm
Signals
Ceiling Analog
Photo Alarm
Signals
Target Sample
Description
Test No.
Ion Photo Min Max Min Max
Max
Radial
Velocity
(m/s)
Max
Temp.
(
o
C)
12123 NAP NAP 16 80 15 65 0.26 40.0
12124 NAP NAP 16 78 15 65 0.30 40.5
12127 164 157 16 74 15 61 0.26 38.0
12146 145 185 16 78 15 65 0.26 39.4
UL 217 Douglas fir
12183 117 173 16 70 15 40 0.28 39.3
12113 NAP NAP 15 38 15 63 0.31 28.0
12122 NAP NAP 15 55 15 65 0.24 27.1
12125 176 179 16 54 15 65 0.28 28.9
12141 87 134 16 45 15 65 0.28 28.4
12144 143 160 16 51 15 65 0.25 29.3
UL 217 Newspaper
12145 126 126 16 47 15 65 0.22 27.4
12112 NAP NAP 17 79 16 59 0.34 30.1
12131 -- 66 16 79 15 49 0.38 31.3
12181 36 70 16 80 15 48 0.33 30.5
12182 34 71 16 80 15 46 0.37 31.4
UL 217 3:1
Heptane/Toluene
mixture
01221 34 72 15 27 15 65 0.31 27.1
12134 210 438 16 78 15 65 0.58 68.3 Coffee maker 12
cup, no carafe
12186 151 334 17 78 15 65 0.53 65.7
Mattress PU foam
100 100 mm
sample
12154 68 ND 15 38 15 39 0.16 26.7
Mattress PU foam
wrapped in CA TB
117 50:50 cotton/
poly sheet 100
100 100 mm foam
12135 DNT DNT 17 36 15 16 0.19 28.6
12142 112 DNT 16 64 15 24 0.30 34.57
Mattress PU foam
wrapped in CA TB
117 50:50 cotton/
poly sheet 150
150 150 mm foam
12156 96 171 16 67 15 27 0.33 34.32
12151 173 221 16 61 15 31 0.20 29.6
12152 162 DNT 16 60 15 29 0.18 28.3
Nylon carpet 100
100 mm sample
12153 137 323 16 61 15 32 0.21 28.0
Notes to Table 23:
NAP = Alarm not present 5
ND = Data not recorded
DNT = Smoke alarm did not trigger

The maximum radial ceiling velocity measured in the flaming test trends with the fire size
measured in the intermediate scale tests. The coffee maker with the peak heat release rate of 10
Smoke Characterization Project Final Report P. 104 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
approximately 100 kW had maximum radial ceiling velocity of approximately 0.5 m/s. The
mattress PU foam and nylon carpet had peak heat release rates of approximately 4 kW in the
intermediate scale tests, and developed maximum ceiling velocity of approximately 0.2 m/s in
the room tests.
5
Non-Flaming Test Results
In Table 24, are presented the obscuration summary for the non-flaming tests using the alarm
activation limits of 0.5 %/ft and 10 %/ft OBS. In this test series, repeat tests were conducted for
PU foam samples.
10
Table 24 Summary of smoke obscuration for non-flaming tests
Time @ UL 217 OBS
Limits (s)
Max. OBS
Target Sample Description Test No.
0.5 %/ft 10.0 %/ft Time (s) (%/ft)
12126 1794 3522 3676 11.42
12132 1767 3770 4128 12.54
12143 2409 NA 4184 8.88
12184 1596 3776 4010 12.17
UL 217 Ponderosa pine
12185 1002 3268 3710 14.94
12133 323 355 440 35.39
12155 323 368 446 33.38
Bread 4 slices
01244 359 405 464 30.56
Polyisocyanurate insulation 150 150
200 mm pieces
12271
5464 NA
6609
0.67
12192 2190 NA 3953 1.82 Mattress PU foam 150 150 50 mm
foam 12193 2337 NA 5267 1.98
12202 2017 NA 3799 8.54
Mattress PU foam 100 125 100 mm
foam with a 25 150 150 mm piece on
two opposing sides
12261
1723 5520
5524
10.57
Mattress PU foam wrapped in CA TB 117
cotton sheet 100 150 200 mm foam
01232
2180 NA
4085
7.03
Mattress PU foam wrapped in CA TB 117
cotton sheet 125 125 300 mm foam
01241
2758 NA
5984
9.33
01233 2885 NA 4225 4.88
Mattress PU foam wrapped in polyester
microfiber sheet 125 125 300 mm
foam
01245
3076 NA
4569
8.63
Nylon carpet 150 150 mm sample 12262 2404 NA 6404 4.27
Polystyrene pellets 69.8 g 12272 3956 NA 5587 5.93
Note to Table 24:
NA = Not attained

Other than bread, only one of the non-UL 217 sample tests resulted in OBS value of 10 %/ft, 15
even though not all of the sample mass was consumed during the tests. For the PU foam tests, the
sample exposed to the hot plate was charred, and this charring reduced the smoke generation
over time. A larger obscuration level was attained when the mass of the PU foam was increased
(see test series 12202, 12261 versus 12192, 12193, and also 01232 versus 01241). This is also
depicted in Figure 75, and Figure 76 respectively. 20
Smoke Characterization Project Final Report P. 105 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The OBS charts for these tests are presented in Figure 72 through Figure 79.

0
2
4
6
8
10
12
14
16
18
20
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time (s)
O
B
S

(
%
/
f
t
)
12126
12132
12143
12184
12185
UL 217 Reference
UL 217 Reference

Figure 72 OBS for Ponderosa pine in non-flaming tests
5


0
5
10
15
20
25
30
35
40
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
O
B
S

(
%
/
f
t
)
01244
12155
12133

Figure 73 OBS for bread in non-flaming tests
Smoke Characterization Project Final Report P. 106 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
O
B
S

(
%
/
f
t
)
12271

Figure 74 OBS for polyisocyanurate foam in non-flaming tests


0
2
4
6
8
10
12
14
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
O
B
S

(
%
/
f
t
)
12192
12193
12261
12202
5
Figure 75 OBS for PU foam in non-flaming tests
Smoke Characterization Project Final Report P. 107 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
1
2
3
4
5
6
7
8
9
10
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
O
B
S

(
%
/
f
t
)
01232
01241

Figure 76 OBS for cotton sheet wrapped PU foam in non-flaming tests


0
2
4
6
8
10
12
14
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
O
B
S

(
%
/
f
t
)
01233
01245
5
Figure 77 OBS for polyester microfiber wrapped PU foam non-flaming tests

Smoke Characterization Project Final Report P. 108 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
1
2
3
4
5
6
7
8
9
10
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
O
B
S

(
%
/
f
t
)
12262

Figure 78 OBS for nylon carpet in non-flaming tests


0
1
2
3
4
5
6
7
8
9
10
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
O
B
S

(
%
/
f
t
)
12272
5
Figure 79 OBS for polystyrene in non-flaming tests

Smoke Characterization Project Final Report P. 109 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The ionization and photoelectric smoke alarm trigger times are summarized in Table 25.

Table 25 Non-flaming mode alarm response times
Analog Signal
Value
Analog Signal
Value
Target Sample Description Test No.
Ion
Alarm
Trigger
Time (s)
MIC
(pA)
Photo
(mV)
Photo
Alarm
Trigger
Time (s)
MIC
(pA)
Photo
(mV)
12126 3244 63.9 71.1 3226 63.9 72.0
12132 DNT -- -- 3318 73.4 76.4
12143 3826 66.0 74.3 3805 68.2 75.0
12184 3547 66.0 70.1 3451 71.6 75.9
UL 217 Ponderosa pine
12185 2894 64.6 73.6 2722 72.3 79.1
12133 319 66.1 98.0 364 45.9 55.5
12155 306 71.5 99.4 371 41.5 45.8 Bread 4 slices
01244 343 75.8 98.5 448 28.4 19.4
Polyisocyanurate insulation
150 150 200 mm pieces
12271 DNT -- -- DNT -- --
12192 DNT -- -- DNT -- -- Mattress PU foam 150 150
50 mm foam 12193 DNT -- -- DNT -- --
12202 DNT -- -- 3149 85.3 77.2
Mattress PU foam 100 125
100 mm foam with a 25
150 150 mm piece on two
opposing sides
12261 5610 63.2 58.5 3032 81.4 68.8
Mattress PU foam wrapped in
CA TB 117 cotton sheet 100
150 200 mm foam
01232 DNT -- -- 3530 83.2 77.5
Mattress PU foam wrapped in
CA TB 117 cotton sheet 125
125 300 mm foam
01241 DNT -- -- 4207 88.5 80.5
01233 DNT -- -- 5353 83.5 79.8
Mattress PU foam wrapped in
polyester microfiber sheet
125 125 300 mm foam
01245 DNT -- -- 4128 90.2 73.6
Nylon carpet 150 150 mm
sample
12262 DNT -- -- 5727 84.4 84.3
Polystyrene pellets 69.8 g 12272 DNT -- -- 5546 82.6 74.5
Note to Table 25:
DNT = Did not trigger 5

For the Ponderosa pine test sample, the photoelectric smoke alarm on an average triggered 2.3 %
faster than the ionization smoke alarm. For bread the ionization smoke alarm was 22 % faster
than the photoelectric smoke alarm. For most of the other test samples the ionization smoke
alarm did not trigger. In each of these cases an OBS of 10%/ft had not been reached. For the one 10
case where the ionization alarm did trigger (PU foam test series 12261), an OBS of 10 %/ft was
attained. In the case of the two tests (polyisocyanurate foam, PU foam) for which neither the
ionization nor the photoelectric alarm triggered, this may be due to the smaller test sample mass.
For the polyisocyanurate foam test the maximum OBS value was calculated to be 0.67 %/ft and
for the two PU foam tests the maximum obscurations were 1.82 and 1.98 %/ft respectively. The 15
PU foam tests were repeated with a larger sample mass (Test series: 12202, 12261).
Smoke Characterization Project Final Report P. 110 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.


