NIST Technical Note 1889v3

CFAST – Consolidated Fire

And Smoke Transport
(Version 7)
Volume 3: Verification and Validation
Guide
Richard D. Peacock
Glenn P. Forney
Paul A. Reneke

This publication is available free of charge from:

[Link]
 NIST Technical Note 1889v3

CFAST – Consolidated Fire

And Smoke Transport
(Version 7)
Volume 3: Verification and Validation
Guide
Richard D. Peacock
Glenn P. Forney
Paul A. Reneke

This publication is available free of charge from:

[Link]

August 2021
CFAST Version 7.7.0
GIT Revision: CFAST7.7.1-2-g0b43aa83

E N T OF C O M
TM M
AR

ER
D EP

CE
ICA
UN
IT

ER

D
E

ST AM
ATES OF

U.S. Department of Commerce

Gina M. Raimondo, Secretary

National Institute of Standards and Technology

James K. Olthoff, Performing the Non-Exclusive Functions and Duties of the Under Secretary of Commerce
for Standards and Technology & Director, National Institute of Standards and Technology
CFAST Developers

The Consolidated Model of Fire and Smoke Transport (CFAST) and Smokeview are the products
of a collaborative effort led by the National Institute of Standards and Technology (NIST). Its
developers and contributors are listed below.

Principal Developers of CFAST

Richard Peacock, NIST

Glenn Forney, NIST
Paul Reneke, NIST
Kevin McGrattan, NIST
Walter Jones

Principal Developer of Smokeview

Glenn Forney, NIST

Contributors

Kristopher Overholt, Continuum Analytics, Austin, Texas, USA

Jason Floyd, Jensen Hughes, Baltimore, Maryland, USA

iii
iv
About the Developers

Richard Peacock is a chemical engineering in the Fire Research Division of NIST. He received a
bachelor of science degree from the Clark School of Engineering of the University of Maryland
in 1973. He joined the NIST staff in 1974 (then the National Bureau of Standards) and has
worked on real-scale testing and the development and validation of fire models, most notably
CFAST.

Glenn Forney is a computer scientist in the Fire Research Division of NIST. He received a bach-
elor of science degree in mathematics from Salisbury State College and a master of science
and a doctorate in mathematics from Clemson University. He joined NIST in 1986 (then the
National Bureau of Standards) and has since worked on developing tools that provide a better
understanding of fire phenomena, most notably Smokeview, an advanced scientific software
tool for visualizing Fire Dynamics Simulation data.

Paul Reneke is a computer scientist in the Fire Research Division of NIST. He received a bachelor
of science degree in mathematical sciences from Clemson Univerity and a master of science
degree in applied mathematics from The Johns Hopkins University. He joined NIST in 1990.
He has worked on the development of user interfaces, graphics and improved numerics in fire
models, notably CFAST. His research interests include sensitivity analysis and validation of
fire models.

Kevin McGrattan is a mathematician in the Fire Research Division of NIST. He received a bach-
elor of science degree from the School of Engineering and Applied Science of Columbia Uni-
versity in 1987 and a doctorate at the Courant Institute of New York University in 1991. He
joined the NIST staff in 1992 and has since worked on the development and validation of fire
models, most notably the Fire Dynamics Simulator.

Walter Jones was a physicist at NIST (now retired). He received a bachelor of arts degree in
physics from Oberlin College and a doctorate in physics from the University of Maryland. He
was the original developer of the CFAST model. In addition to the development of fire models,
he has worked on smart fire alarms and smoke control for naval vessels.

v
vi
Disclaimer

The U. S. Department of Commerce makes no warranty, expressed or implied, to users of CFAST

and associated computer programs, and accepts no responsibility for its use. Users of CFAST
assume sole responsibility under Federal law for determining the appropriateness of its use in any
particular application; for any conclusions drawn from the results of its use; and for any actions
taken or not taken as a result of analyses performed using these tools. CFAST is intended for use
only by those competent in the field of fire safety and is intended only to supplement the informed
judgment of a qualified user. The software package is a computer model which may or may not
have predictive value when applied to a specific set of factual circumstances. Lack of accurate
predictions by the model could lead to erroneous conclusions with regard to fire safety. All results
should be evaluated by an informed user.

National Institute of Standards and Technology Technical Note 1889v3

Natl. Inst. Stand. Technol. Tech. Note 1889v3, 384 pages (August 2021)
CODEN: NTNOEF

This publication is available free of charge from:

[Link]

vii
viii
Intent and Use

The algorithms, procedures, and computer programs described in this report constitute a method-
ology for predicting some of the consequences resulting from a prescribed fire. They have been
compiled from the best knowledge and understanding currently available, but have important lim-
itations that must be understood and considered by the user. The program is intended for use by
persons competent in the field of fire safety and with some familiarity with personal computers. It
is intended as an aid in the fire safety decision-making process.

ix
x
Abstract

This volume of the CFAST Technical Reference Guide provides details of the verification and val-
idation process. It is based in part on the “Standard Guide for Evaluating the Predictive Capability
of Deterministic Fire Models,” ASTM E1355 [1]. ASTM E 1355 defines model evaluation as “the
process of quantifying the accuracy of chosen results from a model when applied for a specific
use.” The model evaluation process consists of two main components: verification and validation.
Verification is a process to check the correctness of the solution of the governing equations. Ver-
ification does not imply that the governing equations are appropriate; only that the equations are
being solved correctly. Validation is a process to determine the appropriateness of the governing
equations as a mathematical model of the physical phenomena of interest. Typically, validation
involves comparing model results with experimental measurement. Differences that cannot be ex-
plained in terms of numerical errors in the model or uncertainty in the measurements are attributed
to the assumptions and simplifications of the physical model.

xi
xii
Acknowledgments

The following individuals and organizations played a role in the validation process of CFAST.

• Continuing support for CFAST is via internal funding at the National Institute of Standards and
Technology (NIST).

• In addition, support is provided by other agencies of the U.S. Federal Government, most no-
tably the Nuclear Regulatory Commission (NRC) Office of Research and the U.S. Department
of Energy. The US Nuclear Regulatory Commission Office of Research has funded key val-
idation experiments, the preparation of the CFAST manuals, and the development of various
sub-models that are of importance in the area of nuclear power plant safety. Special thanks
to Mark Salley, David Stroup, and Jason Dreisbach for their efforts and support. Support to
refine the software development and quality assurance process for CFAST has been provided
by the U.S. Department of Energy (DOE). The assistance of Subir Sen and Debra Sparkman in
understanding DOE software quality assurance programs and the application of the process to
CFAST is gratefully acknowledged. Thanks are also due to Allan Coutts, Washington Safety
Management Solutions for his insight into the application of fire models to nuclear safety ap-
plications and detailed review of the CFAST document updates for DOE.

• Anthony Hamins of NIST directed the NIST/NRC and WTC experiments, and quantified the
experimental uncertainty of these and other experiments used in this study. Alex Maranghides
was the Director of the Large Fire Laboratory at NIST at the time these tests were conducted,
and he helped to design the experiments. Therese McAllister oversaw the instrumentation of
the structural steel during the WTC experiments.

• Anthony Hamins developed the technique of evaluating experimental uncertainty that is used
throughout this Guide.

• Kevin McGrattan and Randy McDermott of the Fire Research Division at NIST developed the
automated process for validation and verification testing of the Fire Dynamics Simulator (FDS)
model. We use this same process for validation and verification testing of CFAST.

• Kevin McGrattan of the Fire Research Division and Blaza Toman of the Statistical Engineering
Division of NIST developed the method of quantifying the model uncertainty.

• For the NIST/NRC tests, Alex Maranghides was in charge of the Large Fire Laboratory at
NIST where these tests were conducted, and helped to design the experiments. Thanks also
to technicians Laurean Delauter, Jay McElroy, and Jack Lee who constructed and conducted
these validation experiments at NIST.

xiii
• Steve Nowlen, formerly at Sandia National Laboratories provided data, documentation and
further information for the NRC-sponsored experiments referred to in this document and the
"FM/SNL Test Series" (Factory Mutual and Sandia National Laboratories conducted these ex-
periments).

• Thanks to VTT, Finland, for their contribution of experimental data, referred to in this docu-
ment as the “VTT Large Hall Experiments.”

• Bryan Klein, currently employed at Thunderhead Engineering, Inc., assisted in the develop-
ment of techniques to automatically generate the plots that are found throughout this Guide.

• David Sheppard, currently of the Bureau of Alcohol, Tobacco and Firearms (ATF), conducted
the experiments referred to as the “UL/NFPRF Test Series” on behalf of the Fire Protection
Research Foundation (then known as the National Fire Protection Research Foundation) while
working at Underwriters Labs in Northbrook, Illinois. Sheppard, along with Bryan Klein,
currently employed at Thunderhead Engineering, Inc., conducted the experiments referred to
as the “ATF Corridors” series in 2008.

• Ken Steckler provided details about the “Steckler Compartment Experiments” of 1979.

• Michael Spearpoint and masters degree student Rob Fleury of the University of Canterbury,
New Zealand, supplied heat flux measurements from propane burner fires (Fleury Heat Flux).

• Taylor Myers, a student at the University of Maryland and a Summer Undergraduate Research
Fellow (SURF) at NIST, analyzed the Vettori Flat sprinkler experiments. Thanks also to Bob
Vettori of the U.S. Nuclear Regulatory Commission and formerly of NIST for his help in
locating the original test data and laboratory notebooks. Taylor Myers also analyzed the NIST
Home Smoke Alarm (Dunes 2000) experiments and the WTC Spray Burner experiments using
CFAST.

• Barbara Hall, a student at the Worcester Polytechnic Institute and a summer intern at NIST,
analyzed the Lawrence Livermore National Laboratory experiments using CFAST.

• Tony Feric, a student at the University of Maryland and a summer intern at NIST, developed
many of the verification tests included in this guide.

• Brianna Gillespie, a student at the Worcester Polytechnic Institute and a summer intern at NIST,
also contributed to the verification tests included in this guide.

xiv
Contents

CFAST Developers iii

About the Developers v

Disclaimer vii

Intent and Use ix

Abstract xi

Acknowledgments xiii

1 Overview 1
1.1 What is Model Validation? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Blind, Specified, and Open Validation Experiments . . . . . . . . . . . . . . . . . 1
1.3 How to Use this Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Survey of Past Verification and Validation Work 5

2.1 Model Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Comparisons of CFAST with Full-Scale Experiments . . . . . . . . . . . . . . . 6
2.2.1 Fire Plumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.2 Multiple Compartments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.3 Large Compartments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.4 Mechanical Ventilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.5 Sprinkler Activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.6 Flashover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 Comparisons of CFAST with Actual Fires . . . . . . . . . . . . . . . . . . . . . 12
2.4 Comparisons of CFAST for Special Applications . . . . . . . . . . . . . . . . . . 12
2.4.1 Nuclear Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4.2 Small Compartments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4.3 Railway and Vehicle Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4.4 Non-Uniform Compartments . . . . . . . . . . . . . . . . . . . . . . . . . 14

3 Verification 15
3.1 Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.1.1 Temperature Equilibrium via Heat Conduction . . . . . . . . . . . . . . . . 16
3.1.2 Temperature Equilibrium via a Window . . . . . . . . . . . . . . . . . . . 17

xv
 3.1.3 Temperature Equilibrium via a Window at a High Elevation . . . . . . . . . 17
3.1.4 A Fire in a Single, Sealed Compartment with a Single Zone . . . . . . . . . 18
3.1.5 A Fire in a Single, Sealed Compartment with Two Zones . . . . . . . . . . 19
3.2 Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.1 A Fire in a Single, Sealed Compartment . . . . . . . . . . . . . . . . . . . 20
3.2.2 A Fire in a Compartment Connected to Another via a Door . . . . . . . . . 21
3.2.3 A Fire in a Compartment Connected to Another via a Ceiling Vent . . . . . 21
3.2.4 A Fire in a Four Compartment Assembly . . . . . . . . . . . . . . . . . . . 22
3.2.5 Product Yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.6 Trace Species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Ventilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.3.1 Air Flow Through a Single Compartment . . . . . . . . . . . . . . . . . . 25
3.3.2 Steady-State Air Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3.3 Non Steady-State Air Flow . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3.4 Velocity Profile of Air Flowing Through a Doorway . . . . . . . . . . . . . 27
3.3.5 Flow Through Different Vent Types . . . . . . . . . . . . . . . . . . . . . 28
3.3.6 Surface Opening Fraction for Vents . . . . . . . . . . . . . . . . . . . . . . 30
3.3.7 Leakage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.4 Heat Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.4.1 Measuring the Temperature Change of a Thermally Thin Target . . . . . . . 32
3.4.2 Measuring the Temperature Change of a Cylindrical Target . . . . . . . . . 33
3.4.3 Measuring Incident Heat Flux to Targets . . . . . . . . . . . . . . . . . . . 34
3.4.4 Total Emissivity and Hemispherical Absorptivitiy . . . . . . . . . . . . . . 35
3.4.5 Adiabatic Target Temperature . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.4.6 Fire Heat Flux to Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.5 Sprinklers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.5.1 Utilizing a Sprinkler to Detect and Suppress a Fire . . . . . . . . . . . . . . 37
3.6 Fire Ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.7 Tenability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.7.1 Gas Tenability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.7.2 Heat Tenability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.8 Summary of Verification Results . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4 Description of Validation Experiments 43

4.1 ATF Corridors Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.2 DelCo Trainer Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.3 Fleury Heat Flux Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.4 FM Four Room Including Corridor Test Series . . . . . . . . . . . . . . . . . . . 47
4.5 FM / SNL Test Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.6 iBMB Compartment Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.7 LLNL Enclosure Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.8 NBS Single Room Tests with Furniture . . . . . . . . . . . . . . . . . . . . . . . 57
4.9 NBS Multi-Compartment Test Series . . . . . . . . . . . . . . . . . . . . . . . . 59
4.10 NIST / NRC Compartment Experiments . . . . . . . . . . . . . . . . . . . . . . 62
4.11 NIST/NRC Corner, Wall, and Cabinet Experiments . . . . . . . . . . . . . . . . . 65

xvi
 4.11.1 Wall and Corner Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.11.2 Cabinet Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.12 NIST Seven-story Hotel Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.13 NIST Smoke Alarm Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.14 NIST Vent Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.15 PRISME Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.16 SP Adiabatic Surface Temperature Experiments . . . . . . . . . . . . . . . . . . 77
4.17 Steckler Compartment Experiments . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.18 UL/NFPRF Sprinkler, Vent, and Draft Curtain Study . . . . . . . . . . . . . . . . 82
4.19 UL/NIJ House Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.20 UL/NIST Vent Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.21 USN High Bay Hangar Experiments . . . . . . . . . . . . . . . . . . . . . . . . 92
4.22 Vettori Flat Ceiling Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.23 VTT Large Hall Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.24 WTC Spray Burner Test Series . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
4.25 Summary of Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

5 Hot Gas Layer Temperature and Depth 103

5.1 Summary of Hot Gas Layer Temperature and Height . . . . . . . . . . . . . . . . 104

6 Fire Plumes, Ceiling Jets, and Device Activation 109

6.1 Flame Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.2 Plume Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.3 Ceiling Jets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
6.4 Device Activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

7 Gas Species and Smoke 115

7.1 Oxygen, CO2 , and CO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
7.2 Smoke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

8 Pressure 119

9 Surface Temperature 121

9.1 Compartment Ceiling, Wall, and Floor Temperature . . . . . . . . . . . . . . . . 121
9.2 Target Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

10 Surface Heat Flux 127

10.1 Heat Flux to Compartment Ceiling, Wall, and Floor Surfaces . . . . . . . . . . . 127
10.2 Heat Flux to Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

11 Summary and Conclusions 133

11.1 Summary of CFAST Model Uncertainty Statistics . . . . . . . . . . . . . . . . . 133
11.2 Normality Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

References 139

xvii
Appendices 147

A Calculation of Layer Height and the Average Upper and Lower Layer Temperatures 149

B Summary of Verification Results 151

C Transient Model / Experiment Comparison Graphs 153

C.1 ATF Corridors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
C.2 DelCo Trainers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
C.3 Fleury Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
C.4 FM Four Room Including Corridor Test Series . . . . . . . . . . . . . . . . . . . 176
C.5 FM/SNL Test Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
C.6 iBMB Compartment Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
C.7 LLNL Enclosure Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
C.8 NBS Single Room Tests with Furniture . . . . . . . . . . . . . . . . . . . . . . . 210
C.9 NBS Multi-Compartment Test Series . . . . . . . . . . . . . . . . . . . . . . . . 214
C.10 NIST/NRC Test Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
C.11 NIST/NRC Corner and Wall Effects Experiments . . . . . . . . . . . . . . . . . . 283
C.12 NIST Smoke Alarm Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 295
C.13 NIST Seven-story Hotel Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
C.14 NIST Vent Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
C.15 PRISME DOOR Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
C.16 SP Adiabatic Surface Temperature Experiments . . . . . . . . . . . . . . . . . . 332
C.17 Steckler Compartment Experiments . . . . . . . . . . . . . . . . . . . . . . . . . 336
C.18 UL/NFPRF Series I Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 344
C.19 UL/NIJ House Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
C.20 UL/NIST Vent Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
C.21 USN High Bay Hangar Experiments . . . . . . . . . . . . . . . . . . . . . . . . 354
C.22 Vettori Flat Ceiling Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
C.23 VTT Large Hall Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
C.24 WTC Test Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368

xviii
List of Figures

2.1 Comparison of correlations, CFAST predictions, and experimental data for the
prediction of flashover in a compartment fire. . . . . . . . . . . . . . . . . . . . . 11

3.1 Results of the test case basic_tempequilib.in . . . . . . . . . . . . . . . . . 16

3.2 Results of the test case basic_tempequilib_window.in . . . . . . . . . . . . 17
3.3 Results of the test case basic_tempequilib_window_elevation.in . . . . 17
3.4 Results of the test case sealed_test.in . . . . . . . . . . . . . . . . . . . . . 19
3.5 Results of the test case sealed_test_2_layers.in . . . . . . . . . . . . . . . 19
3.6 Results of the test case species_mass_1.in . . . . . . . . . . . . . . . . . . . 21
3.7 Results of the test case species_mass_1.in . . . . . . . . . . . . . . . . . . . 21
3.8 Results of the test case species_mass_3.in . . . . . . . . . . . . . . . . . . . 22
3.9 Results of the test case species_mass_4.in . . . . . . . . . . . . . . . . . . . 22
3.10 Results of the test case species_mass_4.in . . . . . . . . . . . . . . . . . . . 23
3.11 Results of the test case methane_flame_simple.in . . . . . . . . . . . . . . . 24
3.12 Results of the test case species_test.in . . . . . . . . . . . . . . . . . . . . 24
3.13 Results of the test case Trace_Species_1.in . . . . . . . . . . . . . . . . . . 25
3.14 Results of the test case Trace_Species_2.in . . . . . . . . . . . . . . . . . . 25
3.15 Results of the test case ventilation_1.in . . . . . . . . . . . . . . . . . . . . 26
3.16 Results of the test case ventilation_2.in . . . . . . . . . . . . . . . . . . . . 26
3.17 Results of the test case ventilation_3.in . . . . . . . . . . . . . . . . . . . . 27
3.18 Results of the test case ventilation_4.in . . . . . . . . . . . . . . . . . . . . 28
3.19 Results of the test case VVent_Tests.in . . . . . . . . . . . . . . . . . . . . . 29
3.20 Geometry of the test case surface_opening_fraction_1.in . . . . . . . . . 30
3.21 Results of the test case surface_opening_fraction_1.in . . . . . . . . . . 31
3.22 Results of the test case Leakage_1.in . . . . . . . . . . . . . . . . . . . . . . . 32
3.23 Results of the test case radiation_1.in . . . . . . . . . . . . . . . . . . . . . 33
3.24 Results of the test case radiation_2.in . . . . . . . . . . . . . . . . . . . . . 34
3.25 Geometry for radiation_3.in . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.26 Results of mixture emission case radiation_3.in . . . . . . . . . . . . . . . . 35
3.27 Results of wall emission case radiation_3.in . . . . . . . . . . . . . . . . . . 35
3.28 Results of Tw = 1226.85 ◦ C case radiation_5.in . . . . . . . . . . . . . . . . 36
3.29 Results of the test case target_1.in . . . . . . . . . . . . . . . . . . . . . . . 37
3.30 Results of the test case target_2.in . . . . . . . . . . . . . . . . . . . . . . . 37
3.31 Results of the test case sprinkler_1.in . . . . . . . . . . . . . . . . . . . . . 38
3.32 Results of the test case Ignition_Test.in . . . . . . . . . . . . . . . . . . . . 39
3.33 Results of the test case Ignition_Test.in . . . . . . . . . . . . . . . . . . . . 39

xix
3.34 Results of the test case gas_tenabilityin . . . . . . . . . . . . . . . . . . . . 40
3.35 Results of the test case heat_tenabilityin . . . . . . . . . . . . . . . . . . . 40

4.1 Geometry of the ATF Corridors Experiments. . . . . . . . . . . . . . . . . . . . . 44

4.2 Instrumentation of the single level DelCo training structure . . . . . . . . . . . . 45
4.3 Instrumentation of the two level DelCo training structure . . . . . . . . . . . . . . 46
4.4 Overview of the Factory Mutual Four Room test series. . . . . . . . . . . . . . . 48
4.5 Geometry of the FM/SNL Experiments . . . . . . . . . . . . . . . . . . . . . . . 50
4.6 Detailed plan, side, and perspective schematic drawings of the iBMB pool fire
experimental arrangement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.7 Detailed plan, side, and perspective schematic drawings of the iBMB cable fire
experimental arrangement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.8 Geometry of the LLNL Enclosure Experiments . . . . . . . . . . . . . . . . . . . 56
4.9 Plan and elevation view schematic of experimental room for NBS single room
tests with furniture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.10 Burning specimen during NBS single room tests with furniture. . . . . . . . . . . 58
4.11 Photo of a 100 kW fire with the burner located against the rear wall of one of the
small compartments in the NBS Multi-Compartment test Series. . . . . . . . . . . 60
4.12 Plan, side and perspective schematic drawings of the NBS experimental arrange-
ment, including the burner. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.13 Photograph of a 1 MW heptane fire seen through the open doorway. Photo pro-
vided by Anthony Hamins, NIST. . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.14 Plan, side and perspective schematic drawings of the NIST NRC experimental
arrangement. The fuel pan and cables B, D, F, and G (dotted lines) are also shown. 64
4.15 Diagram of thermocouple layout for NIST/NRC Corner Effects experiments . . . 67
4.16 Large cabinet drawing, NIST/NRC Corner Effects Experiments . . . . . . . . . . 68
4.17 Medium-sized cabinet drawing, NIST/NRC Corner Effects Experiments . . . . . 69
4.18 Cabinet grill, NIST/NRC Corner Effects Experiments . . . . . . . . . . . . . . . 70
4.19 Overview of the NIST Seven-story hotel test series including smoke control. . . . 71
4.20 Geometry of the NIST Home Smoke Alarm Experiments . . . . . . . . . . . . . 73
4.21 Geometry of the compartment from the NIST Vent Study . . . . . . . . . . . . . 75
4.22 Geometry of the SP/AST compartment for Test 2. . . . . . . . . . . . . . . . . . 78
4.23 Geometry of the Steckler Compartment Experiments. . . . . . . . . . . . . . . . 81
4.24 Plan view of the UL/NFPRF Experiments, Series I. . . . . . . . . . . . . . . . . . 83
4.25 Plan view of the UL/NFPRF Experiments, Series II . . . . . . . . . . . . . . . . 84
4.26 Layout of the one story ranch-style house in the UL/NIJ experiments . . . . . . . 86
4.27 Geometry of UL NIJ Colonial-style House . . . . . . . . . . . . . . . . . . . . . 87
4.28 Geometry of the UL/NIST Experiments. . . . . . . . . . . . . . . . . . . . . . . 91
4.29 Geometry of the Vettori Flat Ceiling compartment . . . . . . . . . . . . . . . . . 93
4.30 Plan, side and perspective schematic drawings of the experimental arrangement
of the VTT large hall fire tests, including the fuel pan . . . . . . . . . . . . . . . 96
4.31 Photo of a 2 MW heptane fire during the VTT large hall tests. Photo provided by
Simo Hostikka, VTT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
4.32 Geometry of the WTC Experiments . . . . . . . . . . . . . . . . . . . . . . . . . 99

xx
5.1 Summary, HGL temperature, natural and forced ventilation . . . . . . . . . . . . 105
5.2 Summary, HGL temperature, unventilated compartments . . . . . . . . . . . . . . 106
5.3 Summary, HGL depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

6.1 Summary, Plume Centerline Temperature . . . . . . . . . . . . . . . . . . . . . . 110

6.2 Summary, Ceiling Jet Temperature . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.3 Summary, Smoke Detector Activation Time . . . . . . . . . . . . . . . . . . . . . 113
6.4 Summary, Sprinkler Activation Time. . . . . . . . . . . . . . . . . . . . . . . . . 114

7.1 Summary, Oxygen Concentration . . . . . . . . . . . . . . . . . . . . . . . . . . 116

7.2 Summary, Carbon Dioxide Concentration . . . . . . . . . . . . . . . . . . . . . . 116
7.3 Summary, Carbon Monoxide Concentration . . . . . . . . . . . . . . . . . . . . . 117
7.4 Summary, Smoke Concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

8.1 Summary, Compartment Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . 120

9.1 Summary, Compartment Surface Temperature . . . . . . . . . . . . . . . . . . . 124

9.2 Summary, Target Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

10.1 Summary, Compartment Heat Flux to Surfaces . . . . . . . . . . . . . . . . . . . 129

10.2 Summary, Heat Flux to Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

C.1 DelCo Trainers, HGL Temperature, Test 2 . . . . . . . . . . . . . . . . . . . . . 159

C.2 DelCo Trainers, HGL Temperature, Tests 3 and 4 . . . . . . . . . . . . . . . . . . 160
C.3 DelCo Trainers, HGL Temperature, Tests 5 and 6 . . . . . . . . . . . . . . . . . . 161
C.4 DelCo Trainers, HGL Temperature, Tests 22-25 . . . . . . . . . . . . . . . . . . 162
C.5 DelCo Trainers, HGL Height, Test 2 . . . . . . . . . . . . . . . . . . . . . . . . 163
C.6 DelCo Trainers, HGL Height, Tests 3 and 4 . . . . . . . . . . . . . . . . . . . . . 164
C.7 DelCo Trainers, HGL Height, Tests 5 and 6 . . . . . . . . . . . . . . . . . . . . . 165
C.8 DelCo Trainers, HGL Height, Tests 22-25 . . . . . . . . . . . . . . . . . . . . . . 166
C.9 DelCo Trainers, ceiling jet temperature, Tests 2-4 . . . . . . . . . . . . . . . . . . 167
C.10 DelCo Trainers, ceiling jet temperature, Tests 5 and 6 . . . . . . . . . . . . . . . 168
C.11 DelCo Trainers, ceiling jet temperature, Tests 22-25 . . . . . . . . . . . . . . . . 169
C.12 Fleury Heat Flux, 100 kW fires. . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
C.13 Fleury Heat Flux, 150 kW fires. . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
C.14 Fleury Heat Flux, 200 kW fires. . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
C.15 Fleury Heat Flux, 250 kW fires. . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
C.16 Fleury Heat Flux, 300 kW fires. . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
C.17 NIST/NRC Corner Effects, HGL temperature and height, corner experiments . . . 284
C.18 NIST/NRC Corner Effects, HGL temperature and height, wall experiments . . . . 285
C.19 NIST/NRC Corner Effects, HGL temperature and height, cabinet experiments 1-4 286
C.20 NIST/NRC Corner Effects, HGL temperature and height, cabinet experiments 5-8 287
C.21 NIST/NRC Corner Effects, HGL temperature and height, cabinet experiments 9-12 288
C.22 NIST/NRC Corner Effects experiments, plume temperature . . . . . . . . . . . . 289
C.23 NIST/NRC Corner Effects experiments, ceiling jet temperature . . . . . . . . . . 290
C.24 NIST/NRC Corner Effects experiments, ceiling jet temperature, large cabinet . . . 291

xxi
 C.25 NIST/NRC Corner Effects experiments, ceiling jet temperature, medium-sized
cabinet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
C.26 NIST/NRC Corner Effects, cabinet temperatures, large cabinet. . . . . . . . . . . 293
C.27 NIST/NRC Corner Effects, cabinet temperatures, medium cabinet. . . . . . . . . 294
C.28 UL/NIJ Experiments, HGL temperature and height, single-story ranch-style House
Tests 1, 2, and 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
C.29 UL/NIJ Experiments, HGL temperature and height, two-story colonial-style House
Tests 1, 4, and 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
C.30 UL/NIJ Experiments, ceiling jet temperature . . . . . . . . . . . . . . . . . . . . 350

List of Tables

4.1 Summary of FM/SNL Experiments . . . . . . . . . . . . . . . . . . . . . . . . . 51

4.2 Summary of LLNL Enclosure Experiments . . . . . . . . . . . . . . . . . . . . . 55
4.3 Summary of NIST/NRC Cabinet Experiments. . . . . . . . . . . . . . . . . . . . 70
4.4 Summary of NIST home smoke alarm experiments selected for model validation. . 74
4.5 Heights of the thermocouples above the floor of each level of the enclosure . . . . 76
4.6 Vent State by Experiment: Time Opened . . . . . . . . . . . . . . . . . . . . . . 77
4.7 Summary of Steckler compartment experiments. . . . . . . . . . . . . . . . . . . 80
4.8 Locations of measurement devices in the UL/NIJ ranch-style house . . . . . . . . 88
4.9 Locations of measurement devices in the UL/NIJ colonial-style house . . . . . . . 88
4.10 Sequence of events for the one story ranch-style house in the UL/NIJ experiments 89
4.11 Sequence of events for the two story colonial-style house in the UL/NIJ experiments 89
4.12 Summary of UL/NIST Vent experiments . . . . . . . . . . . . . . . . . . . . . . 90
4.13 Summary of important experimental parameters. . . . . . . . . . . . . . . . . . . 102

9.1 Wall measurement positions for the NIST/NRC series . . . . . . . . . . . . . . . 122

9.2 Ceiling surface measurement locations for the WTC series . . . . . . . . . . . . . 123

10.1 Heat flux gauge positions relative to the center of the fire pan in the WTC series. . 128

11.1 Summary statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

B.1 Summary of Verification Results . . . . . . . . . . . . . . . . . . . . . . . . . . 152

C.1 Ceiling surface measurement locations for the WTC series . . . . . . . . . . . . . 370

xxii
Chapter 1

Overview

1.1 What is Model Validation?

Key to ensuring the quality of the software is ongoing validation testing of the model. Validation
typically involves comparing model simulations with experimental measurements. To say that the
Consolidated Model of Fire and Smoke Transport (CFAST) is “validated” means that the model
has been shown to be of a given level of accuracy for a given range of parameters for a given
type of fire scenario. Although the CFAST developers periodically perform validation studies, it
is ultimately the end user of the model who decides if the model is adequate for the job at hand.
Thus, this Guide does not and cannot be considered comprehensive for every possible modeling
scenario.
Although there are various definitions of model validation, for example those contained in
ASTM E1355 [1], most define it as the process of determining how well the mathematical model
predicts the actual physical phenomena of interest. Validation typically involves (1) comparing
model predictions with experimental measurements, (2) quantifying the differences in light of un-
certainties in both the measurements and the model inputs, and (3) deciding if the model is appro-
priate for the given application. This Guide only does (1) and (2). Number (3) is the responsibility
of the model user.
The following sections discuss key issues that you must consider when deciding whether or not
CFAST has been validated. It depends on (a) the scenarios of interest, (b) the predicted quantities,
and (c) the desired level of accuracy. Keep in mind that CFAST can be used to model most any
compartment fire scenario and predict quantities of interest, but the prediction may not be accurate
because of limitations in the description of the fire physics, and also because of limited information
about the fuels, geometry, and so on.

1.2 Blind, Specified, and Open Validation Experiments

ASTM E1355 [1] describes three basic types of validation calculations – Blind, Specified, and
Open.

Blind Calculation: The model user is provided with a basic description of the scenario to be mod-
eled. For this application, the problem description is not exact; the model user is responsible

1
 for developing appropriate model inputs from the problem description, including additional
details of the geometry, material properties, and fire description, as appropriate. Additional
details necessary to simulate the scenario with a specific model are left to the judgement of the
model user. In addition to illustrating the comparability of models in actual end-use conditions,
this will test the ability of those who use the model to develop appropriate input data for the
models.
Specified Calculation: The model user is provided with a complete detailed description of model
inputs, including geometry, material properties, and fire description. As a follow-on to the
blind calculation, this test provides a more careful comparison of the underlying physics in the
models with a more completely specified scenario.
Open Calculation: The model user is provided with the most complete information about the sce-
nario, including geometry, material properties, fire description, and the results of experimental
tests or benchmark model runs which were used in the evaluation of the blind or specified cal-
culations of the scenario. Deficiencies in available input (used for the blind calculation) should
become most apparent with comparison of the open and blind calculation.
The calculations presented in this Guide all fall into the Open category. There are several reasons
for this, the first being the most practical:
• All of the calculations presented in this Guide are routinely rerun. The fact that the experiments
have already been performed and the results are known qualify these calculations as Open.
• Some of the calculations described in this Guide did originally fall into the Specified category
because they were first performed before the experiments were conducted. However, in almost
every case, the experiment was not conducted exactly as specified, and the calculation results
were not particularly useful in determining the accuracy of the model.
• None of the calculations were truly Blind, even those performed prior to the experiments. The
purpose of a Blind calculation is to assess the degree to which the choice of input parameters
affects the outcome. However, in such cases it is impossible to discern the uncertainty asso-
ciated from the choice of input parameters from that associated with the model itself. The
primary purpose of this Guide is to quantify the uncertainty of the model itself, in which case
Blind calculations are of little value.

1.3 How to Use this Guide

This guide presents a compilation of past and present validation exercises for the CFAST model.
The structure of the report is as follows:
• Chapter 2 includes a review of published literature by the National Institute of Standards and
Technology (NIST) and others on verification, validation, and sensitivity analysis of earlier
versions of CFAST.
• Chapter 3 discusses verification of the model. A range of analytical calculations are compared
with equivalent model predictions. These are intended to test the correctness of specific model
calculations.

2
 • Chapter 4 provides a summary of the experiments used in the current evaluation for the CFAST
model.

• Chapters 5 through 10 quantifies the comparison of CFAST predictions with experimental

measurements for a number of important model results.

• Chapter 11 summarizes the results of the validation study.

• Appendix A describes calculations used to estimate layer temperatures and interface height
from individual temperature measurements in the experiments.

• Appendix B summarizes the results of verification testing detailed in Chapter 3.

• Appendix C includes graphs of experimental results and model predictions for all comparisons
included in the validation study.

As CFAST continues to develop, it will expand to include new experimental measurements of

newly modeled physical phenomena. With each change in CFAST, the validation and verification
tests are all redone to ensure that changes to the model are consistent with experimental measure-
ments and the overall accuracy of the model is maintained by comparing the results between the
old and new versions of the model. If you are embarking on a validation study, you might want to
consider the following steps:

1. Survey Chapter 2 to learn about past efforts by others to validate the model for applications
similar to yours. Keep in mind that most of the referenced validation exercises have been
performed with older versions of CFAST, and you may want to obtain the experimental data
and the old CFAST input files and redo the simulations with the version of CFAST that you
plan to use.

2. Identify in Chapter 4 experimental data sets appropriate for your application. In particular,
the summary of the experiments found in table 4.13 contains a table listing various non-
dimensional quantities that characterize the parameters of the experiments. For example, the
global equivalence ratio of a compartment fire experiment indicates the degree to which the fire
was over or under-ventilated. To say that the results of a given experiment are relevant to your
scenario, you need to demonstrate that its parameters “fit” within the parameter space outlined
in Table 4.13.

