ABSTRACT
Title of Thesis: INTERACTION OF ACOUSTIC WAVES
WITH A LAMINAR LINE-FLAME
Adam N. Friedman, Master of Science, 2016
Thesis directed by: Professor Stanislav I. Stoliarov
Department of Fire Protection Engineering
A systematic study was conducted to elucidate the effects of acoustic perturbations
on laminar diffusion line-flames and the conditions required to cause acoustically-
driven extinction. Flames were produced from the fuels n-pentane, n-hexane,
n-heptane, n-octane, and JP-8, using fuel-laden wicks. The wicks were housed
inside of a burner whose geometry produced flames that approximated a two
dimensional flame sheet. The acoustics utilized ranged in frequency between 30-50
Hz and acoustic pressures between 5-50 Pa. The unperturbed mass loss rate and
flame height of the alkanes were studied, and they were found to scale in a linear
manner consistent with Burke-Schumann. The mass loss rate of hexane-fueled
flames experiencing acoustic perturbations was then studied. It was found that
the strongest influence on the mass loss rate was the magnitude of oscillatory air
movement experienced by the flame. Finally, acoustic perturbations were imposed
on flames using all fuels to determine acoustic extinction criterion. Using the
data collected, a model was developed which characterized the acoustic conditions
required to cause flame extinction. The model was based on the ratio of an
acoustic Nusselt Number to the Spalding B Number of the fuel, and it was found
that at the minimum speaker power required to cause extinction this ratio was a
constant. Furthermore, it was found that at conditions where the ratio was below
this constant, a flame could still exist; at conditions where the ratio was greater
than or equal to this constant, flame extinction always occurred.
Interaction of Acoustic Waves
with a Laminar Line-Flame
by
Adam N. Friedman
Thesis submitted to the Faculty of the Graduate School of the
University of Maryland, College Park in partial fulfillment
of the requirements for the degree of
Master of Science
2016
Advisory Committee:
Professor Stanislav I. Stoliarov, Chair
Professor Michael J. Gollner
Professor Andre W. Marshall
© Copyright by
Adam N. Friedman
2016
All rights reserved. No part of this work may be reproduced or
transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or by any
information storage and retrieval system, without permission in
writing from the author.
Dedicated to Zaydie. . .
ii
Acknowledgements
This project has required considerable help from many people along the way, and
there isn’t enough room to thank them all. I would especially like to thank though
Dr. Ed Habtour and Mr. Brent Mills from the U.S. Army Research Lab. This
project would not have been possible without them, and their efforts on my behalf
have truly been appreciated. I would also like to thank the Federal Aviation
Administration for providing the initial funding for this project. Lastly, I would
like to give a special thanks to my adviser, Dr. Stanislav Stoliarov. He graciously
indulged my idiosyncrasies while guiding me through the process of producing
quality research and succinct writing.
This research was supported in part by an appointment to the Student Research Participation Program at the
U.S. Army Research Laboratory administered by the Oak Ridge Institute for Science and Education through an
inter-agency agreement between the U.S. Department of Energy and USARL
iii
Contents
Acknowledgements iii
List of Figures vii
List of Tables ix
Physical Constants x
Symbols xi
1 Introduction 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Flame Extinction . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.2 Interaction of Acoustics and Flames . . . . . . . . . . . . . . 5
1.3 Scope of Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2 Testing Apparatus 13
2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Acoustic Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.1 Signal generation . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.2 Sub-Woofer . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3 Burner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 Testing Enclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3 Data Acquisition 23
3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 Acoustic Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.1 Microphone Calibration . . . . . . . . . . . . . . . . . . . . 24