The MIC and Beam response to the PU foam were investigated by comparing the Beam and MIC
signals during these tests with a Ponderosa pine test (Test Series 12132). The Beam vs. MIC
signatures for the other Ponderosa pine tests were similar. 5

In Figure 80 is depicted the Beam vs. MIC response time for the Ponderosa pine sample. The UL
217 limits have been superimposed on the figure with dashed black lines.

Figure 80 Beam vs. MIC response: Ponderosa pine 10




15
60
65
70
75
80
85
90
95
100
60 65 70 75 80 85 90 95 100
Beam Signal
M
I
C

S
i
g
n
a
l
East
Ceiling
West
UL 217 Reference
UL 217 Reference
Smoke Characterization Project Final Report P. 111 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
It was observed that smoldering PU foam by itself has a Beam vs MIC response that also fits
between the UL 217 limits for the Ponderosa pine as shown in Figure 81. In this test (Test Series
12022), the ionization smoke alarm did not trigger.
Figure 81 Beam vs. MIC response for PU foam in non-flaming combustion
5
The data shows that for PU foam heated using the UL 217 hot plate, the Beam vs. MIC response
results in the data falling above the upper limits established for Ponderosa pine. This implies that
there are larger particles in the PU foam smoke that from the smoke generated by Ponderosa pine.

The Beam vs MIC response for PU foam wrapped with cotton fabric is shown Figure 82. It was 10
observed that the effect of the cotton fabric on the Beam vs MIC response is similar to that
observed for PU foam alone.

60
65
70
75
80
85
90
95
100
60 65 70 75 80 85 90 95 100
Beam Signal
M
I
C

S
i
g
n
a
l
East
Ceiling
West
UL 217 Reference
UL 217 Reference
Smoke Characterization Project Final Report P. 112 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
60
65
70
75
80
85
90
95
100
60 65 70 75 80 85 90 95 100
Beam Signal
M
I
C

S
i
g
n
a
l
East
Ceiling
West
UL 217 Reference
UL 217 Reference

Figure 82 Beam vs. MIC response for cotton sheet wrapped PU foam

The Beam vs MIC response for PU foam wrapped in polyester microfiber fabric (Test Series:
01245) is shown in Figure 83. The figure shows that the polyester microfiber fabric has a greater 5
influence on the Beam v. MIC response than PU foam alone.

60
65
70
75
80
85
90
95
100
60 65 70 75 80 85 90 95 100
Beam Signal
M
I
C

S
i
g
n
a
l
East
Ceiling
West
UL 217
Reference
UL 217
Reference

Figure 83 Beam vs MIC response for polyester microfiber wrapped PU foam
10
Smoke Characterization Project Final Report P. 113 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The Beam and MIC response for the polystyrene test is shown in Figure 84.

60
65
70
75
80
85
90
95
100
60 65 70 75 80 85 90 95 100
Beam Signal
M
I
C

S
i
g
n
a
l
Ceiling
West
UL 217 Reference
UL 217 Reference

Figure 84 Beam vs MIC response for Polystyrene in non-flaming combustion
It was observed that similar to the PU foam results, there are relatively more larger smoke 5
particles for polystyrene than UL 217 reference of Ponderosa pine.

From Figure 80 through Figure 84, it may also be observed that, near the end of the test, the
beam signal reduces indicating smaller smoke particle sizes and/or count. This was confirmed by
observations during these tests that over time, there was settling of smoke in the room. In order 10
to further investigate this phenomenon, an obscuration tree consisting light beams and photo-
detectors located at 600, 900, and 1500 mm below the ceiling was used. These obscuration data
complemented the light beam located at the ceiling, and thus provided data on change in smoke
obscuration over the height of the room during the tests. As a comparative reference to flaming
fire, a test with heptane/toluene was also performed. 15

These obscuration data over the height of the room for heptane/toluene mixture is provided in
Figure 85.

Smoke Characterization Project Final Report P. 114 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
5
10
15
20
0 100 200 300 400 500 600
Time
O
B
S

(
%
/
f
t
)
4 in below ceiling
24 in. below ceiling
36 in. below ceiling
60 in. below ceiling

Figure 85 OBS changes in the test room for heptane/toluene mixture
It was observed that for this flaming fire, there was not a significant effect of smoke settling.
This may be due to the higher energy of the smoke, as well as the short duration of the test.
5
The smoke obscuration change over time in the test room for bread is shown in Figure 86. After
peaking at the ceiling the OBS value drops below the 24 inch value at approximately 520
seconds.
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000
Time (sec)
P
e
r
c
e
n
t

O
b
s
c
u
r
a
t
i
o
n

p
e
r

F
o
o
t

(
%
/
f
t
)
4 in below ceiling
24 in.below ceiling
36 in. below ceiling
60 in. below ceiling

Figure 86 OBS changes in the test room for bread 10
Smoke Characterization Project Final Report P. 115 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The OBS change over time in the test room for PU foam wrapped with polyester microfiber
(Test series: 01245) is shown in Figure 87. The OBS value peaks at approximately 4500 s, and
then the OBS at 24 and 36 in. below the ceiling exceed the ceiling values. It may also be
observed that at approximately 5200 s, the OBS 60 in. below the ceiling is greater than at the
ceiling. 5
0
2
4
6
8
10
12
0 1000 2000 3000 4000 5000 6000 7000
Time (sec)
O
B
S

(
%
/
f
t
)


4 in below ceiling
24 in. below ceiling
36 in. below ceiling
60 in below ceiling

Figure 87 OBS changes in the test room for polyester microfiber wrapped PU foam

The OBS changes in the room for cotton fabric wrapped PU foam (Test Series: 01241) is
depicted in Figure 88. 10
0
2
4
6
8
10
12
0 1000 2000 3000 4000 5000 6000 7000
Time (sec)
O
B
S

(
%
/
f
t
)
4 in below ceiling
24 in. below ceiling
36 in. below ceiling
60 in. below ceiling

Figure 88 OBS changes in the test room for cotton fabric wrapped PU foam
Smoke Characterization Project Final Report P. 116 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
In this test, the OBS value 600 mm below the ceiling exceeds 10 %/ft, while the OBS at the
ceiling appears to level off.

The reduction in the smoke obscuration at the ceiling may be due to a number of factors such as
energy loss of the smoke layer at the ceiling, as well gravitational effect on the smoke particles. 5
Because these fires are relatively long in duration, this phenomenon is more pronounced than for
shorter, more intense flaming fires.

A summary of room test signals at an OBS value of 0.5 %/ft is presented in Table 26.
10
Table 26 Observed UL 217 room test signals at ceiling location for non-flaming mode tests at 0.5 % /ft
Target Sample Description
Test
No.
Time
(s)
D
m

(m)
N
m

(cc
-1
)
CO
(ppm)
CO
2

(ppm)
T
(C)
Vel.
(m/s)
12126 1794 0.15 1.58E+05 72 45 23.8 0.05
12132 1767 0.16 1.17E+05 47 13 23.4 0.04
12143 2409 0.16 1.98E+05 124 12 23.6 0.05
12184 1596 0.15 1.18E+05 35 0 22.4 0.03
UL 217 Ponderosa pine
12185 1002 0.17 1.09E+05 19 11 22.2 0.03
12136 323 0.11 1.70E+06 33 49 24.3 0.11
12155 323 0.11 1.66E+06 8 20 25.1 0.08 Bread 4 slices
01244 359 0.10 1.96E+06 6 70 17.8 0.07
Polyisocyanurate insulation
150 150 200 mm pieces
12271 5464 0.10 9.82E+05 14 6 23.5 0.05
12192 2190 0.16 1.14E+05 16 4 NA NA Mattress PU foam 150 150
50 mm foam 12193 2337 0.20 8.94E+04 14 18 NA NA
12202 2017 0.17 1.82E+05 8 4 22.8 0.01
Mattress PU foam 100 125
100 mm foam with a 25
150 150 mm piece on two
opposing sides
12261 1723 0.27 2.76E+04 6 23 22.8 0.03
Mattress PU foam wrapped in
CA TB 117 cotton sheet
100 150 200 mm foam
01232 2180 0.28 1.12E+04 15 0 17.8 0.06
Mattress PU foam wrapped in
CA TB 117 cotton sheet
125 125 300 mm foam
01241 2758 0.16 2.68E+04 10 3 16.5 0.05
01233 2885 0.16 1.26E+04 6 22 17.8 0.06
Mattress PU foam wrapped in
polyester microfiber sheet
125 125 300 mm foam
01245 3076 0.24 1.01E+04 8 11 16.28 0.02
Nylon carpet 150 150 mm
sample
12262 2404 0.21 4.00E+04 23 17 23.1 0.04
Polystyrene pellets 69.8 g 12272 3956 0.22 1.48E+05 1 11 23.3 0.05
Note to Table 26:
NA = Not available

15
Smoke Characterization Project Final Report P. 117 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
A summary of room test signals at OBS value of 10 %/ft is presented in Table 27.