3. Search the Table of Contents to find comparisons of CFAST version 7 simulations with specific
parameters from relevant experiments. For a given experiment, there may be numerous mea-
surements of quantities like the gas temperature, heat flux, and so on. It is a challenge to sort
out all the plots and graphs of all the different quantities and come to some general conclusion.
For this reason, this Guide is organized by output quantity, not by individual experiment or fire
scenario. In this way, it is possible to assess, over a range of different experiments and sce-
narios, the performance of the model in predicting a given quantity. Overall trends and biases
become much more clear when the data is organized this way.

3
The experimental data sets and CFAST input files described in this Guide are all managed via the
on-line project archiving system. You might want to re-run examples of interest to better under-
stand how the calculations were designed, and how changes in the various parameters might affect
the results. This is known as a sensitivity study, and it is difficult to document all the parameter
variations of the calculations described in this report. Thus, it is a good idea to determine which of
the input parameters are particularly important for your application.

4
Chapter 2

Survey of Past Verification and Validation

Work

CFAST has been subjected to extensive validation studies by NIST and others. This chapter pro-
vides a review of select CFAST validation efforts by NIST and others to better understand the
quality of the predictions by the model. Some of the work has been performed at NIST, some by
its grantees and some by engineering firms using the model. Because each organization has its
own reasons for validating the model, the referenced papers and reports do not follow any particu-
lar guidelines. Some of the works only provide a qualitative assessment of the model, concluding
that the model agreement with a particular experiment is “good” or “reasonable.” Sometimes, the
conclusion is that the model works well in certain cases, not as well in others. These studies are
included in the survey because the references are useful to other model users who may have a
similar application and are interested in qualitative assessment. It is important to note that some
of the papers point out flaws in early releases of CFAST that have been corrected or improved in
more recent releases.

2.1 Model Sensitivity

A sensitivity analysis considers the extent to which uncertainty in model inputs influences model
output. For a sensitivity analysis, this uncertainty includes not only that inherent in the input of
data for specific scenarios by the model user, but also uncertainty in empirical data or numerical
parameters in the model such as the time step size used by the model to obtain a solution. Iman
and Helton [2] studied the sensitivity of complex computer models developed to simulate the risk
of severe nuclear accidents which may include fire and other risks. Consistent with the work of
Iman and Helton, ASTM E1355 [1] provides overall guidance on typical areas of evaluation of
the sensitivity of deterministic fire models. These areas may involve one or more of the following
techniques: finite difference or direct analysis methods that provide an explicit solution of the sen-
sitivity equations associated with the governing equations of the model, factorial design or Latin
hypercube sampling studies that investigate the effect of varying the input parameters and conse-
quential interactions between parameters that may be deemed important, and global or response
surface methods that investigate the overall behavior of model outputs for a desired range of inputs.
This section provides a review of the sensitivity studies that have been conducted using CFAST

5
with an emphasis on uncertainty in the input. Other sensitivity investigations of CFAST are also
available [3, 4, 5].
Khoudja [6] has studied the sensitivity of an early version of FAST (the predecessor of CFAST)
with a fractional factorial design involving two levels of 16 different input parameters. The sta-
tistical design, taken from the texts by Box and Hunter [7] and Daniel [8] reduced the necessary
model runs from more than 65,000 to 256 by studying the interactions of input parameters simul-
taneously. The choice of values for each input parameter represented a range for each parameter.
The analysis of the FAST model showed sensitivity to heat loss to the compartment walls and to
the number of compartments in the simulation.
Walker [9] discussed the uncertainties in components of zone models and showed how uncer-
tainty within user-supplied data affects the results of calculations using CFAST as an example. The
study systematically varied inputs related to the fire (heat release rate, heat of combustion, mass
loss rate, radiative fraction, and species yields) and compartment geometry (vent size and ceiling
height). Heat release rate and ceiling height are seen to be the dominant input variables in the
simulations.
Peacock et al. [3] studied the sensitivity of CFAST for a range of input parameters. They
used simple factorial designs for model inputs deemed important to investigate local behavior of
important model outputs along with response surface methods to evaluate overall model behavior.

2.2 Comparisons of CFAST with Full-Scale Experiments

Several studies have been conducted specifically to validate the use of CFAST in building per-
formance design. Dembsey [10] used CFAST version 3.1 to predict the ceiling jet temperatures,
surface heat fluxes and heat transfer coefficients for twenty compartment fire experiments in a
compartment that is similar in size, geometry, and construction to the standard fire test compart-
ment specified in the Uniform Building Code [11]. Results from 330 kW, 630 kW, and 980 kW
fires were used. In general, CFAST made predictions which were higher than the experimental re-
sults. In these cases, the temperature prediction was typically 20 % to 30 % higher than measured
values. Much of this can be attributed to not knowing the species production (soot) and relative
absorption of radiation by the gas layers which highlights the importance of scenario specifica-
tion. This is the most common cause of “over prediction” of temperature by CFAST. A secondary
source of discrepancy is correcting for thermal radiation of thermocouple beads. The authors pro-
vide for this correction, but the corrections cited are not as large as has been reported in other fire
experiments [12].
He et al. [13] describe a series of full-scale fire experiments that were designed to investigate
the validity of two zone models including CFAST version 3.1. The experiments, involving steady
state burning rates and a number of ventilation conditions, were conducted in a four-story building.
Temperature, pressure, flow velocity, smoke density and species concentrations were measured in
various parts of the building. The stack effect and its influence on temperature distribution in a
stair shaft were observed. Comparisons were then made between the experimental results and the
model predictions. Early in the fire there is a few percent difference between the predictions and
measurements; beyond 10 min, there are significant variations. Both the experiment and the model
are internally consistent; that is, higher flow leads to a higher interface height (figure 13 in the
paper). Once again, the difference is about 25 %. The authors discuss the effect of fuel composition

6
and correction for radiation to/from thermocouple beads but did not draw firm conclusions based
on their measurements of fuel products.
A series of experimental results for flaming fires, obtained using realistic fires in a prototype
apartment building were performed by Luo et al. [14]. Fuel configurations in the fire test included
a horizontal plain polyurethane slab, mock-up chair (polyurethane slabs plus a cotton linen cover),
and a commercial chair. CFAST version 3.1 typically over-predicted upper layer temperatures by
10 % to 50 % depending on the test conditions and measurement location in that test. The pre-
dicted and experimental time dependent upper layer temperatures were similar in shape. The time
to obtain peak upper layer temperatures was typically predicted to within 15 % of the experimen-
tal measurements. The authors concluded that CFAST was conservative in terms of life safety
calculations.
Poole et al. [15] reported the results of a cooperative project between the Kitchener, Ontario,
Fire Department and the University of Waterloo aimed at developing design criteria for the con-
struction of a fire fighter training facility. One particular criterion is that realistic training with
respect to temperature, heat release and stratification be provided in such a facility. The purpose
of this paper was to compare existing analytical heat release and upper and lower gas tempera-
ture rise correlations and models with data from actual structures which were instrumented and
burned in collaboration with the Kitchener Fire Department. According to the authors, the CFAST
model was used “successfully” to predict these conditions and will be used in future design of such
facilities.
A report by Bailey et al. [16] compares predictions by CFAST version 3.1 to data from real
scale fire tests conducted onboard ex-USS SHADWELL, a decommissioned naval vessel. The
phenomenon of particular interest in this validation series was the conduction of heat in the vertical
direction through compartment ceilings and floors. As part of this work, Bailey et al. [17] compared
CFAST temperature predictions on the unexposed walls of large metal boxes, driven by steady
state fires. This tested the model’s prediction of radiation and conduction in both the vertical
and horizontal directions. The model and experiment compared well within measurement error
bounds of each. The comparison was particularly good for measurements in the fire compartment
as well as for the compartment and deck directly above it, with predictions typically agreeing with
experiments within measurement uncertainty. The model under-predicted the temperatures of the
compartments and decks not directly adjacent to the fire compartment early in the tests.
Deal [18] reviewed four computer fire models (CCFM [19], FIRST [20], FPETOOL [21] and
FAST [22] version 18.5 (the immediate predecessor to CFAST)) to ascertain the relative perfor-
mance of the models in simulating fire experiments in a small room (about 12 m3 in volume) in
which the vent and fuel effects were varied. Peak fire size in the experiments ranged up to 800 kW.
According to the author, all the models simulated the experimental conditions including tempera-
ture, species generation, and vent flows “quite satisfactorily.”
Matsuyama conducted a series of full-scale experiments [23] using t2 fires 1 . Fire room and
corridor smoke filling processes were measured. The size of the corridors and arrangements of
smoke curtains were varied in several patterns. Comparisons were then made between the exper-
imental results and those predicted by CFAST. The author concludes that while the model does
a “good job” of predicting experimental results, there are systematic differences which could be
1 For a range of fires, the fire growth can be represented with a power law relation of the form q̇ = α t 2 where q̇ is
the heat release rate of the fire, α is the fire intensity coefficient, and t is time

7
reduced with some revision to zone model formulation to include the impact of smoke curtains.

2.2.1 Fire Plumes

Davis compared predictions by CFAST version 5 (and other models) for high ceiling spaces [24].
In this paper, the predictive capability of two algorithms designed to calculate plume centerline
temperature and maximum ceiling jet temperature in the presence of a hot upper layer were com-
pared to measurements from experiments and to predictions using the CFAST ceiling jet algorithm.
The experiments included ceiling heights of 0.58 m to 22 m and heat release rates of 0.62 kW to
33 MW. When compared to the experimental results, the ceiling jet algorithm tended to over-
predict the upper layer temperature by 20 %. With proper adjustment for radiation effects in the
thermocouple measurements, some of this difference disappears. The effect of entrainment of the
upper layer gases was identified for improvement.

2.2.2 Multiple Compartments

Jones and Peacock [25] presented a limited set of comparisons between the FAST model (version
18.5) and a multi-room fire test. The experiment involved a constant fire of about 100 kW in a
three-compartment configuration of about 100 m3 . They observed that the model predicted an
upper layer temperature that was too high by about 20 % with satisfactory prediction of the layer
interface position. These observations were made before the work of Pitts et al. [12] showed that
the thermocouple measurements need to be corrected for radiation effects. Convective heating and
plume entrainment were seen to limit the accuracy of the predictions. A comparison of predicted
and measured pressures in the rooms showed accuracy within 20 %. Since pressure is the driving
force for flow between compartments, this agreement was seen as important.
Levine and Nelson [26] used a combination of full-scale fire testing and modeling to sim-
ulate a fire in a residence. The 1987 fire in a first-floor kitchen resulted in the deaths of three
persons in an upstairs bedroom, one with a reported blood carboxyhemoglobin content of 91 %.
Considerable physical evidence remained. The fire was successfully simulated at full scale in a
fully-instrumented seven-room two-story test structure. The data collected during the test have
been used to test the predictive abilities of two multiroom computer fire models: FAST and HAR-
VARD VI. A coherent ceiling layer flow occurred during the full-scale test and quickly carried high
concentrations of carbon monoxide to remote compartments. Such flow is not directly accounted
for in either computer code. However, both codes predicted the carbon monoxide buildup in the
room most remote from the fire. Prediction of the pre-flashover temperature rise was also ‘good’
according to the authors. Prediction of temperatures after flashover that occurred in the room of
fire origin was seen as ‘less good.’ Other predictions of conditions throughout the seven test rooms
varied from ‘good approximations’ to ‘significant deviations’ from test data. Some of these devia-
tions are believed to be due to combustion chemistry in the hot upper layer not considered in detail
in either of the two models.

2.2.3 Large Compartments

Duong [27] studied the predictions of several computer fire models (CCFM, FAST, FIRST, and BRI
[28]), comparing the models with one another and with large fires (4 MW to 36 MW) in an aircraft

8
hanger (60 000 m3 ). For the 4 MW fire size, he concluded that all the models are ‘reasonably
accurate.’ At 36 MW, however, ‘none of the models did well.’ Limitations of the heat conduction
and plume entrainment algorithms were thought to account for some of the inaccuracies.

2.2.4 Mechanical Ventilation

There have been two papers which have looked at the effectiveness of the mechanical ventilation
system. The first considered a fire chamber of length 4.0 m, width 3.0 m and height 2.8 m with
adjustable ventilation rates [29]. Burning tests were carried out with wood cribs and methanol to
study the preflashover stage of a compartmental fire and the effect of ventilation. The mass loss rate
of fuel, temperature distribution of the compartment and the air intake rate were measured. The
heat release rates of the fuel were calculated and the smoke temperature was used as a validation
parameter. A scoring system was proposed to compare the results predicted by the three models.
According to the author, CFAST does ‘particularly well,’ though there are some differences which
can be attributed to the zone model approach.
A second series of experiments by Luo [30] indicate that the CFAST model (version 3.1) gener-
ally over-predicts the upper layer temperature in the burn room because the two-zone assumption is
likely to break down in the burn room. It was found that the room-averaged temperatures obtained
from CFAST were in ‘good overall agreement’ with the experimental results. The discrepancies
can be attributed to the correction needed for thermocouple measurements. The CO concentra-
tion,however, was inconsistent. CFAST tended to overestimate CO concentration when the air
handling system was in operation. This was seen due to inconsistencies in what is measured (point
measurements) and predicted (global measurements).

2.2.5 Sprinkler Activation

A suppression algorithm [31] was incorporated into CFAST. Chow [32] evaluated the predictive
capability for a sprinkler installed in an atrium roof. There were three main points being consid-
ered: the possibility of activating the sprinkler, thermal response, and water requirement. The zone
model CFAST was used to analyze the possibility of activation of a sprinkler head. Results de-
rived from CFAST were seen to be ‘accurate, that is, providing good agreement with experimental
measurements.’

2.2.6 Flashover
Flashover is a term describing the transition of a relatively localized interior fire to one engulfing
the entire compartment. It is of interest to the fire service because of the danger to fire fighters and
to building designers because of life safety and the attendant impact on occupants. Several papers
have looked at the capability of CFAST to predict the conditions under which flashover can occur.
Chow [33] concluded that FAST correctly predicted the onset of flashover if the appropriate
criteria were used. The criteria were gas temperature near the ceiling, heat flux at the floor level
and flames coming out of the openings. This analysis was based on a series of compartment fires.
A paper by Luo et al. [34] presents a comparison of the results from CFAST version 3 against
a comprehensive set of data obtained from one flashover fire experiment. The experimental results
were obtained from a full-scale prototype apartment building under flashover conditions. Three

9
polyurethane mattresses were used as fuel. It was found that the predicted temperatures from
the CFAST fire model agreed well with the experimental results in most areas, once radiation
corrections are applied to the thermocouple data.
Collier [35] makes an attempt to quantify the fire hazards associated with a typical New Zealand
dwelling with a series of experiments. These tests, done in a three-bedroom dwelling, included
both non-flashover and flashover fires. The predictions by CFAST version 2 were seen by the
author as consistent with the experiments within the uncertainty of each.
Post-flashover fires in shipboard spaces have a pronounced effects on adjacent spaces due to
highly conductive boundaries. CFAST (version 3.1) predictions for the gas temperature and the
cold wall temperature were compared with shipboard fires [36]. The comparisons between the
model and experimental data show ‘conservative predictions’ according to the authors. The authors
attribute this to an overestimation of the average hot wall temperature and an underestimation of
external convective losses due to wind effects.
Finally, a comparison of CFAST with a number of simple correlations was used by Peacock
and Babrauskas [37, 38] to simulate a range of geometries and fire conditions to predict the devel-
opment of the fire up to the point of flashover. The simulations represent a range of compartment
sizes and ceiling heights. Both the correlations and CFAST predictions were seen to provide a
lower bound to observed occurrence of flashover. For very small or very large compartment open-
ings, the differences between the correlations, experimental data, and CFAST predictions was more
pronounced.
The important test of all these prediction methods is in the comparison of the predictions with
actual fire observations. Figure 2.1 (reference [38]) presents estimates of the energy required to
achieve flashover for a range of room and vent sizes. This figure is an extension of the earlier work
of Babrauskas [39] and includes additional experimental measurements from a variety of sources,
most notably the work of Deal and Beyler [40]. For a number of the experimental observations,
values are included that were not explicitly identified as being a minimum value at flashover.
As with some of the experimental data defining flashover as an upper layer temperature reach-
ing 600 ◦ C, many experimental measures were reported as peak values rather than minimum values
necessary to achieve flashover. Thus, ideally all the predictions should provide a lower bound for
the experimental data. Indeed, this is consistent with the graph – the vast majority of the exper-
imental observations√ lie above the correlations and model predictions. For a considerable range
in the ratio AT /A h, the correlations of Babrauskas [39] Thomas [41], and the MQH correla-
tion of McCaffrey et al. [42] provide similar estimates of the minimum energy required to produce √
flashover. The estimates of Hägglund [43] yields somewhat higher estimates for values of AT /A h
greater than 20 m−1/2 .
The results from the CFAST model for this single compartment scenario √ provide similar results
to the√experiments and the correlations for most of the range of AT /A h. For small values of
AT /A h, the CFAST
√ values rise somewhat above the values from the correlations. These small
values of AT√ /A h result from either very small compartments (small AT ) or very large openings
(large AT /A h), both of which stretch the limits of the assumptions inherent in the model. For very
small compartments, radiation from the fire to the compartment surfaces becomes more important,
enhancing the conductive heat losses through the walls. However, the basic two-zone assumption
may break down as the room becomes very small. For very large openings, the calculation of
vent flow via an orifice flow coefficient approach is likely inaccurate. Indeed, for such openings,
this limitation has been observed experimentally [39]. The estimates are close to the range of

10
Figure 2.1: Comparison of correlations, CFAST predictions, and experimental data for the prediction of
flashover in a compartment fire.

11
 √
uncertainty shown by the correlations which also diverge at very small values of AT /A h.
Perhaps most significant in these comparisons is that all the simple correlations provide esti-
mates similar to the CFAST model and all the models are consistent with a wide range of exper-
imental data. For this simple scenario, little is gained with the use of the more complex models.
For more complicated scenarios, the comparison may not be as simple.

2.3 Comparisons of CFAST with Actual Fires

There are numerous cases of CFAST being used to adjudicate legal disputes. Since these are
discussed in courts of law, there is a great deal of scrutiny of the modeling, assumptions, and
results. Most of these simulations and comparisons are not available in the public literature. A few
of the cases which are available are discussed below. The metric for how well the model performed
is its ability to reproduce the time-line as observed by witnesses and the death of occupants or the
destruction of property as was used in evidence in legal proceedings.
As mentioned in section 2.2.2, Levine and Nelson describe the use of FAST for understanding
the deaths of two adults in a residence in Sharon, Pennsylvania in 1987 [26]. The paper compared
the evidence of the actual fire, a full scale mockup done at NIST and the results from FAST (version
18) [44] and Harvard VI [45]. The most notable shortcoming of the models was the lower than
actual temperatures in the bedrooms, caused by loss of heat through the fire barriers. This led to
the improvement in CFAST in the mid-90s to couple compartments together so that both horizontal
and vertical heat transfer can occur to adjacent compartments.
Bukowski used CFAST version 3.1 to analyze a fire in New York City [46] in 1994 which
resulted in the death of three fire fighters. The CFAST model was able to reproduce the observed
conditions and supported the theory as to how the fire began and the cause of death of the three fire
fighters.
Chow describes the use and comparison of CFAST simulations with a 1996 high rise build-
ing fire in Hong Kong [47]. CFAST simulations were performed to help understand the probable
fire environment under different conditions. Three simulations were performed to study the con-
sequences of a fire starting in the lift shaft. Smoke flow in the simulations qualitatively matched
those observed during the incident.
In the early morning hours of March 25,1990 a tragic fire took the lives of 87 persons at a
neighborhood club in the Bronx, New York [48]. The New York City Fire Department requested
the assistance of the NIST Center for Fire Research (CFR) in understanding the factors which
contributed to this high death toll and to develop a strategy that might reduce the risk of a similar
occurrence in the many similar clubs operating in the city. The simulation showed the potential for
development of untenable conditions within the club and particularly in the single exit stairway.

2.4 Comparisons of CFAST for Special Applications

There are several sets of comparisons used in the development of the model or specific applications
beyond those discussed more generally above.

12
2.4.1 Nuclear Facilities

Floyd validated CFAST version 3.1 by comparing the modeling results with measurements from
fire tests at the Heiss-Dampf Reaktor (HDR) facility [49]. The structure was originally the con-
tainment building for a nuclear power reactor in Germany. The cylindrical structure was 20 m in
diameter and 50 m in height topped by a hemispherical dome 10 m in radius. The building was
divided into eight levels. The total volume of the building was approximately 11 000 m3 . From
1984 to 1991, four fire test series were performed within the HDR facility. The T51 test series
consisted of 11 propane gas tests and three wood crib tests. To avoid permanent damage to the
test facility, a special set of test rooms were constructed, consisting of a fire room with a narrow
door, a long corridor wrapping around the reactor vessel shield wall, and a curtained area centered
beneath a maintenance hatch. The fire room walls were lined with fire brick. The doorway and
corridor walls had the same construction as the test chamber. Six gas burners were mounted in the
fire room. The fuel source was propane gas mixed with 10 % air fed at a constant rate to one of the
six burners.

In general, the comparison between CFAST and the HDR results was seen as ‘good’ by the
author, with two exceptions. The first is the over estimate of the temperature of the upper layer,
typically within about 15 % of the experimental measurements. This is common and generally re-
sults from using too low a value for the production of soot, water (hydrogen) and carbon monoxide.
The other exception consists of predictions in spaces where the zone model concept breaks down,
for example in the stairways between levels. In this case, CFAST has to treat the space either in
the filling mode (two layer approximation) or as a fully mixed zone (using the SHAFT option).
Neither is quite correct, and in order to understand the condition in such spaces in detail (beyond
the transfer of mass and energy), a more detailed computational fluid dynamics model must be
used, for example, the Fire Dynamics Simulator (FDS) [50].

The U.S. Nuclear Regulatory Commission (NRC) performed an extensive verification and val-
idation of several fire models commonly used in nuclear power plant applications [51]. These
models included simple spreadsheet calculations, zone models (including CFAST), and CFD mod-
els. The results of this study are presented in the form of relative differences between fire model
predictions and experimental data for fire modeling attributes such as temperature or heat flux that
are important to NPP fire modeling applications. These relative differences are affected by the
capabilities of the models, the availability of accurate applicable experimental data, and the ex-
perimental uncertainty of these data. Evaluation of the two-zone models showed that the models
simulated the experimental results within experimental uncertainty for many of the parameters of
interest. The reason for this may be that the relatively simple experimental configurations selected
for this study conform well to the simple two-layer assumption that is the basis of these models.

While the relative differences sometimes show agreement for many parameters, they also show
both under-prediction and over-prediction in some circumstances, most notably when conditions
vary within a compartment or detailed local conditions are important to accurate prediction (for
example, plume temperature or heat flux near to the fire source). The results and comparisons
included the the NRC study are included in this report for the current version of CFAST.

13
2.4.2 Small Compartments
As an implementation of the zone model concept, CFAST is applicable to a wide range of sce-
narios. One end of this spectrum are small compartments, one to two meters on a side. Several
research efforts have looked at small scale validation. There are three papers by Chow [52, 53, 54]
which examine this issue. The first is the use of an electric heater with adjustable thermal power
output was to verify temperature predictions by CFAST version 3.1. The second was closed cham-
ber fires studied by burning four types of organic liquids, namely ethanol, N-heptane, and kerosene.
The burning behavior of the liquids was observed, and the hot gas temperature measured. These
behaviors along with the transient variations of the temperature were then compared with those
predicted by the CFAST model. Finally, in another series of experiments, three zone models, one
of which was CFAST, were evaluated experimentally using a small fire chamber. Once again,
liquid fires were chosen for having better control on the mass loss rate. The results on the devel-
opment of smoke layer and the hot gas temperature predicted by the three models were compared
with those measured experimentally. According to Chow, ‘fairly good agreement’ was found if the
input parameters were carefully chosen.

2.4.3 Railway and Vehicle Tunnels

Altinakar et al. [55] used a modified version of CFAST for predicting fire development and smoke
propagation in vehicle or railroad tunnels. The two major modifications made to the model dealt
with mixing between the upper and lower layers and friction losses along the tunnel. The model
was tested by simulating several full-scale tests carried out at memorial Tunnel Ventilation Test
Program in West Virginia, and the Offeneg Tunnel in Switzerland. His article compares simulated
values of temperature, opacity and similar sensible quantities with measured values and discusses
the limits of the applicability of zone models for simulating fire and smoke propagation in vehicle
and railroad tunnels.
Peacock et al. [56] compared times to untenable conditions determined from tests in a passen-
ger rail car with those predicted by CFAST for the same car geometry and fire scenarios. For a
range of fire sizes and growth rates, they found agreement that averaged approximately 13 %.

2.4.4 Non-Uniform Compartments

In January 1996, the U.S. Navy began testing how the CFAST model would perform when tasked
with predicting shipboard fires. These conditions include mass transport through vertical vents
(representing hatches and scuttles), energy transport via conduction through decks, improvement
to the radiation transport sub-model, and geometry peculiar to combat ships. The purpose of this
study was to identify CFAST limitations and develop methods for circumnavigating these problems
[57]. A retired ship representing the forward half of a USS Los Angeles class submarine was used
during this test series. Compartments in combat ships are not square in floor area, nor do they have
parallel sides.
Application of CFAST to these scenarios required a direct integration of compartment cross-
sectional area as a function of height to correctly interpret the layer interface position and provide
correct predictions for flow through doors and windows (vertical vents). This required user speci-
fication of the area as a function of height to provide a description for the model to use.

14
Chapter 3

Verification

The terms verification and validation are often used interchangeably to mean the process of check-
ing the accuracy of a numerical model. For many, this entails comparing model predictions with
experimental measurements. However, there is now a fairly broad-based consensus that com-
paring model and experiment is largely what is considered validation. So what is verification?
ASTM E1355 [1], “Standard Guide for Evaluating the Predictive Capability of Deterministic Fire
Models,” defines verification as

The process of determining that the implementation of a calculation method accurately

represents the developer’s conceptual description of the calculation method and the solution
to the calculation method.

and it defines validation as

The process of determining the degree to which a calculation method is an accurate rep-
resentation of the real world from the perspective of the intended uses of the calculation
method.

Simply put, verification is a check of the math; validation is a check of the physics. If the model
predictions closely match the results of experiments, using whatever metric is appropriate, it is
assumed by most that the model suitably describes, via its mathematical equations, what is hap-
pening. It is also assumed that the solution of these equations must be correct. So why do we need
to perform model verification? Why not just skip to validation and be done with it? The reason
is that rarely do model and measurement agree so well in all applications that anyone would just
accept its results unquestionably. Because there is inevitably differences between model and ex-
periment, we need to know if these differences are due to limitations or errors in the numerical
solution, or the physical sub-models, or both.
Whereas model validation consists mainly of comparing predictions with experimental mea-
surements, as documented later in this guide, model verification consists of a much broader range
of activities, from checking the computer program itself, to comparing calculations to analytical
(exact) solutions, to understanding the impact on model outputs from a range of different model
inputs.
A series of verification test cases follow, examining the energy balance, mass balance, venti-
lation, heat transfer and sprinkler effects modeled by CFAST. These test cases are routinely run

15
to ensure the calculations in the model are correct for a range of simple calculations where an-
alytical solutions exist. The energy and mass balance examples test the underlying fundamental
equations that determine conditions within a compartment in the model. The ventilation examples
test the flow of gases between compartments through doors, windows, holes in floors/ceilings, and
mechanical ventilation systems. The heat transfer examples test the flow of energy through walls
of compartments or to user-defined objects within a compartment. The sprinkler example tests the
algorithm used to simulate the response of heat detectors and fire sprinklers in CFAST. For each of
the examples, an analytical solution is presented and compared to the results of a matching CFAST
simulation.

3.1 Energy
For most of the examples presented in this section, the same basic geometry is used, a single 5 m
by 5 m by 5 m compartment.

3.1.1 Temperature Equilibrium via Heat Conduction

As a simple test of the energy balance, raising the external temperature of the base case compart-
ment from an initial condition of 20 ◦ C to 25 ◦ C allows the internal temperature to equilibrate to
the exterior. From the ideal gas law, the pressure inside the compartment is expected to rise to

Tfinal 298.15 K
Pfinal = Pinitial = 101325 Pa × = 103053 Pa (3.1)
Tinitial 293.15 K

or a final pressure rise of 1728 Pa. Figure 3.1 shows the simulated conditions for this test, which
are consistent with the expected results..

30 2000

1500
25

1000

20
500

15 0
0 1 2 3 4 5 0 1 2 3 4 5

Figure 3.1: Interior temperature and pressure in equilibrium with the exterior in the case
basic_tempequilib.in.

16
3.1.2 Temperature Equilibrium via a Window
Now an open window is added to the compartment, with an exterior temperature of 25 ◦ C. Fig-
ure 3.2 shows the interior conditions coming into equilibrium with the exterior, which are consis-
tent with the expected results..

30 1

0.5
25

20
-0.5

15 -1
0 1 2 3 4 5 0 1 2 3 4 5

Figure 3.2: Interior temperature and pressure in equilibrium with the exterior in the case
basic_tempequilib_window.in.

3.1.3 Temperature Equilibrium via a Window at a High Elevation

With the exterior temperature still set to 25 ◦ C, the elevation is raised to 1500 m, approximately
the average elevation of Idaho. Since CFAST calculations are relative to the exterior ambient,
conditions are expected to be identical to the previous examples and equilibrate to those of the
exterior. Figure 3.3 shows the simulated conditions for the test case, which are consistent with the
expected results..

30 1

0.5
25

20
-0.5

15 -1
0 1 2 3 4 5 0 1 2 3 4 5

Figure 3.3: Interior temperature and pressure in equilibrium with the exterior in the case
basic_tempequilib_window_elevation.in.

17
3.1.4 A Fire in a Single, Sealed Compartment with a Single Zone
A 100 kW methane fire burns in a sealed compartment with no ventilation, adiabatic walls, and no
radiative emission. A single zone simulation is run in which it is assumed that the entire volume
is taken up by the upper layer. From Eqs. (2.5) and (2.6) in the CFAST Technical Guide [58], the
governing equation for the pressure and temperature of the single zone compartment are:
dP γ −1
= q̇ (3.2)
dt V
 
dT 1 dP
= q̇ − c p ṁ T +V (3.3)
dt cp m dt
where m is the total mass in the volume, ṁ = ṁf is the mass flow rate of the fuel into the volume,
and q̇ = Q̇ + c p ṁf Tf is the heat release rate of the fire plus the heat flow rate of the fuel flowing
into the volume. Note that ṁ and q̇ are source terms, they are assumed to have this form for this
test problem.
Eq. (3.3) can be shown to be equivalent to the first law of thermodynamics by first substituting
dP/dt from Eq. (3.2) and noting that γ = c p /cv to obtain
dT 1
= (q̇ − c p ṁ T + (γ − 1) q̇) (3.4)
dt cp m
1
= (γ q̇ − c p ṁ T ) (3.5)
cp m
1
= (q̇ − cv ṁ T ) (3.6)
cv m
Next, the first law of thermodynamics can be stated as
d dV
(cv mT ) = q̇ − P = q̇ (3.7)
dt dt
since dV /dt = 0 (for this case). Expanding and solving for dT /dt results in
dT
cv ṁT + cv m = q̇ (3.8)
dt
dT 1
= (q̇ − cv ṁ T ) (3.9)
dt cv m
which is identical Eq. (3.6) and therefore equivalent to Eq. (3.3).
Since the right hand side of Eq. (3.2) is constant, it may be integrated to obtain the solution
γ −1
P(t) = P0 + q̇t (3.10)
V
where P0 is the initial pressure. A solution for Eq. (3.3) may be found by using Eq. (3.10) and the
ideal gas law to obtain
V P(t)
T (t) = (3.11)
R m(t)
where m(t) = m0 + ṁt, m0 = PR0TV0 , is the initial mass, T0 is the initial pressure and R = c p − cv .
Figure 3.4 includes comparisons of the temperature and pressure as predicted by CFAST and
the analytic solutions given in Eqs. (3.10) and (3.11).

18
 1000 300

250
800

200
600
150
400
100

200
50

0 0
0 5 10 15 0 5 10 15

Figure 3.4: Temperature and pressure rise due to a fire in a closed compartment. The case is called
sealed_test.in.

3.1.5 A Fire in a Single, Sealed Compartment with Two Zones

Consider now the same case as in Section 3.1.4, but now with two zones rather than one. The
compartment pressure ought to be the same as before, and the upper layer temperature ought to
converge to the single layer temperature as seen in Fig. 3.5.

1000 300

250
800

200
600
150
400
100

200
50

0 0
0 5 10 15 0 5 10 15

Figure 3.5: Hot gas layer temperature and pressure rise due to a fire in a closed compartment with two zones.
The case is called sealed_test_2_layers.in.

19
3.2 Mass
3.2.1 A Fire in a Single, Sealed Compartment
A methane fire burns in a sealed compartment of dimension 5 m by 6 m by 3 m. The heat release
rate ramps up linearly to 1 kW in 30 s, then remains steady for 5 min, and then ramps down linearly
to 0 in 30 s. The total energy released is 330 kJ, and the total mass of fuel consumed is

330 kJ
= 0.00660 kg (3.12)
50000 kJ/kg

where the heat of combustion of methane is taken to be 50 000 kJ/kg. For complete combustion of
methane, the combustion chemistry is given by

CH4 + 2 O2 → CO2 + 2 H2 O (3.13)

The molecular weight of CH4 is 16 g/mol and CO2 is 44 g/mol; thus, the total mass of CO2
produced by the fire is

44 g/mol
mCO2 = 0.0066 kg × = 0.01815 kg (3.14)
16 g/mol

The molecular weight of H2 O is 18 g/mol; thus, the total mass of H2 O produced by the fire is

2(18) g/mol
mH2 O = 0.0066 kg × = 0.01485 kg (3.15)
16 g/mol

CFAST predicts that the final mole fractions of O2 , CO2 and H2 O in the upper layer are 0.2069,
0.00012 and 0.00024, respectively (calculated by the model consistent with eq. 3.13; see [59] for
details of the combustion chemistry calculations in CFAST). The remainder is N2 , whose mole
fraction is 0.7927. These mole fractions can be converted to mass fractions by

Xk Mk
Yk = N
(3.16)
∑i=1 Xi Mi

The mass of the upper layer can be calculated from the equation of state:

PV γ −1 J
mu = ; R= c p ≈ 289.14 (3.17)
RT γ kg · K

The mass of CO2 and H2 O produced is given by

mk = mu Yk (3.18)

Figure 3.6 shows the resulting product masses, which are consistent with the expected results.

20
 0.02

0.015

0.01

0.005

0
0 5 10 15

Figure 3.6: Expected and predicted masses of CO2 and H2 O for the case species_mass_1.in.

3.2.2 A Fire in a Compartment Connected to Another via a Door

The same natural gas fire described in Section 3.2.1 burns in a compartment of dimension 2 m
by 5 m by 8 m which is connected to another compartment of dimension 5 m by 3 m by 8 m. A
doorway connects the compartments, which has a width of 1 m and a height of 6 m. Because the
fire and the fuel source have not changed, the theoretical calculations for the mass of CO2 and H2 O
produced will remain the same. The remaining portion of the problem is approached in the same
manner, but since there are two compartments, the mass of CO2 and H2 O produced in each layer
of each compartment must be individually calculated and then summed together to produce the net
yields of CO2 and H2 O. Figure 3.7 shows the resulting product masses, which are consistent with
the expected results.

0.02

0.015

0.01

0.005

0
0 5 10 15

Figure 3.7: Expected and predicted masses of CO2 and H2 O for the case species_mass_2.in.

3.2.3 A Fire in a Compartment Connected to Another via a Ceiling Vent

The same natural gas fire described in Section 3.2.1 burns in a compartment of dimension 9 m by
5 m by 4 m which is connected to another compartment of dimension 9 m by 5 m by 2 m. The
compartments are placed such that the second one is located directly above the first one. There
is a square ceiling vent between the compartments that has an area of 4 m2 . This problem is

21
approached in the same exact manner as in Section 3.2.2 because the only difference between the
two scenarios is the specific alignment of the compartments. Figure 3.7 shows the resulting product
masses, which are consistent with the expected results.