3.3 Anemometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.4 High Speed Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.5 Mass Readings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
iv
4 System Characterization 28
4.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.2 System Harmonics . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.2.1 Predicted Resonant Frequencies . . . . . . . . . . . . . . . . 29
4.2.2 Observed Resonant Frequencies . . . . . . . . . . . . . . . . 30
4.2.3 Comparison of Predicted and Observed Resonance . . . . . . 33
4.3 Pressure Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4 Acoustic Pressure Profiles . . . . . . . . . . . . . . . . . . . . . . . 37
4.4.1 Profiles Without Burner . . . . . . . . . . . . . . . . . . . . 37
4.4.2 Profiles with Burner . . . . . . . . . . . . . . . . . . . . . . 37
4.4.3 Comparison of Profiles . . . . . . . . . . . . . . . . . . . . . 39
4.5 Acoustically Induced Flows . . . . . . . . . . . . . . . . . . . . . . 41
4.5.1 PIV Analysis of Acoustic Flow . . . . . . . . . . . . . . . . . 41
5 Free Burn Characterization 47
5.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.2 Mass Loss Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.3 Flame Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.4 Flame Width . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.5 Flame Height Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . 58
6 Burning Rate in an Acoustic Field 61
6.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
6.2 MLR Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
6.3 MLR Profile Comparisons . . . . . . . . . . . . . . . . . . . . . . . 66
6.3.1 Comparable Acoustic Pressures . . . . . . . . . . . . . . . . 66
6.3.2 Comparable Air Speeds . . . . . . . . . . . . . . . . . . . . . 67
6.3.3 Average Mass Loss Rates . . . . . . . . . . . . . . . . . . . . 69
7 Acoustic Extinction 72
7.1 Experimental Summary . . . . . . . . . . . . . . . . . . . . . . . . 72
7.1.1 Acoustic Extinction Results . . . . . . . . . . . . . . . . . . 72
7.1.2 Fan-Driven Extinction Results . . . . . . . . . . . . . . . . . 75
7.2 Proposed Acoustic Extinction Theory . . . . . . . . . . . . . . . . . 76
7.2.1 Comparison of Results . . . . . . . . . . . . . . . . . . . . . 76
7.2.2 Discussion of Results . . . . . . . . . . . . . . . . . . . . . . 77
7.2.3 Heuristic Framework . . . . . . . . . . . . . . . . . . . . . . 80
7.2.4 Proposed Extinction Criterion . . . . . . . . . . . . . . . . . 81
7.3 B - Mass Transfer Number . . . . . . . . . . . . . . . . . . . . . . . 83
7.3.1 Fuel Specific Parameters . . . . . . . . . . . . . . . . . . . . 83
7.3.2 Air Properties . . . . . . . . . . . . . . . . . . . . . . . . . . 86
7.3.3 B Number Calculation . . . . . . . . . . . . . . . . . . . . . 87
7.4 Nusselt Number Correlation . . . . . . . . . . . . . . . . . . . . . . 88
7.4.1 Reynolds Number . . . . . . . . . . . . . . . . . . . . . . . . 88
7.4.2 Mathematical Formulation . . . . . . . . . . . . . . . . . . . 88
7.4.3 Modified Nusselt Number . . . . . . . . . . . . . . . . . . . 89
7.4.4 Exponent of Best Fit . . . . . . . . . . . . . . . . . . . . . . 90
7.5 Calculated Values of Θ′ . . . . . . . . . . . . . . . . . . . . . . . . . 91
7.5.1 Acoustic Results . . . . . . . . . . . . . . . . . . . . . . . . 91
7.5.2 Hypo and Hyper Critical Values . . . . . . . . . . . . . . . . 92
7.5.3 Fan Driven Results . . . . . . . . . . . . . . . . . . . . . . . 94
7.6 Limitations and Considerations . . . . . . . . . . . . . . . . . . . . 96
7.6.1 System Limitations . . . . . . . . . . . . . . . . . . . . . . . 96
7.6.2 Considerations of Applicability . . . . . . . . . . . . . . . . 97
8 Conclusions 101
A Error Analysis 105
A.1 Data Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
A.2 Fuel Parameter Uncertainties . . . . . . . . . . . . . . . . . . . . . 108
A.2.1 Model Uncertainties . . . . . . . . . . . . . . . . . . . . . . 108
A.2.2 Uncertainties in Acoustic Analysis . . . . . . . . . . . . . . . 109
A.2.3 Uncertainties in Fan Driven Analysis . . . . . . . . . . . . . 110
B Product Data Sheets 112
B.1 Agilent 3220A Function Generator . . . . . . . . . . . . . . . . . . 112
B.2 Infinity Reference 860w 8” Woofer . . . . . . . . . . . . . . . . . . . 116