Table 27 Observed UL 217 room test signals at ceiling location for non-flaming mode tests at 10 % Obs/ft
Target Sample Description
Test
No.
Time
(s)
D
m

(m)
N
m

(cc
-1
)
CO
(ppm)
CO
2

(ppm)
T
(C)
Vel.
(m/s)
12126 3522 0.24 6.10E+05 ND ND 24.3 0.05
12132 3770 0.26 7.30E+05 480 140 23.9 0.07
12143 NA NA NA NA NA NA NA
12184 3776 0.25 8.78E+05 429 94 23.3 0.07
UL 217 Ponderosa pine
12185 3268 0.27 7.72E+05 395 102 22.9 0.06
12136 355 0.15 1.81E+06 106 92 24.7 0.11
12155 368 0.17 1.77E+06 42 37 25.1 0.10 Bread 4 slices
01244 405 0.20 2.05E+06 39 90 20.0 0.08
Polyisocyanurate insulation
150 150 200 mm pieces
12271 NA NA NA NA NA NA NA
12192 NA NA NA NA NA NA NA Mattress PU foam 150 150
50 mm foam
12193 NA NA NA NA NA NA NA
12202 NA NA NA NA NA NA NA
Mattress PU foam 100 125
100 mm foam with a 25
150 150 mm piece on two
opposing sides
12261 5609 0.23 5.27E+05 104 60 23.7 0.09
Mattress PU foam wrapped in
CA TB 117 cotton sheet
100 150 200 mm foam
01232
NA NA NA NA NA NA NA
Mattress PU foam wrapped in
CA TB 117 cotton sheet
125 125 300 mm foam
01241
NA NA NA NA NA NA NA
01233
NA NA NA NA NA NA NA
Mattress PU foam wrapped in
polyester microfiber sheet
125 125 300 mm foam
01245 NA NA NA NA NA NA NA
Nylon carpet 150 150 mm
sample
12262 NA NA NA NA NA NA NA
Polystyrene pellets 69.8 g 12272 NA NA NA NA NA NA NA
Notes to Table 27:
NA = Not attained 5
ND = Data not recorded

The mean particle diameter and count for the non- flaming tests are depicted in Figure 89 through
Figure 98.
10
Smoke Characterization Project Final Report P. 118 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
12126
12132
12143
12184
12185




0.00E+00
2.00E+05
4.00E+05
6.00E+05
8.00E+05
1.00E+06
1.20E+06
1.40E+06
1.60E+06
1.80E+06
2.00E+06
0 1000 2000 3000 4000 5000 6000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
12126
12132
12143
12184
12185
5
Figure 89 Mean smoke particle diameter and count for Ponderosa pine in non-flaming tests

Smoke Characterization Project Final Report P. 119 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
12133
12155
01244




0.00E+00
5.00E+05
1.00E+06
1.50E+06
2.00E+06
2.50E+06
3.00E+06
3.50E+06
4.00E+06
4.50E+06
5.00E+06
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
12133
12155
01244
5
Figure 90 Mean smoke particle diameter and count for bread in non-flaming tests


Smoke Characterization Project Final Report P. 120 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
12271




0.00E+00
1.00E+06
2.00E+06
3.00E+06
4.00E+06
5.00E+06
6.00E+06
7.00E+06
8.00E+06
9.00E+06
1.00E+07
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
12271
5
Figure 91 Mean smoke particle diameter and count for polyisocyanurate foam in non-flaming tests

Smoke Characterization Project Final Report P. 121 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
12192
12193




0.00E+00
5.00E+05
1.00E+06
1.50E+06
2.00E+06
2.50E+06
3.00E+06
0 1000 2000 3000 4000 5000 6000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
12192
12193
5
Figure 92 Mean smoke particle diameter and count for PU foam in non-flaming tests



Smoke Characterization Project Final Report P. 122 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
PU Foam (12202)




0.00E+00
1.00E+05
2.00E+05
3.00E+05
4.00E+05
5.00E+05
6.00E+05
7.00E+05
8.00E+05
9.00E+05
1.00E+06
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
PU Foam (12202)
5
Figure 93 Mean smoke particle diameter and count for PU foam in non-flaming tests
(Data from Test 12261 were found to be suspicious and were not plotted)

Smoke Characterization Project Final Report P. 123 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
01232




0.00E+00
2.00E+05
4.00E+05
6.00E+05
8.00E+05
1.00E+06
1.20E+06
1.40E+06
1.60E+06
1.80E+06
2.00E+06
0 1000 2000 3000 4000 5000 6000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
01232
5
Figure 94 Mean smoke particle diameter and count for cotton fabric wrapped PU foam in non-flaming tests

Smoke Characterization Project Final Report P. 124 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
01241




5
Figure 95 Mean smoke particle diameter and count for cotton-poly wrapped PU foam in non-flaming tests

Smoke Characterization Project Final Report P. 125 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
01233
01245




0.00E+00
2.00E+05
4.00E+05
6.00E+05
8.00E+05
1.00E+06
1.20E+06
1.40E+06
1.60E+06
1.80E+06
2.00E+06
0 1000 2000 3000 4000 5000 6000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
01231
01245
5
Figure 96 Mean smoke particle diameter and count for polyester microfiber wrapped PU foam in non-
flaming tests

Smoke Characterization Project Final Report P. 126 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
12262




0.00E+00
2.00E+05
4.00E+05
6.00E+05
8.00E+05
1.00E+06
1.20E+06
1.40E+06
1.60E+06
1.80E+06
2.00E+06
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
12262
5
Figure 97 Mean smoke particle diameter and count for nylon carpet in non-flaming tests
Smoke Characterization Project Final Report P. 127 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
12272




0.00E+00
5.00E+05
1.00E+06
1.50E+06
2.00E+06
2.50E+06
3.00E+06
3.50E+06
4.00E+06
4.50E+06
5.00E+06
0 1000 2000 3000 4000 5000 6000 7000
Time (s)
S
m
o
k
e

P
a
r
t
i
c
l
e

D
e
n
s
i
t
y

(
1
/
c
c
)
12272
5
Figure 98 Mean smoke particle diameter and count for polystyrene in non-flaming tests
Smoke Characterization Project Final Report P. 128 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The ceiling test signatures are summarized in Table 28.

Table 28 UL 217 Fire Test Room ceiling test signatures for non-flaming combustion tests
Alarm Trigger
Time (s)
Ceiling Analog
Ionization Alarm
Signal
Ceiling Analog
Photo Alarm
Signal
Target Sample
Description
Test No.
Ion Photo Min Max Min Max
Max
Radial
Velocity
(m/s)
Max
Temp.
(
o
C)
12126 3244 3226 23 57 36 65 0.09 24.5
12132 NAP 3318 15 61 15 65 0.11 24.7
12143 3826 3805 15 46 15 65 0.10 24.4
12184 3547 3451 16 57 15 65 0.09 23.8
UL 217 Ponderosa
pine
12185 2894 2722 17 67 15 65 0.11 24.0
12133 319 364 17 79 15 65 0.14 26.0
12155 306 371 16 78 15 65 0.15 26.4 Bread 4 slices
01244 343 448 16 80 15 65 0.14 18.8
Polyisocyanurate
insulation 150 150
200 mm pieces
12271 DNT DNT 15 25 15 17 0.11 24
12192 DNT DNT 16 24 15 32 ---
[1]
---
[1]

Mattress PU foam
150 150 50 mm
foam
12193 DNT DNT 16 29 15 34 ---
[1]
---
[1]

12202 DNT DNT 16 33 15 65 0.10 23.8
Mattress PU foam
100 125 100 mm
foam with a 25 150
150 mm piece on two
opposing sides
12261 5610 3032 15 40 15 65 0.11 23.9
Mattress PU foam
wrapped in CA TB 117
cotton sheet 100
150 200 mm foam
01232 DNT 3530 15 28 15 65 0.10 18.6
Mattress PU foam
wrapped in CA TB 117
cotton sheet 125
125 300 mm foam
01241 DNT 4207 16 34 15 65 0.11 17.4
01233 DNT 5353 16 29 15 65 0.10 17.1
Mattress PU foam
wrapped in polyester
microfiber sheet 125
125 300 mm foam
01245 DNT 4128 15 27 15 65 0.12 18.1
Nylon carpet 150
150 mm sample
12262 DNT 5727 15 27 15 62 0.10 24.1
Polystyrene pellets
69.8 g
12272 DNT 5546 15 30 15 65 0.11 24.3
Notes to Table 28:
NAP = Alarm not present 5
DNT = Did not trigger
[1]
Bad velocity and temperature data

It was observed that the maximum radial velocities in the non-flaming tests are on the order of
0.10 m/s. In comparison, the velocity in the UL 217 Sensitivity smoke box test is 0.16 m/s. 10
Smoke Characterization Project Final Report P. 129 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
TASK 4 CORRELATE ANALYTICAL DATA AND PERFORMANCE IN
THE FIRE TEST ROOM

INTRODUCTION
A range of natural, synthetic, and multi-component materials representing the variety of products 5
found in residential settings was evaluated for this investigation.

In this section, the results from the small, intermediate and Fire Test Room tests were analyzed
for specific trends related to the influence of (i) materials and combustion mode, and (ii) mode of
testing on the smoke generated. 10


SMOKE PARTICLE DISTRIBUTION MEASUREMENTS
Light based obscuration systems used in UL 217 operate on a principle of light extinction which
is related to the volume fraction occupied by the scattering particles. Photoelectric alarms are 15
based on light scattering which depends on the amount of particle surface area along with the
particle reflectivity. Ionization field based systems (e.g., MIC, ionization alarms) used in UL 217
however rely equally on the number of particles within the sample chamber as the size of the
particles; hence the specific particle counts are more relevant. These sensor technologies and
particle size and count dependencies are summarized in Table 29. Tests using the WPS 20
spectrometer in the UL 217 Sensitivity Test smoke box confirmed the obscuration and ionization
principles.

Table 29 Theoretical smoke particle dependency for traditional smoke sensor technologies
Sensor Type(s) Principle Smoke Particle Relation
MIC, Ion Alarms Ionization
i
i
d n
Photoelectric Alarms Light scattering
2
i
i
d n
Obscuration Systems Light obscuration
3
i
i
d n
25

INFLUENCE OF MATERIALS AND COMBUSTION MODE: CONE CALORIMETER
The ASTM E1354 cone calorimeter provided a consistent, well- regulated means for evaluating
the smoke generated by different materials under flaming and non-flaming conditions. The
specific extinction area under the two modes of combustion, Figure 99, indicates that most of the 30
materials generate more smoke per unit of consumed mass under non-flaming conditions. The
most significant effect of the combustion mode on smoke production is for the polyurethane and
polyisocyanurate foams, possibly due to the high surface area to volume ratio resulting from
their unique physical structure.
35
Smoke Characterization Project Final Report P. 130 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
C
o
o
k
in
g

o
il
B
re
a
d
N
e
w
s
p
a
p
e
r
D
o
u
g
la
s

fir
P
o
n
d
e
r
o
s
a

p
in
e
C
o
tto
n

b
a
ttin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

(
fa
b
r
ic
)
R
a
y
o
n

(fa
b
ric
)
H
D
P
E
N
y
lo
n

c
a
r
p
e
t
P
o
ly
e
s
te
r
c
a
rp
e
t
P
o
ly
e
s
te
r

fillin
g
P
U

fo
a
m
P
o
ly
is
o
c
y
a
n
u
a
r
a
te

fo
a
m
P
V
C

S
p
e
c
i
f
i
c

E
x
t
i
n
c
t
i
o
n

A
r
e
a

(
m

/
g
)
Flaming
Non-Flaming
UL 217 materials

Figure 99 Specific extinction area for small-scale flaming and non-flaming combustion

The mode of combustion appears to have different effects on the mean size of the generated
smoke particles depending on the material chemistry, Figure 100. Non- flaming combustion 5
generates smaller particles than flaming combustion on natural cellulosic materials but for
synthetic materials the particle sizes were larger in the non-flaming conditions.