0.02

0.015

0.01

0.005

0
0 5 10 15

Figure 3.8: Expected and predicted masses of CO2 and H2 O for the case species_mass_3.in.

3.2.4 A Fire in a Four Compartment Assembly

Four 4 m by 4 m by 4 m compartments are arranged such that two compartments are placed
adjacent to one another and the following two compartments are placed directly on top of the first
two. The same natural gas fire described in Section 3.2.1 burns in the first compartment of this
setup. After 2500 s, the wall between compartments one and two is removed, forcing the gases in
the two rooms to mix. Next, at 5000 s, the wall between compartments three and four is removed.
Lastly, at 7500 s, the ceiling of compartment four is removed, allowing the system to slowly return
to ambient conditions. Figure 3.9 shows the resulting product masses in the structure before the
outside vent is opened, which are consistent with the expected results. Figure 3.10 shows how the
masses of CO2 and H2 O in each compartment change with respect to time.

0.02

0.015

0.01

0.005

0
0 5 10 15

Figure 3.9: Expected and predicted masses of CO2 and H2 O for the case species_mass_4.in.

22
 0.02 0.02

0.015 0.015

0.01 0.01

0.005 0.005

0 0
0 50 100 150 0 50 100 150

0.02 0.02

0.015 0.015

0.01 0.01

0.005 0.005

0 0
0 50 100 150 0 50 100 150

Figure 3.10: Expected and CFAST calculated values for the masses of CO2 and H2 O in each compartment
of a four room test case species_mass_4.in.

3.2.5 Product Yields

A simple 5 kW methane flame is burned inside a closed 5 m x 5 m x 5 m compartment for 100 s.

Initially, the relative humidity within the compartment is zero. The total mass of carbon dioxide
and water produced is calculated. The combustion of a methane is described by the following
single-step reaction:

CH4 + 2 O2 → CO2 + 2 H2 O (3.19)

With a total mass of methane burned of 0.01 kg (5 kW / 50000 MJ/kg × 100 s), 0.027 kg
of CO2 and 0.022 kg of H2 O are produced by the fire. Figure 3.11 compares the mass of the
combustion products to their expected values.
For a more complex test, we add non-zero yields of CO, soot, HCN, and HCl to the reaction.
A 5 kW fire of C12 H9 O3 N2 Cl is burned inside the same 5 m x 5 m x 5 m compartment for 100 s,
again with an initial relative humidity of zero. Yields of soot, CO, and HCN are assumed to be
0.186 kg/kg, 0.063 kg/kg, and 0.05 kg/kg, respectively. All chlorine in the fuel is converted to HCl.
With a total mass of fuel burned of 0.05 kg, 0.057 kg of CO2 , 0.013 kg of H2 O, 0.003 kg of CO,
0.0025 kg of HCN, and 0.007 kg of HCl are produced by the fire. Figure 3.12 compares the mass
of the combustion products to their expected values.

23
 0.05

0.04

0.03

0.02

0.01

0
0 1 2 3 4 5

Figure 3.11: Expected and CFAST calculated values for the masses of CO2 , and H2 O in a one room test
case methane_flame_simple.in.

10 -1

10 -2

10 -3

10 -4
0 1 2 3 4 5 6

Figure 3.12: Expected and CFAST calculated values for the masses of Soot, CO, CO2 , H2 O, and HCN in a
one room test case species_test.in.

3.2.6 Trace Species

With the same 5 kW methane flame as in the previous section, we add a trace species yield of
0.001 kg/kg. With a total mass of methane burned of 0.01 kg (5 kW / 50000 MJ/kg × 100 s),
a total mass of 0.00001 kg of trace species is expected. Figure 3.13 compares the expected and
calculated values.
With the same 5 kW methane flame, we add a second compartment with mechanical ventilation
flowing both directions from room to room. Figure 3.14 compares the expected and calculated
values. Without filtering, one half the trace species mass should ultimately end up in each room
(Figure 3.14a). With 100 % filtering, the trace species mass should trend to zero (Figure 3.14b).

24
 10 -5
1.2

0.8

0.6

0.4

0.2

0
0 1 2 3 4 5

Figure 3.13: Expected and CFAST calculated values for the mass of trace species in a one room test case
Trace_Species_1.in.
10 -5 10 -5
1 1

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 200 400 600 800 1000 1200 1400 0 200 400 600 800 1000 1200 1400

(a) without filtering (b) with 100 % filtering

Figure 3.14: Expected and CFAST calculated values for the mass of trace species in a two room test case
with mechanical ventilation Trace_Species_2.in.

3.3 Ventilation
3.3.1 Air Flow Through a Single Compartment
The setup for this test case includes a 5 m by 5 m by 3 m compartment that contains a mechanical
vent and a wall vent. Air is pumped into a single compartment through a 1 m by 1 m mechanical
vent, at a rate of 0.01 m3 /s. The air is allowed to exit the compartment through a 1 cm by 1 cm
vent located on the opposite side of the compartment. The governing equation for the change in
pressure in the compartment is [60]:
 2
1 ṁ
∆P = (3.20)
2ρ CA
where ρ , ṁ and C take values consistent with CFAST calculations
P
ρ=
; ṁ = ρ∞V̇ ; C = 0.70 (3.21)
RT
Figure 3.15 shows how the pressure changes as air flows through the compartment.

25
 15

10

0
0 5 10 15 20 25 30

Figure 3.15: Expected and CFAST calculated values for pressure of the compartment for the case
ventilation_1.in.

3.3.2 Steady-State Air Flow

A similar setup to the one used in Section 3.2.4 is employed here. A fan blows air into the first
compartment, on the ground floor, at a rate of 1 m3 /s and then continues to the second compartment
after passing through a doorway. The air then travels to the third compartment by passing through a
ceiling vent with an area of 3 m2 and continues to compartment four through another doorway. Air
is finally extracted from compartment four at a rate of 1 m3 /s. Figure 3.16 shows temperature and
pressure as the air flows through the fourth compartment. As expected, the temperature remains at
ambient conditions and flow in the lower compartment matches flow out the upper compartment

21 1.3

20.5 1.25

20 1.2

19.5 1.15

19 1.1
0 2 4 6 8 10 0 5 10 15 20 25 30

Figure 3.16: Expected and CFAST calculated values for temperature and mass flow for the case
ventilation_2.in.

3.3.3 Non Steady-State Air Flow

The same setup that was constructed in Section 3.3.2 is used again in this scenario. All vents
and fans are the same as in the previous case, except the fan in the fourth compartment has been
converted to a round ceiling vent with an area of 4 m2 . Initially, all of the vents are closed and air is
pumped into the first compartment at a rate of 1 m3 /s, for a period of 15 s. After 15 s pass, the fan

26
is shut off. Then, at 200 s, the door between compartments one and two opens completely. Next, at
500 s, the vent between compartments two and three opens completely. At 700 s, the door between
compartments three and four opens completely. Finally, at 1000 s, the ceiling vent in compartment
four opens completely. Figure 3.17 shows how temperature and pressure change as the air flows
through the first compartment.

60 60

50
40
40

30 20

20
0
10

0 -20
0 5 10 15 20 25 0 5 10 15 20 25

Figure 3.17: Expected and CFAST calculated values for pressure and temperature of the first compartment
for the case ventilation_3.in.

3.3.4 Velocity Profile of Air Flowing Through a Doorway

A 200 kW natural gas (methane) fire burns in a 5 m by 5 m by 5 m compartment, where the walls
are assumed to be adiabatic. The heat release rate ramps up linearly to 200 kW in 30 s, then remains
steady for 5.5 min. The compartment contains a 4 m by 1.5 m door that remains completely open
throughout the simulation. The mass flow though each vertical segment of the doorway is given
by: Z t p
ṁ = C 2ρ∆P(z) w dz (3.22)
b
where C is the orifice coefficient taken to be 0.7, ρ is the gas density of the upwind compartment,
w is the width of the doorway and ∆P(z) is the pressure across the interface at elevation z. Finally,
z = b and z = t refer to the heights of the bottom and top of the vertical segment in the doorway
through which the mass flow rate is being calculated. This equation can be analytically integrated
to yield:
|∆Pt |3/2 − |∆Pb |3/2
 
∆Pt + ∆Pb 2 p
ṁ = sign C 2ρ w (t − b) (3.23)
2 3 |∆Pt | − |∆Pb |
The equation for the velocity profile of air traveling through the doorway can be produced after
rearranging equation 3.22:
s
ṁ 2|∆P(z)|
v(z) = = sign(∆P(z))C (3.24)
δAρ ρ
where δ A represents the infinitesimal area of the door segment and ρ corresponds to the density
of air at height z. Figure 3.18 shows how the velocity of air flowing through the doorway changes
from the floor to the top of the doorway.

27
 5

0
-2 0 2 4 6 8

Figure 3.18: CFAST calculated velocity profile of air traveling through the doorway in test case
ventilation_4.in.

3.3.5 Flow Through Different Vent Types

A fan is supplies air into a compartment at a constant rate of 0.1 m3 /s (0.12 kg/s). Vertical and
horizontal natural flow vents connect this compartment to two other compartments stacked verti-
cally. Finally, the structure is vented to the outside by either a vertical or horizontal flow vent in
the rooms removed from the one where air is supplied. Figure 3.19 shows a sample of the mass
flow through the vents.

28
 0.2

0.15

0.1

0.05

0
0 5 10 15

Figure 3.19: Expected and CFAST calculated values for mass flow through vents VVent_Tests.in.

29
3.3.6 Surface Opening Fraction for Vents
This case tests the computation of the surface opening fraction. A three-compartment geometry is
shown in Figure 3.20. Different horizontal, vertical, and/or mechanical vents are included in each
compartment. Given with the dimensions of the compartments, the upper layer thickness of each
compartment, and the locations and dimensions of each vent, the surface opening fraction for a
surface, χ, can be approximated using the following expression

∑N
i=1 Aopening,i
χ= (3.25)
Atotal
where N is the total number of vents, Aopening is the opening area associated with a vent, Atotal is
the total surface area. It should be noted that each compartment has 10 surfaces.
For this verification case, the upper layer thickness for each compartment is set to be 1 m
and the vent opening fraction for all vents is set to be 0, 0.5, and 1 at 0 sec, 0.5 sec, and 1 sec,
respectively. The analytical solutions and CFAST results associated with Compartment 1 (upper
and lower layers), Compartment 2, and Compartment 3 are shown in Figure 3.21. Note that the
vent opening fraction is different than surface opening fraction.

Figure 3.20: Geometry of the test case surface_opening_fraction_1.in.

30
 0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1

0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1

Figure 3.21: Expected and CFAST calculated surface opening fraction values for the case
surface_opening_fraction_1.in.

31
3.3.7 Leakage
Air is supplied to a single compartment (1 m x 1 m x 1 m) at ambient conditions via a single
mechanical vent with a flowrate of 0.01 m3 /s. Floor leakage is specified as 0.01 m2 /m2 . Flow out
the leak should equal the specified flow in through the mechanical vent. The expected mass flow
rate is just the specified volumetric flow (0.01 m3 /s) times the density of air at ambient conditions
(1.2 kg/m3 ) for an expected value of 0.012 kg/s. With a small leak at the floor, Equationp 4.5 from
the Technical Reference Guide [59] can be simplified since ∆Pt u ∆Pb so that V̇ = CA 2∆P/ρ

2
or ∆P = ρ/2 × V̇ /(CA) = 1.2/2 × (0.01/(0.7 × 0.01))2 = 1.22 Pa. In CFAST, the vent flow
coefficient is taken to be 0.7.
The expected results and CFAST results are shown in Figures 3.22.

0.02 2

0.015 1.5

0.01 1

0.005 0.5

0 0
0 5 10 15 0 5 10 15

Figure 3.22: Expected and CFAST calculated wall leakage for the case Leakage_1.in.

3.4 Heat Transfer

3.4.1 Measuring the Temperature Change of a Thermally Thin Target
A constant 10 kW natural gas (methane) fire burns in the center of a 15 m by 15 m by 15 m
compartment. Because the purpose of this case is to test the point source radiation model and
the heating of a thermally-thin target, there is a 20 m2 vent located in the center of the ceiling to
exhaust the smoke from the fire and maintain an ambient temperature lower layer. A 1.5 mm thick
sheet of plain carbon steel is placed 2 m to the right of the fire. It is oriented directly at the fire. The
target is heated by the thermal radiation from the fire, and it is cooled via convective and radiative
loss. The net radiative heat transfer to the target is given by:

χr Q̇
q̇00r = ε + 2εσ (Tg4 − Ts4 ) (3.26)
4π r2

where ε is the emissivity of the target, χr represents the radiative heat fraction of the fire, Q̇ is
the heat release rate of the fire, r is the distance between the fire and the target and σ is the
Stefan-Boltzmann constant, which has a value of 5.67 × 10−11 kW/(m2 ·K4 ). Additionally, Tg is

32
the temperature of the gas surrounding the target and Ts is the temperature of the target’s surface.
The convective heat flux to a solid surface is governed by the following equations:

q̇00c = h(Tg − Ts ) ; h = C|Tg − Ts |1/3 (3.27)

where h is the convective heat transfer coefficient and C is an empirical coefficient determined to
be 1.31 for vertical targets [61]. In order to determine what the temperature of the target’s surface
will be at any given time, the following equation must be integrated [62]:
dT q̇00 + 2q̇00c
= r (3.28)
dt δ ρ Cp

where δ , ρ and Cp are the thickness, density and specific heat capacity of the target. The 2 denotes
that both sides of the target cool via convection. Figure 3.23 shows how the surface temperature of
the target changes as the fire burns in the compartment.

26

25

24

23

22

21

20
0 10 20 30 40 50 60

Figure 3.23: Expected and CFAST calculated target surface temperature values for the case
radiation_1.in.

3.4.2 Measuring the Temperature Change of a Cylindrical Target

A 0.15 cm thick steel plate, a 2.5 cm thick steel plate and a 2.5 cm diameter steel rod are placed
in a furnace1 with uniform temperature of 500 ◦ C. The heat conduction equations normal to the
surface of each plate and the rod, respectively, are:

k ∂ 2T
 
∂T ∂T k 1 ∂ ∂T
= ; = r (3.29)
∂t ρc ∂ x2 ∂t ρc r ∂ r ∂r
where k, ρ and c are the thermal conductivity, density and specific heat capacity of the target. At
the surface of each plate and the rod, the respective boundary conditions are:
∂T ∂T
q̇00 = −k ; q̇00 = k (3.30)
∂x ∂r
1 CFAST has an option to create a uniform temperature environment for diagnostic purposes. The option must be
invoked by adding a special comment to the input file.

33
where q̇00 is the sum of the net radiative and convective heat fluxes:

q̇00r = εσ Tg4 − Ts4 ; q̇00c = h Tg − Ts ; h = C|Tg − Ts |1/3

(3.31)

where ε is the emissivity of the target (taken here to be 1), σ is the Stefan-Boltzmann constant, Tg
is the temperature of the gas surrounding the target, Ts is the temperature of the target’s surface, h
is the convective heat transfer coefficient, and C is 1.31 for vertical targets. Figure 3.24 shows how
the surface temperature for each of the plates and the rod increases with respect to time.

600

500

400

300

200

100

0
0 10 20 30 40 50 60

Figure 3.24: Expected and CFAST calculated surface temperature values for three different types of targets
for the case radiation_2.in.

3.4.3 Measuring Incident Heat Flux to Targets

Consider a 2 m by 2 m by 4 m enclosure with black walls as shown in Figure 3.25. For a pure water
vapor medium at 1 atm with wall and gas temperature being maintained at 726.85 ◦ C, the incident
heat flux to target devices, along the line (X, 2 m, 4 m) on the top surface of the enclosure and
the line (2 m, 1 m, Z) on the right surface, are obtained. Comparison between benchmark results
(exact) reported from the literature [63] and results calculated by CFAST is made. Figure 3.26
shows the incident heat flux to the target devices due to the emission from a hot medium and
Figure 3.27 shows the corresponding results due to the emission from hot walls.

Figure 3.25: Schematic of the considered geometry.

34
 34 34

32 32

30 30

28 28

26 26

24 24

22 22
0 0.5 1 1.5 2 0 1 2 3 4

Figure 3.26: Expected and CFAST calculated target incident heat flux values due to mixture emission for
the case radiation_3.in.

34 34

32 32

30 30

28 28

26 26

24 24

22 22
0 0.5 1 1.5 2 0 1 2 3 4

Figure 3.27: Expected and CFAST calculated target incident heat flux values due to wall emission for the
case radiation_3.in.

3.4.4 Total Emissivity and Hemispherical Absorptivitiy

Consider the identical case being described in Section 3.4.3. For black surfaces, the total emissivity
of the gas medium, radiating to surface Ai , can be simplified to the following form [64]
" #
Qi
; Qi = Ai − ∑ Si Sj (Tg , Tg ) σ Tg4


ε g Tg , Tg = (3.32)
Ai σ Tg4 j6=i

where Qi is the radiative heat transfer to the surface due to the emission from the hot medium from
all surrounding surfaces in the direction of surface Ai . With identical emitting temperature for all
surfaces, the hemispherical absorptivity is given to be

1 

αg Tw , Tg = ∑ Ai Fi−j − Si Sj Tw,j , Tg (3.33)
Ai j6=i

The total emissivity and the hemisperical absorptivity of the gas medium for targets on the top
surface calculated by CFAST are compared with the benchmark results reported by [65, 66]. Two

35
emitting temperatures Tw are considered. Results for Tw = 226.85 ◦ C case and Tw = 1226.85 ◦ C
case are presented in Figure 3.28.

1 1

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2

1 1

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2

Figure 3.28: Expected and CFAST total emissivity and hemispherical absorptivity of the gas medium to the
targets located at top surface for the case radiation_5.in.

3.4.5 Adiabatic Target Temperature

A target is located at the center of a 4 m by 4 m by 4 m enclosure. Assuming adiabatic conditions
for the target, an adiabatic surface temperature2 can be evaluated by solving the following energy
balance:
ε q̇00inc,rad − σ TAST
4


+ h (Tgas − TAST ) = 0 (3.34)

where, q̇00inc,rad is the incident radiative heat flux onto the surface, ε is the surface emissivity, h is
the convective heat transfer coefficient, and Tgas is the surrounding gas temperature. The closed
form equation can be obtained from Malendowsky [67].

2 CFAST has an option to create a target with adiabatic surface conditions. The option must be invoked by adding

a specification directly to the input file; it is not an input in CEdit.

36
 100

80

60

40

20

0
0 5 10 15 20

Figure 3.29: Expected and CFAST calculated values for the adiabatic surface temperature of a target in the
case target_1.in.

3.4.6 Fire Heat Flux to Targets

Consider a 6 m by 6 m by 8 m enclosure with a constant 50 kW fire located at a fixed position3
(3 m, 3 m, 4 m). The fire is assumed to be radiating uniformly in all directions in the enclosure.
Eight black target devices are located right above the fire at different vertical positions. The vertical
positions of each target are 4.1 m, 4.2 m, 4.3 m, 4.4 m, 4.5 m, 5 m, 6 m, and 8 m, respectively.
Figure 3.30 shows the fire heat flux to different target devices.

35

30

25

20

15

10

0
4 5 6 7 8

Figure 3.30: Expected and CFAST calculated values for the radiative heat flux of a fire to different target
devices in the case target_2.in.

3.5 Sprinklers
3.5.1 Utilizing a Sprinkler to Detect and Suppress a Fire
A constant 10 kW natural gas (methane) fire burns in the center of a 4 m by 4 m by 4 m compart-
ment. The compartment is sealed and a sprinkler is located at the center of the ceiling, with an
3 In CFAST, the fire can be specified accordingly such that the emission location of the fire remains at a fixed
vertical location.

37
activation temperature of 40 ◦ C and a spray density of 0.07 mm/s. The sprinkler link temperature
can be calculated by integrating the following equation [68]:
√
dTL v

= Tg − TL (3.35)
dt RTI
where TL and Tg are the link and gas temperatures, v is the gas speed, and RTI (Response Time
Index) is a measure of the sensor’s thermal inertia. After the sprinkler link temperature reaches the
activation temperature, t > tact , the sprinkler will begin to suppress the fire by diminishing its heat
release rate at the exponential rate described by [69]:

Q̇(t) = Q̇(tact ) e−(t−tact )/τ ; −1.8

τ = 3 uw (3.36)

where uw is the water spray density, expressed in units of mm/s. Figure 3.31 shows the sprinkler
link temperature rising as the fire burns in the compartment and the heat release rate decreasing
exponentially once the sprinkler has been activated.

150
10

8
100
6

50 4

0 0
0 20 40 60 80 10 20 30 40 50 60 70 80

Figure 3.31: Expected and CFAST calculated values for the sprinkler link temperature and the heat release
rate of the fire for the case sprinkler_1.in.

3.6 Fire Ignition

Two small fires are located in a single compartment some distance away from a larger initiating
fire burning in the center of the compartment. The two smaller fires are set to ignite at an incident
heat flux of 6 kW/m2 and 200 ◦ C, respectively. Figures 3.32 and 3.33 show the heat release rate,
incident heat flux, and surface temperature for the two fires overlayed with the calculated ignition
time.

38
 2500 12

10
2000

8
1500
6
1000
4

500
2

0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12

Figure 3.32: Heat release rate of the fire and incident heat flux to the fire location for the case
Ignition_Test.in.

2500 400

2000
300

1500
200
1000

100
500

0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12

Figure 3.33: Heat release rate of the fire and surface temperature at the fire location for the case
Ignition_Test.in.

3.7 Tenability

3.7.1 Gas Tenability

For the same scenario as the example in section 3.2.5, gas tenability can be calculated. A steady-
state concentration of 0.0223 % by volume of CO2 , and 0.00137 % by volume of CO is reached.
A 1 s increment of asphyxiant gas FED can by calculated by

νCO2 = eϕCO2 / 5 = e0.0223 / 5 ≈ 1 (3.37)

ϕCO · νCO2 13.7 · 1 1
∆χFED = ∆t = = 0.0000066 (3.38)
35000 35000 60

for a 1 s increment. The following figure compares the calculated FED to its expected value.

39
 10 -5
1

0.8

0.6

0.4

0.2

0
0 5 10 15

Figure 3.34: Expected and CFAST calculated values for fractional effective dose due to asphyxiant gases in
a one room test case gas_tenability.in.

3.7.2 Heat Tenability

A compartment is maintained at a constant 100 ◦ C and a target is placed in the center of the
compartment. Heat flux to the target is less than 2.5 kW/m2 . Heat tenability can be calculated by
1
∆χFED = ∆t = 0.00068 (3.39)
4.1 × 108 T −3.61
for a 1 s increment. The following figure compares the calculated FED to its expected value.

10 -3
1

0.8

0.6

0.4

0.2

0
0 5 10 15

Figure 3.35: Expected and CFAST calculated values for fractional effective dose due to convected heat in a
one room test case heat_tenability.in.

3.8 Summary of Verification Results

The tests included in this chapter test the energy and mass balance equations in CFAST and various
heat transfer (through walls, targets, and fire detectors/sprinklers) and mass transfer (flow through
vents) calculations included in the model with comparisons to analytical calculations. These com-
parisons are an integral part of a process known as regression testing, which aims to evaluate the
accuracy of outputs predicted by a numerical software package. The results from this chapter can

40
be compared to another version of CFAST as a way of measuring the quality of results between
versions. These results also help determine the impact of code additions and modifications on the
overall accuracy of CFAST and can help identify errors in the CFAST source code. Appendix B
presents the results of the verification tests for the current version of CFAST.

41
42
Chapter 4

Description of Validation Experiments

This chapter contains a brief description of the experiments that were used for model validation.
Only enough detail is included here to provide a general understanding of the model simulations.
Anyone wishing to use the experimental measurements for validation ought to consult the cited
test reports or other publications for a comprehensive description.
Much of the content follows the FDS Validation Guide [70] and the verification and validation
guide for nuclear power plants developed by the US NRC [51]. This study focused on the pre-
dicted results of the CFAST fire model and did not include an assessment of the user interface for
the model. However, all input files used for the simulations were prepared using the CFAST graph-
ical user interface (GUI) and reviewed for correctness prior to the simulations. The comparisons
between the experiments and model predictions were characterized as open calculations, i.e., the
modelers were given detailed descriptions of the test conditions, test geometry, and fire source, but
did not modify model inputs from these given conditions to improve model predictions. As such,
the comparisons in this report provide an assessment of the predictive capability of the model, but
not an assessment of the ability of different modelers to develop appropriate model inputs.

4.1 ATF Corridors Experiments

A series of eighteen experiments were conducted in a two-story structure with long hallways and
a connecting stairway in the large burn room of the ATF Fire Research Laboratory in Ammendale,
Maryland, in 2008 [71]. The test enclosure consisted of two 17.0 m long hallways connected by a
stairway consisting of two staircases and an intermediary landing. There was a door at the opposite
end of the first floor hallway, which was closed during all tests. The end of the second floor hallway
was open with a soffit near the ceiling.
The walls and ceilings of the test structure were constructed of 1.2 cm gypsum wallboard. The
flooring throughout the structure, including the stairwell landing floor, consisted of one layer of
1.3 cm thick cement board on one layer of 1.9 cm thick plywood supported by wood joists. The
first set of stairs, which had eight risers, led from the first floor up to the landing area. The second
set of stairs, which had nine risers, led from the landing area up to the second floor. The stairs
were constructed of 2.5 cm thick clear pine lumber. The two set of stairs were separated by an
approximately 0.42 m wide gap in the middle of the stairwell. This gap was separated from the
stairs by a 0.91 m tall barrier constructed of a single piece of gypsum board. The flue space was

43
open to the first floor. The flue space was separated from the second floor by a 0.9 m tall barrier
constructed of gypsum board. There was a metal exterior type door at the end of the first floor near
the burner. The door was closed during all experiments.
The fire source was a natural gas diffusion flame burner. The burner surface was horizontal,
square and 0.45 m on each side, its surface was 0.37 m above the floor, and it was filled with gravel.
The burner was located near the end of the first floor away from the stairs. A diagram of the test
structure is displayed in Figure 4.1.

Figure 4.1: Geometry of the ATF Corridors Experiments.

44
4.2 DelCo Trainer Experiments
The NIST Fire Fighting Technology Group conducted a series of experiments in two structures of
similar design located at the Delaware County (“DelCo”) Emergency Services Training Center in
Sharon Hill, Pennsylvania [72]. Three propane burners were used to provide the fire source for all
experiments, and various sensors were used to collect gas temperature, gas velocity, heat flux, and
gas concentration measurements throughout the structure.
The single level structure was instrumented with five bare-bead thermocouple arrays and two
gas sample inlet pipes at the locations shown in Fig. 4.2. Both floors of the two level structure
were instrumented with three bare-bead thermocouple arrays and one gas sample inlet pipe at the
locations shown in Fig. 4.3.

12.2 m

9.1 m
5.5 m

0.6 m N
A4
A2 A10

A3
A8
Roof Vent
A7 1.8 m
7.3 m
5.5 m
A9
4.8 m
4.6 m 4.4 m
3.0 m
A5
3.5 m
A1
2.6 m
1.6 m
1.5 m 1.3 m 1.6 m 1.5 m

Figure 4.2: Instrumentation of the single level DelCo training structure. The thermocouple arrays are de-
noted by crossed circles and the gas sampling measurement locations are denoted by hexagons at locations
A1 and A4. The burner is denoted by three cross-hatched squares.

45
 11.2 m

N
7.8 m
A7 A8

2.3 m A13

6.4 m
A14
4.6 m
A11 5.8 m A9 3.5 m
A16 2.6 m
1.6 m 0.5 m
A10 1.5 m

0.3 m A17

11.9 m

0.6 m N

A5
5.2 m A2 A3
7.0 m
A6
3.6 m 2.9 m
3.4 m
7.1 m
2.4 m
0.4 m

A1
0.4 m 0.6 m

Figure 4.3: Instrumentation of the second floor (top) and first floor (bottom) of the two level DelCo train-
ing structure. The thermocouple arrays are denoted by crossed circles and the gas sampling measurement
locations are denoted by hexagons at locations A1 and A10. The burner is denoted by three cross-hatched
squares.

46
4.3 Fleury Heat Flux Measurements
Rob Fleury, a masters degree student at the University of Canterbury in Christchurch, New Zealand,
measured the heat flux from a variety of propane fires [73]. The objective of the work was to eval-
uate a variety of empirical heat flux calculation methods. For the measurements, heat flux gauges
were mounted on moveable dollies that were placed in front of, and to the side of, burners with
dimensions of 0.3 m by 0.3 m (1:1 burner), 0.6 m by 0.3 m (2:1 burner), and 0.9 m by 0.3 m (3:1
burner). The heat release rates were set to 100 kW, 150 kW, 200 kW, 250 kW, and 300 kW. The
gauges were mounted at heights of 0 m, 0.5 m, 1.0 m, and 1.5 m relative to the top edge of the
burner.

4.4 FM Four Room Including Corridor Test Series

This data set describes a series of tests conducted in a multiple room configuration with more
complex gas burner fires than the previous data set. This study [74] was included because, in many
ways, it is similar to the smoke movement study performed at NBS [75], and permits comparisons
between two different laboratories. In addition, it expands upon that data set by providing larger
a time-varying gas burner fires in a room-corridor configuration. Fire size was about up to 1 MW
with a total volume of 200 m3 .
This study was performed to collect data allowing for variations in fire source, ventilation, and
geometry in a multi-compartment structure, especially for situations with closed doors. This test
program was carried out at Factory Mutual Research Corporation (FMRC) in West Glocester, RI,
in which 60 fire experiments were conducted in a multiple-room enclosure to furnish validation
data for theoretical fire models.
Figure 4.4 shows a diagram of the basic facility with indications of instrumentation location.
The facility was built on the floor of FMRC’s fire test building, using part of the 67 m by 76 m test
building where the ceiling height is 18.3 m. The layout in figure 25 shows a burn room and two
target rooms connected to a corridor. The corridor was 2.43 m wide x 18.89 m long x 2.43 m high.
The burn room measured 3.63 m deep x 3.64 m wide x 2.45 m high; a sealable window opening,
measuring 0.85 m square, was centered on the rear wall, 0.34 m down from the top, and a door,
measuring 0.92 m by 2.05 m high, was centered on the front wall (opening to the corridor). For
closed window experiments, the wood-framed calcium silicate board window cover was pressed
against a bead of caulking around the steel window frame and held by drop bars positioned into
slots on the outside wall.
Room 3, located opposite the burn room, measured 3.65 m deep x 3.64 m wide x 2.45 m high;
a door, measuring 0.88 m by 2.02 m high, was centered on the front wall (opening to the corridor).
Room 4, located at the opposite end of the corridor, measured 3.65 m deep x 3.65 m wide x 2.43 m
high and had a 0.88 m by 2.02 m high door centered on the front wall (opening to the corridor); an
observation alcove, measuring 1.28 m by 0.86 m by 1.99 m high, was located in the front corner of
room 4. Each room was equipped with a 102 mm inside diameter vent tube with a 61 mm inside
diameter orifice meter and thermocouple, with option of exhaust fan (tube centered 0.27 m from
the floor and 0.17 m from the closest parallel wall). An inlet vent (0.29 m2 ) used with exhaust fans
was centered 0.43 m above the floor at the end of the corridor between the burn room and room 3.
When not in use, the inlet vent was sealed with a gypsum board cover taped in place.

47
 V

TTSGS

TGS TTSS TTSGS TTSS IV

V
Fire source, gas burner
T Gas temperature array
TS Surface temperature (ceiling)
G Gas concentration TGS
S Smoke obscuration
Differential pressure array
V Vent tube (102 mm ID) fan V
IV Inlet vent used with exhaust fans (0.27 m2)

Figure 4.4: Overview of the Factory Mutual Four Room test series.

The target room doors were commercial fire doors (wood-faced composite doors with calcium
silicate cores, 14 h rated) mounted on 16 gage steel frames. The burn room door was fabricated
from 12.7 mm calcium silicate, mounted in a steel frame lined with calcium silicate. Details of the
doors and the spacings (cracks) are given in the original reference [74].
Gypsum wallboard, 12.7 mm thick, on wood studs was used throughout the experimental fa-
cility. In addition, the walls and ceiling of the burn room were overlaid with calcium silicate, also
12.7 mm thick, to harden against repeated fire exposure. The existing concrete floor of the test
building was used.
Two types of fire sources were used: 1) steady propylene fires at 56 kW on a 0.30 m diameter
sand burner and 522 kW on a 0.91 m diameter burner and 2) propylene fires on the 0.91 m diameter
burner programmed under computer control to grow with the square of time, exceeding 1 MW in
1, 2, 4, or 8 min.
The 0.91 m diameter, 0.58 in high propylene burner was used for most of the tests. Its design
consisted of a 12 gage steel container with a gas distributor near the bottom, filled with gravel to
a 67 percent height, where there was wire mesh screen, and coarse sand to the full height of the
burner. The 0.30 m diameter burner was a scaled-down version of similar design.

4.5 FM / SNL Test Series

The Factory Mutual and Sandia National Laboratories (FM/SNL) test series consists of 25 com-
partment fire experiments conducted in 1985 for the U.S. Nuclear Regulatory Commission (NRC)
by Factory Mutual Research Corporation (FMRC), under the direction of Sandia National Labora-
tories (SNL) [76, 77]. The primary purpose of these experiments was to provide data with which
to validate computer models for various types of compartments typical of nuclear power plants.
The experiments were conducted in an enclosure measuring approximately 18 m long by 12 m
wide by 6 m high, constructed at the FMRC fire test facility in Rhode Island. A drawing is in-
cluded in Fig. 4.5. All of the experiments included forced ventilation to simulate typical power

48
plant conditions. Six of the experiments were conducted with a full-scale control room mock-up
in place. Parameters varied during the experiments included fire intensity, enclosure ventilation
rate, and fire location. Only data from nineteen experiments (Tests 1-17, 21, and 22) is used in the
current study. In these experiments, the fires were fueled by a propylene gas burner, and heptane
and methanol liquid pools. In the experiments not selected, the heat release was not reported and
could not be estimated with confidence. Table 4.1 lists the test parameters.
The following information was provided by the test director, Steve Nowlen of Sandia National
Laboratory. In particular, Tests 4, 5, and 21 were given extra attention.

Heat Release Rate: The HRR was determined using oxygen consumption calorimetry in the ex-
haust stack with a correction applied for the carbon dioxide in the upper layer of the compart-
ment. The uncertainty of the fuel mass flow was not documented. Several tests selected for
this study had the same target peak heat release rate of 516 kW following a 4 min “t-squared”
growth profile. The test report contains time histories of the measured HRR, for which the
average, sustained HRR following the ramp up for Tests 4, 5, and 21 have been estimated as
510 kW, 480 kW, and 470 kW, respectively. Once reached, the peak HRR was maintained es-
sentially constant during a steady-burn period of 6 min in Tests 4 and 5, and 16 min in Test 21.
Note that in Test 21, Nowlen reports a “significant” loss of effluent from the exhaust hood that
could lead to an under-estimate of the HRR towards the end of the experiment.

Radiative Fraction: The radiative fraction was not measured during the experiment, but in this
study it is assumed to equal 0.35, which is typical for a smoky hydrocarbons. It was further
assumed that the radiative fraction was about the same in Test 21 as the other tests, as fuel
burning must have occurred outside of the electrical cabinet in which the burner was placed.

Measurements: Four types of measurements were conducted during the FM/SNL test series that
are used in the current model evaluation study, including the HGL temperature and depth, and
the ceiling jet and plume temperatures. Aspirated thermocouples (TCs) were used to make
all of the temperature measurements. Generally, aspirated TC measurements are preferable to
bare-bead TC measurements, as systematic radiative exchange measurement error is reduced.

HGL Depth and Temperature: Data from all of the vertical TC trees were used when reducing
the HGL height and temperature. For the majority of the tests, Sectors 1, 2, and 3 were used,
all weighted evenly. For Tests 21 and 22, Sectors 1 and 3 were used, evenly weighted. Sector
2 was partially within the fire plume.