B.3 AE Techron 8102 Amplifier . . . . . . . . . . . . . . . . . . . . . . 118
B.4 Kaowool Ceramic Insulation . . . . . . . . . . . . . . . . . . . . . . 120
B.5 BSWA MPA 231 Microphone . . . . . . . . . . . . . . . . . . . . . 122
B.6 BSWA MC102 Signal Conditioner . . . . . . . . . . . . . . . . . . . 123
B.7 BSWA CA111 Calibrator . . . . . . . . . . . . . . . . . . . . . . . . 124
B.8 Tektronix 2024C Digital Oscilloscope . . . . . . . . . . . . . . . . . 125
B.9 Extech Termo-Anemometer . . . . . . . . . . . . . . . . . . . . . . 128
B.10 Phantom High Speed Camera . . . . . . . . . . . . . . . . . . . . . 129
B.11 Mettler Toledo Precision Balance . . . . . . . . . . . . . . . . . . . 132
Bibliography 134
List of Figures
2.1 Testing Apparatus Overview . . . . . . . . . . . . . . . . . . . . . . 14
2.2 Plan and View of Woofer Sabot . . . . . . . . . . . . . . . . . . . . 16
2.3 Speaker Sabot Placement . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 Burner Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5 Burner Assembly Sequence . . . . . . . . . . . . . . . . . . . . . . . 20
2.6 Views of Flame Sheet . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.7 Plan of Testing Enclosure . . . . . . . . . . . . . . . . . . . . . . . 21
2.8 Views of Testing Enclosure . . . . . . . . . . . . . . . . . . . . . . . 22
2.9 Equipment Placement . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.1 Burner Balance Stand . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.1 1st Harmonic Tube Interior . . . . . . . . . . . . . . . . . . . . . . . 31
4.2 2nd Harmonic Tube Interior . . . . . . . . . . . . . . . . . . . . . . 31
4.3 3rd Harmonic Tube Interior . . . . . . . . . . . . . . . . . . . . . . 32
4.4 1st Harmonic Tube Exterior . . . . . . . . . . . . . . . . . . . . . . 34
4.5 2nd Harmonic Tube Exterior . . . . . . . . . . . . . . . . . . . . . . 34
4.6 3rd Harmonic Tube Exterior . . . . . . . . . . . . . . . . . . . . . . 35
4.7 Exterior Acoustic Pressure Decay . . . . . . . . . . . . . . . . . . . 36
4.8 Pressure Profiles Without Burner . . . . . . . . . . . . . . . . . . . 38
4.9 Pressure Profile Placement Schematic . . . . . . . . . . . . . . . . . 39
4.10 Pressure Profiles With Burner . . . . . . . . . . . . . . . . . . . . . 40
4.11 Air Speed vs Acoustic Pressure . . . . . . . . . . . . . . . . . . . . 42
4.12 PIVLab Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.13 PIV Calculated Velocity Components . . . . . . . . . . . . . . . . . 45
4.14 PIV vs. Anemometer . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.1 Alkane MLR Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.2 Flame Height Video Sample . . . . . . . . . . . . . . . . . . . . . . 51
5.3 Flame Height ROI . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.4 Hexane Flame Luminous Intensity Profiles . . . . . . . . . . . . . . 53
5.5 Flame Height Determination . . . . . . . . . . . . . . . . . . . . . . 54
5.6 Alkane Flame Height Profiles . . . . . . . . . . . . . . . . . . . . . 55
vii
5.7 Flame Width Determination . . . . . . . . . . . . . . . . . . . . . . 57
5.8 MLR and Flame Height . . . . . . . . . . . . . . . . . . . . . . . . 58
5.9 Flame Height vs. HRR′ . . . . . . . . . . . . . . . . . . . . . . . . . 59
6.1 Burner on Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
6.2 Hexane MLR Profiles with Acoustics . . . . . . . . . . . . . . . . . 64
6.3 Hexane MLR Profiles with Fan Driven Flow . . . . . . . . . . . . . 65
6.4 MLR Profiles at Comparable Pressures . . . . . . . . . . . . . . . . 67
6.5 MLR Profiles at Comparable Air Speeds . . . . . . . . . . . . . . . 68
6.6 Average MLR’s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
7.1 Average Values of P and UA at Extinction . . . . . . . . . . . . . . 74
7.2 Heptane Extinction Boundaries . . . . . . . . . . . . . . . . . . . . 75
7.3 Average Values of UF at Extinction . . . . . . . . . . . . . . . . . . 76
7.4 Comparison of Acoustic and Fan Extinction Flows . . . . . . . . . . 78
7.5 Fan-Drive Flow Visualization . . . . . . . . . . . . . . . . . . . . . 79
7.6 Flame Displacement Sequence . . . . . . . . . . . . . . . . . . . . . 82
7.7 Coefficient of Variation for Model . . . . . . . . . . . . . . . . . . . 91
7.8 Calculated Values of Θ′Aext . . . . . . . . . . . . . . . . . . . . . . . 92