0.00
0.05
0.10
0.15
0.20
0.25
0.30
C
o
o
k
in
g

o
il
B
re
a
d
N
e
w
s
p
a
p
e
r
D
o
u
g
la
s

fir
P
o
n
d
e
ro
s
a

p
in
e
C
o
tto
n

b
a
ttin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

(
fa
b
r
ic
)
R
a
y
o
n
(fa
b
ric
)
H
D
P
E
N
y
lo
n

c
a
r
p
e
t
P
o
ly
e
s
te
r c
a
rp
e
t
P
o
ly
e
s
te
r
fillin
g
P
U

fo
a
m
P
o
ly
is
o
c
y
a
n
u
a
r
a
te

fo
a
m
P
V
C

M
e
a
n

P
a
r
t
i
c
l
e

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Flaming
Non-Flaming
UL 217 materials

Figure 100 Mean particle diameter for small-scale flaming and non-flaming combustion 10
Smoke Characterization Project Final Report P. 131 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Measured specific particle counts plotted in Figure 101 does not indicate any material
independent trends for the effects of combustion mode on the number of particles generated per
unit consumed mass.

0.E+00
5.E+05
1.E+06
2.E+06
2.E+06
3.E+06
3.E+06
4.E+06
C
o
o
k
in
g

o
il
B
re
a
d
N
e
w
s
p
a
p
e
r
D
o
u
g
la
s

fir
P
o
n
d
e
ro
s
a

p
in
e
C
o
tto
n

b
a
ttin
g
C
o
tto
n
/P
o
ly
e
s
te
r

b
le
n
d

(
fa
b
r
ic
)
R
a
y
o
n

(fa
b
ric
)
H
D
P
E
N
y
lo
n

c
a
rp
e
t
P
o
ly
e
s
te
r
c
a
rp
e
t
P
o
ly
e
s
te
r

fillin
g
P
U

fo
a
m
P
o
ly
is
o
c
y
a
n
u
a
ra
te

fo
a
m
P
V
C

S
p
e
c
i
f
i
c

P
a
r
t
i
c
l
e

C
o
u
n
t

(
1
/
c
c
/
g
)
Flaming
Non-Flaming
UL 217 materials
5
Figure 101 Specific particle count for small-scale flaming and non-flaming combustion

INFLUENCE OF MATERIALS AND COMBUSTION MODE: FIRE TEST ROOM
The cone calorimeter was used to characterize the inherent material products of combustion (e.g.
heat, smoke and effluent gases generated) under consistent, well-regulated conditions. The 10
continuous removal of smoke and other combustion products via the cone calorimeter exhaust
flow prohibits smoke concentration build-up and potential smoke particle aggregation that would
be expected in relatively stagnant air spaces such as a residential settings. Smoke build-up in a
given air space depends on the volume of the air space, the inherent smoke particulate rate
formation and consequently the size and geometry of the involved burning material, and the 15
mode of combustion. Therefore comparison of combustion products generated by the more
complex test targets evaluated in the stagnant air Fire Test Room is more appropriate at a set
obscuration level as opposed to a set time. As seen in Figure 102, larger smoke particles were
generally observed for non- flaming combustion than for flaming combustion. These results
parallel results obtained on the cone calorimeter. 20

Smoke Characterization Project Final Report P. 132 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.00
0.05
0.10
0.15
0.20
0.25
D
o
u
g
la
s

fir
P
o
n
d
e
ro
s
a

p
in
e
N
e
w
s
p
a
p
e
r
H
e
p
ta
n
e
/T
o
lu
e
n
e
P
o
ly
s
ty
r
e
n
e
C
o
ffe
e

m
a
k
e
r
P
U

fo
a
m
P
U

fo
a
m

in

C
o
tto
n
/P
o
ly
P
U

fo
a
m

in

C
o
tto
n
P
U

fo
a
m

in

P
o
ly
N
y
lo
n

c
a
r
p
e
t
P
o
ly
is
o
c
y
a
n
u
r
a
te
B
re
a
d
M
e
a
n

P
a
r
t
i
c
l
e

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Flaming
Non-flaming

Figure 102 Mean particle diameters at an obscuration of 0.5 %/ft in the Fire Test Room

Measured MIC, analog ionization alarm, obscuration, and analog photo alarm signals are plotted
against respective particle size and count data in Figure 103 through Figure 110. Individual test 5
results support the predicted relationships described in Table 29. Comparison of tests for
different materials, however, indicate that there is a material effect on the respective signal in
addition to the predicted particle size and count relationship. This material dependency effect is
more evident for ionization and scattering sensor technologies than light obscuration because the
smoke particulate size and count does account for either the propensity of the particulate to 10
ionize or its reflectivity.

Categorical evaluation of the data for combustion mode response indicates that the scattering
sensor technology is more sensitive to combustion mode than either obscuration or ionization
technologies. 15


Smoke Characterization Project Final Report P. 133 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
10
20
30
40
50
60
70
80
90
100
0.0E+00 2.0E+05 4.0E+05 6.0E+05 8.0E+05 1.0E+06
(nidi )
M
I
C

S
i
g
n
a
l

C
h
a
n
g
e

(
p
A
)
Coffee Maker
Coffee Maker
Nylon carpet
Nylon carpet
Nylon carpet
PU Foam
PU Foam + cotton/poly
PU Foam + cotton/poly
PU Foam + cotton/poly
PU Foam + cotton/poly

Figure 103 MIC signal versus particle size data for Fire Test Room flaming tests
MIC signal response for flaming (Figure 103) and non-flaming (Figure 104) tests demonstrate
the linear relationship predicted for particle size and count. Variation in signal responsiveness
between materials however, indicates a material-soot chemistry dependency that is not addressed 5
by the model such as soot-air ionization potential () and ion diffusivity (D). The flaming and
non- flaming combustion data suggests that ionization technology is sensitive to the mode of
combustion.
0
10
20
30
40
50
60
70
80
90
100
0.0E+00 2.0E+05 4.0E+05 6.0E+05 8.0E+05 1.0E+06
(ni di )
M
I
C

S
i
g
n
a
l

C
h
a
n
g
e

(
p
A
)
Bread
Bread
Bread
Nylon Carpet
Polyisocyanurate
Foam
Polystyrene
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
PU Foam
PU Foam
PU Foam
PU + cotton
PU + cotton
PU + poly
PU + Poly

Figure 104 MIC signal versus particle size data for Fire Test Room non-flaming tests 10
Smoke Characterization Project Final Report P. 134 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
10
20
30
40
50
60
70
0.0E+00 2.0E+05 4.0E+05 6.0E+05 8.0E+05
(ni di )
A
n
a
l
o
g

I
o
n

S
i
g
n
a
l

C
h
a
n
g
e
Coffee Maker
Coffee Maker
Douglas Fir
Douglas Fir
Douglas Fir
Heptane/Toluene
Heptane/Toluene
Heptane/Toluene
Heptane/Toluene
Newspaper
Newspaper
Newspaper
Nylon Carpet
Nylon Carpet
Nylon Carpet
PU Foam
PU Foam + cotton/poly
PU Foam + cotton/poly
PU Foam + cotton/poly
PU Foam + cotton/poly

Figure 105 Analog ion signal versus particle size data for Fire Test Room flaming tests
Analog ion signal responses for flaming (Figure 105) and non- flaming (Figure 106) tests parallel
the observed MIC signal response: linear relationship with particle size and count, material/soot
chemistry dependency, and sensitivity to the mode of combustion. 5

0
10
20
30
40
50
60
70
0.0E+00 2.0E+05 4.0E+05 6.0E+05 8.0E+05 1.0E+06
(nidi)
A
n
a
l
o
g

I
o
n

S
i
g
n
a
l

C
h
a
n
g
e
Bread
Bread
Bread
Nylon Carpet
Isocyanuarate Foam
Polystyrene
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
PU Foam
PU Foam
PU Foam
PU Foam
PU Foam + cotton
PU Foam + poly
PU Foam + poly

Figure 106 Analog ion signal versus particle size data for Fire Test Room non-flaming tests

Smoke Characterization Project Final Report P. 135 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
5
10
15
20
25
30
35
40
0.0E+00 2.0E+04 4.0E+04 6.0E+04 8.0E+04 1.0E+05 1.2E+05
(n idi
3
)
O
B
S

(
%
/
f
t
)
Coffee Maker
Coffee Maker
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Heptane/Toluene
Heptane/Toulene
Heptane/Toluene
Newspaper
Newspaper
Nylon Carpet
Nylon Carpet
Nylon Carpet
PU Foam
PU foam + cotton/poly
PU Foam + cotton/poly
PU Foam + cotton/poly

Figure 107 Obscuration versus particle size data for Fire Test Room flaming tests
Obscuration responses for flaming (Figure 107) and non-flaming (Figure 108) tests demonstrate
the predicted linear relationship with particle count and third order relationship with particle size.
Variation in signal responsiveness between materials indicates a material/soot chemistry 5
dependency that is not addressed by the model such as refractive index and soot particle density.
The flaming and non-flaming combustion data suggests that obscuration technology is relatively
insensitive to the mode of combustion.
0
5
10
15
20
25
30
35
40
0.0E+00 2.0E+04 4.0E+04 6.0E+04 8.0E+04 1.0E+05 1.2E+05
(n idi
3
)
O
B
S