49
Figure 4.5: Geometry of the FM/SNL Experiments.

50
 Table 4.1: Summary of FM/SNL Experiments. ACH stands for Air Changes per Hour.

Test Fuel Nominal Peak Fire Ventilation Room

No. Type HRR (kW) Position Rate (ACH) Configuration
1 Propylene Burner 516 Center 10 Empty
2 Propylene Burner 516 Center 10 Empty
3 Propylene Burner 2000 Center 10 Empty
4 Propylene Burner 516 Center 1 Empty
5 Propylene Burner 516 Center 10 Empty
6 Heptane Pool 500 Wall 1 Empty
7 Propylene Burner 516 Center 1 Empty
8 Propylene Burner 1000 Center 1 Empty
9 Propylene Burner 1000 Center 8 Empty
10 Heptane Pool 1000 Wall 4.4 Empty
11 Methanol Pool 500 Wall 4.4 Empty
12 Heptane Pool 2000 Wall 4.4 Empty
13 Heptane Pool 2000 Wall 8 Empty
14 Methanol Pool 500 Wall 1 Empty
15 Heptane Pool 1000 Wall 1 Empty
16 Heptane Pool 500 Corner 1 Empty
17 Heptane Pool 500 Corner 10 Empty
21 Propylene Burner 500 Cabinet 1 Furnished
22 Propylene Burner 1000 Cabinet 1 Furnished

4.6 iBMB Compartment Tests

A series of small compartment kerosene pool fire experiments, conducted at the Institut für Baustoffe,
Massivbau und Brandschutz (iBMB) of Braunschweig University of Technology in Germany in
2004 [78]. These experiments involved relatively large fires in a relatively small (3.6 m by 3.6
m by 5.7 m) concrete enclosure. Figure 4.6 shows plan, side and perspective schematic drawings
of the experimental arrangement, including the location of the fuel pan, which was located at the
center of the compartment.
A second series of fire experiments in 2004, conducted under the International Collaborative
Fire Model Project (ICFMP) involved realistically routed cable trays inside the same concrete en-
closure at iBMB [79]. The compartment was configured slightly differently with a ceiling height
of 5.6 m. A schematic diagram from plan, side, and perspective views of the experimental arrange-
ment is shown in figure 4.7.

51
 
Figure 4.6: Detailed plan, side, and perspective schematic drawings of the iBMB pool fire experimental
arrangement.

52
 
Figure 4.7: Detailed plan, side, and perspective schematic drawings of the iBMB cable fire experimental
arrangement.

53
4.7 LLNL Enclosure Tests
Sixty-four enclosure fire tests were conducted by Lawrence Livermore National Laboratory (LLNL)
in 1986 to study the effects of ventilation on enclosure fires [80]. The test enclosure was 6 m long,
4 m wide, and 4.5 m high. It contained a methane rock burner which was placed in the center of
the space. For most of the tests the burner was placed on the floor. The fires varied in size from
50 kW to 400 kW. The burner was 0.57 m in diameter and 0.23 m height.
The door (2.06 m high by 0.76 m wide) was closed and sealed for most tests, and air was
pulled through the space at rates varying from 100 to 500 g/s. In some tests the enclosure included
a plenum space, where make-up air could be injected from above or below. The test matrix is listed
in Table 4.2.

54
 Table 4.2: Summary of LLNL Enclosure Experiments.

Test Room h0 Q̇ ṁ T∞ Test Room h0 Q̇ ṁ T∞

No. Config. m kW g/s ◦C No. Config. m kW g/s ◦C

1 TL 0 200 0 23 33 PH 0 100 200 23

2 TL 0 200 0 27 34 PH 0 100 300 34
3 TL 0 400 0 27 35 PH 0 100 400 22
4 TL 0 300 0 24 36 PH 0 100 500 29
5 TL 0 50 0 28 37 PH 0 200 100 20
6 TL 0 100 0 29 38 PH 0 200 300 29
7 TL 0 100 0 35 39 PH 0 250 100 18
8 TL 0 200 0 35 40 PH 0 200 400 28
9 TL 0 200 500 33 41 PH 0 150 100 20
10 TL 0 200 100 28 42 PHE 2 200 180 30
11 TL 0 200 200 18 43 PHE 2 200 0 32
12 TL 0 200 300 21 44 PHE 1 200 180 19
13 TL 0 200 400 28 45 PHE 1 200 0 30
14 TL 0 200 400 28 46 PHE 0.6 200 180 19
15 TL 0 100 300 24 47 PHE 0.6 200 0 19
16 TL 0 200 300 21 48 PHE 0.3 200 0 21
17 PL 0 200 500 26 49 PHE 0.3 200 180 26
18 PL 0 200 400 21 50 PHE 1 200 180 21
19 PL 0 200 300 18 51 PNE 1 200 NAT 33
20 PL 0 200 200 16 52 PN 0 200 NAT 23
21 PL 0 200 100 23 53 PHGS 0 200 185 33
22 PH 0 200 190 30 54 PHGS 0 200 215 21
23 PH 0 200 215 28 55 PN 0 100 NAT 31
24 PH 0 200 205 26 56 PHGW 0 200 190 20
25 PH 0 200 205 25 57 PHGW 0 200 215 29
26 PH 0 200 500 24 58 PHX 0 200 190 18
27 PH 0 200 100 23 59 PHXE 1 200 190 24
28 PH 0 150 150 31 60 PN 0 400 NAT 22
29 PH 0 250 250 28 61 TN 0 200 NAT 31
30 PH 0 250 300 34 62 TN 0 400 NAT 22
31 PH 0 250 500 36 63 TN 0 50 NAT 28
32 PH 0 100 100 33 64 TN 0 100 NAT 17

T full compartment N natural ventilation (door open)

P plenum configuration X 3 ft extension on inlet opening
L low inlet duct GS grate on inlet, north/south configuration
H high inlet duct GW grate on inlet, east/west configuration
E elevated fire, h0

55
Figure 4.8: Geometry of the LLNL Enclosure Experiments.

56
4.8 NBS Single Room Tests with Furniture
These data describe a series of room fire tests using upholstered furniture items in a room of fixed
size but with varying opening sizes and shapes [39] conducted by the National Bureau of Standards
(NBS, former name of NIST). It was selected for its well characterized and realistic fuel sources
in a simple single-room geometry. In addition, the wide variation in opening size should provide
challenges for current zone fire models. Peak fire size was about 2.9 MW with a total room volume
of 21 m3 . A series of four single-room fire tests were conducted using upholstered furniture items
for comparison with their free burning behavior, previously determined in a furniture calorimeter.
The experiments were conducted in a single room enclosure; ventilation to the room was provided
by window openings of varying sizes. The room was equipped with an instrumented exhaust
collection system outside the window opening.
A second similar test series also utilized a single-room fire test with furniture as the fire source
[81]. It expanded upon the first data set by adding the phenomenon of wall burning. Peak fire
size was about 7 MW. The room size was similar to the first test series. Figure 4.9 illustrates the
configuration for the two test series.
The test furniture included a 28.3 kg armchair or a similar 40.0 kg love seat for the first test
series. Both were of conventional wood frame construction and used polyurethane foam padding,
made to minimum California State flammability requirements, and polyolefin fabric. A single
piece of test furniture and igniting wastebasket were the only combustibles in the test room.
For the second test series, room furnishings consisted of a 1.37 m wide x 1.91 m long x 0.53 m
high double bed, a 2.39 m X 0.89 m high headboard, and 0.51 m wide x 0.41 m deep x 0.63 m high
night table. Both headboard and night table were fabricated from 12.7 mm thick plywood. The
bedding was comprised of two pillows, two pillow cases, two sheets, and one blanket. The pillows
had a polypropylene fabric with a polyester filling. The pillow cases and sheets were polyester-
cotton. The blanket was acrylic material. The bedding was left in a "slept in" condition which was
duplicated to the degree possible in each test. In all of the tests, the fire was started with match
flame ignition of a 0.34 kg (240 mm x 140 mm x 240 mm high) wastebasket, filled with 0.41 kg of
trash, positioned adjacent to the bed.

57
 1.13 m
3.94 m

2.31 m
2.0 m

Open
Vent
2.26 m

T H TVG

1.5 m
2.31 m
2.0 m

Fire source, specimen mass loss 1.29 m

2.31 m
Fire source, gas burner

2.0 m
T Gas temperature array
G Gas concentration (CO, CO2, O2)
V Gas velocity array

Figure 4.9: Plan and elevation view schematic of experimental room for NBS single room tests with furni-
ture. Note dotted lines on burning specimen indicates vertical surface for wall burning experiments. Speci-
men and instrumentation placement are approximate.

Figure 4.10: Burning specimen during NBS single room tests with furniture.

58
4.9 NBS Multi-Compartment Test Series
The National Bureau of Standards (NBS, which is now called the National Institute of Standards
and Technology, NIST) Multi-Room Experiments consisted of 45 fire tests representing 9 different
sets of conditions were conducted in a three-room suite (see Figs. 4.11 and 4.12). The experiments
were conducted in 1985 and are described in detail in Ref. [75]. The suite consisted of two rel-
atively small rooms, connected via a relatively long corridor. The fire source, a gas burner, was
located against the rear wall of one of the small compartments. Fire tests of 100 kW, 300 kW and
500 kW were conducted. For the current study, only three 100 kW fire experiments have been
used, including Test 100A from Set 1, Test 100O from Set 2, and Test 100Z from Set 4. These
tests were selected because they had been used in prior validation studies, and because these tests
had the steadiest values of measured heat release rate during the steady-burn period.
Following is additional information provided by the test director, Richard Peacock of NIST:

Heat Release Rate: In the two tests for which the door was open, the HRR during the steady-
burn period measured via oxygen consumption calorimetry was 110 kW with an uncertainty of
about 15 %, consistent with the replicate measurements made during the experimental series
and the uncertainty typical of oxygen consumption calorimetry. It was assumed that the closed
door test (Test 100O) had the same HRR as the open door tests.

Radiative Fraction: Natural gas was used as the fuel in Test 100A. In Tests 100O and 100Z,
acetylene was added to the natural gas to increase the smoke yield, and as a consequence, the
radiative fraction increased. The radiative fraction of natural gas has been studied previously,
whereas the radiative fraction of the acetylene/natural gas mixture has not been studied. The
radiative fraction for the natural gas fire was assigned a value of 0.20, whereas a value of 0.30
was assigned for the natural gas/acetylene fires.

Measurements: Only two types of measurements conducted during the NBS test series were used
in the evaluation considered here, because there was less confidence in the other measurements.
The measurements considered here were the HGL temperature and depth, in which bare bead
TCs were used to make these measurements. Single point measurements of temperature within
the burn room were not used in the evaluation of plume or ceiling jet algorithms. This is
because the geometry was not consistent in either case with the assumptions used in the model
algorithms of plumes or jets. Specifically, the burner was mounted against a wall, and the room
width-to-height ratio was less than that assumed by the various ceiling jet correlations.

59
Figure 4.11: Photo of a 100 kW fire with the burner located against the rear wall of one of the small
compartments in the NBS Multi-Compartment test Series.

60
Figure 4.12: Plan, side and perspective schematic drawings of the NBS experimental arrangement, including
the burner.
61
4.10 NIST / NRC Compartment Experiments
These experiments, sponsored by the US NRC and conducted at NIST, consisted of 15 large-
scale experiments performed in June 2003. All 15 tests were included in the validation study.
The experiments are documented in Ref. [82]. The fire sizes ranged from 350 kW to 2.2 MW in a
compartment with dimensions 21.7 m by 7.1 m by 3.8 m high, designed to represent a compartment
in a nuclear power plant containing power and control cables. A photo of the fire seen through the
compartment doorway is shown in figure 4.13. A diagram of the test structure is displayed in
Figure 4.14.
The walls and ceiling were covered with two layers of marinate boards, each layer 0.0125 m
thick. The floor was covered with one layer of gypsum board on top of a layer of plywood.
Thermo-physical and optical properties of the marinate and other materials used in the compart-
ment are given in Ref. [82]. The room had one door and a mechanical air injection and extraction
system. Ventilation conditions, the fire size, and fire location were varied. Numerous measure-
ments (approximately 350 per test) were made including gas and surface temperatures, heat fluxes
and gas velocities.
Following are some notes provided by Anthony Hamins, who conducted the experiments:
Natural Ventilation: The compartment had a 2 m by 2 m door in the middle of the west wall.
Some of the tests had a closed door and no mechanical ventilation (Tests 2, 7, 8, 13, and 17),
and in those tests the measured compartment leakage was an important consideration. The
test report lists leakage areas based on measurements performed prior to Tests 1, 2, 7, 8, and
13. For the closed door tests, the leakage area used in the simulations was based on the last
available measurement. The chronological order of the tests differed from the numerical order.
For Test 4, the leakage area measured before Test 2 was used. For Tests 10 and 16, the leakage
area measured before Test 7 was used.
Mechanical Ventilation: The mechanical ventilation and exhaust was used during Tests 4, 5, 10,
and 16, providing about 5 air changes per hour. The door was closed during Test 4 and open
during Tests 5, 10, and 16. The supply duct was positioned on the south wall, about 2 m off
the floor. An exhaust duct of equal area to the supply duct was positioned on the opposite wall
at a comparable location. The flow rates through the supply and exhaust ducts were measured
in detail during breaks in the testing, in the absence of a fire. During the tests, the flows were
monitored with single bi-directional probes during the tests themselves.
Heat Release Rate: A single nozzle was used to spray liquid hydrocarbon fuels onto a 1 m by 2 m
fire pan that was about 0.1 m deep. The test plan originally called for the use of two nozzles to
provide the fuel spray. Experimental observation suggested that the fire was less unsteady with
the use of a single nozzle. In addition, it was observed that the actual extent of the liquid pool
was well-approximated by a 1 m circle in the center of the pan. For safety reasons, the fuel flow
was terminated when the lower-layer oxygen concentration dropped to approximately 15 % by
volume. The fuel used in 14 of the tests was heptane, while toluene was used for one test. The
HRR was determined using oxygen consumption calorimetry. The recommended uncertainty
values were 17 % for all of the tests.
Radiative Fraction: The values of radiative fraction and its uncertainty were reported as 0.44 ± 0.07
and 0.40 ± 0.09 for heptane and toluene, respectively.

62
Figure 4.13: Photograph of a 1 MW heptane fire seen through the open doorway. Photo provided by Anthony
Hamins, NIST.

Soot Yield: The values of the soot yield and its uncertainty were reported as 0.0149 kg/kg ± 0.0033 kg/kg
and 0.195 kg/kg ± 0.052 kg/kg for heptane and toluene, respectively.

63
Figure 4.14: Plan, side and perspective schematic drawings of the NIST NRC experimental arrangement.
The fuel pan and cables B, D, F, and G (dotted lines) are also shown.
64
4.11 NIST/NRC Corner, Wall, and Cabinet Experiments
In the summer of 2017, experiments were conducted in a large compartment in the NIST large
fire laboratory on behalf of the U.S. Nuclear Regulatory Commission. There were two sets of
experiments. In the first set, conducted in July, 2017, a natural gas burner was positioned either
in a corner or against a wall, and gradually moved outward. In the second set of experiments,
conducted in September, 2017, a natural gas burner was placed inside one of two steel cabinets
meant to represent typical industrial-scale electrical enclosures.
The compartment for all experiments was 11 m long, 7 m wide, and 3.8 m high. The long
dimension of the compartment ran east-west. A 1.8 m wide, 2.4 m high door was centered on the
east (short) wall.
All of the fires were fueled by one or more 30.5 cm (1 ft) square natural gas burners. Each
burner was essentially a steel box, 30.5 cm square in plan and 15 cm deep, fueled from below. The
lip of the burner was 2.5 cm (1 in) wide. A 2.5 cm thick piece of Kaowool insulation was placed
under a steel mesh to form the surface of the burner.

4.11.1 Wall and Corner Effects

Six large compartment experiments were conducted in July, 2017, where four natural gas burners
were positioned (1) in a corner and (2) against a wall, and then moved outward in stages until the
corner or wall effect became negligible. The quad burner was 60 cm by 60 cm and the burner
surface was 54 cm above the floor. The corner fire was located in the southwest corner of the large
compartment. The wall fire was centered on the south (long) wall.
The experiments began with the quad burner in the corner or against the wall for the first
30 min. At 30 min, the burner was moved so that its edge(s) was 10 cm away from the wall(s). It
remained for 15 min, after which it was moved to 20 cm, 30 cm, 50 cm, 100 cm, and 160 cm, each
time remaining 15 min for a total experiment time of 2 h.
A three-dimensional array of thermocouples was positioned on a track mounted to the ceiling
above the burner. The purpose of this array was to measure maximum plume temperatures at
heights of 2.1 m, 2.7 m, and 3.4 m above the floor. As the burner moved, the thermocouple array
moved with it. For the corner fire experiments, when the burner was at the 0 cm, 10 cm, and
20 cm positions, the thermocouple array overhead remained at its original location in the corner.
As the burner moved beyond 20 cm, the thermocouple array was moved the same amount so that
the burner was always below the array in the same position. In other words, for the corner fire
experiments, after the center point of the burner reached the point directly below the position 18
on the diagram below, the burner and array moved together, maintaining their relative position.
The experimental data consists primarily of thermocouple measurements. The key to the col-
umn names are as follows:

• TC-AG-01 through TC-AG-29 are the thermocouples at the top of the cage, 46 cm below the
ceiling (see pattern below).

• TC-BG-01 through TC-BG-29 are the thermocouples at the mid-level of the cage, 107 cm
below the ceiling (see pattern below).

65
 • TC-CG-01 through TC-CG-29 are the thermocouples at the bottom of the cage, 168 cm below
the ceiling (see pattern below).

• TC-WT-01 through TC-WT-13 are the thermocouples of the vertical array called the West
Tree. The array was 2.75 m from the west (short) wall and 3.5 m from the south (long) wall.
TC-WT-01 was located 2 cm below the ceiling, and the rest were spaced 30 cm apart.

• TC-ET-01 through TC-ET-13 are the thermocouples of the East Tree. The array was 2.75 m
from the east (short) wall and 3.5 m from the south (long) wall. TC-ET-01 was located 2 cm
below the ceiling, and the rest were spaced 30 cm apart.

• TC-C-01 through TC-C-11 are the thermocouples 2 cm from the corner above the corner fire.
TC-C-01 was located 2 cm below the ceiling, and the rest were spaced 30 cm apart.

• TC-W-01 through TC-W-11 are the thermocouples 2 cm from the wall above the wall fire.
TC-W-01 was located 2 cm below the ceiling, and the rest were spaced 30 cm apart.

• HRR (cal) is the heat release rate of the fire as measured using oxygen consumption calorime-
try. HRR (NG) is the heat release rate determined from the mass flow rate of natural gas.

The East and West Tree thermocouples were used to estimate the height of the hot gas layer
(HGL), and the average temperatures of the upper and lower layers. Also, the three horizontal
arrays of thermocouples above the burner were processed by first taking a 2 min running average
of each TC, and then choosing the maximum value for each of the three elevations above the fire.
These were taken as approximate centerline plume temperatures at each height. These experimen-
tal files are labelled with “HGL” and “Plume”, respectively.

4.11.2 Cabinet Effects

In this second series of experiments, conducted in September, 2017, two different mock steel
cabinets were used. Each cabinet was constructed of 12 gauge (2.8 mm or 7/64 in) steel plate with
openings as shown in Figs. 4.16 and 4.17. The large cabinet was nominally 0.9 m by 0.9 m by
2.1 m and the medium size cabinet was 0.6 m by 0.6 m by 2.1 m. The openings near the top of
each cabinet were sometimes covered with a steel grill, shown in Fig. 4.18.
For the first set of experiments (1-6), the large cabinet was positioned with its front opening
facing eastward towards the opening of the test compartment. Its left side was 1.8 m from the south
wall and its front side was 5.8 m from the east wall. Two 0.3 m by 0.3 m natural gas burners were
placed side by side in the cabinet from the perspective of the cabinet front opening. The top of
the burner was 50 cm above the floor of the cabinet. For Tests 1-4, the front door of the cabinet
was closed, and the heat release rate was initially set to 50 kW for 30 min, then it was increased to
100 kW for 15 min, 200 kW for 15 min, and 400 kW for 15 min. For Tests 5-6, the front door was
opened, and the heat release rate was set to 200 kW, 400 kW, and 700 kW for 15 min each, and
then 1000 kW for 5 min, a total of 50 min.
In the second set of experiments (7-10), the medium-sized cabinet was positioned so that its
front was the same distance from the east wall as the large cabinet, and its left side was 2.0 m (6.5
ft) from the south wall. A single 30 cm by 30 cm gas burner was centered within. For the closed

66
 6
25 26 27 28 29
t t t t t

24
t

19 20 21 22 23
t t t t t

18
t

0.91 m (36 in)

13 14 15 16 17
t t t t t

12
t

7 8 9 10 11
t t t t t

6
t

1 2 3 4 5
t t t t t ?

South Wall

Figure 4.15: Diagram of thermocouple layout for NIST/NRC Corner Effects experiments.

door tests, the heat release rate was 25 kW, 50 kW, 100 kW, and 200 kW, each for 15 min. For the
open door tests, the heat release rate was 40 kW, 80 kW, 200 kW, and 325 kW, each for 15 min.
In the third set of experiments (11-12), the cabinet was removed, and two 30 cm by 30 cm
burners were spaced 0.9 m (3 ft) apart, side to side. One of the burners was centered under the
array of thermocouples. Both burners were 2.0 m from the south wall. These experiments used the
same heat release rate sequence as the open and closed door large cabinet experiments.
The data files for these experiments are labelled, NIST_NRC_Cabinet_Test_n.csv. These
files contain the same measurement positions as the corner and wall experiments, with the follow-
ing additional measurements:

• PT-1 through PT-8 are plate thermometers positioned 0.6 m (2 ft) from each side of the cabinet
at heights of 0.8 m (2.5 ft) and 1.4 m (4.5 ft). PT-1 is the upper plate on the left side. PT-2 is
lower left. PT-3 is upper back. PT-4 is lower back. PT-5 is upper front. PT-6 is lower front.
PT-7 is upper right. PT-8 is lower right.

• STC-1 through STC-6 are sheathed thermocouples within the cabinet, 15 cm (6 in) from the
left side, centered. STC-1 is 6 cm (2.5 in) from the top. STC-2 through STC-6 are 30 cm,
60 cm, 90 cm, 120 cm, and 150 cm from the top, respectively.

67
Figure 4.16: Large cabinet drawing, NIST/NRC Corner Effects Experiments.

68
Figure 4.17: Medium-sized cabinet drawing, NIST/NRC Corner Effects Experiments.

69
 Figure 4.18: Cabinet grill, NIST/NRC Corner Effects Experiments.

• TC-Cab is a single 24 gauge Type K thermocouple welded to the center of the back side on the
outside of the cabinet. For Test 11, this TC was placed just under the Kaowool surface of the
burner, and for Test 12, it was placed just above the surface.

The three dimensional array of thermocouples used in the wall and corner experiments was posi-
tioned over the front of the cabinet, such that TC positions 1, 7, 13, 19, and 25 in Fig. 4.15 were
just above the upper front edge of the cabinet.
The test matrix is as follows:

Table 4.3: Summary of NIST/NRC Cabinet Experiments.

Test Cabinet Front Door Top Vents Upper Side Vents HRR (kW)
1 Large Closed Closed Grill 50, 100, 200, 400
2 Large Closed All open Grill 50, 100, 200, 400
3 Large Closed Closed Front open, all others closed 50, 100, 200, 400
4 Large Closed Closed Front and back open, others closed 50, 100, 200, 400
5 Large Open Closed Front and back open, others closed 200, 400, 700, 1000
6 Large Open Open All open 200, 400, 700, 1000
7 Medium Closed Closed Grill 25, 50, 100, 200
8 Medium Closed Closed Open 25, 50, 100, 200
9 Medium Open Closed Open 40, 80, 200, 325
10 Medium Open Closed Closed 40, 80, 200, 325
11 None N/A N/A N/A 200, 400, 700, 1000
12 None N/A N/A N/A 50, 100, 200, 400

70
 Exhaust fan on 2nd floor

7th floor

T
6th floor

5th floor Stairwell

T T
4th floor
Elevator

Supply
4.25 m3/s
T TGS
3rd floo (pressurized at 6 air changes per hour)r

2nd floor (exhausted at 6 air changes per hour)

Fire source, gas burner
T Gas temperature array
Ground Level
G Gas concentration
S Smoke obscuration Ground floor (pressurized at 6 air changes per hour)
Differential pressure (2nd floor hallway to stairway
and 2nd floor hallway ceiling to 3rd floor
Basement Open Exterior
Door
Supply
fan on
1st and 3rd
floors

Figure 4.19: Overview of the NIST Seven-story hotel test series including smoke control.

4.12 NIST Seven-story Hotel Tests

By far the most complex test, this data set is part of a series of full-scale experiments conducted to
evaluate zoned smoke control systems, with and without stairwell pressurization citeKlote:1990. It
was conducted in a seven story hotel with multiple rooms on each floor and a stairwell connecting
all floors. This data set was chosen because it would challenge the scope of most current fire
models. Measured temperatures and pressure differences between the rooms and floors of the
building are extensive and consistent. Peak fire size was 3 MW with a total building volume of
140 000 m3 .
Smoke movement and the performance of smoke control systems were studied in a seven story
hotel building with smoke generated from wood fires and theatrical smoke. A total of 12 single
experiments were conducted under a variety of conditions: two different fire sizes; sprinklered vs
non-sprinklered wood fires; zoned smoke control on or off; stairwell pressurization on or off; with
and without ventilation to the outside; and open and closed doors.
The Plaza Hotel building was a masonry structure consisting of two wings, one three stories
and the other seven stories tall. The two wings were built at different times. The wings were
connected to each other at only one location on each floor. The connections between the wings
at each floor were sealed off, and the fires were set on the second floor of the seven-story wing,
using the shorter wing as an instrumentation area. Areas of the second floor were fire hardened to
minimize structural damage to the building.
The smoke control systems were designed using the methods presented in the smoke control
manual of the American Society of Heating, Refrigerating, and Air-Conditioning Engineers [83],
and the design analysis is discussed in detail by Klote [84]. The minimum design pressure differ-
ence was 25 Pa, meaning that the system should be able to maintain at least this value without a
fire. The Plaza Hotel building had no central forced air heating, ventilating, and air-conditioning
(HVAC) system, so a dedicated system of fans and ducts was installed for zoned smoke control and
stairwell pressurization. The smoke control system consisted of the three 0.944 m3 /s centrifugal
fans shown in 4.19, plus another centrifugal fan (not shown) located outside and supplying 4.25
m3 /s of pressurization air to the stairwell at the first floor. The smoke control system is illustrated

71
in figure 4.19. All the test fires were located in the second floor smoke zone. This smoke was ex-
hausted at about six air changes per hour. The first and second floors were pressurized at about six
air changes per hour. When the stairwell pressurization system was activated, the exterior stairwell
door was open. This approach is intended to minimize fluctuations due to opening and closing
doors.

4.13 NIST Smoke Alarm Experiments

A series of experiments was conducted by NIST to measure the activation time of ionization and
photoelectric smoke alarms in a residential setting [85]. Tests were conducted in actual homes
with representative sizes and floor plans, utilized actual furnishings and household items for fire
sources, and tested actual smoke alarms sold in retail stores at that time. Thirty-six tests were
conducted in two homes; 27 in a single-story manufactured home, and 8 in a two-story home.
Figure 4.20 shows a diagram of the layout and instrumentation in the manufactured home. The
primary partitioning of the 84.7 m2 floor plan consisted of three bedrooms: one full bathroom,
one kitchen/dining area, one living room, and two hallways. For testing, the doors to bedroom 3
and the bathroom were always closed. The ceiling was peaked on the long axis, reaching a height
of 2.4 m. The outside walls were approximately 2.1 m in height. The slope of the ceiling was
approximately 8.4◦ .
An upholstered chair, mattress, or pan of cooking oil was used as the fire source in each test.
There were three primary ignition sources: flaming, smoldering, and cooking. The flaming ignition
source used a moderate flame source to quickly ignite the fuel package. Groups of smoke alarms
were located in the room of fire origin, at least one bedroom, and in a central location. Five stations
(A-E) containing smoke alarm arrays were mounted parallel to the ceiling.
Nine experiments that were conducted in the single-story manufactured home were selected
for model validation. Only tests that used a flaming ignition source with a couch or mattress fuel
package were considered; the cooking oil fires and tests that used a smoldering ignition source
were not considered.
Although a load cell was used in the experiments to measure the mass loss rate of the fuel
package, the mass loss data were not reliable enough to reconstruct the HRR curves for each test.
Instead, the HRR curves were determined by approximating the fire growth using a t 2 ramp, as
in Eq. (4.1). The parameters for the t 2 ramp were calibrated in FDS by using the temperature
measured at the highest thermocouple in the tree (2 cm below the ceiling) in the fire room.
 t 2
Q̇ = Q̇0 (4.1)
τ
A time offset was used to align the predicted ceiling thermocouple temperatures with the measured
temperatures. This offset is reported as the time at which the t 2 ramp begins. The t-squared
calibration parameters and time offsets for the HRR ramps are shown in Table 4.4. Additionally, the
ignition source had a small effect on the measured ceiling thermocouple temperatures. Therefore,
the size of the ignition source was approximated as either 3 kW or 7 kW, and the time offset of the
ignition source was also calibrated by using the measured ceiling thermocouple temperatures. The
resulting HRR curve was input into CFAST.
A summary of the nine tests selected for model validation is shown in Table 4.4.

72
Figure 4.20: Geometry of the NIST Smoke Alarm Experiments.

73
 Table 4.4: Summary of NIST home smoke alarm experiments selected for model validation.

Test No. Fire Source Fire Location Q̇0 (kW) τ (s) Time Offset (s)
SDC02 Chair Living Room 150 180 20
SDC05 Mattress Bedroom 200 180 20
SDC07 Mattress Bedroom 350 180 50
SDC10 Chair Living Room 150 180 40
SDC15 Chair Living Room 225 400 180
SDC33 Chair Living Room 100 180 10
SDC35 Chair Living Room 100 180 10
SDC38 Mattress Bedroom 120 180 25
SDC39 Mattress Bedroom 200 180 25

4.14 NIST Vent Study

A series of 15 reduced-scale enclosure experiments were conducted during the summer of 2017 by
Summer Undergraduate Research Fellows (SURF) Fateema Farzana and Cory Schovanec. There
is no test report or paper describing these experiments; only what is included here.

Enclosure Geometry

A drawing of the enclosure is given in Fig. 4.21. The enclosure consisted of two compartments
stacked one on top of the other. The interior lateral dimensions of each compartment were 119 cm
by 121 cm. The height of the lower compartment was 59 cm, and the upper was 61 cm. The
enclosure was located within a vented laboratory space that was approximately 3 m by 3 m by
2.8 m high. The floor of this space was tiled, but a single sheet of gypsum board served as the
lower floor of the test enclosure.
The front door was open in all experiments. For some portion of some of the experiments, the
second floor was completely sealed. The leakage area was approximately 18 cm2 , measured at an
over-pressure of 25 Pa.

Material Properties

The walls, ceiling and floor of each compartment was 1.6 cm (5/8 in) Type X gypsum board. Wood
studs formed the exterior frame. Thermo-physical properties of the gypsum board were taken from
in Ref. [86] and manufacturer literature. It was assumed that the specific heat was 1.089 kJ/(kg ·
K), thermal conductivity 0.15 W/m/K, and density 673 kg/m3 . Additionally, a layer of kaowool,
shown in Fig. 4.21 was used to seal the gap between the removable roof and the second story walls.
Aluminum tape was used to seal all other seams.

74
 DETAIL C wood stud
PLAN VIEW SCALE 1/3 5
8 in gypsum board
Left Ceiling Vent Right Ceiling Vent
(dimensions vary) 1
2 in kaowool

30 cm 10 cm
10 cm
84 cm
121 cm

10 cm 87 cm
C
10 cm

30 cm 30 cm

84 cm 17 cm

119 cm

Indicates position of both burner

and 2nd floor ceiling vent thermocouples

FRONT VIEW RIGHT VIEW

Indicates position of
identical windows
on either side

10 cm

59 cm 20 cm 79 cm 61 cm

30 cm 30 cm

20 cm
59 cm

30 cm
49 cm

2 cm
Note that door sill begins
above height of floor due
to external wood stud

Figure 4.21: Geometry of the compartment from the NIST Vent Study

75
Burner
For all experiments, a 10 cm square propane burner fueled at a rate of 1.65 L/min generated a
2.5 kW fire according to the following calculation:

L 1 min 1 m3 kg kJ
1.65 × × × 1.967 3 × 46, 300 = 2.50 kW (4.2)
min 60 s 1000 L m kg
Note that the mass flow controller (Sierra Instruments SmartTrak 50) assumed standard conditions
to be 0 ◦ C and 101325 Pa. For Tests 13-15, a Dwyer flow meter was used in place of the mass flow
controller. The flow meter had a flow range of 4 L/min air.

Thermocouples
Eight Type-K thermocouples were inserted at each level to measure the vertical temperature profile.
The thermocouples formed a vertical array at 84 cm from the left wall of the enclosure, and 17 cm
from the front wall. TC-1 was defined as the uppermost thermocouple, with heights defined as the
vertical distance from the compartment specific floor.

Table 4.5: Heights of the thermocouples above the floor of each level of the enclosure

Floor 2 TC’s 1 2 3 4 5 6 7 8
Height (cm) 56.5 50.8 45.5 41.0 36.0 29.8 19.8 10.5
Floor 1 TC’s 9 10 11 12 13 14 15 16
Height (cm) 51.8 47.0 40.64 35.6 30.48 25.7 16.5 7.0

Test Procedure
Each experiment lasted 100 min with a 5 min cool down period. Table 4.6 indicates the times
when vents and windows were opened after the start of each experiment. For Test No. 1-4, two
trials were performed. In each case, the difference in temperature remained within 3 percent, a
difference of less than 1 ◦ C. For this reason, no further replicates were conducted.

76
 Table 4.6: Vent State by Experiment: Time Opened

Test Front Left Right Left Right Left Right Right Roof
No. Window Window Window Vent Vent Vent Area Vent Area Vent Vent
(min) (min) (min) (min) (min) (cm2 ) (cm2 ) Shape (min)
1 0 0 0 Closed 0 0 100 Square Closed
2 0 0 0 Closed 20 0 100 Square Closed
3 60 40 Closed Closed 20 0 100 Square Closed
4 0 0 0 Closed 0 0 400 Square Closed
5 0 0 0 Closed 20 0 400 Square Closed
6 60 40 Closed Closed 20 0 400 Square Closed
7 0 0 0 0 0 100 400 Square Closed
8 0 0 0 20 40 100 400 Square Closed
9 80 60 Closed 20 40 100 400 Square Closed
10 0 0 0 Closed 0 0 100 Circle Closed
11 0 0 0 Closed 20 0 100 Circle Closed
12 60 40 Closed Closed 20 0 100 Circle Closed
13 0 0 0 0 0 100 400 Square 0
14 0 0 0 20 40 100 400 Square 60
15 Closed Closed Closed 20 40 100 400 Square 60

4.15 PRISME Project

PRISME is the name of a fire test program conducted under the auspices of the Organization
for Economic Cooperation and Development, Nuclear Energy Agency (OECD/NEA). The exper-
iments were conducted at the French Institut de radioprotection et de sûreté nucléaire (IRSN) at
Cadarache. A variety of experiments were conducted to study ventilation effects, electrical cable
failure, and leakage. The test reports are not publicly available, but an entire edition of Fire Safety
Journal documented various experimental and modeling studies [87].
The PRISME DOOR series consisted of six experiments, five of which involving two compart-
ments connected by an open door (Tests 1-5) and one involving a third compartment (Test 6). The
compartments were 5 m by 6 m by 4 m high. A well-instrumented ventilation system supplied
air and exhausted combustion products at specified rates, but the thermal expansion of the gases
caused these rates to change, a phenomenon that was intended to test the ventilation capabilities of
the models.