7.9 Hypo and Hyper Critical Extinction Data Points . . . . . . . . . . 94
7.10 Calculated Values of Θ′F . . . . . . . . . . . . . . . . . . . . . . . . 95
7.11 Speaker and Acoustic Wave Power . . . . . . . . . . . . . . . . . . 97
7.12 Ratio of acoustic power to speaker power . . . . . . . . . . . . . . . 97
7.13 Ratio of ` to wb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
List of Tables
4.1 Calculated Resonant Frequencies . . . . . . . . . . . . . . . . . . . 30
4.2 Predicted and Measured Resonant Frequencies . . . . . . . . . . . . 32
4.3 Pressure Profile Comparison . . . . . . . . . . . . . . . . . . . . . . 41
5.1 Flame Height Data Structure . . . . . . . . . . . . . . . . . . . . . 52
7.1 Acoustic Flame Extinction Test Results . . . . . . . . . . . . . . . . 73
7.2 Fan-Driven Flame Extinction Test Results . . . . . . . . . . . . . . 76
7.3 Selected Properties of Fuels and Gases . . . . . . . . . . . . . . . . 85
7.4 Selected Properties of Air for Different Fuels . . . . . . . . . . . . . 87
7.5 Calculated Fuel Properties . . . . . . . . . . . . . . . . . . . . . . . 87
7.6 Analysis of Hypo and Hyper Critical Extinction Conditions . . . . . 93
A.1 Uncertainty Analysis Data Structure . . . . . . . . . . . . . . . . . 106
ix
Physical Constants
Constant Name Symbol Constant Value
Speed of Sound cs 340.3 m/s
Ambient Air Temperature T∞ 298 K
Ambient Air Density ρ∞ 1.23 kg/m3
x
Symbols
Symbol Name Unit
A Flam Height Scaling Factor m/W
aE Extinction Strain Rate 1/s
cp Specific Heat J/kg·K
D Diameter m
D Diffusivity m2/s
dB Decibels
e Correction Factor m
∆F Frame Smoothing Kernel
hv Heat of Vaporization J/kg
∆Hc Heat of Combustion J/mol
∆hc Heat of Combustion J/kg
∆h◦f,liq Liquid Heat of Formation J/kg
∆h◦f,gas Gaseous Heat of Formation J/kg
I Pixel Intensity
k Calibration Constant Pa/mV
L Latent Heat of Gassification J/kg
Lf Flame Height m
Lo Length of Acoustic Column m
` Characteristic Length m
m˙ Mass Loss Rate kg/s
m˜ Average Mass Loss Rate kg/s
N Number of Elements in Set
n Harmonic Mode
P Pressure Pa
xi
Q˙ Heat Release Rate W
r Stoichiometric Fuel to Air Ratio
S Uncertainty
T Temperature K
Tb Boiling Temperature K
UA Acoustic rms air speed m/s
UF Fan Driven Air Speed m/s
u u Velocity Component m/s
V Voltage V
v v Velocity Component m/s
Wf Flame Width m
wb Burner Width m
x Distance m
Y Mass Fraction
y Distance m
Z Specific Acoustic Impedance Pa·s/m
z Distance m
γ Reynolds Number Exponent
δ Prandtl Number Exponent
Θ Acoustic Extinction Criterion
λ Wavelength m
µ Arithmetic Mean
ν Kinematic Viscosity m2/s
ρ Density kg/m3
σ Standard Deviation
τ Characteristic Time s
ω Frequency Hz
ωRn Resonant Frequency of n
th mode Hz
B Spalding Mass Transfer Number
Da Damko¨holer Number
Nu Nusselt Number
Pr Prandtl Number
Re Reynolds Number
Chapter 1
Introduction
1.1 Motivation
Since the early 1900’s halon has been used to effectively extinguish fires [1]. By
the 1960’s, halon began to see widespread use in the U.S. Military and quickly
become one of its primary means for suppressing fires and explosions [2]. Due
to the ozone depleting properties of halon though, the United States became a
signatory to the Montreal Protocol in 1987, which effectively ended the production
of halon worldwide [3]. In order to meet its present needs for halon, the U.S.
Military maintains a reserve that is supplemented with product acquired from
decommissioned systems.
With no new halon available, finding a suitable halon replacement technology has
become an active area of research for the U.S. Military [2]. In August 2013, the
1
Introduction
U.S. Army Research Lab (ARL) approached the Department of Fire Protection
Engineering at the University of Maryland, College Park, to initiate a joint research
project in the field of novel fire suppression. The objective of this research was to
explore the feasibility of using acoustics to achieve flame extinction. Prior research
funded by the Defense Advanced Projects Research Agency (DARPA) had explored
acoustic flame suppression using gaseous fuel sources [4], and the proposed research
was meant to be a continuation of DARPA’s work. Of particular interest to ARL
was whether acoustics could be used to suppress flames from a liquid fuel source
and, if so, the development of a model that could predict the conditions required.
The following work is the fulfillment of that research goal.