(
%
/
f
t
)
Bread
Bread
Bread
Nylon Carpet
Polyisocyanurate
Foam
Polystyrene
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
PU Foam
PU Foam
PU Foam
PU Foam + cotton
PU Foam + cotton
PU Foam + Poly
PU Foam + Poly

Figure 108 Obscuration versus particle size data for Fire Test Room non-flaming tests 10
Smoke Characterization Project Final Report P. 136 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0
10
20
30
40
50
60
0.0E+00 5.0E+04 1.0E+05 1.5E+05 2.0E+05 2.5E+05
(n idi
2
)
A
n
a
l
o
g

P
h
o
t
o

S
i
g
n
a
l

C
h
a
n
g
e
Coffee Maker
Coffee Maker
Douglas Fir
Douglas Fir
Douglas Fir
Heptane/Toluene
Heptane/Toulene
Heptane/Toluene
Newspaper
Newspaper
Newspaper
Nylon Carpet
Nylon Carpet
Nylon Carpet
PU Foam
PU Foam + cotton/poly
PU Foam + cotton/poly
PU Foam + cotton/poly

Figure 109 Analog photo (scattering) signal versus particle size data for Fire Test Room flaming tests
Scattering responses for flaming (Figure 109) and non- flaming (Figure 110) tests demonstrate
the predicted linear relationship with particle count and second order relationship with particle
size. Variation in signal responsiveness between materials indicates a material/soot chemistry 5
dependency that is not addressed by the model such as particle reflectivity and refractive index.
The flaming and non-flaming combustion data suggests that scattering technology is more
sensitive to the mode of combustion than obscuration. This difference may be attributed to
variations in smoke color, i.e. reflectivity.
0
10
20
30
40
50
60
0.0E+00 5.0E+04 1.0E+05 1.5E+05 2.0E+05 2.5E+05
(n
i
d
i
2
)
A
n
a
l
o
g

P
h
o
t
o

S
i
g
n
a
l

C
h
a
n
g
e
Bread
Bread
Bread
Nylon Carpet
Isocyanurate Foam
Polystyrene Foam
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
PU Foam
PU Foam
PU Foam
PU Foam + cotton
PU Foam + poly
PU foam + poly
10
Figure 110 Analog photo (scattering) signal versus particle size data for Fire Test Room non-flaming tests
Smoke Characterization Project Final Report P. 137 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Comparison of ionization and photoelectric alarm trigger times for the materials under different
modes of combustion indicated that ionization alarms responded faster for flaming combustion
tests whereas photoelectric alarms responded faster for the less energetic, non- flaming tests,
Table 30.
5
Table 30 Fire Test Room alarm trigger times
Alarm Trigger Time (s) Alarm Trigger Time (s)
Flaming Tests
Ion Photo
Non-Flaming Tests
Ion Photo
Douglas fir 142 172 Ponderosa pine 3378 3304
Newspaper 133 150 Polyisocyanurate DNT DNT
Heptane/Toluene 35 70 PU foam 5610 3032
Coffee maker 181 386 PU foam in Cotton DNT 3870
PU foam 68 DNT PU foam in Poly DNT 4741
PU foam in Cotton/Poly 104 171 Nylon carpet DNT 5727
Nylon carpet 157 272 Polystyrene DNT 5546
Bread 323 394
Notes to Table 30:
DNT = Did not trigger

It was observed that both PU foam and cotton/polyester blend fabric have relatively low particle 10
size but have relatively high particle density. This may explain why the photoelectric smoke
alarm did not trigger in the room tests (more receptive to larger particles), where as the
ionization smoke alarm triggered (more receptive to larger particle counts).

The non- flaming decomposition was observed to be dependant on the mode of heat provided to 15
the sample.


INFLUENCE OF TESTING METHOD
In this investigation, testing was performed on the small-scale using the cone calorimeter, on the 20
intermediate-scale using ULs product calorimeters, and in ULs Fire Test Room.

The mean smoke diameter data obtained during the cone calorimeter and intermediate
calorimeter tests are presented in Table 31.
25
Smoke Characterization Project Final Report P. 138 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Table 31 Influence of scale on mean smoke diameter
Mean Diameter D
m

(m)
Test Sample
Small-Scale
Cone Calorimeter
Intermediate
Calorimeter
3:1 Heptane/Toluene
[1]
0.26 0.28
Heptane
[2]
0.19 0.23
Newspaper
[1]
0.04 0.09
Douglas fir
[1]
0.06 0.07
Cotton Batting
[3]
0.09 0.05
PU Foam
[3]
0.05 ---
Nylon Carpet 0.12 0.15
Notes to Table 31:
[1]
Sample tested using UL 217 assembly in intermediate scale
[2]
Sample ignited using a lighter
[3]
Sample tested using a TB 604 burner for ignition 5

It was observed that the mean smoke particle sizes for the flaming mode were similar between
the cone calorimeter and the intermediate-scale test even the ignition methods were different.
The small increase in the diameter observed in the intermediate calorimeter tests may be due to
higher aggregation of smoke in the intermediate scale tests prior to sampling. A larger increase in 10
intermediate scale test was observed for the newspaper sample. This is anticipated as there were
different packing conditions between the two tests and that would have resulted in different
combustion conditions for burning. The initial diameter data from the room tests are in good
agreement with the data mean diameter data from the cone calorimeter.
15
A limited amount of testing was conducted on how the mode of heating influences the smoke
characteristics. However, the results in Table 32 show a significant difference in particle size and
count for the PU foam. This has also been documented by T.J. Ohlemiller
12
.

Table 32 Influence of heating mode on smoke characteristics: non-flaming 20
Test sample Heating Mode
Mean particle
size, D
m
(m)
Average Count
Density (1/cc)
Radiant heating
(15 kW/m
2
)
0.083 8.82E+05
PU Foam
Hot plate
(UL 217 controller)
0.118 7.50E+06
Radiant Heating
(15 kW/m
2
)
0.100 3.30E+06
Bread
Electric Toaster 0.135 2.94E+06

The PU foam non- flaming tests in Fire Test Room tests were conducted with the hot plate with
the temperature controlled according to UL 217 Smoldering Test protocol. The larger mean
particle size observed in the intermediate-scale tests may explain why the photoelectric alarm
triggered sooner than the ionization smoke alarm for Test 12261 (3032 versus 5610 s 25
respectively).
Smoke Characterization Project Final Report P. 139 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Comparisons of smoke release rates measured on the small- and intermediate-scale calorimeters
to obscuration values measured in the Fire Test Room for flaming PU foam, heptane/toluene
mixture, nylon carpet, and the coffee maker are presented in Figure 111 through Figure 114.

0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.018
0.020
0 60 120 180 240 300
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
0
2
4
6
8
10
O
B
S

(
%
/
f
t
)
Cone Calorimeter
Fire Test Room
5
Figure 111 Small-scale smoke release rate versus Fire Test Room obscuration for flaming PU foam tests


0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0 30 60 90 120 150 180 210 240
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
0
2
4
6
8
10
12
14
O
B
S

(
%
/
f
t
)
Intermediate-Scale
Fire Test Room

Figure 112 Intermediate-scale smoke release rate versus Fire Test Room obscuration for flaming 10
heptane/toluene mixture tests
Smoke Characterization Project Final Report P. 140 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0 60 120 180 240 300 360 420 480 540 600
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
0
1
2
3
4
5
6
7
8
9
10
O
B
S

(
%
/
f
t
)
Intermediate-Scale
Fire Test Room

Figure 113 Intermediate-scale smoke release rate versus Fire Test Room obscuration for flaming nylon
carpet tests

5
0
1
2
3
4
5
6
7
0 120 240 360 480 600 720 840 960 1080
Time (s)
S
m
o
k
e

R
e
l
e
a
s
e

R
a
t
e

(
m
2
/
s
)
0
10
20
30
40
50
60
70
O
B
S

(
%
/
f
t
)
Intermediate-Scale
Fire Test Room

Figure 114 Intermediate-scale smoke release rate versus Fire Test Room obscuration for flaming coffee
maker tests
These plots illustrate how obscuration behavior measured in the Fire Test Room reflects smoke
release rate. This relationship is more evident during the early stages of the experiments than the 10
Smoke Characterization Project Final Report P. 141 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
latter stages because smoke accumulates throughout the Fire Test Room tests but not the smoke
release rate measurements.

Particle size data from the IMO and Fire Test Room tests were compared to study the influence
of particulate aggregation in the test room and are presented in Figure 115 through Figure 119. 5
For each material data set compared, the trends appear to be similar but the Fire Test Room
results indicate a time lag. Presumably this time lag is associated with the time for particles to be
transported from the source to the sampling location and the propensity of the material to
produce smoke particulate matter.
10
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 50 100 150 200 250 300 350 400
Time (s)
P
a
r
t
i
c
l
e

M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Smoke Room 12112
Smoke Room 12131
Smoke Room 12181
Smoke Room 12182
Smoke Room 01221
IMO 12067
IMO 12066

Figure 115 IMO and Fire Test Room smoke particle mean diameter for flaming heptane/toluene mixture
tests
Even though the initial mean diameters are similar for heptane/toluene, the particle sizes at the
sampling point in the room remain higher due to accumulation and smoke aggregation. 15


Smoke Characterization Project Final Report P. 142 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.

Figure 116 IMO and Fire Test Room smoke particle mean diameter for flaming Douglas fir tests
The mean particle diameter data for Douglas fir in the Fir Tests Room tests are similar to the
IMO data except they appear to be shifted in time. The reduction in mean diameter in both the
room and the IMO tests are from the charring of wood. A reduction in mean particle diameter 5
was observed in the cone calorimeter tests.


0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 50 100 150 200 250 300 350 400
Time (s)
P
a
r
t
i
c
l
e

M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Smoke Room 12113
Smoke Room 12122
Smoke Room 12125
Smoke Room 12141
Smoke Room 12144
Smoke Room 12145
IMO 12068
MO 12069

Figure 117 IMO and Fire Test Room smoke particle mean diameter for flaming newspaper tests 10
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 100 200 300 400 500 600 700
Time (s)
P
a
r
t
i
c
l
e

M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Smoke Room 12123
Smoke Room 12124
Smoke Room 12127
Smoke Room 12146
Smoke Room 12183
IMO 120610
IMO 120611
Smoke Characterization Project Final Report P. 143 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
There is a greater variation in the mean particle diameter for the newspaper both in the IMO and
Fire Test Room tests. This variation is from the specific combustion conditions developed based
upon the packing of the newspaper in test sample assembly.