4.16 SP Adiabatic Surface Temperature Experiments

In 2008, three compartment experiments were performed at SP Technical Research Institute of
Sweden under the sponsorship of Brandforsk, the Swedish Fire Research Board [88]. The objective
of the experiments was to demonstrate how plate thermometer measurements in the vicinity of a

77
simple steel beam can be used to supply the boundary conditions for a multi-dimensional heat
conduction calculation for the beam. The adiabatic surface temperature was derived from the plate
temperatures and used by TASEF, a finite-element thermal-structural program.
The experiments were performed inside a standard compartment designed for corner fire testing
(ISO 9705). The compartment is 3.6 m deep, 2.4 m wide and 2.4 m high and includes a door
opening 0.8 m by 2.0 m. The room was constructed of 20 cm thick light weight concrete blocks
with a density of 600 kg/m3 ± 100 kg/m3 . The heat source was a gas burner run at a constant
power of 450 kW. The top of the burner, with a square opening 30 cm by 30 cm, was placed 65 cm
above the floor, 2.5 cm from the walls. A single steel beam was suspended 20 cm below the ceiling
along the centerline of the compartment. There were three measurement stations along the beam at
lengths of 0.9 m (Position A), 1.8 m (Position B), and 2.7 m (Position C) from the far wall where
the fire was either positioned in the corner (Tests 1 and 2), or the center (Test 3). The beam in Test
1 was a rectangular steel tube filled with an insulation material. The beam in Tests 2 and 3 was an
I-beam. A diagram of the room used in Test 2 is displayed in Figure 4.22.

Figure 4.22: Geometry of the SP/AST compartment for Test 2.

78
 A second series of experiments involving plate thermometers was carried out in 2011 [89]. A
6 m long, 20 cm diameter vertical steel column was positioned in the center of 1.1 m and 1.9 m
diesel fuel and 1.1 m heptane pool fires. Gas, plate thermometer, and surface temperatures were
measured at heights of 1 m, 2 m, 3 m, 4 m, and 5 m above the pool surface. These experiments are
notable because the column is partially engulfed in flames.

4.17 Steckler Compartment Experiments

Steckler, Quintiere and Rinkinen performed a set of 55 compartment fire tests at NBS in 1979
[90]. The compartment was 2.80 m by 2.80 m by 2.13 m high1 , with a single door of various
widths, or alternatively a single window with various heights. A 30 cm diameter methane burner
was used to generate fires with heat release rates of 31.6 kW, 62.9 kW, 105.3 kW and 158 kW.
Vertical profiles of velocity and temperature were measured in the doorway, along with a single
vertical profile of temperature within the compartment. A full description and results are reported
in Reference [90]. The basic test matrix is listed in Table 4.7. Note that the test report does
not include a detailed description of the compartment. However, an internal report2 by the test
sponsor, Armstrong Cork Company, reports that the compartment floor was composed of 19 mm
calcium silicate board on top of 12.7 mm plywood on wood joists. The walls and ceiling consisted
of 12.7 mm ceramic fiber insulation board over 0.66 mm aluminum sheet attached to wood studs.
A diagram of the compartment is displayed in Fig. 4.23.

1 The test report gives the height of the compartment as 2.18 m. This is a misprint. The compartment was 2.13 m
high.
2 Technical
Research Report, Fire Induced Flows Through Room Openings - Flow Coefficients, Project 203005-
003, Armstrong Cork Company, Lancaster, Pennsylvania, May, 1981.

79
 Table 4.7: Summary of Steckler compartment experiments.

Opening Opening HRR Burner Opening Opening HRR Burner

Test Width Height Q̇ Location Test Width Height Q̇ Location
(m) (m) (kW) (m) (m) (kW)
10 0.24 1.83 62.9 Center 224 0.74 0.92 62.9 Back Corner
11 0.36 1.83 62.9 Center 324 0.74 0.92 62.9 Back Corner
12 0.49 1.83 62.9 Center 220 0.74 1.83 31.6 Back Corner
612 0.49 1.83 62.9 Center 221 0.74 1.83 105.3 Back Corner
13 0.62 1.83 62.9 Center 514 0.24 1.83 62.9 Back Wall
14 0.74 1.83 62.9 Center 544 0.36 1.83 62.9 Back Wall
18 0.74 1.83 62.9 Center 512 0.49 1.83 62.9 Back Wall
710 0.74 1.83 62.9 Center 542 0.62 1.83 62.9 Back Wall
810 0.74 1.83 62.9 Center 610 0.74 1.83 62.9 Back Wall
16 0.86 1.83 62.9 Center 510 0.74 1.83 62.9 Back Wall
17 0.99 1.83 62.9 Center 540 0.86 1.83 62.9 Back Wall
22 0.74 1.38 62.9 Center 517 0.99 1.83 62.9 Back Wall
23 0.74 0.92 62.9 Center 622 0.74 1.38 62.9 Back Wall
30 0.74 0.92 62.9 Center 522 0.74 1.38 62.9 Back Wall
41 0.74 0.46 62.9 Center 524 0.74 0.92 62.9 Back Wall
19 0.74 1.83 31.6 Center 541 0.74 0.46 62.9 Back Wall
20 0.74 1.83 105.3 Center 520 0.74 1.83 31.6 Back Wall
21 0.74 1.83 158.0 Center 521 0.74 1.83 105.3 Back Wall
114 0.24 1.83 62.9 Back Corner 513 0.74 1.83 158.0 Back Wall
144 0.36 1.83 62.9 Back Corner 160 0.74 1.83 62.9 Center∗
212 0.49 1.83 62.9 Back Corner 163 0.74 1.83 62.9 Back Corner∗
242 0.62 1.83 62.9 Back Corner 164 0.74 1.83 62.9 Back Wall∗
410 0.74 1.83 62.9 Back Corner 165 0.74 1.83 62.9 Left Wall∗
210 0.74 1.83 62.9 Back Corner 162 0.74 1.83 62.9 Right Wall∗
310 0.74 1.83 62.9 Back Corner 167 0.74 1.83 62.9 Front Center∗
240 0.86 1.83 62.9 Back Corner 161 0.74 1.83 62.9 Doorway∗
116 0.99 1.83 62.9 Back Corner 166 0.74 1.83 62.9 Front Corner∗
122 0.74 1.38 62.9 Back Corner ∗ Raised burner

80
Figure 4.23: Geometry of the Steckler Compartment Experiments.

81
4.18 UL/NFPRF Sprinkler, Vent, and Draft Curtain Study
In 1997, a series of 34 heptane spray burner experiments was conducted at the Large Scale Fire
Test Facility at Underwriters Laboratories (UL) in Northbrook, Illinois [91]. The experiments were
divided into two test series. Series I consisted of 22 4.4 MW experiments. Series II consisted of
12 10 MW experiments. The objective of the experiments was to characterize the temperature and
flow field for fire scenarios with a controlled heat release rate in the presence of sprinklers, draft
curtains, and smoke & heat vents.
The Large Scale Fire Test Facility at UL contains a 37 m by 37 m (120 ft by 120 ft) main fire
test cell, equipped with a 30.5 m by 30.5 m (100 ft by 100 ft) adjustable height ceiling. The layout
of the experiments is shown in Figs. 4.24 and 4.25.

Ceiling: The ceiling was raised to a height of 7.6 m and instrumented with thermocouples and
other measurement devices. The ceiling was constructed of 0.6 m by 1.2 m by 1.6 cm UL
fire-rated Armstrong Ceramaguard (Item 602B) ceiling tiles. The manufacturer reported the
thermal properties of the material to be: specific heat 753 J/(kg·K), thermal conductivity
0.0611 W/(m·K), and density 313 kg/m3 .

Draft Curtains: Sheet metal, 1.2 mm thick and 1.8 m deep, was suspended from the ceiling for
16 of the 22 Series I tests, enclosing an area of about 450 m2 and 49 sprinklers. The curtains
were in place for all of the Series II tests.

Sprinklers: Central ELO-231 (Extra Large Orifice) uprights were used for all the tests. The ori-
fice diameter of this sprinkler is reported by the manufacturer to be nominally 1.6 cm (0.64 in),
the reference actuation temperature is reported by the manufacturer to be 74◦ C (165◦ F). The
RTI (Response Time Index) and C-factor (Conductivity factor) were reported by UL to be
1 1
148 (m·s) 2 and 0.7 (m/s) 2 , respectively [91]. When installed, the sprinkler deflector was lo-
cated 8 cm below the ceiling. The thermal element of the sprinkler was located 11 cm below
the ceiling. The sprinklers were installed with nominal 3 m by 3 m (exact 10 ft by 10 ft) spac-
ing in a system designed to deliver a constant 0.34 L/(s·m2 ) (0.50 gpm/ft2 ) discharge density
when supplied by a 131 kPa (19 psi) discharge pressure

Vent: UL-listed double leaf fire vents with steel covers and steel curb were installed in the ad-
justable height ceiling in the position shown in Figs. 4.24 and 4.25. The vent is designed to
open manually or automatically. The vent doors were recessed into the ceiling about 0.3 m
(1 ft).

Heat Release Rate: The heptane spray burner consisted of a 1 m by 1 m square of 1.3 cm pipe
supported by four cement blocks 0.6 m off the floor. Four atomizing spray nozzles were used
to provide a free spray of heptane that was then ignited. For all but one of the Series I tests, the
total heat release rate from the fire was manually ramped up following a “t-squared” curve to a
steady-state in 75 s (150 s was used in Test I-16). The fire was ramped to 10 MW in 75 s for the
Series II tests. The fire growth curve was followed until a specified fire size was reached or the
first sprinkler activated. After either of these events, the fire size was maintained at that level
until conditions reached roughly a steady state, i.e., the temperatures recorded near the ceilings
remained steady and no more sprinkler activations occurred. The heat release rate from the

82
 North Wall 6

6 Adjustable Height Ceiling 6

260

 200 1100 - 670 100 - 120 -

?
6 Draft Curtains

r104 r97 r90 r65 r61 r57 r53

70 B 71 1200
r103 r96 r89 r64 r60 r56 r52
pp 43
p
Appp
44
69 C 45 1000
p 46
47
r102 r95 r88 r63 r59 r55 r51 48
49

68 67 66

r101 r94 r87 r62 r58 r54 r50

710 200
D
r
100 r93 r
86 r83 r80 r77 r74

r99 r92 r85 r82 r79 r

76 r73
100 - 
20 1100
r98 r91 r84
? r81 r78 r75 r72
?
6 ?
6 20 1000 20

Figure 4.24: Plan view of the UL/NFPRF Experiments, Series I. The sprinklers are indicated by the solid
circles and are spaced 3 m apart. The number beside each sprinkler location indicates the channel number
of the nearest thermocouple. The vent dimensions are 4 ft by 8 ft. The boxed letters A, B, C and D indicate
burner positions. Corresponding to each burner position is a vertical array of thermocouples. Thermocou-
ples 1–9 hang 7, 22, 36, 50, 64, 78, 92, 106 and 120 in from the ceiling, respectively, above Position A.
Thermocouples 10 and 11 are positioned above and below the ceiling tile directly above Position B, followed
by 12–20 that hang at the same levels below the ceiling as 1–9. The same pattern is followed at Positions C
and D, with thermocouples 21–31 at C and 32–42 at D.

83
 North Wall 6

r r r r r r r r r r

r r r r r r r r r r

F
r r r r B r r r r r r

r r r r A r r r r r r
1200

r r r r r C r r r r r
1000
E
r r r r r r D r r r r

r r r r r r r r r r

r r r r r r r r r r

r r r r r r r r r r

r r r r r r r r r r ?

South Wall ?

Figure 4.25: Plan view of the UL/NFPRF Experiments, Series II. The boxed letters A, B, C, D, E and F
indicate burner positions. The sprinklers are indicated by the solid circles and are spaced 10 ft apart. The
branch lines run north to south. The vents are 4 ft by 8 ft.

84
 burner was confirmed by placing it under the large product calorimeter at UL, ramping up the
flow of heptane in the same manner as in the tests, and measuring the total and convective heat
release rates. It was found that the convective heat release rate was 0.65 ± 0.02 of the total.

Instrumentation: The instrumentation for the tests consisted of thermocouples, gas analysis equip-
ment, and pressure transducers. The locations of the instrumentation are referenced in the plan
view of the facility (Fig. 4.24). Temperature measurements were recorded at 104 locations.
Type K 0.0625 in diameter Inconel sheathed thermocouples were positioned to measure (i)
temperatures near the ceiling, (ii) temperatures of the ceiling jet, and (iii) temperatures near the
vent.

4.19 UL/NIJ House Experiments

The Firefighter Safety Research Institute (FSRI) of Underwriters Laboratory (UL), with support
from the National Institute of Justice (NIJ), conducted fire experiments in a one-story ranch-style
house and a two-story colonial-style house in the large fire facility at Northbrook, Illinois, in 2016
and 2017 [92]. The goal of the experiments was to determine the effect of ventilation on fire
patterns in fully-furnished homes. As part of the project, a natural gas burner was positioned
within the unfurnished structures to evaluate changes in flow patterns as various doors and windows
were opened and closed. The test structures were instrumented with thermocouple trees, pressure
transducers, and bi-directional probes. The floor plan for the single story ranch-style house is
included in Fig. 4.26, and plans for the two story colonial-style house in Fig. 4.27. The ceiling was
2.4 m above the floor in the ranch-style house. The colonial-style house featured an atrium with a
5.6 m ceiling height. The first story ceiling in the colonial-style house was 2.5 m above the floor,
and the second story ceiling was 2.8 m above the floor.
The leakage area in the single story structure was measured to be 0.08 m2 at 10 Pa. The leakage
area in the two story structure was measured to be 0.18 m2 at 10 Pa.
The natural gas burner in each set of experiments had dimensions of 0.6 m by 0.6 m, and the
surface of the burner was approximately 0.6 m above the floor. The heat release rate for the single
story ranch house was 250 kW, and 500 kW for the two story colonial. The locations of the burner
in each house are indicated in the floor plans. The labels for the measurement devices in the one
and two story houses are listed in Tables 4.8 and 4.9, respectively. Three experiments with unique
sequences of events (e.g. opening doors and windows) were conducted in each style of house. All
experiments began with all exterior doors and windows closed and the times at which each door
and window was opened are listed Table 4.10 and 4.11.

85
 0.38 m 1.78 m 2.32 m 0.85 m 1.07 m 0.79 m 0.86 m 1.07 m 0.94 m 0.86 m 0.94 m

A B 0.30 m 0.30 m C
Back

Instrumentation
1.83 m 1.83 m Bedroom 3 Bedroom 2
Kitchen

3.58 m 1.37 m 3.92 m

2.13 m 2.13 m 1.45 m

0.10 m 0.41 m
1.42 m
0.76 m 1.26 m 1.30 m
1.75 m 0.47 m

0.30 m 1.98 m 1.98 m 1.37 m

Instrumentation
0.84 m Bedroom 1
3.96 m
1.35 m 2.74 m D 0.86 m

Living Room
Dining Room
1.78 m
1.40 m
G Front F 0.30 m
E

0.47 m 1.78 m 0.47 m 0.89 m 1.01 m 2.67 m 0.84 m 1.49 m 0.85 m 1.45 m

Icon Instrumentation
Thermocouple Array
Gas Velocity
Pressure Tap

Figure 4.26: Layout of the one story ranch-style house in the UL/NIJ experiments.

86
 0.21 m 0.85 m 3.11 m 1.78 m 1.77 m 0.91 m 1.77 m 0.62 m 1.08 m 0.85 m 1.09 m

0.20 m
A Back 0.51 m B C D
0.85 m 0.30 m
H 0.30 m 0.30 m
Kitchen 2.39 m 2.03 m 2.39 m
Family Room

4.85 m
3.05 m
3.80 m 1.52 m
2.44 m

Den
0.30 m

5.84 m
0.76 m
Living Room
Dining Room Foyer
1.60 m
0.46 m 5.13 m

Laundry Room
1.42 m 1.85 m

2.39 m 1.83 m
2.18 m
0.30 m 0.30 m 1.73 m
1.22 m
G F Front E

1.01 m 0.85 m 1.00 m 0.65 m 1.77 m 0.88 m 0.60 m 0.81 m 1.77 m 1.18 m

1.07 m 1.77 m 1.09 m 1.08 m 0.85 m 1.09 m

I J
Bedroom 2
Master Bedroom Open To Below
3.05 m

4.57 m

1.75 m
1.91 m
1.06 m 0.30 m

10.13 m 2.62 m

0.30 m 3.33 m
1.83 m 0.75 m Hallway Instrumentation
0.84 m 1.07 m 0.46 m

0.30 m
0.84 m
0.30 m
Open To
1.65 m Below
1.75 m

3.33 m 3.02 m
Bedroom 4
1.19 m Bedroom 3
M L K

1.00 m 0.85 m 1.00 m 0.66 m 1.77 m 0.60 m 4.05 m 0.81 m 1.77 m 1.18 m

Figure 4.27: Geometry of UL NIJ Colonial-style House

87
Icon Instrumentation
Thermocouple Array
 Temp. Location Pressure Location Velocity Location
1TC Bedroom 1 1PT Bedroom 1 1BDP Window E
2TC Bedroom 2 2PT Bedroom 2 2BDP Window D
3TC Bedroom 3 3PT Bedroom 3 3BDP Window B
4TC Hallway 4PT Living Room 4BDP Hallway
5TC Living Room 5PT Kitchen 5BDP Window F
6TC Dining Room 6PT Dining Room 6BDP Front Door
7TC Kitchen (East) 7BDP Window A
8TC Kitchen (West)

Table 4.8: Locations of measurement devices in the UL/NIJ ranch-style house.

Temp. Location Pressure Location Velocity Location

1TC Master Bedroom 1PT Master Bedroom 1BDP Front Door
2TC Bedroom 2 2PT Bedroom 2 2BDP Window J
3TC Bedroom 3 3PT Bedroom 3 3BDP Window K
4TC Bedroom 4 4PT Bedroom 4 4BDP Window L
5TC 2nd Floor Hallway (East) 5PT Front Corridor 5BDP Window E
6TC 2nd Floor Hallway (Middle) 6PT Living Room 6BDP Window A
7TC Family Room (West Corner) 8PT Den
8TC Family Room (Center) 9PT Family Room
9TC Den 10PT Kitchen
10TC Living Room
11TC Under 2nd Floor Hallway
12TC Foyer
13TC Dining Room
14TC Laundry Room
15TC Kitchen

Table 4.9: Locations of measurement devices in the UL/NIJ colonial-style house.

88
 Event Test 1 Test 2 Test 5
Front Door Open 300 s 1200 s 1290 s
Back Door Open 1275 s 1860 s 300 s
Window A Open 1260 s 1845 s -
Window B Open 1245 s 1830 s -
Window C Open 1230 s 900 s -
Window D Open 1215 s - 600 s
Window E Open 600 s 600 s -
Window F Open 1200 s 1815 s 1200 s
Window G Open 1300 s 1875 s -
Burner Off 900 s 1500 s 900 s

Table 4.10: Sequence of events for the one story ranch-style house in the UL/NIJ experiments.

Event Test 1 Test 4 Test 6

Front Door Open 600 s 600 s 1200 s
Back Door Open - - -
Window A Open - - 900 s
Window B Open - 900 s -
Window C Open - - -
Window D Open - - -
Window E Open - - -
Window F Open - - -
Window G Open - - -
Window H Open - - -
Window I Open - - -
Window J Open - - -
Window K Open 900 s 1260 s 600 s
Window L Open 1200 s 1320 s -
Window M Open - - -
Burner Off 1500 s 1200 s 1500 s

Table 4.11: Sequence of events for the two story colonial-style house in the UL/NIJ experiments.

89
4.20 UL/NIST Vent Experiments
In 2012, the Fire Fighting Technology Group at NIST conducted experiments at Underwriters
Laboratories (UL) in Northbrook, Illinois, to assess the change in compartment temperature due
to the opening of one or two 1.2 m square ceiling vents [93]. Four experiments were conducted
using a natural gas burner in a 6.1 m by 4.3 m by 2.4 m compartment with a single door open-
ing. The fires ranged in size from 500 kW to 2 MW, and the vents were opened and closed such
that during the four experiments there were 31 discrete time intervals in which model predictions
could be compared to quasi-steady conditions. The compartment contained two vertical arrays of
thermocouples, and the door and vents were instrumented with thermocouples and bi-directional
velocity probes. Only the thermocouple data has been used in the validation study. A diagram of
the compartment is displayed in Figure 4.28. The major test parameters are listed in Table 4.12.

Table 4.12: Summary of UL/NIST Vent experiments. Note that the 31 “experiments” are actually
discrete time intervals during the course of four separate fires.

Exp. End Time HRR No. of Exp. End Time HRR No. of
No. (s) (kW) Vents No. (s) (kW) Vents
Fire 1 Fire 3
1 1215 430 0 14 453 476 0
2 1840 430 1 15 816 476 1
3 2168 430 2 16 1153 476 2
4 2474 430 0 17 1640 1002 0
5 2955 1011 0 18 1936 1002 1
6 3170 1011 1 19 2233 1002 2
7 3604 1011 2 Fire 4
8 3840 1011 0 20 519 1011 0
9 4153 2188 0 21 967 1011 1
10 4284 2188 1 22 1325 1011 2
Fire 2 23 1559 470 2
11 565 2144 0 24 1653 470 1
12 833 2144 1 25 2013 470 0
13 931 2144 2 26 2411 470 1
27 2910 470 2
28 3399 2188 2
29 3586 2188 0
30 3803 2188 1
31 4035 2188 2

90
Figure 4.28: Geometry of the UL/NIST Experiments.

91
4.21 USN High Bay Hangar Experiments
The U.S. Navy sponsored a series of 33 tests within two hangars examining fire detection and sprin-
kler activation in response to spill fires in large enclosures [94, 95]. Experiments were conducted
using JP-5 and JP-8 fuels in two Navy high bay aircraft hangars located in Naval Air Stations in
Barber’s Point, Hawaii and Keflavik, Iceland.
The Hawaii tests were conducted in a 15 m high hangar measuring 97.8 m in length and 73.8 m
in width. Of the 13 tests conducted in the facility 11 were conducted in pans ranging from .09 m2
to 4.9 m2 in area with heat release rates varying from 100 kW to 7.7 MW. The burner was placed
in the center of the room on a scale that continuously recorded the pans weight. The facility
was equipped with a number of detection devices including thermocouples, electronic smoke and
spot heat detectors, projected beam smoke detectors, combination UV/IR optical flame detectors,
line-type heat detectors, as well as sprinklers. Measurements were recorded at a large number of
locations allowing for a through profile of compartment behavior.
It was suspected that fire plume behavior and response of detection devices in a cold building
may not have been well replicated by the experiments held in the warm hangar in Hawaii. The
Iceland tests were conducted under a 22 m barrel vaulted ceiling in a hangar measuring 45.7 m by
73.8 m. 22 tests in total were conducted. The majority of these tests fires burned JP-5 fuel with the
remainder burning JP-8. The jet fuel fires ranged in size from .06 m2 to 20.9 m2 and in heat release
rate from 100 kW to approximately 33 MW. The facility was equipped similarly to the Hawaii
hangar.

4.22 Vettori Flat Ceiling Experiments

Vettori [96] analyzed a series of 45 experiments conducted at NIST that were intended to compare
the effects of different ceiling configurations on the activation times of quick response residen-
tial pendent sprinklers. The two ceiling configurations used consisted of an obstructed ceiling,
with parallel beams 0.038 m wide by 0.24 m deep placed 0.41 m on center, and a smooth ceiling
configuration, in which the beams were covered by a sheet of gypsum board. In addition to the
two ceiling configurations, there were also three fire growth rates and three burner locations used
– a total of 18 test configurations. The fire growth rate was provided by a computer controlled
methane gas burner to mimic a standard t-squared fire growth rate with either a slow, medium, or
fast ramp up. The burner was placed in a corner of the room, then against an adjacent wall, and
then in a location removed from any wall. Measurements were taken to record sprinkler activation
time, temperatures at varying heights and locations within the room, and the ceiling jet velocities
at several other locations. A diagram of the test structure is displayed in Figure 4.29.

92
Figure 4.29: Geometry of the Vettori Flat Ceiling compartment.

93
4.23 VTT Large Hall Tests
The experiments are described in Ref. [97]. The series consisted of 8 experiments, but because of
replicates only three unique fire scenarios. The experiments were undertaken to study the move-
ment of smoke in a large hall with a sloped ceiling. The tests were conducted inside the VTT
Fire Test Hall, with dimensions of 19 m high by 27 m long by 14 m wide. Each test involved a
single heptane pool fire, ranging from 2 MW to 4 MW. Four types of predicted output were used
in the present evaluation – the HGL temperature and depth, average flame height and the plume
temperature. Three vertical arrays of thermocouples (TC), plus two thermocouples in the plume,
were compared to CFAST predictions. The HGL temperature and height were reduced from an
average of the three TC arrays using the standard algorithm described in Chapter 5. The ceiling
jet temperature was not considered, because the ceiling in the test hall is not flat, and the standard
model algorithm is not appropriate for this geometry.
The VTT test report lacks some information needed to model the experiments, which is why
some information was based on private communications with the principal investigator, Simo
Hostikka.

Surface Materials: The walls and ceiling of the test hall consist of a 1 mm thick layer of sheet
metal on top of a 5 cm layer of mineral wool. The floor was constructed of concrete. The report
does not provide thermal properties of these materials.

Natural Ventilation: In Cases 1 and 2, all doors were closed, and ventilation was restricted to
infiltration through the building envelope. Precise information on air infiltration during these
tests is not available. The scientists who conducted the experiments recommend a leakage area
of about 2 m2 , distributed uniformly throughout the enclosure. By contrast, in Case 3, the
doors located in each end wall (Doors 1 and 2, respectively) were open to the external ambient
environment. These doors are each 0.8 m wide by 4 m high, and are located such that their
centers are 9.3 m from the south wall.

Mechanical Ventilation: The test hall has a single mechanical exhaust duct, located in the roof
space, running along the center of the building. This duct had a circular section with a diameter
of 1 m, and opened horizontally to the hall at a distance of 12 m from the floor and 10.5 m
from the west wall. Mechanical exhaust ventilation was operational for Case 3, with a constant
volume flow rate of 11 m3 /s drawn through the exhaust duct.

Heat Release Rate: Each test used a single liquid fuel pan with its center located 16 m from the
west wall and 7.4 m from the south wall. For all tests, the fuel was heptane in a circular steel
pan that was partially filled with water. The pan had a diameter of 1.17 m for Case 1 and 1.6 m
for Cases 2 and 3. In each case, the fuel surface was 1 m above the floor. The trays were placed
on load cells, and the HRR was calculated from the mass loss rate. For the three cases, the fuel
mass loss rate was averaged from individual replicate tests. In the HRR estimation, the heat
of combustion (taken as 44,600 kJ/kg) and the combustion efficiency for n-heptane was used.
Hostikka suggests a value of 0.8 for the combustion efficiency. Tewarson reports a value of
0.93 for a 10 cm pool fire [98]. For the calculations reported in the current study, a combustion
efficiency of 0.85 is assumed. In general, an uncertainty of 15 % has been assumed for the
reported HRR of most of the large scale fire experiments used.

94
Radiative Fraction: The radiative fraction was assumed to be 0.35, similar to many smoky hy-
drocarbons.

A diagram of the test structure is displayed in Figure 4.30. Figure 4.31 is a photo of a 2 MW fire.

95
Figure 4.30: Plan, side and perspective schematic drawings of the experimental arrangement of the VTT
large hall fire tests, including the fuel pan
96
Figure 4.31: Photo of a 2 MW heptane fire during the VTT large hall tests. Photo provided by Simo
Hostikka, VTT.

4.24 WTC Spray Burner Test Series

As part of its investigation of the World Trade Center disaster, the Building and Fire Research Lab-
oratory at NIST conducted several series of fire experiments to both gain insight into the observed
fire behavior and to validate FDS for use in reconstructing the fires. The first series of experiments
involved a relatively simple compartment with a liquid spray burner and various structural elements
with varying amounts of sprayed fire-resistive materials (SFRM). A diagram of the compartment
is shown in figure 4.32. A complete description of the experiments can be found in the NIST WTC
report NCSTAR 1-5B [99]. The overall enclosure was rectangular, as were the vents and most of
the obstructions. The compartment walls and ceiling were made of 2.54 cm thick marinite. The
manufacturer provided the thermal properties of the material used in the calculation. The density
was 737 kg/m3 , conductivity 0.12 W/m/K, and the specific heat ranged from 1.17 kJ/kg/K at 93 ◦ C
to 1.42 kJ/kg/K at 425 ◦ C. This value was assumed for higher temperatures. The steel used to
construct the column and truss flanges was 0.64 cm thick. The density of the steel was assumed to
be 7,860 kg/m3 ; its specific heat 0.45 kJ/kg/K.
Two fuels were used in the tests. The properties of the fuels were obtained from measurements
made on a series of unconfined burns that are referenced in the test report. The first fuel was
a blend of heptane isomers, C7 H16 . Its soot yield in CFAST was set at a constant 1.5 %. The
second fuel was a mixture (40 % - 60 % by volume) of toluene, C7 H8 , and heptane. Because a fire
in CFAST only considers the burning of a single hydrocarbon fuel, the mixture was taken to be
C7 H12 with a soot yield of 11.2 %. The radiative fraction for the heptane blend was 0.44; for the

97
heptane/toluene mixture it was 0.39. The heat release rate of the simulated burner was set to that
which was measured in the experiments with the fire placed on the floor at the center of the fire
pan.

98
Figure 4.32: Geometry of the compartment used for the WTC Experiments.

99
4.25 Summary of Experiments
Experiments were chosen for inclusion based on several criteria: 1) Are the experiments well-
documented in publicly-available publications in terms of physical parameters such as compart-
ment geometry, ventilation, and fire characteristics? 2) Is the experimental data publicly available?
3) Do the experiments extend the range of one or more of the test parameters described below to
broaden the scope of the validation?
Table 4.13 presents a summary of all the experiments described in this chapter in terms of
parameters commonly used in fire protection engineering. This “parameter space” outlines the
range of applicability of the validation studies performed to date. In other words, if this guide is
to be cited as justification for using CFAST to simulate a given fire scenario, that scenario must
be similar to these experiments in the sense of having comparable physical parameters. These
parameters are explained below:

Heat Release Rate, Q̇, is the range of peak heat release rates of the fires in the test series.
p
Fire Diameter, D, is the equivalent diameter of the base of the fire, calculated D = 4A/π, where
A is the area of the base.

Ceiling Height, H, is the distance from floor to ceiling.

Fire Froude Number, Q̇∗ , is a useful non-dimensional quantity for plume correlations and flame
height estimates.
Q̇
Q̇∗ = √ (4.3)
ρ∞ c p T∞ gDD2
It is essentially the ratio of the fuel gas exit velocity and the buoyancy-induced plume velocity.
Jet fires are characterized by large Froude numbers. Typical accidental fires have a Froude
number near unity.

Flame Height relative to Ceiling Height, Lf /H, is a convenient way to express the physical size
of the fire relative to the size of the room. The height of the visible flame, based on Heskestad’s
correlation, is estimated by:
 
∗ 2/5
Lf = D 3.7 (Q̇ ) − 1.02 (4.4)

Global Equivalence Ratio, φ , is the ratio of the mass flux of fuel to the mass flux of oxygen into
the compartment, divided by the stoichiometric ratio.
 1 √
ṁf Q̇ (kW) 2 0.23 A0 H0 : Natural Ventilation
φ= ≡ ; ṁO2 =
r ṁO2 13, 100 (kJ/kg) ṁO2 0.23 ρ V̇ : Mechanical Ventilation
(4.5)
Here, r is the stoichiometric ratio, A0 is the area of the compartment opening, H0 is the height
of the opening, ρ is the density of air, and V̇ is the volume flow of air into the compartment.
If φ < 1, the compartment is considered “well-ventilated” and if φ > 1, the compartment is
considered “under-ventilated.”

100
Compartment Aspect Ratios, W /H and L/H, indicate if the compartment is shaped like a hall-
way, typical room, or vertical shaft.

Relative Distance along the Ceiling, rcj /H, indicates the distance from the fire plume of a sprin-
kler, smoke detector, etc., relative to the compartment height, H.

Relative Distance from the Fire, rrad /D, indicates whether a “target” is near or far from the fire.

101
 Table 4.13: Summary of important experimental parameters.

Q̇ D H
Test Series Q̇∗ Lf /H φ W /H L/H rcj /H rrad /D
(kW) (m) (m)
ATF Corridors 50 – 500 0.5 2.4 0.3 – 3.3 0.3 – 0.9 0.0 – 0.1 0.8 7.1 0.8 – 6.0 N/A
Fleury Heat Flux 100 – 300 0.3 – 0.6 Open 0.3 – 5.5 Open Open Open Open Open 1.7 – 3.3
FM/SNL 470 – 516 0.9 6.1 0.6 – 2.4 0.3 – 0.6 0.0 – 0.2 2.0 3.0 0.2 – 0.3 N/A
iBMB* 3500, 400 1.13, 0.79 5.7, 5.6 2.4, 0.7 0.6, 0.4 0.6, 0.1 0.6 0.6 N/A
LLNL Enclosure 50 – 400 0.6 4.5 0.2 – 1.5 0.1 – 0.4 0.1 – 0.4 0.9 1.3 0.3 – 1.0 N/A
NBS Multi-Room 110 0.3 2.4 1.5 0.5 1.0 5.1 N/A
NBS Single-Compartment 2900 – 7000 1.1 – 1.7 2.4 1.7 – 2.3 1.1 1.0 1.5 N/A
NIST Seven-Story 1130 0.7 2.6 2.2 1.1 0.7 5.6 N/A
NIST/NRC 350 – 2200 1.0 3.8 0.3 – 2.0 0.3 – 1.0 0.0 – 0.3 1.9 5.7 0.3 – 2.1 2.0 – 4.0
NIST/NRC Cabinet 200 – 400 0.3 – 0.5 2.1 0.3 – 3.7 0.2 – 0.9 1.3 – 12 0.3 0.4 N/A 1.2 – 2.0
NIST/NRC Corner 200 – 400 0.7 3.8 0.4 – 0.9 0.3 – 0.5 <0.1 1.8 2.9 0.5 – 2.3 N/A
NIST Smoke Alarm 100 – 350 1.0 2.4 0.1 – 0.3 0.2 – 0.5 1.7 8.3 1.3 – 8.3 N/A

102
PRISME 480 – 1600 0.7 – 1.1 4.0 1.1 0.5 – 0.8 0.5 1.3 1.5 0.0 – 0.5 2.3 – 5.7
SP AST 450 0.3 2.4 6.1 1.1 0.1 1.0 1.5 N/A
Steckler 31.6 – 158 0.3 2.1 0.8 – 3.8 0.3 – 0.7 0.0 – 0.6 1.3 1.3 N/A
UL/NFPRF 4400 – 10000 1.0 7.6 4.0 – 9.1 0.7 – 1.0 Open 4.9 4.9 0.6 – 3.9 N/A
UL/NIST Vents 500 – 2000 0.9 2.4 0.7 – 2.6 0.8 – 1.6 0.2 – 0.6 1.8 2.5 1.0 – 2.3
USN Hawaii 100 – 7700 0.3 – 2.5 15 0.7 – 1.3 0.1 – 0.4 Open 4.9 6.5 0 – 1.2 N/A
USN Iceland 100 – 15700 0.3 – 3.4 22 0.7 – 1.3 0.0 – 0.3 Open 2.1 3.4 0 – 1.0 N/A
Vettori Flat 1055 0.7 2.6 2.5 1.1 0.3 2.1 3.5 0.8 – 2.9 N/A
VTT Large Hall 1860 – 3640 1.4 – 1.8 19 0.7 0.2 0 1.0 1.4 0 – 0.6 N/A
WTC 1970 – 3240 1.6 3.8 0.6 – 0.9 0.8 – 1.1 0.3 – 0.5 0.9 1.8 0.0 – 0.8 0.3 – 1.3
Chapter 5

Hot Gas Layer Temperature and Depth

CFAST simulated all of the chosen experiments. Details of the comparisons with experimental
data, are provided in Appendix B. The results are organized by quantity as follows:

• hot gas layer (HGL) temperature and height

• ceiling jet temperature

• plume temperature

• flame height

• oxygen and carbon dioxide concentration

• smoke concentration

• compartment pressure

• radiation heat flux, total heat flux, and target temperature

• wall heat flux and surface temperature

The measure of model accuracy used throughout this study is related to experimental uncer-
tainty. In brief, the accuracy of a measurement, for example, a gas temperature, is related to the
measurement device, a thermocouple. In addition, the accuracy of the model prediction of the gas
temperature is related to the simplified physical description of the fire and the accuracy of the input
parameters, especially the specified heat release rate. Ideally, the purpose of a validation study is
to determine the accuracy of the model in the absence of any errors related to the measurement of
both its inputs and outputs. Because it is impossible to eliminate experimental uncertainty, at the
very least a combination of the uncertainty in the measurement of model inputs and output can be
used as a yardstick for the model predictions. Black dotted lines in figure 5.1 show this combined
uncertainty estimate for the prediction of HGL temperature. Corresponding estimates are included
for other quantities in later chapters. If the numerical prediction falls within the range of uncer-
tainty attributable to both the measurement of the input parameters and the output quantities, it is
not possible to further quantify its accuracy. At this stage, it is said that the prediction is within

103
experimental uncertainty. Reference [51] discusses this issue in detail. Red dotted lines in figure
5.1 show the calculated uncertainty in the model results compared to the experimental results.
Note that the calculation of relative difference is based on the temperature rise above ambient,
and the layer depth, that is, the distance from the ceiling to where the hot gas layer descends. Where
the model over-predicts the HGL temperature or the depth of the HGL, the relative difference is
a positive number. This convention is used throughout this report where the model over-predicts
the severity of the fire, the relative difference is positive; where it under-predicts, the difference is
negative.
Arguably the most frequent question asked about a fire is, “How hot did it become?” Average
temperature in the upper layer of a compartment is an obvious indicator to answer this question.
Peak temperature, time to peak temperature, or time to reach a chosen temperature tenability limit
are typical values of interest. Quality of the prediction (or measurement) of layer interface po-
sition is more difficult to quantify. Although observed in a range of experiments, the two-layer
assumption is in many ways just a convenience for modeling. In experimental measurements, tem-
perature is typically measured with an array of thermocouples from floor to ceiling. This floor to
ceiling temperature profile can then used to estimate a hot gas layer height and the temperature of
the upper and lower gas layers [100] [101] consistent with the two-zone assumption. Appendix A
provides details of the calculation.
From a standpoint of hazard, time of descent of the hot gas layer to a chosen level may be a
reasonable criterion The position of this gas layer interface represents the break between a clean
atmosphere near the floor and the hazardous upper layer gases. Minimum values may also be used
to indicate general agreement. For the single-room tests with furniture or wall-burning, these are
appropriate indicators to judge the comparisons between model and experiment. For the more-
closely steady-state tests , a steady-state average may better characterize the experimental results.
A good prediction of the HGL height is largely a consequence of a good prediction of its
temperature because smoke and heat are largely transported together and most numerical models
describe the transport of both with the same type of algorithm. Typically, CFAST slightly over-
predicts the HGL temperature, most often within experimental uncertainty. Hot gas layer height
is typically within experimental uncertainty for well-ventilated tests and near floor level for under-
ventilated tests where compartments are closed to the outside. For HGL height, both open- and
closed-door tests are included. For closed-door tests, visual observations typically show that the
HGL fills the entire compartment volume from floor to ceiling, inconsistent with the calculated
results for the experimental data. Thus, the comparisons with experimental values of HGL height
for closed-door tests are expected to have larger differences that those for open-door tests.
Figures 5.1 and 5.3 shows a comparison of predicted and measured values for HGL temperature
and depth. Appendix B provides individual graphs of model and experimental values.