1.2 Background
1.2.1 Flame Extinction
In a general sense, diffusion flame extinction occurs when heat losses from a
flame become sufficiently large that the heat released during combustion can no
longer maintain a temperature which will sustain chemical kinetics [5]. In this
context, Quintiere and Rangwala have proposed that for every flame there exists
a critical temperature, below which flame extinction will occur [6]. Using this
framework, a qualitative understanding of the conditions that cause flame extinction
can be gained. For example, the introduction of a diluent reduces the oxidizer’s
2
Introduction
concentration, thereby reducing the chemical reaction and heat release rates (HRR)
[7]; the application of liquid water not only cools a fuel source and inhibits pyrolysis,
but also causes significant heat losses from the flame during the conversion process
to steam [8]. Flame extinction occurs in both these examples because heat losses
cause chemical kinetics to slow, which reduces the HRR, thereby further lowering
temperatures in the flame and retarding chemical kinetics.
Chemical kinetics alone though do not fully explain extinction phenomenon [9].
Closely coupled with kinetics are heat and mass transport processes, and these
are needed to create an environment where combustion chemistry can occur. The
amount of strain (or stretch) experienced by the flame is of particular interest when
considering transport processes, and there are multiple ways to conceptualize this.
If a diffusion flame is conceptualized in the context of a reactive flow, then it is
useful to consider strain as a measure of the rate of deformation in the flow [10]. As
strain in the reactant flow increases, the residence time of the reactants decreases
[11]. This in turn has the effect of lowering both the reaction and HRR [11, 12].
If a diffusion flame is conceptualized in the context of a “cellular” entity with a
surface, then it may be useful to consider strain in the context of flame stretch. In
this context, flame stretch is a measure of curvature in the flame’s surface. As flame
stretch increases, the symmetry between the fuel and oxidizer sides of the flame
decreases, and this causes imbalances in heat and mass fluxes on either side of
the flame sheet [13]. For steady state diffusion flames in particular, Lin˜a´n showed
3
Introduction
through theoretical analysis that increased flame stretch caused increased heat
losses from the flame on both sides of the reaction zone [14].
For unsteady flames it may also be useful to conceptualize flame stretch as the
proportional rate of change in the flame’s surface area with respect to time [15].
Using this conceptualization, Katta et al showed that the effects of stretch on the
unsteady flame also created transport imbalances. In their study of an unsteady
opposing jet flame, they found that as flame stretch increased, the amount of
reactant that was able to enter the reaction zone also increased. The net effect
was that while the HRR increased, there was also an increase in the amount of
products from incomplete combustion. These products acted as a heat sink within
the flame and actually lowered the flame’s temperature [16].
With its multiple physical interpretations and strong effects on flame chemistry,
flame stretch has become a commonly used criterion for predicting flame extinc-
tion [5, 9, 11, 12, 17–19]. Underlying all these interpretations, though, is the
understanding that increased flame stretch enhances transport processes, which in
turn competes with combustion chemical kinetics. Using an Asymptotic Energy
Analysis (AEA), Lecoustre et al showed in fact that as flame stretch increased,
the temperature required to sustain a diffusion flame also increased. Since the
effects of transport processes and chemical kinetics are so closely coupled, it is
often desirable to represent them in relationship to each other. Such a comparison
is commonly done through the use of a Damko¨holer number.
4
Introduction
The Damko¨hler number (Da) of a flame is generally defined as the ratio of a char-
acteristic mixing (transport or residence) time (τmix) to a characteristic chemistry
time (τchem) [5, 17–20]. Mathematically, this is expressed as Da=τmix/τchem. Rates,
though, are inversely proportional to time, and it is sometimes useful to express
the Damko¨hler number as Da = rxn rate/mix rate. For large values of Da, it is
expected that the effects of slow transport processes will dominate, and flame
chemistry will occur at a faster rate. As values of Da become smaller, though, the
effects of increased transport rates and slower chemical kinetics begin to dominate
until the system becomes non-reactive [20]. Therefore, for every flame there exists
a critical value of Da, below which flame extinction will occur [5, 18, 21].
1.2.2 Interaction of Acoustics and Flames
The interaction of acoustic waves and flames has been a field of study since the
1960’s [22]. The primary focus of this early research was the effects of acoustics
on droplet burning in turbine engines and combustion chambers; this continues
to be an active field of research [5, 22–24]. The results of this research showed
that acoustics do influence droplet combustion by altering the rates of heat and
mass transfer [24]. Of particular interest in this field of study are the instabilities
that form within a combustion chamber [5]. These instabilities can often lead to
combustion inefficiencies and damage to the chamber [5, 25]. The causes of these
phenomena are due to disturbances in the reaction flow field and are reviewed in
detail by O’Connor et al [26].