5
Figure 118 IMO and Fire Test Room smoke particle mean diameter for flaming PU foam tests
There appear to be significant influence of smoke aggregation for the PU foam test sample in the
Fire Test Room tests.

10
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
P
a
r
t
i
c
l
e

M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Smoke Room 12134
Smoke Room 12186
IMO 12085
IMO 12084

Figure 119 IMO and Fire Test Room smoke particle mean diameter for flaming coffee maker tests
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
P
a
r
t
i
c
l
e

M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Smoke Room 12135
IMO 12074
IMO 12075
Smoke Room 12142
(6x6x6)
Smoke Characterization Project Final Report P. 144 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
The mean particle diameter history appears to trend very well with data from the IMO tests. It
may be due to heat release profile (the coffee maker had a peak heat release rate of
approximately 100 kW in the IMO tests). The higher energy fire would result in faster ceiling
jets. This would tend to replenish smoke particles at the smoke sampling location more quickly
than other fires. The higher mean diameter size later into the test is from accumulation and 5
aggregation of smoke at the ceiling.

Both the intermediate scale and Fire Test Room non- flaming Ponderosa pine test (UL 217
smoldering Ponderosa pine) were conducted in the same room using the same heat source (UL
217 hot plate). In the intermediate scale test, the smoke was sampled approximately 0.4 m above 10
the hot plate, whereas in the Fire Test Room tests, the smoke was sampled 5.4 m away at the
ceiling in vicinity of the MIC instrument. Despite the longer transport times expected for the
tests in which the smoke was sampled at the ceiling, the mean smoke particle diameters remain
similar, Figure 120. There is insignificant smoke aggregation as evidenced by the relatively
constant particle diameter in the Fire Test Room tests until approximately 2400 seconds (40 15
minutes).

0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 1000 2000 3000 4000 5000
Time (s)
P
a
r
t
i
c
l
e

M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Smoke Room 12126
Smoke Room 12132
Smoke Room 12143
Smoke Room 12184
Smoke Room 12185
Int. Scale

Figure 120 Intermediate-scale and Fire Test Room smoke particle mean diameter for non-flaming
Ponderosa pine tests 20



Smoke Characterization Project Final Report P. 145 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
P
a
r
t
i
c
l
e

M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Smoke Room 12133
Smoke Room 12155
Smoke Room 01244
IMO 12083
IMO 12082

Figure 121 IMO and Fire Test Room smoke particle mean diameter for non-flaming bread tests
The mean particle diameters for bread appear to be in good agreement between the IMO and the
Fire Test Room tests. This indicates that there is not a significant effect of particle aggregation.
5

0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0 100 200 300 400 500 600
Time (s)
P
a
r
t
i
c
l
e

M
e
a
n

D
i
a
m
e
t
e
r

(
m
i
c
r
o
n
)
Smoke Room 12151
Smoke Room 12152
Smoke Room 12153
IMO 12078
IMO 12079

Figure 122 IMO and Fire Test Room smoke particle mean diameter for flaming nylon carpet tests
The mean smoke diameter results from the Fire Test Room tests appear to trend with the data
from IMO tests. There is a time shift that may result from the transport time for the smoke to 10
travel from the source to the sampling location.
Smoke Characterization Project Final Report P. 146 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
TASK 5 - IDENTIFY FUTURE CONSIDERATIONS
In this section, future considerations derived from the results of this Smoke Characterization
Project are identified as follows:

1. The addition of other test materials such as polyurethane foam in the flaming and non- 5
flaming combustion modes in UL 217.

Rationale
- Currently PU foam is prevalent in residential furniture and bedding products.
- Tests in the small-scale and intermediate-scale showed that PU foam generated 10
smoke that is different in particle size and count than the UL 217 test materials.
- Some of the evaluated flaming and non-flaming test scenarios triggered one but
not both the photoelectric and ionization smoke alarms within the alarm response
criteria specified in UL 217.
15
2. Whether a smoke alarm, once triggered, should remain activated unless deactivated
manually.

Rationale
- In the non- flaming tests, it was found that there was stratification of the smoke 20
over time. This led to a smoke alarm that had triggered to deactivate once the
smoke at the ceiling had cleared below the activation level.

3. Requiring the use of combination ionization and photoelectric alarms for residential use
in order to maximize responsiveness to a broad range of fires. 25

Rationale
- Some of the evaluated flaming and non-flaming test scenarios triggered one but
not both the photoelectric and ionization smoke alarms within the alarm response
criteria specified in UL 217. Thus, a combination unit may maximize 30
responsiveness of each technology to a non-specific fire.

4. Characterize materials described in UL 217 using cone calorimeter, smoke particle
spectrometer and analytical testing.
35
Rationale
- The results from this research showed that the cone calorimeter augmented by the
WPS particle spectrometer provided useful data on the combustibility and smoke
characteristics of materials. This in conjunction with FTIR for material chemistry,
and the TGA may be used to characterize the materials used in UL 217. 40

Smoke Characterization Project Final Report P. 147 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
SUMMARY OF FINDINGS
The findings from this research investigation are presented herein.

Gas Analysis and Smoke Characterization Measurement
1. Physical Smoke Particle Characterization - The particle spectrometer provides data on 5
smoke particle size and count distribution over a size range of 0.01 to 10 microns
whereas traditional techniques to quantify smoke such as obscuration and ionization are
limited to 0.05 to 1 micron and 0.1 to 10 microns respectively.
2. Relationship of Smoke Particle Characterization to Traditional Methods - Linear
relationships between the smoke particle data and the traditional techniques were 10
demonstrated such that:
a. Particle size and number count are linearly related to MIC signal change:
MIC ~ d
m
n
m
(Eq. 12, Figure 7)
b. Number count is linearly related to scattering while particle size exhibits a second
order relationship:
2
i i
d n s

(Figure 110) 15
c. Number count is linearly related to obscuration while particle size exhibits a third
order relationship:
3
i i
d n
OD

l
(Eq. 3, Figure 6).
3. Smoke Particle Aggregation - Tests conducted in the UL 217 Sensitivity Test smoke box
and the UL 217/UL 268 Fire Test Room indicate an aggregation of smaller smoke
particles to form larger particles as evidenced by the increase in smoke particle 20
concentrations in conjunction with increasing fractions of larger smoke particles (Figure
5, Figure 115 Figure 120). This was more evident for non- flaming fires than flaming
fires.
4. Smoke Gas Effluent Composition - Gas effluent analysis showed the dominant gas
components were water vapor, carbon dioxide and carbon monoxide (Appendices C 25
through H).

Influence of Material Chemistry
1. Combustion Behavior: Synthetic and Natural Materials - Cone calorimeter tests indicate
synthetic materials (e.g. polyethylene, polyester, nylon, polyurethane) generate higher 30
heat (Figure 11) and smoke release rates (Figure 12) than the natural materials (e.g. wood,
cotton batting). This is anticipated to be primarily due to the modes of degradation and
chemical structure of synthetic versus natural materials.
2. Charring Effects - Materials exhibiting charring behavior such as wood alter the size and
amount of smoke particles generated as the combustion process progresses (Figure 15). 35
3. Influence on Smoke Particle Size - In general, the synthetic materials tested generated
larger mean smoke particle sizes than natural materials in flaming mode (Figure 13).

Mode of Combustion
1. Flaming Combustion - Flaming combustion tends to create smaller mean particle sizes 40
than non-flaming combustion (Figure 100). This is primarily due to the more efficient
conversion of high molecular weight polymers to low molecular weight combustion
products and ultimately CO, CO
2
and H
2
O instead of organic by-products and soot.
Smoke Characterization Project Final Report P. 148 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
2. Non-Flaming Combustion - Non-flaming combustion tends to generate more smoke for a
given consumed mass than flaming combustion (Figure 99).

Small-Scale and Intermediate-Scale Test
1. Cone Calorimeter Test - The cone calorimeter provided combustibility, smoke 5
characteristics and gas effluent data in flaming and non-flaming modes for a range of
materials studied. The smoke characterization data revealed the influences of material
chemistry, physical sample structure, and the mode of combustion. The data were found
to be repeatable. In the non- flaming mode, the heat and smoke release rates were lower
than the resolution of the cone calorimeter measurement system for several materials 10
investigated. However, the smoke particle spectrometer provided repeatable data on
smoke size and count distribution for both flaming and non- flaming modes.
2. Intermediate-Scale Test - The intermediate scale test provided a platform to scope
combustion scenarios, and provided data on the heat and smoke release rates as well as
smoke size and count distribution for test samples subsequently used in the UL 217/UL 15
268 Fire Test Room. The tests also identified test samples with heat and smoke
characteristics that varied from UL 217 fire test samples such as Douglas fir, newspaper,
heptane/toluene mixture, and Ponderosa pine. In the non- flaming mode, the method used
for heating the test sample was observed to influence the smoke characteristics. The
heating by a hot plate provided larger particle size as compared to radiant heating. 20