5.1 Summary of Hot Gas Layer Temperature and Height

The two-zone assumption inherent in CFAST, modeled as a series of ordinary differential equations
that describe mass and energy conservation of flows in a multiple-compartment structure typically
provide prediction of gas layer temperature and layer height for the applications studied.
• The CFAST predictions of the HGL temperature and height are, with exceptions, are clus-
tered near experimental uncertainty. The CFAST predictions are typical of those found in other

104
103

102

101
101 102 103

102

1
10
1 2
10 10

Figure 5.1: Summary, HGL temperature, natural and forced ventilation.

105
 102

101
101 102

Figure 5.2: Summary, HGL temperature, unventilated compartments.

studies where the HGL temperature is typically somewhat over-predicted and HGL height
somewhat lower (HGL depth somewhat thicker) than experimental measurements. These dif-
ferences are likely attributable to simplifications in the model dealing with mixing between the
layers, entrainment in the fire plume, and flow through vents.

• Calculation of HGL temperature and height has higher uncertainty in rooms remote from the
fire compared to those in the fire compartment. Most likely, this is due to a combination of
the simplified vent flow predictions (based on idealized Bernoulli flow) and the assumption of
constant compartment surface thermal properties that are assumed independent of temperature.

• It is worth noting that the UL/NIST Vents tests which include large vents in the ceiling of the
test compartment show particularly higher HGL temperatures and corresponding smaller upper
layer volume compared to experimental values. Without these tests, the model uncertainties
would be significantly lower. This is likely due to the large size of the vents relative to those
included in the original experimental correlations used by CFAST.

106
101

100

10-1
10-1 100 101

101

100

-1
10
-1 0 1
10 10 10

Figure 5.3: Summary, HGL depth.

107
108
Chapter 6

Fire Plumes, Ceiling Jets, and Device

Activation

6.1 Flame Height

Flame height is recorded by visual observations, photographs, or video footage. Videos from the
NIST/NRC test series and photographs from the VTT Large Hall Test Series were used to estimate
flame height. It is difficult to precisely measure the average flame height, but the photos and videos
allow one to make estimates relative to a known burner diameter for the tests.

VTT Large Hall Test Series

The height of the visible flame in photographs has been estimated to be between 2.4 and 3 pan
diameters (3.8 m to 4.8 m). From the CFAST calculations, the estimated flame height is 4.3 m.

NIST/NRC Test Series

CFAST estimates the peak flame height to be 2.8 m, consistent with the roughly 3 m flame height
observed through the doorway during the test. The test series was not designed to record accurate
measurements of flame height.

NIST/Navy High Bay Hangar Test Series

For the 9 Iceland tests, CFAST predicts flame height within 25 % of the experimentally reported
values, with the largest relative differences for the smaller heat release rate fires. Uncertainty in
the flame height measurements for the experiments was reported to be ± 0.5 m, approaching 30 %
of the experimental values for the lower heat release rate fires.

6.2 Plume Temperature

As with the ceiling jet, CFAST includes a specific plume temperature model based on the model of
Alpert and Heskestad [102] to account for presence of higher gas temperatures near a target located

109
 103

102

101
101 102 103

Figure 6.1: Summary, Plume Centerline Temperature.

at the centerline of the fire plume. The correlation has been subjected to validation efforts by [24]
and shown to provide predictions within about 30 % of a wide range of experimental results [24].
In the model, this increased temperature has the effect of increasing the convective heat transfer to
the target. Only two of the six test series (VTT and FM/SNL) included measurements of plume
centerline temperature.
Figure 6.1 shows a comparison of predicted and measured values for plume temperature. Ap-
pendix B provides individual graphs of model and experimental values. All of the comparisons are
to the surrounding gas temperature predicted by CFAST. Comparisons to the target surface tem-
perature or target center temperature would be expected to have a smaller relative difference since
all the predictions of surrounding gas temperature are higher than experimental measurements.
Following is a summary of the predictions in the two test series.

6.3 Ceiling Jets

CFAST includes an algorithm to account for the presence of the higher gas temperatures near the
ceiling surfaces in compartments involved in a fire. In the model, this increased temperature has the
effect of increasing the convective heat transfer to ceiling surfaces. The temperature and velocity
of the ceiling jet are also available from the model by placing a heat detector at the specified loca-
tion. The ceiling jet algorithm is based on the model of Alpert and Heskestad [102], with details
described in the CFAST Technical Reference Guide [59]. The algorithm predicts gas temperature

110
 103

102

101
101 102 103

Figure 6.2: Summary, Ceiling Jet Temperature.

and velocity under a flat, unconstrained ceiling above a fire source. Only two of the six test series
(NIST/NRC and FM/SNL) involved relatively large flat ceilings.
Figure 6.2 shows a comparison of predicted and measured values for ceiling jet temperature.
Appendix B provides individual graphs of model and experimental values. Following is a summary
of the accuracy assessment for the ceiling jet predictions.

6.4 Device Activation

Smoke detector, heat detector, and sprinkler activations are all treated similarly in CFAST. Device
activation is modeled using temperatures and velocities from the ceiling jet. For rooms without
a fire (which do not have a ceiling jet), the upper layer temperature (and a default velocity of
0.1 m/s) is used. Devices are described with a characteristic activation temperature and response
time index (RTI). The RTI a measure of the sensor’s sensitivity to temperature change (thermal
inertia). For heat detectors and sprinklers, the activation temperature and RTI values are part of the
device specification. For smoke detectors, a temperature rise of 10 ◦ C and RTI of 5 (m s)1/2 were
used, consistent with values in NUREG 1805 [103]. With these inputs, the characteristic detector
temperature is modeled using the differential equation [104]
p
dTL v(t)
= (Tg (t) − TL (t)) , TL (0) = Tg (0) (6.1)
dt RT I

111
 where TL and Tg are the link and gas temperatures, v is the gas velocity, and RT I.
Figure 6.3 shows a comparison of predicted and measured values for activation times. Ap-
pendix B provides individual graphs of model and experimental values.

112
102

101
101 102

102

1
10
1 2
10 10

Figure 6.3: Summary, Smoke Detector Activation Time.

113
102

101
101 102

Figure 6.4: Summary, Sprinkler Activation Time.

114
Chapter 7

Gas Species and Smoke

CFAST simulates a fire as a mass of fuel that burns at a prescribed rate and releases both energy and
combustion products. CFAST calculates species production based on these user-defined production
yields, and both the mass burning rate and the resulting energy and species generation may be
limited by the oxygen available for combustion. Mass and species concentrations, assumed to be
homogeneous throughout each layer, are tracked by the model as gases flow through openings in a
structure to other compartments in the structure or to the outdoors.
The fire chemistry scheme in CFAST is essentially a species balance from user-prescribed
species yields and the oxygen available for combustion. For a given scenario, the user specifies the
fuel that is burned, and the yield of CO and soot. Once generated, it is a matter of bookkeeping to
track the mass of species throughout the various control volumes in a simulated building. It does,
however, provide a check of the flow algorithms within the model. Since the major species (CO
and CO2 ) are generated only by the fire, the relative accuracy of the predicted values throughout
multiple rooms of a structure should be comparable.
Gas sampling data are available from a number of the experimental tests. Species yields for
CFAST simulations were taken from the experimental reports.

7.1 Oxygen, CO2, and CO

Generation of oxygen and CO2 are calculated by CFAST based on the user-specified fuel and a
simple combustion reaction. CFAST treats CO like all other combustion products, with an overall
mass balance dependent on interrelated user-specified species yields for major combustion species.
To model CO, the user prescribes the CO yield relative to the mass burning rate. Details are
available in the CFAST Technical Reference Guide [59].
Figure 7.1 shows a comparison of predicted and measured values for oxygen and carbon diox-
ide concentrations, along with a summary of the relative difference for the tests.

115
10-1

10-2
10-2 10-1

Figure 7.1: Summary, Oxygen Concentration.

10-1

-2
10
-2 -1
10 10

Figure 7.2: Summary, Carbon Dioxide Concentration.

116
 10-2

10-3

10-4
10-4 10-3 10-2

Figure 7.3: Summary, Carbon Monoxide Concentration.

7.2 Smoke
CFAST treats smoke like all other combustion products, with an overall mass balance dependent
on interrelated user-specified species yields for major combustion species. To model smoke, the
user prescribes the smoke yield relative to the mass burning rate. A simple combustion chemistry
scheme in the model then determines the smoke particulate concentration in the form of an optical
density. Figure 7.4 shows a comparison of predicted and measured values for smoke concentration
along with a summary of the relative difference for the tests.

117
102

101
101 102

Figure 7.4: Summary, Smoke Concentration.

118
Chapter 8

Pressure

Comparisons between measurement and prediction of compartment pressure for the NIST/NRC
test series and two of the NBS furniture tests are shown in of Appendix B. Figure 8.1 shows a
comparison of predicted and measured values for compartment pressure, along with a summary of
the relative difference for the tests.
For those tests in which the door to the compartment is open, the magnitude of the pressures
are only a few Pascals; however, when the door is closed, the over-pressures are several hundred
Pascals. For both the open- and closed-door tests, CFAST predicts the pressure to within experi-
mental uncertainty with exceptions. The most notable exception is Test 16, which involved a large
(2.3 MW) fire with the door closed and the ventilation on. By contrast, Test 10 involved a 1.2 MW
fire with comparable geometry and ventilation. There is considerable uncertainty in the magnitude
of both the supply and return mass flow rates for Test 16. Compared to Test 16, Test 10 involves
a greater measured supply velocity and a lesser measured exhaust velocity. This is probably the
result of the higher pressure caused by the larger fire in Test 16. CFAST does not adjust the venti-
lation rate based on the compartment pressure until a specified cutoff pressure is reached. This is
also the most likely explanation for the over-prediction of compartment pressure in Test 16.
In general, prediction of pressure in CFAST in closed compartments is critically dependent
on correct specification of the leakage from the compartment. Compartments are rarely entirely
sealed, and small changes in the leakage area can produce significant changes in the predicted
over-pressure. Information on leakage is not typically available for experimental data sets. For
closed compartment tests, which typically included mechanical ventilation systems, leakage was
estimated by matching the experimental pressure prior to ignition by inserting a small vent in
the wall of each compartment and comparing the measured and calculated compartment pressure
without a fire in the simulation.

119
103

2
10

101

100
100 101 102 103

Figure 8.1: Summary, Compartment Pressure.

120
Chapter 9

Surface Temperature

Solid surfaces in CFAST are treated as one-dimensional transient conduction problems. This chap-
ter divides solid surfaces into two major categories – compartment linings (i.e., walls, ceiling, floor)
and targets (i.e., anything that is not a wall, ceiling, or floor). Heat transfer to the inside surface
of compartment linings and the front and rear faces (as specified by the user) of targets consists
of convection (through the use of empirical correlations) and radiation (calculated by the model
using view factors for the fire, gas layers, and compartment surfaces). Heat conduction into a solid
surface is calculated via a one-dimensional solution of the heat equation in cartesian or cylindri-
cal coordinates. The latter is particularly useful for predicting the thermal response of electrical
cables.
For compartment linings, the “outside” surface is, by default, exposed to the exterior ambient
temperature with convection and radiation calculated in a similar manner to the inside surface.
The “outside” boundary condition can also be specified as a constant temperature (i.e., the outside
surface can be at ambient temperature) or can be connected to the “outside” surface of part or all
of a second compartment. For targets, the back surface is simply pointed in a direction opposite
that of the front surface with convection and radiation calculated in a similar manner to the front
surface.

9.1 Compartment Ceiling, Wall, and Floor Temperature

In the NIST/NRC and WTC tests, thermocouples and heat flux gauges were positioned at various
locations on all four walls of the test compartments, plus on the ceilings and floors. Over the
course of 15 experiments, a number of the thermocouples and gauges failed, but because over half
of the measurement points were in roughly the same relative location to the fire, the faulty data was
discarded based on examining replicate experiments or locations on the opposite wall. Table 9.1
lists the measurement locations for each test. For each test, eight locations are used for comparison,
two on the long (mainly north) wall, two on the short (east) wall, two on the floor, and two on the
ceiling. Of the two locations for each panel, one is considered in the far-field, relatively remote
from the fire; one is in the near-field, relatively close to the fire. How close or far varied from
test to test, depending on the availability of working flux gauges. The two short wall locations are
equally remote from the fire; thus, one location is in the lower layer, one in the upper.
The WTC test measured ceiling temperatures, both at the surface and beneath a layer of mari-

121
Table 9.1: Wall thermocouple and heat flux gauge positions for the NIST/NRC series. The origin of
the coordinate system lies on the floor in the southwest corner of the compartment. The designation
“U” and “C” is irrelevant, and the last digit “2” indicates that the thermocouple is measuring the
wall temperature rather than the heat flux gauge temperature.

Name x y z Name x y z
TC North U-1-2 3.85 7.04 1.49 TC South U-1-2 3.86 0 1.49
TC North U-2-2 3.86 7.04 3.71 TC South U-2-2 3.86 0 3.82
TC North U-3-2 9.48 7.04 1.86 TC South U-3-2 9.54 0 1.86
TC North U-4-2 12.07 7.04 1.88 TC South U-4-2 12.08 0 1.86
TC North U-5-2 17.69 7.04 1.49 TC South U-5-2 17.69 0 1.50
TC North U-6-2 17.69 7.04 3.69 TC South U-6-2 17.74 0 3.70
TC East U-1-2 21.66 1.52 1.12 TC West U-1-2 0 1.59 1.12
TC East U-2-2 21.66 1.52 2.40 TC West U-2-2 0 1.59 2.42
TC East U-3-2 21.66 5.68 1.13 TC West U-3-2 0 5.70 1.12
TC East U-4-2 21.66 5.70 2.42 TC West U-4-2 0 5.70 2.42
TC Floor U-1-2 3.08 3.51 0 TC Ceiling U-1-2 3.04 3.60 3.82
TC Floor U-2-2 9.08 1.94 0 TC Ceiling C-2-2 8.99 2.00 3.82
TC Floor U-3-2 9.06 5.97 0 TC Ceiling C-3-2 9.03 5.97 3.82
TC Floor U-4-2 10.86 2.38 0 TC Ceiling C-4-2 10.79 2.38 3.82
TC Floor C-5-2 10.93 5.20 0 TC Ceiling C-5-2 10.79 5.20 3.82
TC Floor U-6-2 13.13 1.99 0 TC Ceiling C-6-2 13.00 2.07 3.82
TC Floor U-7-2 13.00 5.92 0 TC Ceiling C-7-2 12.84 5.98 3.82
TC Floor U-8-2 18.63 3.54 0 TC Ceiling U-8-2 18.71 3.54 3.82

122
nite board. Table 9.2 below lists the coordinates of the measurement locations relative to the center
of the fire pan. Names with “IN” appended are measurements made between two marinite boards
which lined the compartment surfaces.

Table 9.2: Locations of ceiling surface temperature measurements relative to the fire pan in the
WTC series.

Name x (m) y (m) z (m)

TCC 0.62 0.07 3.82
TCN3 0.62 0.67 3.82
TCS3 0.62 -0.53 3.82
TCE7 2.18 0.07 3.82
TCW7 -1.15 0.07 3.82
TCCIN 0.62 0.07 3.83
TCN3IN 0.62 0.67 3.83
TCS3IN 0.62 -0.53 3.83
TCE4IN 1.28 0.07 3.83
TCW4IN 0.05 0.07 3.83

Figure 9.1 shows a comparison of predicted and measured values for total heat flux. Appendix
B provides comparisons of heat flux and surface temperature on cable and surface targets.

123
103

102

101
101 102 103

Figure 9.1: Summary, Compartment Surface Temperature.

124
9.2 Target Temperature
Target temperature and heat flux data are available from the NIST/NRC test series. In the NIST/NRC
tests, the targets are different types of cables in various configurations: horizontal, vertical, in trays,
or free-hanging. Since these tests are intended to represent electrical cables, they are modeled as
cylindrical targets. Targets in the SP AST and WTC tests, intended to represent various compo-
nents of the building structure, are modeled as normal thermally-thick targets in CFAST.
The SP Adiabatic Surface Temperature Experiments included measurements of gas, plate ther-
mometer, and steel temperatures for compartment and pool fire experiments conducted at SP, Swe-
den. Only the compartment fire experiments are included in the CFAST comparisons. Three
additional experiments were conducted at SP, Sweden, in 2011, in which a 6 m long, 20 cm diame-
ter vertical column was positioned in the middle of 1.1 m and 1.9 m diesel and 1.1 m heptane pool
fires [89]. Gas, plate, and steel surface temperature measurements were made at heights of 1 m,
2 m, 3 m, 4 m, and 5 m above the pool surface. At heights of 1 m, 3 m, and 5 m, these measure-
ments were made at only one angular position. However, at 2 m and 4 m, the measurements were
made at four positions. At these heights, two conventional plates thermometers were positioned
approximately 10 cm from the column surface, along with two special plate thermometers (SPT)
that were installed flush with the column surface. At each height, comparable predictions were
made with CFAST using normal, thermally-thick targets.
The compartment for the WTC experiments contained a hollow box column roughly 0.5 m
from the fire pan, two trusses over the top of the pan, and one or two steel bars resting on the
lower truss flanges. In Tests 1, 2 and 3, the steel was bare, and in Tests 4, 5 and 6, the steel was
coated with various thicknesses of sprayed fire-resistive materials. The column was instrumented
near its base (about 0.5 m from the floor, middle (1.5 m), and upper (2.5 m). Four measurements
of steel (and insulation) temperatures were made at each location, for each of its four sides. These
elements were modeled using thin sheet obstructions with a resolution of 10 cm. In addition to
the steel structural elements, five cylinders (“slugs”) of nickel 200 (≥ 99 % nickel), 25.4 cm long
and 10.2 cm in diameter, were positioned 50 cm north of the centerline in the WTC experiments.
Slugs 1 through 5 were located 2.92 m, 1.82 m, 0.57 m, 0.05 m, and 1.56 m, respectively, from the
longitudinal axis of the fire pan. All the slugs were 50 cm north of the lateral axis. The fire pan
measured 2 m by 1 m. Four thermocouples were inserted into each slug at various locations. All
four temperatures for each slug were virtually [Link] were modeled with normal
thermally-thick targets in CFAST.
Figure 9.2 shows a comparison of predicted and measured values for total heat flux. Appendix
B provides comparisons of heat flux and surface temperature on cable and surface targets.

125
103

102

101
101 102 103

Figure 9.2: Summary, Target Temperature.

126
Chapter 10

Surface Heat Flux

CFAST includes calculation of heat flux via convection, radiation, and conduction to compartment
surfaces and targets. Heat transfer to the inside surface of compartment linings and the front and
rear faces (as specified by the user) of targets consists of convection (through the use of empirical
correlations) and radiation (calculated by the model using view factors for the fire, gas layers, and
compartment surfaces). Heat conduction into a solid surface is calculated via a one-dimensional
solution of the heat equation in cartesian or cylindrical coordinates. The latter is particularly useful
for predicting the thermal response of electrical cables.
For compartment linings, the “outside” surface is, by default, exposed to the exterior ambient
temperature with convection and radiation calculated in a similar manner to the inside surface.
The “outside” boundary condition can also be specified as a constant temperature (i.e., the outside
surface can be at ambient temperature) or can be connected to the “outside” surface of part or all
of a second compartment. For targets, the back surface is simply pointed in a direction opposite
that of the front surface with convection and radiation calculated in a similar manner to the front
surface. This chapter contains a variety of heat flux measurements, ranging from less than 1 kW/m2
from very small gas burners to more than 100 kW/m2 in full-scale compartment fires.

10.1 Heat Flux to Compartment Ceiling, Wall, and Floor Sur-

faces
In the NIST/NRC tests, heat flux gauges and thermocouples were positioned at various locations
on the walls, floor, and ceiling of the fire compartments. The locations are given in Table 9.1. The
heat flux gauges were not water cooled; thus, they measured the net rather than the gauge heat flux.
However, the net heat flux is a function of the temperature of the heat flux gauge itself, which is
not something that is modeled. To better compare model and measurement, the measured net heat
flux is converted into a gauge heat flux using the following formula:
 
00 00 4 4
kW/m2


q̇gauge = q̇net + σ Tgauge − T∞ + h Tgauge − T∞ (10.1)

where σ = 5.67 × 10−11 kW/m2 /K4 and h = 0.005 kW/m2 /K.

Also, over the course of 15 experiments, numerous heat flux gauges failed, most often due to
loss of contact with the wall or faulty thermocouples. All of the measurements from Test 13 and

127
16 were found to be flawed.
In the WTC tests, there were a variety of heat flux gauges installed in the test compartment.
Most were within 2 m of the fire. Their locations and orientations are listed in Table 10.1. This
section contains the measurements at the floor and ceiling.

Table 10.1: Heat flux gauge positions relative to the center of the fire pan in the WTC series.

Name x (m) y (m) z (m) Orientation Location

H2FU 0.64 0.63 3.30 +z Truss Support
H2RU 0.64 0.51 3.30 +z Truss Support
H2FD 0.64 0.30 3.15 −z Truss Support
H2RD 0.64 0.42 3.15 −z Truss Support
HCoHF -0.90 0.84 3.46 +x Column, facing fire
HCoHW -0.97 0.92 3.27 +y Column, facing north
HCoLF -0.90 0.84 0.92 +x Column, facing fire
HCoLW -0.97 0.92 1.02 +y Column, facing north
HF1 1.06 0.13 0.13 +z Floor
HF2 1.56 0.10 0.13 +z Floor
HCe1 -0.45 0.35 3.82 −z Ceiling
HCe2 0.05 0.35 3.82 −z Ceiling
HCe3 0.80 0.35 3.82 −z Ceiling
HCe4 2.56 0.35 3.82 −z Ceiling

Figure 10.1 shows a comparison of predicted and measured values for total heat flux. Appendix
B provides comparisons of heat flux and surface temperature on cable and surface targets.

128
102

101

100

10-1
10-1 100 101 102

Figure 10.1: Summary, Compartment Heat Flux to Surfaces.

129
10.2 Heat Flux to Targets
In the NIST/NRC tests, cables in various types (power and control), and configurations (horizontal,
vertical, in trays or free-hanging), were installed in the test compartment. For each of the four
cable targets considered, measurements of the radiative and total heat flux were made with gauges
positioned near the cables themselves. In the WTC tests, There were a variety of heat flux gauges
installed in the test compartment. Most were within 2 m of the fire. Their locations and orientations
are listed in Table 10.1. Figure 10.2 shows a comparison of predicted and measured values for
total heat flux. Appendix B provides comparisons of heat flux and surface temperature on cable
and surface targets.

130
102

101

100

10-1
10-1 100 101 102

Figure 10.2: Summary, Heat Flux to Targets.

131
132
Chapter 11

Summary and Conclusions

How to best quantify the comparisons between model predictions and experiments is not obvious.
The necessary and perceived level of agreement for any variable is dependent upon both the typical
use of the variable in a given simulation, the nature of the experiment, and the context of the
comparison in relation to other comparisons being made. For instance, the user may be interested
in the time it takes to reach a certain temperature in the room, but have little or no interest in peak
temperature for experiments that quickly reach a steady-state value. Insufficient experimental data
and understanding of how to compare the numerous variables in a complex fire model prevent a
complete validation of the model.

11.1 Summary of CFAST Model Uncertainty Statistics

A true validation of a model would involve proper statistical treatment of all the inputs and outputs
of the model with appropriate experimental data to allow comparisons over the full range of the
model. Thus, the comparisons of the differences between model predictions and experimental data
discussed here are intentionally simple and vary from test to test and from variable to variable
due to the changing nature of the tests and typical use of different variables. The following table
lists the summary statistics for the different quantities of interest examined in this Guide. the table
lists the bias and relative standard deviation of the predicted values. It also lists the total number
of experimental data sets on which these statistics are based, as well as the total number of point
to point comparisons. Obviously, the more data sets and the more points, the more reliable the
statistics.
For further details about model uncertainty and the meaning of these statistics, see the Refs.
[70, 51].
CFAST predictions in this validation study were consistent with numerous earlier studies,
which show that the use of the model is appropriate over a range of conditions for a variety of
fire scenarios. The CFAST model has been subjected to extensive evaluation studies by NIST and
others (see, for example Ref. [51] and Chapter 2. Although differences between the model and
the experiments were evident in these studies, most differences can be explained by limitations
of the model and the experiments. Like all predictive models, the best predictions consider the
limitations of the model and of the inputs provided to perform the calculations.

133
 Table 11.1: Summary statistics for all quantities of interest

Quantity Section Datasets Points σeE σeM Bias

HGL Temperature 5.1 17 373 0.07 0.30 1.06
HGL Temperature: Forced Ventilation 5.1 7 132 0.07 0.21 1.14
HGL Temperature: Natural Ventilation 5.1 13 216 0.07 0.35 1.05
HGL Temperature: No Ventilation 5.1 3 24 0.07 0.16 0.97
HGL Depth 5.1 15 262 0.05 0.26 1.01
HGL Depth: Open Compartments 5.1 14 228 0.05 0.19 0.95
HGL Depth: Closed Compartments 5.1 4 34 0.05 0.24 1.45
Ceiling Jet Temperature 6.3 13 503 0.07 0.42 1.05
Plume Temperature 6.2 6 75 0.07 0.23 1.08
Oxygen Concentration 7.1 5 60 0.08 0.26 1.03
Carbon Dioxide Concentration 7.1 5 54 0.08 0.27 0.89
Carbon Monoxide Concentration 7.1 2 29 0.19 0.65 1.04
Smoke Concentration 7.2 1 15 0.19 0.68 3.44
Compartment Over-Pressure 8 4 66 0.23 0.52 1.51
Target Temperature 9.2 4 423 0.07 0.49 1.26
Surface Temperature 9.1 6 403 0.07 0.21 0.99
Target Heat Flux 10.1 3 244 0.11 0.60 0.97
Surface Heat Flux 10.2 7 241 0.11 0.23 0.91
Smoke Alarm Activation Time (Temperature Surrogate) 6.4 1 142 0.34 0.43 1.23
Smoke Alarm Activation Time (Smoke Obscuration) 6.4 1 142 0.34 0.51 0.56
Sprinkler Activation Time 6.4 4 101 0.06 0.20 1.01

134
11.2 Normality Tests
The histograms on the following pages display the distribution of the quantity ln(M/E), where M
is a random variable representing the Model prediction and E is a random variable representing
the Experimental measurement. From the development of the statistics used to compare model and
experimental values [70], ln(M/E) is assumed to be normally distributed. To test this assumption
for each of the quantities of interest listed in Table 11.1, Spiegelhalter’s normality test has been
applied [105]. This test examines a set of values, x1 , ..., xn whose mean and standard deviation are
computed as follows:
n
1 n
x̄ = ∑ xi ; σ 2 =
n − 1 ∑ (xi − x̄)2 (11.1)
i=1 i=1
Spiegelhalter tests the null hypothesis that the sample xi is taken from a normally distributed pop-
ulation. The test statistic, S, is defined:
n
N − 0.73 n xi − x̄
S= √ ; N = ∑ Zi2 ln Zi2 ; Zi = (11.2)
0.9 n i=1 σ

Under the null hypothesis, the test statistic is normally distributed with mean 0 and standard devi-
ation of 1. If the p-value   
 S 
p = 1 − erf √ 
 (11.3)
2
is less than 0.05, the null hypothesis is rejected.
The flaw in most normality tests is that they tend to reject the assumption of normality when the
number of samples is relatively large. As can be seen in some of the histograms on the following
pages, some fairly “normal” looking distributions fail while decidedly non-normal distributions
pass. For this reason, the p-value is less important than the qualitative appearance of the histogram.
A best-fit Gaussian curve is also shown in the figures. If the histogram exhibits the typical bell-
shaped curve, this adds confidence to the statistical treatment of the data. If the histogram is not
bell-shaped, this might cast doubt on the statistical treatment for that particular quantity.

135
 50
160

140 40

120
30
100

80
20
60

40
10
20

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

100
7

80 6

5
60
4

40 3

2
20
1

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

100
70

80 60

50
60
40

40 30

20
20
10

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

10 250

8 200

6 150

4 100

2 50

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

136
 25

20
20

15
15

10
10

5 5

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

16
10
14
8
12

10
6
8

6 4

4
2
2

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

0
1 2 3 4 5 6 7 8 9 10

137
138
References

[1] American Society for Testing and Materials, West Conshohocken, Pennsylvania. ASTM E
1355-04, Standard Guide for Evaluating the Predictive Capabilities of Deterministic Fire
Models, 2004. xi, 1, 5, 15

[2] R. L. Iman and J. C. Helton. An Investigation of Uncertainty and Sensitivity Analysis

Techniques for Computer Models. Rick Analysis, 8(1):71–90, 1988. 5

[3] R. D. Peacock, P. A. Reneke, C. L. Forney, and M. M. Kostreva. Issues in Evaluation of

Complex Fire Models. Fire Safety Journal, 30:103–136, 1988. 6

[4] A. Beard. Evaluation of Fire Models: Part I – Introduction. Fire Safety Journal, 19:295–306,
1992. 6

[5] K. A. Notarianni. The Role of Uncertainty in Improving Fire Protection Regulation. PhD
thesis, Carnegie Mellon University, Pittsburgh, PA, 2000. 6

[6] N. Khoudja. Procedures for Quantitative Sensitivity and Performance Validation of a De-
terministic Fire Safety Model. PhD thesis, Texas A & M University, 1988. 6

[7] G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters, An Introduction

to Design, Data Analysis and Model Building. John Wiley and Sons, New York, 1978. 6

[8] C. Daniel. Applications of Statistics to Industrial Experimentation. John Wiley and Sons,
New York, 1976. 6

[9] A. M. Walker. Uncertainty Analysis of Zone Fire Models. Research Report 97/8, University
of Canterbury, New Zealand, 1997. 6

[10] N. A. Dempsey, P. J. Pagni, and R. B. Williamson. Compartment Fire Experiments: Com-

parison With Models. Fire Safety Journal, 25(3):187, 1995. 6

[11] International Conference of Building Officials, Whittier, CA. 1997 Uniform Building Code,
1997. 6

[12] W. M. Pitts, E. Braun, R. D. Peacock, H. E. Mitler, E. L. Johnsson, P. A. Reneke, and L. G.