5
Introduction
Although acoustics can lead to inefficiencies during droplet combustion, they can
also be used to enhance combustion. For example, during spray processes fuel
droplets often break up into even smaller droplets, and some of these sub-droplets
may be inhomogeneous with lower boiling temperatures. Rapid boiling of these
sub-droplets can create micro-explosions which cause further droplet breakup and
leads to instabilities within the reactor. Miglani et al found that the application of
acoustics in the narrow bandwidth of 80 to 120 Hz could stabilize the fuel droplets
and reduce the number of sub-droplets formed [27].
Acoustics can also be used to modulate the burning rate and combustion chemistry
of fuel droplets. Sevilla-Esparza et al studied the droplet combustion of ethanol,
methanol, JP-8, and a synthetic fuel at various frequencies within a standing wave.
The acoustics were produced with two movable speakers, which allowed them to
also study the droplets at different phase angles within the standing wave. Their
results showed that the burning rate of each fuel was sensitive to both frequency
and phase angle, and they attributed this sensitivity to the deflection angle of the
droplet within the wave. By measuring OH∗ chemiluminescence, they were also
able to quantify the HRR for methanol droplets at different frequencies. They
found that at low acoustic frequency, there was a strong coupling between the
relative OH∗ concentration and the acoustic pressure. As the acoustic frequency
was increased, the strength of this coupling was seen to diminish. They concluded
that the coupling was attributable to the magnitude of the velocity perturbations
experienced by the droplet, which decreased as the acoustic frequency increased
6
Introduction
[24].
Within the context of this research, there have also been investigations into the
acoustically-driven extinction of droplet flames. McKinney and Dunn-Rankin
studied this phenomenon using a streaming flow of methanol droplets. Droplets of
various sizes were injected into a resonating tube and exposed to acoustic waves
at various frequencies and pressures to identify extinction criteria. They found
that at the same frequency, the acoustic pressure required to cause extinction
increased with droplet size. They also found that for droplets of the same size,
the acoustic pressure required to cause extinction increased with frequency. The
authors determined that extinction occured when the flame was displaced far
enough from the droplet that evaporation was shut down. Key to their findings
was that the magnitude of displacement had to be at least the radius of the droplet
[23].
More recently, there has been a growing body of research on the interaction of
acoustics with both premixed and diffusion flames using gaseous fuel sources
[4, 21, 28–34]. The breadth of this research has included a myriad of topics such
as pollutant reduction [32], combustion instabilities [34], and even acoustic flame
extinction [4, 21, 31]. Key to this research has been the need to understand the
effects of an oscillatory strain rate on a flame. Of particular interest are the effects
when a flame is near its extinction limit, where flames show an increased sensitivity
to acoustic excitations [28].
7
Introduction
The response of a flame to acoustic excitations can be classified as either linear
or non-linear with respect to the excitation frequency [33]. Kim and Williams
studied linear responses and acoustic extinction criteria by applying a theoretical
analysis to a counter-flow diffusion flame. In their analysis, they adopted the model
developed by Lin˜a´n, where the reactive layer is shifted to the oxidizer side of the
stagnation plane and sandwiched between two convective-diffusive layers [14]. They
then considered the effects of acoustic perturbations on the reactive layer in the
frequency range of 103-104 Hz, which is on the same order of magnitude as the
extinction strain rate for most hydrocarbon fuels. To evaluate their results, the
authors used a Rayleigh criterion, which states that acoustic instabilities become
greatest when the acoustic pressure and flame’s HRR are in phase [35]. The results
of their analysis showed that linear responses in the flame’s position, HRR, and
values of Da were caused by oscillations in the position of the reaction sheet and
magnitude of field variables (e.g. pressure, velocity, density) in the transport zone.
Near the flame’s extinction limit they found that it was the oscillations in the
reaction sheet which were dominant and produced the most dramatic effect. When
closer to equilibrium conditions though, they found that oscillations in the field
variables that were dominant, although these produced much less dramatic effects
[21].
Wang et al studied the non-linear effects of acoustics on the puffing frequency and
flame height of a buoyant diffusion methane flame. In their study, they used a
bluff-body stabilized burner and introduced acoustics of varying frequency and
8
Introduction
amplitude into the fuel flow prior to its exit. The frequencies tested ranged from
6-100 Hz, and the acoustic pressures tested ranged from 1.1-90 Pa. A high speed
camera was used to measure the flame’s height, from which the puffing frequency
could be determined. Their results showed that while acoustics produced effects
over the entire frequency range tested, the effects were particularly pronounced in
the range of 6-20 Hz. Within this range, they found that the puffing frequency of
the flame was half the excitation frequency, which they attributed to sequential
bulges in the flame’s natural puffing cycle being merged into one. At higher
frequencies, they found there was a “doubling” effect on the flame’s puffing, which
they attributed to breakdowns in the flow structures occurring at a faster rate.