UL 217/UL 268 Fire Test Room Tests
1. Smoke Particle Size and Count Distribution - The tests provided smoke particle size and
count distribution data in conjunction with traditional obscuration and Measuring
Ionization Chamber data. PU foams in the flaming mode produced the smallest particle 25
sizes of all materials tested (Table 21).
2. Combustion Mode Effects - Changes in the combustion mode (flaming versus non-
flaming) resulted in different smoke particle size and count distributions that influenced
the response of photoelectric and ionization smoke alarms. The particle size distribution
for the non- flaming fires yielded larger mean smoke particle diameter than the flaming 30
mode fires. The ionization alarm responded quicker to flaming fires; the photoelectric
responded quicker to non- flaming fires (Table 30).
3. Smoke Alarm Response to Flaming Fires - In all but one flaming test the ion alarm
activated first (Table 20, Table 30). Both alarm types activated within the 4 minute time
limit specified in UL 217 for the three UL 217 flaming test targets (Douglas fir, 35
heptane/toluene mixture, and newspaper). In one of two flaming tests involving PU foam
with cotton/poly fabric the photoelectric smoke alarm did not activate, however the
ionization alarm did activate in both tests. In a flaming PU foam with cotton/poly fabric
test using a smaller sample size neither alarm type activated. It should be noted that the
maximum obscuration in these PU foam tests was less than for Douglas fir, 40
heptane/toluene mixture, and newspaper test samples.
4. Smoke Alarm Response to Non-Flaming Fires - The photoelectric alarm activated first in
the non-flaming tests with the exception of the higher energy bread/toaster test in which
the ion alarm activated first (Table 25, Table 30). The UL 217 smoldering Ponderosa pine
test triggered both the ionization and photoelectric smoke alarms. For many of the other 45
materials, the ionization smoke alarm did not trigger. In each of these cases, the
Smoke Characterization Project Final Report P. 149 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
obscuration value was less than the 10 %/ft limit specified in UL 217. It was also found
that there was settling of the smoke particles in the test room over time. Measurements
from several non- flaming tests showed that the obscuration values at the ceiling dropped
over time, and the maximum obscuration values were observed at the 2 feet measurement
location below the ceiling. 5
5. Smoke Stratification - Non- flaming fires result in changes in the smoke build up over
time, such that stratification of smoke below the ceiling occurs. This time-dependent
phenomenon results in less obscuration at the ceiling than below the ceiling (Figure 85 to
Figure 88). This caused both detection technologies to drift out of alarm.
10

Smoke Characterization Project Final Report P. 150 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
APPENDIX A: Material Chemistry

Table A1 Chemistry of Natural Materials
Material or
Substance Type
Reference
Code
Chemistry
Bread N1 Composed primarily of starch, sugar, fats and oils.
Butter N2
Composed largely of glycerides of oleic (C
18
unsaturated),
stearic (C
18
saturated) and palmitic (C
16
saturated) acids.
Elemental composition C, H, O.
Carbohydrates N3
A compound of carbon, hydrogen and oxygen that contains the
saccharose group (R-CHOH-CO-R). It is the building block
for essentially all natural products.
Cotton N4
Staple fiber consisting primarily of cellulose (88-96%) with
other natural-derived aliphatic organic compounds (C, H, O).
Cellulose is a natural carbohydrate polymer (polysaccharide)
consisting of anhydroglucose units joined by an oxygen linkage
to form essentially linear high molecular weight chains.
Cellulose N5
A natural carbohydrate consisting of anhydroglucose units joined
by oxygen linkages to form long, high molecular chains that are
essentially linear. Elemental composition C, H, O; polymer
structure aliphatic
Glycerides N6
An ester of glycerol and fatty acids in which one or more of the
hydroxyl groups of the glycerol have been replaced with acid
radicals. Mono and triglycerides are commonly found in food
and cosmetic products and other compounded products.
Linen N7 Thread and fabric made from the fibers of the flax plant.
Paper N8
A processed product of cellulosic fibers primarily made from
softwoods.
Silk N9
A natural fiber secreted as a continuous filament by the
silkworm. Silk consists essentially of a the protein fibroin and,
in the raw state, is coated with a gum, which is usually removed
before spinning.
Starch N10
Anhydroglucose C
6
H
10
O
5
. This aliphatic ring compound with
hydroxyl groups (and its derivatives) is the common building
block for many of the products produced by natural processes
(photosynthesis).
Sugar N11
Carbohydrate product of photosynthesis and comprised by one,
two or more saccharose groups. Chief among the
monosaccharides are glucose (dextrose) and fructose (general
formula C
6
H
10
O
5
).
Triglyceride N12
Any naturally occurring ester of a normal fatty acid and glycerol.
Fatty acids are composed of a chain of alkyl groups (R-CH
2
-R)
containing 4 to 22 carbon atoms with a terminal carboxylic acid
(R-COOH)

Smoke Characterization Project Final Report P. 151 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Material or
Substance Type
Reference
Code
Chemistry
Vegetable Oil N13
Edible oils extracted from the seeds, fruit or leaves of plants.
Generally considered to be mixtures of glycerides (safflower,
sunflower, peanut, walnut, etc.).
Wool N14
Staple fibers from the fleece of sheep. Chemically, wool consists
essentially of protein chains (keratin) bound together by
disulfide cross- linkages. Elemental composition C, H, O, N, S;
polymer structure essentially aliphatic.
Wood N15
Wood is typically composed of 40-60% cellulose and 20-40%
lignin, together with gums, resins, variable amounts of water and
inorganic matter.


Table A2 Chemistry of Synthetic Materials
Material or
Polymer Type
Reference
Code
Chemistry, Structure and Related Information
ABS S1
An engineering thermoplastic copolymer composed of
acrylonitrile, butadiene and styrene monomers. ABS is often
used in appliance and enclosure housings. Elemental
composition - C, H, N; structure aliphatic and aromatic. See
Acrylonitrile, Butadiene, Polystyrene.
Acrylic S2
Generic term used for materials composed of acrylic acid (R-
CH
2
CHCOOH-R) or acrylic acid esters (R-CH
2
CHCOOR-R).
Acrylic fibers however, are prepared from acrylonitrile (see
Acrylonitrile). Acrylic resins are thermoplastic polymers or
copolymers of acrylic acid, methacrylic acid (R-C(CH
3
)-
CHCOOH-R), esters of these acids or acrylonitrile. Elemental
composition - C, H, O, and N (when acrylonitrile present),
polymer structure typically aliphatic.
Acrylonitrile S3
Commonly referred to as vinyl cyanide or propenenitrile
(CH
2
=CHCN). As a monomer, acrylonitrile is often used to
modify other plastics such as: ABS, acrylic or modacrylic fibers,
nitrile rubbers or cotton fibers. Elemental composition C, N;
polymer structure - aliphatic
Butadiene S4
As with acrylonitrile, butadiene (CH
2
=CHCH=CH
2
) is a
monomer that can be polymerized into polybutadiene or modify
other polymers through copolymerization, such as ABS and
nitrile elastomers. Elemental composition C, H; polymer
structure typically aliphatic
Heptane S5 Linear hydrocarbon chain of 7 carbons - aliphatic
Noryl S6
Engineering thermoplastic sold by of General Electric. Noryl is
an engineering thermoplastic copolymer alloy of polyphenylene
oxide (PPO) and polystyrene (PS). Elemental composition C,
H, O; structure aromatic.

Smoke Characterization Project Final Report P. 152 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Material or
Polymer Type
Reference
Code
Chemistry, Structure and Related Information
Nylon S7
Generic name for a family of polyamide polymers characterized
by the presence of an amide group (R-CONH-R) where R can be
various hydrocarbon groups. As with polyesters, nylons are used
in various applications, such as textiles and structural housings.
The nylon properties are dictated by the various monomers used
in the polymerization and subsequent compounded fillers that
may be incorporated into the structure in post processing steps.
Typical aliphatic nylons for textile applications include Nylon 6
(formed from the homopolymerization of caprolactam and Nylon
6,6 with the copolymerization of adipic acid and hexamethylene
diamene. Aromatic nylons are often found in high strength and
high temperature fibers (Kevar, or Nomex), or engineering
thermoplastic housings.
Polyacrylates S8
Polymers produced by the homopolymerization or
copolymerization of acrylic acid or methacrylic acid on their
esters. Elemental composition C, H, O; polymer structure
aliphatic.
Polycarbonate
(PC)
S9
Engineering thermoplastic with unique impact and high
temperature properties. PC is often used in appliance and
enclosure housings and injection molded articles. PC is
produced by various companies; particularly one sold by General
Electric under the trade name Lexan. Polycarbonate is
produced by the polymerization of bisphenol A and phosgene.
Elemental composition C, H, O; structure aromatic.
Polyester S10
A generic term for commercially available textile and
thermoplastic products based upon ester polymers with the
characteristic linkage (R-COO-R) where R can be various
hydrocarbon groups. Ester polymers are produced by either the
condensation reaction of dicarboxylic acids with dihydroxy
alcohols or the reaction of lactones or hydroxyl-carboxylic acids.
Polyester textiles are usually composed of PET polyethylene
terephthalate. PET is formed by the reaction of terephthalic acid
(aromatic compound) and ethylene glycol (aliphatic compound).
Another common polyester in this class is PBT, where ethylene
glycol is replaced with butane diol. Thermoplastic polyesters are
also found in appliance housings. These polymers use modified
acids and alcohols with fillers incorporated and possible
crosslinking agents for specific property modification (modulus,
impact, temperature resistance, etc.). Elemental composition C,
H, O; structure either aliphatic or aromatic.