Blevins. Temperature Uncertainties for Bare-Bead and Aspirated Thermocouple Measure-
ments in Fire Environments. In Thermal Measurements: The Foundation of Fire Standards,
Special Technical Publication 1427, West Conshohocken, PA, 2001. American Society for
Testing and Materials. 6, 8

139
[13] Y. He and V. Beck. Smoke Spread Experiment in a Multi-Storey Building and Computer
Modeling. Fire Safety Journal, 28(2):139, 1997. 6
[14] M. Luo, Y. He, and V. Beck. Comparison of Existing Fire Model Predictions With Ex-
perimental Results From Real Fire Scenarios. Journal of Applied Fire Science, 6(4):357,
1996/1997. 7
[15] G. M. Poole, E. J. Weckman, and A. B. Strong. Fire Growth Rates in Structural Fires.
NISTIR 5499, National Institute of Standards and Technology, 1994. 7
[16] J. Bailey and P. Tatem. Validation of Fire/Smoke Spread Model (CFAST) Using Ex-USS
SHADWELL Internal Ship Conflagration Control (ISCC) Fire Tests. Technical Report
NRL/MR/6180-95-7781, Naval Research Laboratory, 1995. 7
[17] J. L. Bailey, P. A. Tatem, W. W. Jones, and G. P. Forney. Development of an Algo-
rithm to Predict Vertical Heat Transfer Trough Ceiling/Floor Conduction. Fire Technology,
34(2):139, 1998. 7
[18] S. Deal. A Review of Four Compartment Fires with Four Compartment Fire Models. In
Proceedings of the Annual Meeting of the Fire Retardant Chemicals Association, volume
Fire Safety Developments and Testing, pages 33–51, Ponte Verde Beach, Florida, October
21-24 1990. Fire Retardant Chemicals Association. 7
[19] L. Y. Cooper and G. P. Forney. The Consolidated Compartment Fire Model (CCFM) Com-
puter Application CCFM-VENTS – Part I: Physical Reference Guide. NISTIR 4342, Na-
tional Institute of Standards and Technology, 1990. 7
[20] H. E. Mitler and J. A. Rockett. Users’ Guide to FIRST, A Comprehensive Single-Room Fire
Model. NBSIR 87-3595, National Institute of Standards and Technology, September 1987.
7
[21] H. E. Nelson. FPETOOL: Fire Protection Engineering Tools for Hazard Estimation. NISTIR
4380, National Institute of Standards and Technology, 1990. 7
[22] W. W. Jones and R. D. Peacock. Technical Reference Guide for FAST Version 18. Technical
Note 1262, National Institute of Standards and Technology, 1989. 7
[23] K. Matsuyama, T. Wakamatsu, and K. Harada. Systematic Experiments of Room and Cor-
ridor Smoke Filling for Use in Calibration of Zone and CFD Fire Models. NISTIR 6588,
National Institute of Standards and Technology, 2000. 7
[24] W. D. Davis. Comparison of Algorithms to Calculate Plume Centerline Temperature and
Ceiling Jet Temperature With Experiments. Journal of Fire Protection Engineering, 12:9,
2002. 8, 110
[25] W. W. Jones and R. D. Peacock. Refinement and Experimental Verification of a Model for
Fire Growth and Smoke Transport. In T. Wakamatsu, Y. Hasemi, A. Sekizawa, P. G. Seeger,
P. J. Pagni, and C. E. Grant, editors, Fire Safety Science. Proceedings. 2nd International
Symposium, pages 897–906, Tokyo, Japan, June 13-17 1988. International Association for
Fire Safety Science, Hemisphere Publishing Corporation. 8

140
[26] R. S. Levine and H. E. Nelson. Full Scale Simulation of a Fatal Fire and Comparison of
Results with Two Multiroom Models. NISTIR 90-4168, National Institute of Standards and
Technology, 1990. 8, 12

[27] D. Q. Duong. The Accuracy of Computer Fire Models: Some Comparisons with Experi-
mental Data from Australia. Fire Safety Journal, 16:415, 1990. 8

[28] T. Tanaka. Model of Multiroom Fire Spread. Fire Science and Technology, 3(2):105–121,
1983. 8

[29] W. Chow. Use of Zone Models on Simulating Compartmental Fires With Forced Ventilation.
Fire and Materials, 14:466, 1995. 9

[30] M. Luo. One Zone or Two Zones in the Room of Fire Origin During Fires? Journal of Fire
Sciences, 15:240, 1997. 9

[31] D. Madrzykowski and R. Vettori. A Sprinkler Fire Suppression Algorithm for the GSA
Engineering Fire Assessment System. NISTIR 4833, National Institute of Standards and
Technology, 1992. 9

[32] W. Chow. Performance of Sprinkler in Atria. Journal of Fire Sciences, 14:466, 1996. 9

[33] W. Chow. Predicability of Flashover by Zone Models. Journal of Fire Sciences, 16:335,
September/October 1988. 9

[34] M. Luo, Y. He, and V. Beck. Application of Field Model and Two-Zone Model to Flashover
Fires in a Full-Scale Multi-Room Single Level Building. Fire Safety Journal, 29:1, 1997. 9

[35] P. Collier. Fire in a Residential Building: Comparisons Between Experimental Data and a
Fire Zone Model. Fire Technology, 32:195, August/September 1996. 10

[36] D. White, C. Beyler, J. Scheffrey, and F. Williams. Modeling the Impact of Post-Flashover
Shipboard Fires on Adjacent Spaces. Journal of Fire Protection Engineering, 10, 2000. 10

[37] R. D. Peacock, P. A. Reneke, R. W. Bukowski, and V. Babrauskas. Defining Flashover for

Fire Hazard Calculations. Fire Safety Journal, 32(4):331–345, 1999. 10

[38] V. Babrauskas, R. D. Peacock, and P. A. Reneke. Defining Flashover for Fire Hazard Cal-
culations: Part II. Fire Safety Journal, 38:613–622, 2003. 10

[39] V. Babrauskas. Upholstered Furniture Room Fires–Measurements, Comparison with Fur-

niture Calorimeter Data, and Flashover Predictions. Journal of Fire Sciences, 2(1):5–19,
January/February 1984. 10, 57, 210

[40] S. Deal and C. Beyler. Correlating Preflashover Room Fire Temperatures. Journal of Fire
Protection Engineering, 2(2):33–48, 1990. 10

[41] P. H. Thomas. Testing Products and Materials for Their Contribution to Flashover in Rooms.
Fire and Materials, 5:103–111, 1981. 10

141
[42] B. J. McCaffrey, J. G. Quintiere, and M. F. Harkleroad. Estimating Room Temperatures and
the Likelihood of Flashover Using Fire Test Data Correlations. Fire Technology, 17(2):98–
119, 1981. 10

[43] B. Hägglund. Estimating Flashover Potential in Residential Rooms. FOA Rapport C

202369-A3, Forsvarets Forkningsanstalt, 1980. 10

[44] W. W. Jones. Multicompartment Model for the Spread of Fire, Smoke and Toxic Gases.
Fire Safety Journal, 9(1):172, 1985. 12

[45] J. A. Rockett and M. Morita. The NBS Harvard VI Multi-room Fire Simulation. Fire
Science and Technology, 5(2):159–164, 1985. 12

[46] R. W. Bukowski. Modeling a Backdraft Incident: The 62 Watts Street (New York) Fire. Fire
Engineers Journal, 56(185):14–17, 1996. 12

[47] W. Chow. Preliminary Studies of a Large Fire in Hong Kong. Journal of Applied Fire
Science, 6(3):243–268, 1996/1997. 12

[48] R. W. Bukowski. Analysis of the Happyland Social Club Fire with HAZARD I. Fire and
Arson Investigator, 42:36, 1992. 12

[49] J. Floyd. Comparison of CFAST and FDS for Fire Simulation With the HDR T51 and T52
Tests. NISTIR 6866, National Institute of Standards and Technology, 2002. 13

[50] K.B. McGrattan, S. Hostikka, J.E. Floyd, H.R. Baum, and R.G. Rehm. Fire Dynamics Sim-
ulator (Version 5), Technical Reference Guide. NIST Special Publication 1018-5, National
Institute of Standards and Technology, Gaithersburg, Maryland, October 2007. 13

[51] M.H. Salley and A. Lindeman. Verification and Validation of Selected Fire Models for
Nuclear Power Plant Applications, Supplement 1. NUREG 1824, U.S. Nuclear Regulatory
Commission, Rockville, Maryland, 2015. 13, 43, 104, 133

[52] G. Lui and W. Chow. A Short Note on Experimental Verification of Zone Models with
an Electric Heater. International Journal on Engineering Performance-Based Fire Codes,
5:30, 2003. 14

[53] W. Chow. Studies on Closed Chamber Fires. Journal of Fire Sciences, 13:89, 1995. 14

[54] W. Chow. Experimental Evaluation of the Zone Models CFAST, FAST and [Link].
Journal of Applied Fire Science, 2:307, 1992-1993. 14

[55] M. Altinakar, A. Weatherhill, and P. Nasch. Use of a Zone Model in Predicting Fire and
Smoke Propagation in Tunnels. In J. R. Gillard, editor, 9th International Symposium on
Aerodynamics and Ventilation of Vehicle Tunnels, pages 623–639, Aosta Valley, Italy, Octo-
ber 6-8 1997. BHR Group. 14

142
[56] R. D. Peacock, J. D. Averill, D. Madrzykowski, D. W. Stroup, P. A. Reneke, and R. W.
Bukowski. Fire Safety of Passenger Trains; Phase III: Evaluation of Fire Hazard Analysis
Using Full-Scale Passenger Rail Car Tests. NISTIR 6563, National Institute of Standards
and Technology, 2004. 14

[57] J. Hoover and P. A. Tatem. Application of CFAST to Shipboard Fire Modeling. Part 3.
Guidelines for Users. NRL/ML 6180-01-8550, Naval Research Laboratory, 2001. 14

[58] R. D. Peacock, G. P. Forney, and P. A. Reneke. CFAST – Consolidated Model of Fire

Growth and Smoke Transport (Version 6): Technical Reference Guide. Special Publication
1026r1, National Institute of Standards and Technology, Gaithersburg, Maryland, October
2011. 18

[59] R.D. Peacock, K.B. McGrattan, G.P. Forney, and P.A. Reneke. CFAST – Consolidated Fire
And Smoke Transport (Version 7) Volume 1: Technical Reference Guide. Technical Note
1889v1, National Institute of Standards and Technology, Gaithersburg, Maryland, Novem-
ber 2015. 20, 32, 110, 115

[60] H.W. Emmons and T. Tanaka. SFPE Handbook of Fire Protection Engineering, chapter
Vent Flows. National Fire Protection Association, Quincy, Massachusetts, fourth edition,
2008. 25

[61] J. P. Holman. Heat Transfer. McGraw-Hill Book Company, 1990. 33

[62] W. F. Moss and G. P. Forney. Implicitly coupling heat conduction into a zone fire model.
NISTIR 4886, National Institute of Standards and Technology, 1992. 33

[63] F. Liu. Numerical Solutions of Three-Dimensional Non-Grey Gas Radiative Transfer Using
the Statistical Narrow-Band Model. Journal of Heat Transfer, 121:200–203, 1999. 34

[64] W. W. Yuen, W. C. Tam, and W. K. Chow. Assessment of radiative heat transfer charac-
teristics of a combustion mixture in a three-dimensional enclosure using RAD-NETT (with
application to a fire resistance test furnace). International Journal of Heat and Mass Trans-
fer, 68:383–390, 2014. 35

[65] W. C. Tam and W. W. Yuen. Assessment of Radiation Solver of Fire Simulation Models
Using RADNNET-ZM. In 11th Asia-Oceania Symposium on Fire Science and Technology
(AOSFST), December 2018. 35

[66] W. W. Yuen and W. C. Tam. RADNNET-MBL: A Neural Network Approach for Evalua-
tion of Absorptivity and Emissivity of Non-gray Combustion Gas Mixture Between Finite
Areas and Volumes. In Extended Abstracts of The Second Pacific Rim Thermal Engineering
Conference, December 2019. 35

[67] M. Malendowski. Analytical solution for adiabatic surface temperature (AST). Fire tech-
nology, 53:413–420, 2017. 36

143
[68] R.L.P. Custer, B.J. Meacham, and R.P. Schifiliti. SFPE Handbook of Fire Protection En-
gineering, chapter Design of Detection Systems. National Fire Protection Association,
Quincy, Massachusetts, fourth edition, 2008. 38

[69] D. D. Evans. Sprinkler Fire Suppression for HAZARD. NISTIR 5254, National Institute of
Standards and Technology, 1993. 38

[70] K.B. McGrattan, S. Hostikka, R. McDermott, J.E. Floyd, C. Weinschenk, and K. Overholt.
Fire Dynamics Simulator, Technical Reference Guide, Volume 3: Validation. NIST Special
Publication 1018 Sixth Edition, National Institute of Standards and Technology, Gaithers-
burg, Maryland, September 2014. 43, 133, 135

[71] D.T. Sheppard and B.W. Klein. Burn Tests in Two Story Structure with Hallways. Technical
report, ATF Laboratories, Ammendale, Maryland, 2009. 43

[72] J.M. Willi. Propane Gas Fire Experiments in Residential Scale Structures. NIST Technical
Note 1953, National Institute of Standards and Technology, Gaithersburg, Maryland, 2016.
45

[73] R. Fleury. Evaluation of Thermal Radiation Models for Fire Spread Between Objects. Mas-
ter’s thesis, University of Canterbury, Christchurch, New Zealand, 2010. 47

[74] G. Heskestad and J. P. Hill. Experimental Fires in Multiroom/Corridor Enclosures. NBS-

GCR 86-502, National Institute of Standards and Technology, 1986. 47, 48, 176

[75] R. D. Peacock, S. Davis, and B. T. Lee. Experimental Data Set for the Accuracy Assessment
of Room Fire Models. NBSIR 88-3752, National Bureau of Standards, 1988. 47, 59, 176,
214

[76] S.P. Nowlen. Enclosure Environment Characterization Testing for the Baseline Validation
of Computer Fire Simulation Codes. NUREG/CR-4681 (SAND86-1296), Sandia National
Laboratory, Albuqueque, New Mexico, March 1987. Work performed under contract to the
US Nuclear Regulatory Agency, Washington DC. 48, 181

[77] J.M. Chavez and S.P. Nowlen. An Experimental Investigation of Internally Ignited Fires
in Nuclear Power Plant Control Cabinets, Part II: Room Effects Tests. NUREG/CR-4527
(SAND86-0336), Sandia National Laboratory, Albuqueque, New Mexico, November 1988.
Work performed under contract to the US Nuclear Regulatory Agency, Washington DC. 48,
181

[78] W. Klein-Helβ ling and M. Röwenkamp. Evaluation of Fire Models for Nuclear Power
Plant Applications: Fuel Pool Fire Inside a Compartment. Technical report, Gesellschaft
für Anlagen-und Reaktorsicherheit (GRS), Köln, Germany, May 2005. 51, 193

[79] O. Riese and D. Hosser. Evaluation of Fire Model for Nuclear Power Plant Applications:
Flame SPread in Cable Tray Fires. Draft Version Revision 1, Gesellschaft für Anlagen-und
Reaktorsicherheit (GRS), June 2004. 51, 193

144
[80] K.L. Foote. 1986 LLNL Enclosure Fire Tests Data Report. Technical Report UCID-21236,
Lawrence Livermore National Laboratory, August 1987. 54

[81] B. T. Lee. Effect of Wall and Room Surfaces on the Rates of Heat, Smoke, and Carbon
Monoxide Production in a Park Lodging Bedroom Fire. NBSIR 85-2998, National Bureau
of Standards, 1985. 57, 210

[82] A. Hamins, A. Maranghides, E. L. Johnsson, M. K. Donnelly, J. C. Yang, G. W. Mullhol-

land, and R. L. Anleitner. Report of Experimental Results for the International Fire Model
Benchmarking and Validation Exercise #3. Special Publication 1013-1, National Institute
of Standards and Technology, 2005. 62, 218

[83] J. H. Klote and Jr. Fothergill, J. W. Design of Smoke Control Systems for Buildings. Ameri-
can Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc, 1983. 71

[84] J. H. Klote. Project Plan for Full Scale Smoke Movement and Smoke Control Tests. NBSIR
88-3800, National Bureau of Standards, 1988. 71

[85] R.W. Bukowski, R.D. Peacock, J.D. Averill, T.G. Cleary, N.P. Bryner, W.D. Walton, P.A.
Reneke, and E.D. Kuligowski. Performance of Home Smoke Alarms Analysis of the Re-
sponse of Several Available Technologies in Residential Fire Settings. NIST Technical Note
1455-1, National Institute of Standards and Technology, Gaithersburg, Maryland, February
2008. 72

[86] S.L. Manzello, S.H. Park, T. Mizukami, and D.P. Benz. Measurement of Thermal Properties
of Gypsum Board at Elevated Temperatures. In Proceedings of the Fifth International Con-
ference on Structures in Fire (SiF’08), pages 656–665, Nanyang Technological University,
Singapore, 2008. 74

[87] L. Audouin, L. Rigollet, H. Prétrel, W. LeSaux, and M. Röwekamp. OECD PRISME

project: Fires in confined and ventilated nuclear-type multi-compartments–Overview and
main experimental results. Fire Safety Journal, 62:80–101, 2013. 77

[88] U. Wickström, R. Jansson, and H. Tuovinen. Verification fire tests on using the adiabatic
surface temperature for predicting heat transfer. Technical Report 2009:19, SP Technical
Research Institute of Sweden, Bo̊ras, Sweden, 2009. 77

[89] J. Sjöström, A. Byström, and U. Wickström. Large scale test on thermal exposure to steel
column exposed to pool fires. Technical Report 2012:04, SP Technical Research Institute of
Sweden, Bo̊ras, Sweden, 2012. ISBN: 978-91-87017-21-6. 79, 125, 334

[90] K. D Steckler, J. G. Quintiere, and W. J. Rinkinen. Flow Induced by Fire in a Compartment.

NBSIR 82-2520, National Bureau of Standards, 1982. 79, 336

[91] D.T. Sheppard and D.R. Steppan. Sprinkler, Heat & Smoke Vent, Draft Curtain Project –
Phase 1 Scoping Tests. Technical report, Underwriters Laboratories, Northbrook, Illinois,
May 1997. 82

145
 [92] D. Madrzykowski and C. Weinschenk. Impact of Fixed Ventilation on Fire Damage Patterns
in Full-Scale Structures. Technical report, Underwriters Laboratories Firefighter Safety
Research Institute, Columbia, MD, April 2019. 85

[93] K.M. Opert. Assesment of Natural Vertical Ventilation for Smoke and Hot Gas Layer Con-
trol in a Residential Scale Structure. Master’s thesis, University of Maryland, 2012. 90

[94] J. E. Gott, D. L. Lowe, K. A. Notarianni, and W. D. Davis. Analysis of High Bay Hangar Fa-
cilities for Fire Detector Sensitivity and Placement. Technical Note 1423, National Institute
of Standards and Technology, Gaithersburg, Maryland, February 1997. 92

[95] W. D. Davis, K. A. Notarianni, and P. Z. Tapper. An Algorithm for Calculating the Plume
Centerline Temperature in the Presence of a Hot Upper Layer. Journal of Fire Protection
Engineering, 10(3):23–31, 2000. 92

[96] R. Vettori. Effect of an Obstructed Ceiling on the Activation Time of a Residential Sprinkler.
NISTIR 6253, National Institute of Standards and Technology, Gaithersburg, Maryland,
November 1998. 92

[97] S. Hostikka, M. Kokkala, and J. Vaari. Experimental Study of the Localized Room Fires,
NFDC2 Test Series. VTT Research Notes 2104, VTT Building and Transport, 2001. 94,
365

[98] A. Tewarson. Generation of Heat and Chemical Compounds in Fires, chapter 4-3. National
Fire Protection Association and The Society of Fire Protection Engineers, Quincy, MA, 3rd
edition, 2003. 94

[99] A. Hamins, A. Maranghides, K.B. McGrattan, E. Johnsson, T. Ohlemiller, M. Donnelly,

J. Yang, G. Mulholland, K. Prasad, S. Kukuck, R. Anleitner, and T. McAllister. Federal
Building and Fire Safety Investigation of the World Trade Center Disaster: Experiments
and Modeling of Structural Steel Elements Exposed to Fire. NIST NCSTAR 1-5B, National
Institute of Standards and Technology, Gaithersburg, Maryland, September 2005. 97

[100] M.L. Janssens and H.C. Tran. Data Reduction of Room Tests for Zone Model Validation.
Journal of Fire Sciences, 10:528–555, 1992. 104, 149

[101] K.B. McGrattan and G. P. Forney. Fire Dynamics Simulator (Version 4), User’s Guide.
NISTIR 7013, National Institute of Standards and Technology, June 2003. 104

[102] R.L. Alpert. SFPE Handbook of Fire Protection Engineering, chapter Ceiling Jet Flows.
National Fire Protection Association, Quincy, Massachusetts, fourth edition, 2008. 109,
110

[103] N. Iqbal and M.H. Salley. Fire Dynamics Tools (FDTs), Quantitative Fire Hazard Analysis
Methods for the U.S. Nuclear Regulatory Commission Fire Protection Inspection Program.
NUREG 1805, U. S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Re-
search, Rockville, MD, 2004. 111

146
[104] G. Heskestad and H. F. Smith. Investigation of a New Sprinkler Sensitivity Approval Test:
the Plunge Test. Technical Report 22485 2937, Factory Mutual Research Corporation, Nor-
wood, MA, 1976. 111

[105] D. J. Spiegelhalter. Diagnostic test of distributional shape. Biometrika, 70(2):401–409,

1983. 135

[106] K.B. McGrattan, S. Hostikka, J.E. Floyd, H.R. Baum, and R.G. Rehm. Fire Dynamics
Simulator, Technical Reference Guide, Volume 1: Mathematical Model. NIST Special Pub-
lication 1018-1, National Institute of Standards and Technology, Gaithersburg, Maryland,
2015. 149

[107] Y. He, A. Fernando, and M. Luo. Determination of Interface Height from Measured Param-
eter Profile in Encolsure Fire Experiment. Fire Safety Journal, 31:19–38, 1998. 149

[108] Fire Modeling Code Comparison. TR 108875, Electric Power Research Institute (EPRI),
Palo Alto, CA. 214

[109] J. H. Klote. Fire Experiments of Zoned Smoke Control at the Plaza Hotel in Washington,
DC. NISTIR 90-4253, National Institute of Standards and Technology, 1990. 297

147
148
Appendix A

Calculation of Layer Height and the

Average Upper and Lower Layer
Temperatures

Fire protection engineers often need to estimate the location of the interface between the hot,
smoke-laden upper layer and the cooler lower layer in a burning compartment. Zone fire models
such as CFAST compute this quantity directly, along with the average temperature of the upper
and lower layers. In an experimental test or a computational fluid dynamics (CFD) model like
FDS [106], there are not two distinct zones, but rather a continuous profile of temperature. Nev-
ertheless, methods have been developed to estimate layer height and average temperatures from
a continuous vertical profile of temperature. One such method [100] is as follows: Consider a
continuous function T (z) defining temperature T as a function of height above the floor z, where
z = 0 is the floor and z = H is the ceiling. Define Tu as the upper layer temperature, Tl as the lower
layer temperature, and zint as the interface height. Compute the quantities:
Z H
(H − zint ) Tu + zint Tl = T (z) dz = I1 (A.1)
0
Z H
1 1 1
(H − zint ) + zint = dz = I2 (A.2)
Tu Tl 0 T (z)
Solve for zint :

Tl (I1 I2 − H 2 )
zint = (A.3)
I1 + I2 Tl2 − 2 Tl H
Let Tl be the temperature in the lowest mesh cell or lowest measurement point and, using
Simpson’s Rule, perform the numerical integration of I1 and I2 . Tu is defined as the average upper
layer temperature via
Z H
(H − zint ) Tu = T (z) dz (A.4)
zint
For experimental test data or CFD model output, the integral function of temperature as a
function of height can be estimated empirically from a number of discrete data points. Further
discussion of similar procedures can be found in Ref. [107].

149
150
Appendix B

Summary of Verification Results

This appendix summarizes the accuracy of results from cases in the CFAST verification suite.
These metrics are an integral part of a process known as regression testing, which aims to evaluate
the accuracy of output predicted by a numerical software package. The results from this appendix
can be compared to another version of CFAST as a way of measuring the quality of results between
revisions. These statistical metrics also help determine the impact of code additions on the overall
accuracy of CFAST and can help identify errors in the CFAST source code.

151
 Table B.1: Summary of Verification Results
Case Name Section Expected Expected Predicted Predicted Type of Error Error Within
Quantity Value Quantity Value Error Tolerance Tol.
energy_balance 3.1 ULT 8.20e+02 ULT_1 8.24e+02 Relative 5.54e-03 1.00e-01 Yes
energy_balance 3.1 ULT 8.20e+02 ULT_1 8.24e+02 Relative 5.40e-03 1.00e-01 Yes
energy_balance 3.1 ULT 5.00e+00 ULT_1 5.00e+00 Relative 4.00e-11 1.00e-01 Yes
energy_balance 3.1 LLT 5.00e+00 LLT_1 5.00e+00 Relative 4.00e-11 1.00e-01 Yes
energy_balance 3.1 PRS 1.73e+03 PRS_1 1.73e+03 Relative 4.71e-07 1.00e-01 Yes
energy_balance 3.1 ULT 5.00e+00 ULT_1 5.00e+00 Relative 2.42e-04 1.00e-01 Yes
energy_balance 3.1 LLT 5.00e+00 LLT_1 5.00e+00 Relative 2.40e-04 1.00e-01 Yes
energy_balance 3.1 ULT 5.00e+00 ULT_1 5.00e+00 Relative 3.46e-04 1.00e-01 Yes
energy_balance 3.1 LLT 5.00e+00 LLT_1 5.00e+00 Relative 3.57e-04 1.00e-01 Yes
mass_balance 3.2 MassCO2Prod 1.81e-02 mCO2 1.81e-02 Relative 1.10e-04 1.00e-01 Yes
mass_balance 3.2 MassH2OProd 1.48e-02 mH2O 1.48e-02 Relative 1.35e-04 1.00e-01 Yes
mass_balance 3.2 MassCO2Prod 1.81e-02 mCO2 1.81e-02 Relative 1.66e-04 1.00e-01 Yes
mass_balance 3.2 MassH2OProd 1.48e-02 mH2O 1.48e-02 Relative 2.02e-04 1.00e-01 Yes
mass_balance 3.2 MassCO2Prod 1.81e-02 mCO2 1.81e-02 Relative 3.31e-04 1.00e-01 Yes
mass_balance 3.2 MassH2OProd 1.48e-02 mH2O 1.48e-02 Relative 4.05e-04 1.00e-01 Yes
mass_balance 3.2 MassCO2Prod 1.81e-02 mCO2 1.81e-02 Relative 0.00e+00 1.00e-01 Yes
mass_balance 3.2 MassH2OProd 1.48e-02 mH2O 1.48e-02 Relative 0.00e+00 1.00e-01 Yes
mass_balance 3.2 mCO21 1.81e-02 mCO21 1.81e-02 Relative 2.24e-05 1.00e-01 Yes
mass_balance 3.2 mH2O1 1.48e-02 mH2O1 1.48e-02 Relative 2.06e-05 1.00e-01 Yes
mass_balance 3.2 mCO22 9.05e-03 mCO22 9.04e-03 Relative 1.39e-03 1.00e-01 Yes
mass_balance 3.2 mH2O2 7.41e-03 mH2O2 7.40e-03 Relative 1.36e-03 1.00e-01 Yes
mass_balance 3.2 mCO23 6.04e-03 mCO23 5.92e-03 Relative 1.93e-02 1.00e-01 Yes
mass_balance 3.2 mH2O3 4.94e-03 mH2O3 4.85e-03 Relative 1.93e-02 1.00e-01 Yes
mass_balance 3.2 mCO24 4.53e-03 mCO24 4.20e-03 Relative 7.13e-02 1.00e-01 Yes
mass_balance 3.2 mH2O4 3.71e-03 mH2O4 3.44e-03 Relative 7.13e-02 1.00e-01 Yes

152
mass_balance 3.2 Temp 5.94e+00 ULT_1 5.98e+00 Relative 6.42e-03 1.00e-01 Yes
mass_balance 3.2 Gas_FED 6.56e-06 TRGDFEDG_1 6.56e-06 Relative 5.81e-04 1.00e-01 Yes
mass_balance 3.2 Heat_FED 6.75e-04 TRGDFEDH_1 6.75e-04 Relative 7.44e-07 1.00e-01 Yes
mass_balance 3.2 mco21 2.74e-02 ULMCO2_1 2.74e-02 Relative 1.67e-03 1.00e-01 Yes
mass_balance 3.2 mh2o1 2.25e-02 ULMH2O_1 2.24e-02 Relative 1.67e-03 1.00e-01 Yes
mass_balance 3.2 msoot 9.30e-03 ULMOD_1 9.25e-03 Relative 5.00e-03 1.00e-01 Yes
mass_balance 3.2 mco2 5.67e-02 ULMCO2_1 5.64e-02 Relative 5.00e-03 1.00e-01 Yes
mass_balance 3.2 mh2o 1.28e-02 ULMH2O_1 1.27e-02 Relative 5.00e-03 1.00e-01 Yes
mass_balance 3.2 mco 3.15e-03 ULMCO_1 3.13e-03 Relative 5.00e-03 1.00e-01 Yes
mass_balance 3.2 mhcn 2.50e-03 ULMHCN_1 2.49e-03 Relative 5.00e-03 1.00e-01 Yes
mass_balance 3.2 mhcl 6.89e-03 ULMHCL_1 6.85e-03 Relative 5.00e-03 1.00e-01 Yes
mass_balance 3.2 Trace1 1.00e-05 ULMTS_1 9.98e-06 Relative 1.67e-03 1.00e-01 Yes
ventilation 3.3 Pressure 1.10e+04 PRS_1 1.10e+04 Relative 2.10e-03 1.00e-01 Yes
ventilation 3.3 M_Outside_1_1 1.20e+00 M_4_Outside_2 1.20e+00 Relative 7.53e-11 1.00e-01 Yes
ventilation 3.3 ULT_1 2.23e+01 ULT_1 2.29e+01 Relative 2.68e-02 1.00e-01 Yes
ventilation 3.3 MFLOW 1.20e-02 L_Room_1_Outside_1 1.20e-02 Relative 3.83e-03 1.00e-01 Yes
ventilation 3.3 MFLOW 1.20e-02 L_Room_1_Outside_1 1.20e-02 Relative 3.83e-03 1.00e-01 Yes
ventilation 3.3 P_FLOOR 1.22e+00 PRS_1 1.22e+00 Relative 3.84e-03 1.00e-01 Yes
radiation 3.4 Temp 4.31e+00 TRGSURT_1 4.21e+00 Relative 2.55e-02 1.00e-01 Yes
radiation 3.4 rod temp 4.80e+02 TRGSURT_1 4.80e+02 Relative 2.08e-12 1.00e-01 Yes
radiation 3.4 plate temp 4.79e+02 TRGSURT_2 4.80e+02 Relative 2.09e-03 1.00e-01 Yes
radiation 3.4 thin plate temp 4.80e+02 TRGSURT_3 4.80e+02 Relative 7.82e-08 1.00e-01 Yes
radiation 3.4 INCIDENT HEAT FLUX (GAS) 7.35e+00 GAS(TOP) 5.87e-01 Relative 9.20e-01 1.00e-01 Out of Tolerance
radiation 3.4 INCIDENT HEAT FLUX (WALL) 7.35e+00 WALL(TOP) 1.23e+00 Relative 8.33e-01 1.00e-01 Out of Tolerance
radiation 3.4 INCIDENT HEAT FLUX (GAS SIDE) 7.80e+00 GAS(SIDE) 1.16e+00 Relative 8.52e-01 1.00e-01 Out of Tolerance
radiation 3.4 INCIDENT HEAT FLUX (WALL SIDE) 7.80e+00 WALL(SIDE) 2.12e+00 Relative 7.28e-01 1.00e-01 Out of Tolerance
radiation 3.4 EMISSIVITY (TOP) 5.30e-02 EMIS(TOP) 3.94e-02 Relative 2.56e-01 1.00e-01 Out of Tolerance
radiation 3.4 HEMISPHERICAL ABSORPTIVITY (TOP) 1.54e-01 ABSO(TOP) 3.94e-02 Relative 7.44e-01 1.00e-01 Out of Tolerance
radiation 3.4 EMISSIVITY (TOP) 5.30e-02 EMIS(TOP) 3.94e-02 Relative 2.56e-01 1.00e-01 Out of Tolerance
radiation 3.4 HEMISPHERICAL ABSORPTIVITY (TOP) 1.36e-01 ABSO(TOP) 3.94e-02 Relative 7.10e-01 1.00e-01 Out of Tolerance
sprinklers 3.5 TempLinkEx 1.15e+02 SENST_1 1.15e+02 Relative 3.46e-05 1.00e-01 Yes
sprinklers 3.5 HRREx 9.95e+03 HRR_1 1.00e+04 Relative 5.19e-03 1.00e-01 Yes
Appendix C

Transient Model / Experiment Comparison

Graphs

This appendix contains a series of plots of experimental measurements and model predictions for
each of the individual comparisons included in the scatter plots in Chapters 5 through 10 in this
report.

C.1 ATF Corridors

The ATF Corridors experiments consisted of two corridors one on top of the other and connected
by a stairwell. The fire, a natural gas sand burner, was located on the first level at the end of the
corridor away from the stairwell. The corridor was closed at this end, and open at the same position
on the second level. Two-way flow occurred on both levels because make-up air flowed from the
opening on the second level down the stairs to the first. The only opening to the enclosure was the
open end of the second-level corridor.
Temperatures were measured with seven thermocouple trees. Tree A was located fairly close
to the fire on the first level. Tree B was located halfway down the first-level corridor. Tree C was
close to the stairwell entrance on the first level. Tree D was located in the doorway of the stairwell
on the first level. Tree E was located roughly along the vertical centerline of the stairwell. Tree F
was located near the stairwell opening on the second level. Tree G was located near the exit at the
other end of the second-level corridor. The graphs on the following pages show the top and bottom
TC from each tree for the given fire sizes of 50 kW, 100 kW, 250 kW, 500 kW, and a mixed HRR
“pulsed” fire.
HGL temperature and depth reductions were carried out using three arrays of thermocouples in
the lower corridor (Trees A, B, and C) and two arrays in the upper corridor (Trees G and H). Ceiling
jet temperatures were compared using the top thermocouple in the three downstairs thermocouple
trees (Trees A, B, and C). Since CFAST only computes ceiling jet temperatures in compartments
with a fire, no comparisons were made for the upper compartment.

153
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

154
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

155
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

156
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 0 5 10 15 20

157
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
500 500

400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
500 500

400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
500 500

400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 0 5 10 15 20

158
C.2 DelCo Trainers
The DelCo Trainer experiments were conducted in two different structures. Tests 2-6 were con-
ducted in a single level structure consisting of three rooms. Rooms 1 and 3 had two thermocouple
trees and Room 2 had one. Tests 22-25 were conducted in a two level structure. Floors 1 and 2
each had three thermocouple arrays. See Section 4.2 for their exact locations.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
700 500

600
400
500

400 300

300 200
200
100
100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
300

250

200

150

100

50

0
0 5 10 15 20 25 30

Figure C.1: DelCo Trainers, HGL Temperature, Test 2.

159
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
700 500

600
400
500

400 300

300 200
200
100
100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
300

250

200

150

100

50

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
700 500

600
400
500

400 300

300 200
200
100
100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
300

250

200

150

100

50

0
0 5 10 15 20 25 30

Figure C.2: DelCo Trainers, HGL Temperature, Tests 3 and 4.

160
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
700 500

600
400
500

400 300

300 200
200
100
100

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83
300

250

200

150

100

50

0
0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
700 500

600
400
500

400 300

300 200
200
100
100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
300

250

200

150

100

50

0
0 5 10 15 20 25 30

Figure C.3: DelCo Trainers, HGL Temperature, Tests 5 and 6.

161
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 200

300 150

200 100

100 50

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 200

300 150

200 100

100 50

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 200

300 150

200 100

100 50

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 200

300 150

200 100

100 50

0 0
0 5 10 15 20 0 5 10 15 20

Figure C.4: DelCo Trainers, HGL Temperature, Tests 22-25.

162
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
2.5

1.5

0.5

0
0 5 10 15 20 25 30

Figure C.5: DelCo Trainers, HGL Height, Test 2.

163
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
2.5

1.5

0.5

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
2.5

1.5

0.5

0
0 5 10 15 20 25 30

Figure C.6: DelCo Trainers, HGL Height, Tests 3 and 4.

164
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83
2.5

1.5

0.5

0
0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
2.5

1.5

0.5

0
0 5 10 15 20 25 30

Figure C.7: DelCo Trainers, HGL Height, Tests 5 and 6.

165
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 0 5 10 15 20

Figure C.8: DelCo Trainers, HGL Height, Tests 22-25.

166
 The plots below and on the following pages display comparisons of ceiling jet temperatures
for the DelCo Trainer experiments. Tests 2-6 were conducted in a single level house mock-up
with three rooms adjacent to one another. Locations A1 and A2 were in the fire room, A3 was
in an adjacent room, and A4 and A5 were in a room next to the adjacent room. Tests 22-25 were
conducted in a two level house mock-up. Locations A1, A2, and A3 were 2 cm below the ceiling of
the first level, and A7, A8, and A9 were 2 cm below the ceiling of the second level. See Section 4.2
for their exact locations.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 1000

800
600

600
400
400

200
200

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 1000

800
600

600
400
400

200
200

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 1000

800
600

600
400
400

200
200

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Figure C.9: DelCo Trainers, ceiling jet temperature, Tests 2-4.

167
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 1000

800
600

600
400
400

200
200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 1000

800
600

600
400
400

200
200

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Figure C.10: DelCo Trainers, ceiling jet temperature, Tests 5 and 6.