[30].
Complimenting the work of Wang et al was a study by Chen et al, who also
examined the effects of acoustics on a buoyant diffusion flame. In their study, they
placed a propane burner at the approximate midpoint of a 1.1 m long glass tube
with a square cross section and speaker mounted at the bottom. Acoustic waves at
frequencies of 90, 150, and 200 Hz were then produced and Schlieren imaging was
used to study the effects on the flame’s flow field. They also noted that the most
dramatic effects on the flame’s height occured at the lowest frequency, and that
at the highest frequency there was a non-linear response in the flame’s flickering
frequency [33].
Interspersed throughout this body of work have been several studies that explored
acoustic extinction for gaseous-fueled flames[4, 21, 31, 36]. Although Kim and
9
Introduction
Williams did identify acoustic extinction criteria in their theoretical work, their
results only applied to oscillations of the flame’s reaction zone and its affects on
flame chemistry [21]. Other authors have examined the phenomenon for flames
in the context of a buoyant flow field. For example, Hardalupas and Selbach
studied acoustic extinction for a methane flame [31], and Whiteside studied the
phenomenon for methane along with several other gaseous fuels [4].
Hardalupas and Selbach used a co-axial swirl stabilized burner with acoustics
introduced in the fuel and oxidizer flows below the flame. Using frequencies of 200-
920 Hz, they created a flame in a lifted state, from which they determined conditions
for reattachment. They found that at certain frequencies the acoustic perturbations
could cause a lifted flame to reattach. They attributed this phenomenon to the
creation of vortex rings by the waves and the impulses they impressed on the flame
as they shed from the flow. They also explored flame extinction at 200 Hz and 350
Hz, and concluded that the mechanism of flame extinction was blow off. While
the work of Wang [30] and Chen [33] indicated that the flame’s response was more
sensitive to acoustic frequency than amplitude, Hardalupas and Selbach came to a
different finding. They concluded that it was the acoustic amplitude which has the
strongest effect, since that caused the largest movement of air [31].
Whiteside looked specifically at acoustic flame extinction from a burner using
methane, ethanol, hexane, and heptane. In contrast to the work of other authors,
Whiteside used acoustics that propagated in a transverse direction to the flame.
Frequencies ranging from 35-150 Hz were employed, with acoustic pressures ranging
10
Introduction
from 0.2-112 Pa. Whiteside’s data showed that as the molar mass of the fuel
increased, so too did the acoustic pressure required to cause flame extinction.
The extinction pressure though for each fuel was independent of the burner’s
cross-sectional area. The author concluded that there was a minimum acoustic
velocity required to cause extinction for each fuel, and that acoustic extinction
could be achieved at any frequency provided the acoustic pressure was high enough
to achieve that velocity. Whiteside also concluded that blow off alone did not fully
explain the extinction mechanism, since the flame could exist in a lifted state for
short periods [4].
1.3 Scope of Work
While other authors have explored acoustic extinction criterion for fuel droplet
flames [23] and gaseous flames from a burner [4, 31], there has been no work in this
context on flames fueled by a stagnant liquid. The flame from a stagnant liquid
though represents the most realistic scenario from a fire-protection perspective.
In addition, the governing phenomena of the observed extinctions, especially in
the case of Whiteside’s work, are not fully understood. An investigation into
acoustically-driven flame extinction, especially for flames with liquid fuel sources,
is therefore ripe for inquiry.
An apparatus was constructed that produced collimated acoustic waves which
could interact with a laminar diffusion line-flame that approximated a flame sheet.
11
Introduction
The flame was fed by a liquid fuel source, and to limit the transient effects of
heat feedback, the fuel was supplied through a wick. The fuels chosen for testing
were n-pentane, n-hexane, n-heptane, n-octane, and JP-8. By modulating the
frequency and amplitude of the acoustics produced, the conditions required to
cause acoustic extinction of flames from each fuel could be determined. The
work was supplemented with studies of the alkanes’ burning rate and flame height
unperturbed, and the study of a hexane flame’s burning rate experiencing acoustic
perturbation.