Smoke Characterization Project Final Report P. 153 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Material or
Polymer Type
Reference
Code
Chemistry, Structure and Related Information
Polyethylene and
copolymers (PE)
S11
Polymers based on the polymerization of ethylene (CH
2
=CH
2
)
and other unsaturated monomers. PE polymers and copolymers
can take many forms due to factors, such as cross- link density,
molecular weight, degree of branching, incorporation of co-
monomers, etc. Elemental composition essentially C, H
depending upon type and percentage of co- monomers; structure
aliphatic.
Polyolefin S12
A class or group of thermoplastic polymers (or copolymers)
derived from simple olefins; such as ethylene, propylene, butane,
and isoprene. Essentially these polymers only contain
hydrocarbon monomers (C, H) without any oxygen in the
polymer structure.
Polyphenylene
oxide (PPO)
S13
Engineering thermoplastic polymer with exceptional dielectric
and high temperature properties. Produced by the oxidative
polymerization of 2, 6-dimethyl phenol. Elemental composition
C, H, O; structure aromatic.
Polypropylene
and copolymers
(PP)
S14
Polymers based on the polymerization of propylene
(CH
2
=CHCH
3
) and other unsaturated monomers. PP polymers
and copolymers can take many forms due to factors, such as
cross- link density, molecular weight, degree of branching,
incorporation of co-monomers, etc. Elemental composition
essentially C, H depending upon type and percentage of co-
monomers; structure aliphatic.
Polyurethane
(PU)
S15
A broad class of thermoplastic or thermosetting polymers based
upon the urethane linkage (R-NH-COOR-R). Polyurethanes are
produced by the condensation reaction of a polyisocyanates and
hydroxyl-containing materials. The range of properties and
physical appearance (morphology) is dictated by the isocyanate
and hydroxyl precursors. Depending upon the reactive materials
used, polyurethanes can be flexible foams, coatings, elastomers
and/or moldable resins (see below). Elemental composition C,
H, O, N; structure primarily aromatic.
Polyurethane,
flexible
S16
Flexible PU foams are produced by the reaction of toluene
diisocyanate and polyhydroxy materials in the presence of
blowing agents and catalyst. The polyhydroxy compounds are
often referred to as polyols, which are low molecular weight
aliphatic compounds with ether (R-C-O-R) or ester (R-
COOR-R) linkages. Polyurethane foams (unless flame
retarded) are lightly cross- linked and readily decomposed by
heat or open flame resulting in liquefaction, polymer chain
scission and release of low molecular weight fragments. The
sensitivity of flexible PU foams to degradation is dictated by the
physical structure (thin-wall, open cells) and chemical structure
(aromatic, ether and/or ester content).
Smoke Characterization Project Final Report P. 154 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Material or
Polymer Type
Reference
Code
Chemistry, Structure and Related Information
Polyurethane,
rigid
S17
In contrast to flexible PU foams, rigid PU foams have a high
cross- link density. Crosslinking is achieved by the ratio of co-
monomers and reactive group functionality. One example of
rigid foam is produced by MDI (diphenyl methane diisocyanate),
water, catalyst and blowing agents. Water readily reacts with
isocyanates to form amine groups, which further react to form
urea linkages (R-NH-CO-NH-R) in the polymer structure. Rigid
foams typically have a close-cell structure and more resistant to
degradation (liquefaction) due to the high cross- link density.
Polystyrene (PS) S18
PS is formed by the free radical reaction of styrene monomer
(vinyl benzene) in the presence of catalysts. Depending upon the
reaction conditions, PS can take the form of a transparent, hard
solid or cellular expanded foam structure. PS is sensitive to UV
degradation and solvents and is combustible and non self-
extinguishing. Elemental composition C, H; structure
aromatic.
Polyvinyl
chloride (PVC)
S19
PVC is produced by the polymerization of vinyl chloride
(CH
2
=CHCl). Once polymerized, PVC has the appearance of a
white powder or granular salt. PVC has a huge range of
properties due to its ability to incorporate plasticizers, fillers
and ability to be expanded with blowing agents (see below).
PVC has excellent resistance to UV degradation, is combustible,
but self-extinguishing. Elemental composition C, Cl; structure
aliphatic or aromatic depending upon modification.
PVC, flexible S20
Flexible PVC is produced by the incorporation of 20-60% w/w
aromatic or aliphatic ester plasticizers in the PVC powder. This
plasticization produces materials with exceptional elastomeric
properties, toughness and weatherability. Typical aromatic
plasticizers are based upon terephthalic acid (di-carboxylic acid)
or trimellitic acid (tri-carboxylic acid). Alcohols used in these
plasticizers usually contain from 8 to 16 carbon atoms.
Elemental composition C, H, O; structure aromatic or
aliphatic depending upon modification. Typical applications are
for electrical insulation, tubing, coatings, gaskets, etc.
PVC, rigid S21
Rigid PVC differs from flexible PVC products by the ingredients
compounded into the PVC resin. Rigid PVC has high
percentages of inorganic fillers and additives and can be
expanded with the use of blowing agents. Rigid PVC is widely
used as pipe, gutters, siding and in many structural applications.
Polyvinylidine
chloride (PVDC)
S22
Polyvinylidine chloride is produced by the polymerization of
vinylidine chloride (CH=CCl
2
) or with or lesser amounts of
unsaturated compounds. PVDC is used in numerous packaging
film products and commonly known under the trade name
Saran.
Smoke Characterization Project Final Report P. 155 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Material or
Polymer Type
Reference
Code
Chemistry, Structure and Related Information
Rayon S23
Generic name for a manufactured fiber composed of regenerated
cellulose in which >15% of hydroxyl substituents have been
replaced by chemical modification (for example by acetate
groups). The fiber ignites and burns readily. Chemical
composition C, H, O; structure - aliphatic
Toluene S24
Toluene (methyl benzene) is a 7-carbon aromatic hydrocarbon liquid
composed of a 6-membered aromatic ring (benzene C
6
H
6
)

with an
attached methyl (-CH
3
) group. Toluene is a main ingredient in paint
thinner.
Wax (candle) S25
A low melting organic mixture or compound composed of
hydrocarbons, esters or fatty acids or alcohols. Candle waxes
typically contain aliphatic hydrocarbons that readily melt and
burn when ignited.

Smoke Characterization Project Final Report P. 156 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
APPENDIX B: Test Sample Documentation and Characterization
PU Foam: FTIR (top) and TGA (bottom)


5
Smoke Characterization Project Final Report P. 157 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Cotton Batting: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 158 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Cotton Sheet: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 159 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Cotton/Polyester Sheet: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 160 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Polyester Microfiber Sheet: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 161 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Pillow Stuffing: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 162 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Rayon Sheet: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 163 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Nylon Carpet: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 164 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Polyester Carpet: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 165 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Polyisocyanurate Foam: FTIR (top) and TGA (bottom)



5
Smoke Characterization Project Final Report P. 166 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
HDPE: FTIR (top) and TGA (bottom)


-1
0
1
2
3
D
e
r
i
v
.

W
e
i
g
h
t

(
%
/

C
)
-20
0
20
40
60
80
100
120
W
e
i
g
h
t

(
%
)
0 200 400 600 800 1000
Temperature (C)
Sample: HDPE
Size: 5.2090 mg
Method: Q500 TGA 40-850C;20C/MIN
Comment: Smoke detector project - HDPE
TGA
File: HDPE.001
Operator: TF,DA,100
Run Date: 2006-03-29 12:40
Instrument: TGA Q500 V6.3 Build 189
Universal V4.1D TA Instruments

5
Smoke Characterization Project Final Report P. 167 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Polypropylene: FTIR (top) and TGA (bottom)


-0.5
0.0
0.5
1.0
1.5
2.0
2.5
D
e
r
i
v
.

W
e
i
g
h
t

(
%
/

C
)
0
20
40
60
80
100
120
W
e
i
g
h
t

(
%
)
0 200 400 600 800 1000
Temperature (C)
Sample: Polypro
Size: 5.2000 mg
Method: Q500 TGA 40-850C;20C/MIN
Comment: Smoke detector project - polypro
TGA
File: polypro.001
Operator: TF,DA,,100
Run Date: 2006-03-29 11:24
Instrument: TGA Q500 V6.3 Build 189
Universal V4.1D TA Instruments

5
Smoke Characterization Project Final Report P. 168 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.
Coffee Maker: FTIR


Smoke Characterization Project Final Report P. 169 of 169
This Report cannot be modified or reproduced, in part, without the prior written permission of Underwriters Laboratories Inc.
Copyright 2007 Underwriters Laboratories Inc.


REFERENCES
1
R. W. Bukowski, W. J. Christian, and T.E. Waterman, Detector Sensitivity and Siting Requirements for
Dwellings. Final Technical Report, IITRI Project J6340, Underwriters Laboratories Inc. File USNC-62, Project
74NK6752 (August 1975). Prepared for US Bureau of Standards, Center for Fire Research.
2
S. W. Harpe, T.E. Waterman, and W. J. Christian, Detector Sensitivity and Siting Requirements for Dwellings -
Phase 2. Final Report, IITRI Project J6340, Underwriters Laboratories Inc File USNC-62, Project 75NK7701
(July 1976). Prepared for US Bureau of Standards, Center for Fire Research, Washington, D.C.
3
UL 217 - Single and Multiple Station Smoke Alarms, Underwriters Laboratories Inc., 333 Pfingsten Road,
Northbrook, IL, 60062.
4
NFPA 72 National Fire Alarm Code, National Fire Protection Association, One Batterymarch Park, Quincy,
MA, 02169.
5
M. Ahrens, U.S. Experience with Smoke Alarms , NFPA Fire Analysis & Research Division, One
Batterymarch Park, Quincy, MA, (November 2004).
6
Smoke Alarm Performance in Residential Structure Fires, U. S. Fire Administration Topical Fire Research
Series, Vol. 1, Issue 15, (March 2001).
7
R.W. Bukowski, et. al, Performance of Home Smoke Alarms. Analysis of Response of Several Available
Technologies in Residential Fires, NIST Technical Note 1455, National Institute of Standards and Technology,
Technology Administration, U.S. Department of Commerce, Washington DC, 20402 (2004).
8
G.G. Hawley, The Condensed Chemical Dictionary, 8th Edition, Van Nostrand Reinhold Company (1971).
G. Odian, Principles of Polymerization, McGraw Hill Company (1970).
F. Billmeyer, Jr., Textbook of Polymer Science, Wiley-Interscience (1970).
J.B. Hendricks, D.J. Cram and G.S. Hammond, Organic Chemistry, McGraw-Hill Book Company (1970).
9
C.D. Litton, K.R. Smith, R. Edwards, T. Allen, Combined Optical and Ionization Techniques for Inexpensive
Characterization of Micrometer and Submicrometer Aerosols , J. Sci. and Tech., 38 1054 (2004).
10
V. Babrauskas and G. Mulholland, Smoke and Soot Determinations in the Cone Calorimeter, Mathematical
Modeling of Fires, American Society for Testing and Materials (1987).
11
G. Mulholland, ICFRE Conference, Chicago, IL (October 1999).
12
T.J. Ohlemiller, Smoldering Combustion, SFPE Handbook of Fire Protection Engineering, 2
nd
Edition, pp. 2-
171, Society of Fire, Boston, Massachusetts (1995).