168
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 300

500 250

400 200

300 150

200 100

100 50

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 300

500 250

400 200

300 150

200 100

100 50

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 300

500 250

400 200

300 150

200 100

100 50

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 300

500 250

400 200

300 150

200 100

100 50

0 0
0 5 10 15 20 0 5 10 15 20

Figure C.11: DelCo Trainers, ceiling jet temperature, Tests 22-25.

169
C.3 Fleury Experiments
The plots on the following pages contain comparisons of predicted and measured heat fluxes from
a series of propane burner fires. Heat flux gauges were mounted on moveable dollies that were
placed in front of, and to the side of, burners with dimensions of 0.3 m by 0.3 m (1:1 burner),
0.6 m by 0.3 m (2:1 burner), and 0.9 m by 0.3 m (3:1 burner). The heat release rates were set to
100 kW, 150 kW, 200 kW, 250 kW, and 300 kW. The gauges were mounted at heights of 0 m,
0.5 m, 1.0 m, and 1.5 m relative to the top edge of the burner. Each page contains the results for a
given HRR.

170
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
10 10

8 8

6 6

4 4

2 2

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
10 10

8 8

6 6

4 4

2 2

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
10 10

8 8

6 6

4 4

2 2

0 0
0.5 1 1.5 2 0.5 1 1.5 2

Figure C.12: Comparison of predicted (lines) and measured (circles) heat flux for the 100 kW Fleury fires.

171
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

Figure C.13: Comparison of predicted (lines) and measured (circles) heat flux for the 150 kW Fleury fires.

172
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

Figure C.14: Comparison of predicted (lines) and measured (circles) heat flux for the 200 kW Fleury fires.

173
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
25 25

20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
25 25

20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
25 25

20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

Figure C.15: Comparison of predicted (lines) and measured (circles) heat flux for the 250 kW Fleury fires.

174
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
30 30

25 25

20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
30 30

25 25

20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
30 30

25 25

20 20

15 15

10 10

5 5

0 0
0.5 1 1.5 2 0.5 1 1.5 2

Figure C.16: Comparison of predicted (lines) and measured (circles) heat flux for the 300 kW Fleury fires.

175
C.4 FM Four Room Including Corridor Test Series
This data set describes a series of tests conducted in a multiple room configuration with more
complex gas burner fires than the previous data set. This study [74] was included because, in many
ways, it is similar to the smoke movement study performed at NBS [75], and permits comparisons
between two different laboratories. In addition, it expands upon that data set by providing larger
a time-varying gas burner fires in a room-corridor configuration. Fire size was about up to 1 MW
with a total volume of 200 m3 . This study was performed to collect data allowing for variations
in fire source, ventilation, and geometry in a multi-compartment structure, especially for situations
with closed doors.

176
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 2.5

2
300

1.5
200
1

100
0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
120 2.5

100
2

80
1.5
60
1
40

0.5
20

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 2.5

80 2

60 1.5

40 1

20 0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 2.5

80 2

60 1.5

40 1

20 0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10

177
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 2.5

2
300

1.5
200
1

100
0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
120 2.5

100
2

80
1.5
60
1
40

0.5
20

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 2.5

80 2

60 1.5

40 1

20 0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 2.5

80 2

60 1.5

40 1

20 0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10

178
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.25

0.2 0.2

0.15 0.15

0.1 0.1

0.05 0.05

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.25

0.2 0.2

0.15 0.15

0.1 0.1

0.05 0.05

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83
0.25

0.2

0.15

0.1

0.05

0
0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83
0.25

0.2

0.15

0.1

0.05

0
0 2 4 6 8 10

179
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.25

0.2 0.2

0.15 0.15

0.1 0.1

0.05 0.05

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.25

0.2 0.2

0.15 0.15

0.1 0.1

0.05 0.05

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83
0.25

0.2

0.15

0.1

0.05

0
0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83
0.25

0.2

0.15

0.1

0.05

0
0 2 4 6 8 10

180
C.5 FM/SNL Test Series
The Factory Mutual and Sandia National Laboratories (FM/SNL) Test Series was a series of 25
fire tests conducted in 1985 for the NRC by Factory Mutual Research Corporation (FMRC), under
the direction of Sandia National Laboratories (SNL). The primary purpose of these tests was to
provide data with which to validate computer models for various types of NPP compartments. The
experiments were conducted in an enclosure measuring 18 m x 12 m x 6 m, constructed at the
FMRC fire test facility in Rhode Island. The FM/SNL test series is described in detail, including
the types and locations of measurement devices, as well as some results in References [76, 77].

181
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

182
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

183
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

184
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

185
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 6

5
150
4

100 3

2
50
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

186
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 0 5 10 15

187
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 600

500
300
400

200 300

200
100
100

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 400

500
300
400

300 200

200
100
100

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 0 5 10 15

188
 CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 5 10 15

CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 5 10 15

CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 5 10 15

189
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

190
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

191
 CFAST7.7.1-2-g0b43aa83
250

200

150

100

50

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
250

200

150

100

50

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
250

200

150

100

50

0
0 5 10 15 20 25 30

192
C.6 iBMB Compartment Tests
A series of small compartment kerosene pool fire experiments, conducted at the Institut für Baustoffe,
Massivbau und Brandschutz (iBMB) of Braunschweig University of Technology in Germany in
2004 [78]. The results from Test 1 were considered here. These experiments involved relatively
large fires in a relatively small (3.6 m x 3.6 m x 5.7 m) concrete enclosure.
A second series of fire experiments in 2004, conducted under the International Collaborative
Fire Model Project (ICFMP) involved realistically routed cable trays inside the same concrete
enclosure at iBMB [79]. The compartment was configured slightly differently with a ceiling height
of 5.6 m.
Temperature measurements conducted during the test series were used to estimate the HGL
temperature and depth.

193
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 6

5
600
4

400 3

2
200
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 6

5
300
4

200 3

2
100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

194
C.7 LLNL Enclosure Series
The plots on the following pages compare predicted and measured layer temperatures from the
LLNL Enclosure test series. In the experiments, fifteen thermocouples were evenly spaced from
floor to ceiling on either side of the burner. The measured temperatures were reported as averages
of the lower, middle, and upper five TCs. Some of the experiments were conducted with a separated
plenum space in the top one-third of the overall compartment (Tests 17-60). In these cases, the
upper five TCs are a measure of the average plenum temperature.
In the figures on the following pages, the black circles represent the average of the five upper-
most TC measurements. The lines represent the simulation. The red circles represent the average
of the middle five TC measurements. For the plenum tests, these TCs are located immediately
beneath the plenum and their average temperature is typically greater than that of the plenum. Note
that in a number of tests, the fuel flow was stopped or the fire self-extinguished. The simulations
last only as long as the reported measurements.

195
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 10 20 30 40 50 0 10 20 30 40 50

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 10 20 30 40 50 0 10 20 30 40 50

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 10 20 30 40 50 0 10 20 30 40 50

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 10 20 30 40 50 0 10 20 30 40 50

196
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

197
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

198
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

199
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

200
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

201
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

202
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100

203
 The test report of the LLNL Enclosure experiments lists the mass flow rate, ṁ, through the
exhaust duct at different times during the tests. It also lists the compartment pressures, ∆p, at these
same times. From the simple leak formula:
s
ṁ 2 ∆p
=A (C.1)
ρ0 ρ0

the leakage area, A, is estimated to be 0.018 m2 , based on the initial exhaust rate and pressure.
For modeling purposes, the “leakage area” is assumed to be the sum of the inlet duct area plus
any actual compartment leakage area. The mass flow rate through the exhaust duct is specified
explicitly in the model. The test report does not provide enough information about the ventilation
system to model the fan and filtration system within the exhaust duct.
In the figures on the following pages, the open circles represent the measured pressure; the line
represents the predicted pressure. The predicted pressures are time-averaged over a time interval
that is one-tenth the total simulation time. In general, the short-duration pressure spike that is
typical of fires within relatively tight compartments has been smoothed over in the reported test
data. Depending on the simulation, it often appears in the simulation data. The comparison of
measurement and prediction is based on the final few pressure points, not the initial spike.
Only some of the 64 experiments are included. In some cases, the fan was turned off and there
is not enough information in the test report to determine the pressure losses through the duct. In
other cases, there are only two pressure measurements reported; one at the start of the test, the
other either before or after extinction. In various other cases, either there is not enough data or
the data is inconsistent with the reported conditions. For cases where the door to the compartment
was open, the measured gauge pressures at the start of the experiment ranged from 0 Pa to 10 Pa.
There is not enough information in the test report to explain why the starting pressures were not
0 Pa; thus, the measured pressures were adjusted so that the starting pressure is 0 Pa.

204
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
-250 100

-300 50

-350 0

-400 -50

-450 -100
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0 -100

-50 -150

-100 -200

-150 -250

-200 -300
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
-100 0

-150 -50

-200 -100

-250 -150

-300 -200
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0 -250

-50 -300

-100 -350

-150 -400

-200 -450
0 20 40 60 80 100 0 20 40 60 80 100

205
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
-100 0

-150 -50

-200 -100

-250 -150

-300 -200
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

50
50
0

0 -50

-100
-50
-150

-100 -200
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
-250 0

-300 -50

-350 -100

-400 -150

-450 -200
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0 -250

-50 -300

-100 -350

-150 -400

-200 -450
0 20 40 60 80 100 0 20 40 60 80 100

206
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
50 100

50

0 -50

-100

-150

-50 -200
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0 -100

-50 -150

-100 -200

-150 -250

-200 -300

-250 -350

-300 -400
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
-300 0

-400
-50
-500

-600
-100
-700

-800
-150
-900

-1000 -200
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
-100 0

-20
-150

-40
-200
-60

-250
-80

-300 -100
0 20 40 60 80 100 0 20 40 60 80 100

207
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0 0

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
50 50

40 40

30 30

20 20

10 10

0 0

-10 -10
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0 30

-20
20

-40
10
-60

0
-80

-100 -10
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0 0

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 20 40 60 80 100 0 20 40 60 80 100

208
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0 30

-20
20

-40
10
-60

0
-80

-100 -10
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
30 30

20 20

10 10

0 0

-10 -10
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
30 30

20 20

10 10

0 0

-10 -10
0 20 40 60 80 100 0 20 40 60 80 100

209
C.8 NBS Single Room Tests with Furniture
These data describe a series of room fire tests using upholstered furniture items in a room of fixed
size but with varying opening sizes and shapes [39] conducted by the National Bureau of Standards
(NBS, former name of NIST). It was selected for its well characterized and realistic fuel sources
in a simple single-room geometry. In addition, the wide variation in opening size should provide
challenges for current zone fire models. Peak fire size was about 2.9 MW with a total room volume
of 21 m3 . A series of four single-room fire tests were conducted using upholstered furniture items
for comparison with their free burning behavior, previously determined in a furniture calorimeter.
The experiments were conducted in a single room enclosure; ventilation to the room was provided
by window openings of varying sizes. The room was equipped with an instrumented exhaust
collection system outside the window opening.
A second similar test series also utilized a single-room fire test with furniture as the fire source
[81]. It expanded upon the first data set by adding the phenomenon of wall burning. Peak fire size
was about 7 MW. The room size was similar to the first test series.

210
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 2.5

800 2

600 1.5

400 1

200 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 2.5

800 2

600 1.5

400 1

200 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 2.5

2
600

1.5
400
1

200
0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 2.5

1000
2

800
1.5
600
1
400

0.5
200

0 0
0 2 4 6 8 10 0 2 4 6 8 10

211
 CFAST7.7.1-2-g0b43aa83
0.25

0.2

0.15

0.1

0.05

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
0.25

0.2

0.15

0.1

0.05

0
0 5 10 15 20 25 30

212
 CFAST7.7.1-2-g0b43aa83
1

-1

-2

-3

-4

-5
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
1

-1

-2

-3

-4

-5
0 2 4 6 8 10

213
C.9 NBS Multi-Compartment Test Series
The National Bureau of Standards (NBS, former name of NIST) Multi-Compartment Test Series
consisted of 45 fire tests representing 9 different sets of conditions were conducted in a three-room
suite. The experiments were conducted in 1985 and are described in detail in reference [75]. The
suite consisted of two relatively small rooms, connected via a relatively long corridor. Total volume
of the structure was approximately 100 m2 . The fire source, a gas burner, was located against the
rear wall of one of the small compartments . Fire tests of 100 kW, 300 kW and 500 kW were
conducted. For the current study, three 100 kW fire experiments have been used, including Test
100A from Set 1, Test 100O from Set 2, and Test 100Z from Set 4. For the NBS Multi-room series,
Tests 100A, 100O and 100Z were selected for study, because they were constructively used in a
previous validation study [[108], and because these tests had the steadiest values of measured heat
release rate during the steady burning period. The selected data are also available in Reference
[108].

214
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 2.5

2
300

1.5
200
1

100
0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

215
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 2.5

2
300

1.5
200
1

100
0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

216
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 2.5

2
300

1.5
200
1

100
0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 2.5

100
1.5

1
50

0.5

0 0
0 5 10 15 20 0 5 10 15 20

217
C.10 NIST/NRC Test Series
These experiments, sponsored by the US NRC and conducted at NIST, consisted of 15 large-
scale experiments performed in June 2003. All 15 tests were included in the validation study.
The experiments are documented in Ref. [82]. The fire sizes ranged from 350 kW to 2.2 MW in a
compartment with dimensions 21.7 m by 7.1 m by 3.8 m high, designed to represent a compartment
in a nuclear power plant containing power and control cables. The room had one door and a simple
mechanical ventilation system. Ventilation conditions, the fire size, and fire location were varied.
Numerous measurements (approximately 350 per test) were made.

218
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

219
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

220
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Open Door Tests to follow

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

221
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

222
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

223
 CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

224
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

225
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

226
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Open Door Tests to follow

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

227
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

228
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

229
 CFAST7.7.1-2-g0b43aa83
3000

2500

2000

1500

1000

500

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

230
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

50 50

0 0

-50 -50

-100 -100
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

200 200

0 0

-200 -200

-400 -400
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

100 100

0 0

-100 -100

-200 -200
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

200 200

0 0

-200 -200

-400 -400
0 5 10 15 20 25 30 0 5 10 15 20 25 30

231
 CFAST7.7.1-2-g0b43aa83
400

200

-200

-400
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1 1

0 0

-1 -1

-2 -2

-3 -3

-4 -4

-5 -5
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1 1

0 0

-1 -1

-2 -2

-3 -3

-4 -4

-5 -5
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1 1

0 0

-1 -1

-2 -2

-3 -3

-4 -4

-5 -5
0 5 10 15 20 25 30 0 5 10 15 20 25 30

232
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
3 3

2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

233
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

234
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

235
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
10 10

8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

236
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

237
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

238
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 400

800
300

600
200
400

100
200

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
80 10

8
60

6
40
4

20
2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

239
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
3 3

2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

240
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
10 10

8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

241
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
10 10

8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

242
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

243
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
10 10

8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

244
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
10 10

8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

245
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 10

8
15

6
10
4

5
2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

246
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

247
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

248
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

249
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

250
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

251
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

252
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 400

600 300

400 200

200 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
30 30

25 25

20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

253
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
3 3

2.5 2.5

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

254
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

255
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

256
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

257
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 8

6 6

4 4

2 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

258
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
8 15

6
10

5
2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

259
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

260
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

261
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

262
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Experimental Heat Flux Data Not Available Experimental Heat Flux Data Not Available

263
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

264
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

265
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

266
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

267
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

268
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Experimental Heat Flux Data Not Available Experimental Heat Flux Data Not Available

269
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

270
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

271
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
5

0
0 5 10 15 20 25 30

Experimental Heat Flux Data Not Available

272
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 600

500 500

400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Experimental Heat Flux Data Not Available Experimental Heat Flux Data Not Available

273
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

274
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 600

500 500

400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

275
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
5

0
0 5 10 15 20 25 30

Experimental Heat Flux Data Not Available

276
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

277
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

278
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Experimental Heat Flux Data Not Available Experimental Heat Flux Data Not Available

279
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

280
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 200

250
150
200

150 100

100
50
50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

281
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

282
C.11 NIST/NRC Corner and Wall Effects Experiments
The plots on the following pages compare conditions in a large compartment where corner, wall,
and cabinet effects experiments were conducted. The corner and wall experiments involved a
60 cm by 60 cm natural gas burner with heat release rates of 200 kW, 300 kW, and 400 kW. The
burner was either set in a corner or against a wall. The cabinet experiments involved gas burners
set in one of two mock steel cabinets, with a variety of heat release rates.
In all experiments, two vertical thermocouple arrays were placed along the centerline of the
room, each one-third of the room length from each respective short wall. The arrays each had 13
bare-bead thermocouples. The first was 2 cm below the ceiling, and the rest were spaced 30 cm
apart.

283
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 4

250
3
200

150 2

100
1
50

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 4

250
3
200

150 2

100
1
50

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 4

250
3
200

150 2

100
1
50

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

Figure C.17: NIST/NRC Corner Effects experiments, HGL temperature and height, corner experiments.

284
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 4

250
3
200

150 2

100
1
50

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 4

250
3
200

150 2

100
1
50

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 4

250
3
200

150 2

100
1
50

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

Figure C.18: NIST/NRC Corner Effects experiments, HGL temperature and height, wall experiments.

285
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 20 40 60 80 0 20 40 60 80

Figure C.19: NIST/NRC Corner Effects experiments, HGL temperature and height, cabinet experiments
1-4.
286
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 4

300 3

200 2

100 1

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 4

200
3

150
2
100

1
50

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 4

200
3

150
2
100

1
50

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

Figure C.20: NIST/NRC Corner Effects experiments, HGL temperature and height, cabinet experiments
5-8.
287
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 4

200
3

150
2
100

1
50

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 4

200
3

150
2
100

1
50

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

Figure C.21: NIST/NRC Corner Effects experiments, HGL temperature and height, cabinet experiments
9-12.

288
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 800

1000
600
800

600 400

400
200
200

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 800

1000
600
800

600 400

400
200
200

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 800

1000
600
800

600 400

400
200
200

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

Figure C.22: NIST/NRC Corner Effects experiments, plume temperature.

289
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120

Figure C.23: NIST/NRC Corner Effects experiments, ceiling jet temperature, wall and corner tests.

290
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
500 500

400 400

300 300

200 200

100 100

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

Figure C.24: NIST/NRC Corner Effects experiments, ceiling jet temperature, large cabinet.

291
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

Figure C.25: NIST/NRC Corner Effects experiments, ceiling jet temperature, medium-sized cabinet.

292
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 600

500 500

400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 600

500 500

400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 600

500 500

400 400

300 300

200 200

100 100

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

Figure C.26: NIST/NRC Corner Effects, cabinet temperatures, large cabinet.

293
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 600

500 500

400 400

300 300

200 200

100 100

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
600 600

500 500

400 400

300 300

200 200

100 100

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

Figure C.27: NIST/NRC Corner Effects, cabinet temperatures, medium cabinet.

294
C.12 NIST Smoke Alarm Experiments
The primary purpose of the NIST smoke alarm experiments was to measure smoke alarm activa-
tion times in residential settings. In the single-story manufactured home tests that were selected
for validation, five smoke alarm measurement stations (A-E) were located in different areas of the
manufactured home. Included is a scatterplot of measured and predicted smoke alarm activation
time. Thermocouple trees were also located at each measurement station. The highest thermocou-
ple in the tree can be compared to ceiling jet temperature predictions. The plots on the following
page show the measured and predicted ceiling jet temperatures for the five measurement stations
in each test.

295
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 50 100 150 0 50 100 150

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 50 100 150 0 50 100 150

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 50 100 150 0 50 100 150

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 50 100 150 0 50 100 150

296
C.13 NIST Seven-story Hotel Tests
This data set is part of a series of full-scale experiments conducted to evaluate zoned smoke control
systems, with and without stairwell pressurization [109]. It was conducted in a seven story hotel
with multiple rooms on each floor and a stairwell connecting all floors. This data set was chosen
because it would challenge the scope of most current fire models. Measured temperatures and
pressure differences between the rooms and floors of the building are extensive and consistent.
Peak fire size was 3 MW with a total building volume of 140 000 m3 . The hotel was a masonry
structure consisting of two wings, one three stories and the other seven stories tall. The fires were
set on the second floor of the seven-story wing.

297
 CFAST7.7.1-2-g0b43aa83
200

150

100

50

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
500

400

300

200

100

0
0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
50

40

30

20

10

0
0 5 10 15 20 25 30

298
 CFAST7.7.1-2-g0b43aa83
0.25

0.2

0.15

0.1

0.05

0
0 5 10 15 20 25 30

C.14 NIST Vent Study

The experiments were done in a reduced-scale two story enclosure and the main focus was on
the flow through different ceiling vents. The fire size was 2.5 kW and each compartment had a
total volume of approximately 0.89 m3 . In the experiments, there was a vertical array of thermo-
couples from floor to ceiling on the front-right of the enclosure with eight thermocouples in each
compartment.

299
120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

300
120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

301
120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

302
120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

303
120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

120 1

100
0.8

80
0.6
60
0.4
40

0.2
20

0 0
0 20 40 60 80 100 0 20 40 60 80 100

304
C.15 PRISME DOOR Experiments
The compartments in the PRISME DOOR experiments contained vertical arrays of thermocouples
to measure the HGL temperature and depth. Each array contained 18 TCs and each compartment
included three arrays. The array above the fire was excluded from the calculation of the HGL
temperature and depth.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

305
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
500 5

400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
500 5

400 4

300 3

200 2

100 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

306
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

307
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 5

4
300

3
200
2

100
1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

308
 In the PRISME DOOR experiments, the uppermost TC in the vertical arrays were used to
measure the ceiling jet temperature. These TCs were approximately 10 cm below the ceiling.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
500 500

400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
500 500

400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
700 700

600 600

500 500

400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

309
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

310
 Each compartment in the PRISME DOOR experiments contained an oxygen and carbon diox-
ide measurement in the upper (haut) and lower (bas) layers.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

311
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

312
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

313
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

314
 The PRISME experiments were conducted in a relatively well-sealed set of compartments with
a well-controlled ventilation system. Supply air was forced into and exhaust products extracted
from the test compartments via two fans and a fairly extensive ventilation network. The air flow
rates and nodal pressures were measured throughout the system. The CFAST simulations repre-
sented the ventilation system as a simple matching nominal flow into and out of each compartment
taken from the test reports for the experiments.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

500 500

0 0

-500 -500

-1000 -1000
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

500 500

0 0

-500 -500

-1000 -1000
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1500 1000

1000 500

500 0

0 -500

-500 -1000
0 5 10 15 20 25 30 0 5 10 15 20 25 30

315
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

500 500

0 0

-500 -500

-1000 -1000
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

500 500

0 0

-500 -500

-1000 -1000
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1500 1000

1000 500

500 0

0 -500

-500 -1000
0 5 10 15 20 25 30 0 5 10 15 20 25 30

316
 For the PRISME Door experiments, the temperatures of surrogate cables were predicted di-
rectly from the predicted thermal environment of the entire compartment. The measurement points
in these experiments were labelled, for example, TCA_L2_HE_SURF, meaning thermocouple of
the “analytical” cable, compartment 2, haut (high), east, surface. BW means bas (low) west, for
example. Thermocouples were positioned on the cable surface (SURF), halfway towards center
(INTER), and center (CENTRE).

317
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

318
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

319
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
200

150

100

50

0
0 5 10 15 20 25 30

320
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

321
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

322
 Thermocouples were positioned at various points on the walls in the PRISME Door exper-
iments. Each room contained a vertical array labelled, for example, TP_L1_NE265. The TP
indicates a surface temperature measurement, L1 indicates compartment 1, which is where the fire
was located, NE indicates northeast corner of the room, and 265 indicates the number of centime-
ters above the floor. In addition, each room contained four measurement points centered on each
wall at a height of approximately 260 cm. These points are labelled, for example, TP_L2_SC265,
compartment 2, center of south wall, 265 cm high.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

323
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

324
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 5 10 15 20 25 30 0 5 10 15 20 25 30

325
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 5 10 15 20 25 30 0 5 10 15 20 25 30

326
 Total and radiative heat flux gauges were positioned at various points on the walls. Each room
contained a vertical array labelled, for example, FLT_L1_NC265. The FLT indicates a surface total
heat flux measurement, L1 indicates compartment 1, which is where the fire was located, NC indi-
cates north wall center, and 265 indicates the number of centimeters above the floor. In addition,
each room contained four measurement points centered on each wall at a height of approximately
260 cm. These points are labelled, for example, FLT_L2_SC265, compartment 2, center of south
wall, 265 cm high.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

327
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

328
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
20 20

15 15

10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

329
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

330
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

331
C.16 SP Adiabatic Surface Temperature Experiments
Three experiments were conducted in a standard compartment, 3.6 m long by 2.4 m wide by
2.4 m high, with a 0.8 m wide by 2.0 m high door centered on the narrow wall. A single beam was
suspended 20 cm below the ceiling lengthwise along the centerline of the compartment. There were
three measurement stations along the beam at distances of 0.9 m (Station A), 1.8 m (Station B),
and 2.7 m (Station C) from the far wall where the fire was either positioned in the corner (Tests 1
and 2), or the center (Test 3). The gas temperatures reported here were measured 10 cm away from
all four sides of the beam at Station A, and 10 cm away from the two lateral sides at Stations B and
C. In the figure legends, the measurement station is denoted A, B, or C, and the position is denoted
1, 2, 3, or 4. Position 1 is 10 cm above the beam. Position 2 is 10 cm from the side of the beam
facing away from the fire, Position 3 is 10 cm below the beam, and Position 4 is 10 cm away from
the side of the beam facing the fire.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

332
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

333
 Three experiments were conducted at SP, Sweden, in 2011, in which a 6 m long, 20 cm diameter
vertical column was positioned in the middle of 1.1 m and 1.9 m diesel fuel and 1.1 m heptane pool
fires [89]. Gas, plate, and steel surface temperature measurements were made at heights of 1 m,
2 m, 3 m, 4 m, and 5 m above the pool surface. Gas temperatures were measured with 0.25 mm and
0.50 mm bead thermocouples. The results are very similar and only the 0.25 mm values are used.
In the experiments, the fire was reported to lean. The lean was significant for the 1.9 m diesel fuel
fire. In that case, only data from 1 m and 2 m above the pool are used. The average temperature
between 10 min and 15 min is the basic of comparison.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1000

1000
800

800
600
600
400
400

200
200

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 600

500
600
400

400 300

200
200
100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 5 10 15 20 25 30

334
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1200

1000 1000

800 800

600 600

400 400

200 200

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1000

1000
800

800
600
600
400
400

200
200

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
800 600

500
600
400

400 300

200
200
100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 5 10 15 20 25 30

335
C.17 Steckler Compartment Experiments
Steckler et al. [90] mapped the doorway/window flows in 55 compartment fire experiments. The
test matrix is presented in Table 4.7. Shown on the following pages are the temperature profiles
inside the compartment compared with model predictions. To quantify the difference between
prediction and measurement, the maximum temperatures were compared.

336
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

337
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

338
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

339
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

340
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

341
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

342
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83

2 2

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300

CFAST7.7.1-2-g0b43aa83

1.5

0.5

0
0 50 100 150 200 250 300

343
C.18 UL/NFPRF Series I Experiments
The primary purpose of the UL/NFPRF experiments was to measure sprinkler activation times for
a series of heptane spray burner fires. To determine activation times, thermocouples were affixed
to each sprinkler, and a sudden drop in temperature indicated activation. These same thermocouple
temperatures can be compared to ceiling jet temperature predictions. Referring to Fig. 4.24, the
chosen measurement locations are 56, 68, 86, and 98, providing comparisons as close to, and as
far away from, the fire as possible.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

344
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 200

250
150
200

150 100

100
50
50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

345
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 2 4 6 8 10 0 2 4 6 8 10

346
C.19 UL/NIJ House Experiments
Details of the experiments can be found in Section 4.19
The HGL temperature and height for three experiments conducted in the ranch-style house
were calculated from two vertical arrays of eight thermocouples. The thermocouple arrays were
located in the hallway (4TC) and in the living room (5TC).
The HGL temperature and height for three experiments conducted in the two-story colonial-
style house were calculated from a vertical array of thermocouples in the center of the family room
(8TC). The uppermost TC was 2.5 cm below the ceiling. The second TC was 30 cm (1 ft) below
the ceiling, and the rest were spaced evenly by 30 cm (1 ft).

347
 300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

300 3

250 2.5

200 2

150 1.5

100 1

50 0.5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Figure C.28: UL/NIJ Experiments, HGL temperature and height, single-story ranch-style House Tests 1, 2,
and 5

348
 300 6

250 5

200 4

150 3

100 2

50 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

300 6

250 5

200 4

150 3

100 2

50 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

300 6

250 5

200 4

150 3

100 2

50 1

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Figure C.29: UL/NIJ Experiments, HGL temperature and height, two-story colonial-style House Tests 1, 4,
and 6

349
 The following plots compare the uppermost thermocouple measurements with corresponding
model predictions for the ranch-style and colonial-style houses.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
400 400

300 300

200 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30

Figure C.30: UL/NIJ Experiments, ceiling jet temperature

C.20 UL/NIST Vent Experiments

The HGL temperature and height for the four experiments was calculated from two vertical arrays
of eight thermocouples each. The arrays were centered on the long central axis of the compartment
and 90 cm from each short size wall. The 2.4 m by 1.2 m double vent was 90 cm from each array.

350
The uppermost TC was 2.5 cm below the ceiling. The second TC was 30 cm (1 ft) below the
ceiling, and the rest were spaced evenly by 1 ft.
The ceiling jet temperatures were measured at two locations, 90 cm from the short ends of the
2.4 m by 1.2 m double vent.

351
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 2.5

800 2

600 1.5

400 1

200 0.5

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 2.5

800 2

600 1.5

400 1

200 0.5

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 2.5

800 2

600 1.5

400 1

200 0.5

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 2.5

800 2

600 1.5

400 1

200 0.5

0 0
0 20 40 60 80 0 20 40 60 80

352
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1000

1000
800

800
600
600
400
400

200
200

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1000

1000
800

800
600
600
400
400

200
200

0 0
0 20 40 60 80 0 20 40 60 80

353
C.21 USN High Bay Hangar Experiments
A large number of plume temperature measurements are available from the US Navy experiments
conducted at Keflavik, Iceland, and Barber’s Point, Hawaii. The hangars were very large in size
(22 m high in Iceland and 15 m high in Hawaii) and the heat release rates varied from 100 kW to
33 MW. All experiments made use of a fuel pan filled with either JP-5 or JP-8 jet fuel, positioned
in the center of the hangar.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
25 25

20 20

15 15

10 10

5 5

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
40 40

30 30

20 20

10 10

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
50 50

40 40

30 30

20 20

10 10

0 0
0 5 10 15 0 5 10 15

354
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
25
70

60 20

50
15
40

30 10

20
5
10

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
50 50

40 40

30 30

20 20

10 10

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
50
70

40 60

50
30
40

20 30

20
10
10

0 0
0 5 10 15 0 5 10 15

355
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 250

200

100
150

100
50

50

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
250 100

200 80

150 60

100 40

50 20

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 250

200

100
150

100
50

50

0 0
0 5 10 15 0 5 10 15

356
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
50 100

40 80

30 60

20 40

10 20

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 150

80

100
60

40
50

20

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 200

150 150

100 100

50 50

0 0
0 5 10 15 0 5 10 15

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
200 50

40
150

30
100
20

50
10

0 0
0 5 10 15 0 5 10 15

357
C.22 Vettori Flat Ceiling Experiments
For these experiments, the measured and predicted thermocouple temperature at the location of
the first two activating sprinklers are compared. The experiments consisted of either Smooth or
Obstructed ceilings; Slow, Medium or Fast fires; and a burner in the Open, at the Wall, or in the
Corner. The experiments included three replicates of each of the smooth ceiling configurations and
two replicates of each of the obstructed ceiling configurations.

358
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 0 20 40 60 80

359
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 120 140 0 50 100 150 200

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 50 100 150 200 0 10 20 30 40 50

360
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 10 20 30 40 50 0 10 20 30 40 50

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 0 50 100 150

361
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 50 100 150 0 20 40 60 80 100 120 140

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 50 100 150 200 250 0 50 100 150 200 250

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 10 20 30 40 0 10 20 30 40

362
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 10 20 30 40 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 0 10 20 30 40 50

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 10 20 30 40 50 0 10 20 30 40 50

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140

363
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 0 20 40 60 80 100

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
300 300

250 250

200 200

150 150

100 100

50 50

0 0
0 20 40 60 80 100 0 50 100 150 200

CFAST7.7.1-2-g0b43aa83
300

250

200

150

100

50

0
0 50 100 150 200

364
C.23 VTT Large Hall Tests
The experiments are described in reference [97]. The series consisted three unique fire scenarios
with replications for a total of 8 experiments. The experiments were undertaken to study the
movement of smoke in a large hall with a sloped ceiling. The tests were conducted inside the VTT
Fire Test Hall, with dimensions of 19 m high by 27 m long by 14 m wide. Each test involved a
single heptane pool fire, ranging from 2 MW to 4 MW.

365
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
140

120
15
100

80
10
60

40 5
20

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
140

120
15
100

80
10
60

40 5
20

0 0
0 2 4 6 8 10 0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
140

120
15
100

80
10
60

40 5
20

0 0
0 2 4 6 8 10 0 2 4 6 8 10

366
 CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 2 4 6 8 10

CFAST7.7.1-2-g0b43aa83
400

300

200

100

0
0 2 4 6 8 10

367
C.24 WTC Test Series
The HGL temperature and height for the WTC experiments were calculated from two TC trees,
one that was approximately 3 m to the west and one 2 m to the east of the fire pan (see Fig. 4.32).
Each tree consisted of 15 thermocouples, the highest point being 5 cm below the ceiling.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 4

800
3

600
2
400

1
200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 4

800
3

600
2
400

1
200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 4

800
3

600
2
400

1
200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

368
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 4

800
3

600
2
400

1
200

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 4

800
3

600
2
400

1
200

0 0
0 20 40 60 80 0 20 40 60 80

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 4

800
3

600
2
400

1
200

0 0
0 20 40 60 80 0 20 40 60 80

369
 The following pages contain comparisons of predicted and measured ceiling temperatures,
both at the surface and beneath a layer of marinite board. Table C.1 below lists the coordinates of
the measurement locations relative to the center of the fire pan. Names with “IN” appended are
measurements made under the marinite board.

Table C.1: Locations of ceiling surface temperature measurements relative to the fire pan in the
WTC series.

Name x (m) y (m) z (m)

TCC 0.62 0.07 3.82
TCN3 0.62 0.67 3.82
TCS3 0.62 -0.53 3.82
TCE7 2.18 0.07 3.82
TCW7 -1.15 0.07 3.82
TCCIN 0.62 0.07 3.83
TCN3IN 0.62 0.67 3.83
TCS3IN 0.62 -0.53 3.83
TCE4IN 1.28 0.07 3.83
TCW4IN 0.05 0.07 3.83

370
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1200

1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

371
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1200

1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

372
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1200

1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

373
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1200

1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

374
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1200 1200

1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

375
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

376
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

377
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

378
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

379
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
1000 1000

800 800

600 600

400 400

200 200

0 0
0 20 40 60 80 0 20 40 60 80

380
 Heat flux to the ceiling and the floor were measured with a series of Schmidt-Boelter, water-
cooled heat flux gauges.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 0
0 20 40 60 80 0 20 40 60 80

381
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 150

100 100

50 50

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 150

100 100

50 50

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 150

100 100

50 50

0 0
0 20 40 60 80 0 20 40 60 80

382
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
100 100

80 80

60 60

40 40

20 20

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
150 150

100 100

50 50

0 0
0 20 40 60 80 0 20 40 60 80

383
 The following pages present comparisons of oxygen and carbon dioxide concentration predic-
tions and measurements for the WTC experiments. There was only one measurement of each made
near the ceiling of the compartment roughly 2 m from the fire.
CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

384
 CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

CFAST7.7.1-2-g0b43aa83 CFAST7.7.1-2-g0b43aa83
0.25 0.1

0.2 0.08

0.15 0.06

0.1 0.04

0.05 0.02

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60

385