12
Chapter 2
Testing Apparatus
2.1 Overview
A testing apparatus was designed that facilitated the study of a laminar flame
experiencing acoustic perturbation. The primary objective of the design was to
create a line-flame that approximated a flame sheet, and which could interact with
a planar acoustic wave front simultaneously across the flame’s entire surface. Other
key design features included minimizing the transient effects of heat feedback into
the fuel and errant air flows around the flame.
The apparatus involved three main components: an acoustic source and collimator,
burner, and testing enclosure. The acoustic source and collimator consisted of a
sub-woofer mounted inside a 2.54×10−1 m diameter, 3.05 m long PVC tube. In
some experiments, the sub-woofer was replaced with a fan to study the effects of
13
Testing Apparatus
a forced flow on the flame. At the opposing end, a line-flame was created from a
fuel laden wick mounted inside the burner. The burner was housed in a screened
enclosure which was large enough to accommodate both it and the data acquisition
equipment (DAQ). Depending on the experiment being conducted, the burner
was either supported on a fixed stand or a mass balance. A schematic of the
testing apparatus is shown in Fig. (2.1). Detailed descriptions of the acoustic
source, burner, and testing enclosure are provided in Sec.’s (2.2), (2.3), and (2.4)
respectively.
PVC Tube
3.05 m Long
0.25 m ?
Screened
Enclosure
Speaker or Fan
Burner on
Stand
Burner on
Balance
Figure 2.1: Testing Apparatus Overview
14
Testing Apparatus
2.2 Acoustic Source
The acoustics used for testing were generated using a 2.03×10−1 m diameter sub-
woofer which was mounted on a sabot and placed inside the tube. The signal sent
to the sub-woofer originated at a signal generator, which was used to modulate
both the frequency and amplitude of the signal. The signal was then sent to an
amplifier, which increased the power of the signal to sub-woofer’s operational range.
2.2.1 Signal generation
Signals for the sub-woofer were generated using the Agilent 3220A, 20 MHz signal
generator; it’s data sheet is included in Appendix (B.1). The signals used for
testing ranged in frequency from 30-50 Hz, and voltages from 50-1500 mVrms.
Upon generation, the signals were sent along 16-gauge stranded copper wire to
an AE Techron 8102 Amplifier, where the voltage was increased by a factor of
20; the amplifier’s data sheet is included in Appendix (B.3). The electrical power
developed by the amplifier ranged between 2.50×10−1 W and 2.25×102 W, which
was also delivered along 16-gauge stranded copper wire to the sub-woofer.
2.2.2 Sub-Woofer
The sub-woofer used was an Infinity Reference 860w; it’s data sheet is included
in Appendix (B.2). The outside diameter of the sub-woofer measured 2.03×10−1
15
Testing Apparatus
m, which was 5.1×10−2 m less than the inside diameter of the tube. A sabot was,
therefore, constructed which held the sub-woofer in place at the tube’s center.
The sabot was made from two annular rings cut from 3.18×10−3 m thick aluminum
plating. The rings had an outside diameter of 2.54×10−1 m and an inside diameter
2.03×10−1 m. The rings were spaced 4.19×10−1 m apart using three 1.27×10−2 m
diameter threaded rods. A schematic and picture of the sabot with the speaker
are shown in Fig.’s (2.2a) and (2.2b) respectively.
5.08 E-1 m
4.19 E-1 m
?2.54 E-1 m
?2.03 E-1 m
(a) Plan of woofer sabot (b) View of woofer sabot
Figure 2.2: Plan and view of woofer sabot
The sabot was placed inside the tube with the sub-woofer’s face located 6.1×10−1 m
from the tube opening, as shown in Fig. (2.3). Rubber door stops (not shown) were
wedged between the protruding sections of the threaded rods and the sidewall of
the tube to help dampen the sabot’s vibrations. A foam disk measuring 7.62×10−2
m thick (not shown) was used to plug the back of the tube.
16
Testing Apparatus
6.10 E-1 m
1.97 E-1 m
Figure 2.3: Speaker Sabot Placement
2.3 Burner
The burner was constructed from 3.18×10−3 m thick steel sheet metal, cut by high
pressure water jets, and welded at the component interfaces. The burner consisted
of three main components: support rails, a base plate, and a lid. Sandwiched
between the lid and the base plate was insulation, the wick, and two sheets of
borosilicate glass spaced 5×10−3 m apart. A schematic of the burner’s overall
design is shown in Fig. (2.4).
For consistency, a new wick holding 3.5 mL of fuel was placed in the burner for
every trial. The material used for both the wick and surrounding insulation was
Type PM Kaowool®; its product data sheet is included in Appendix (B.4). An
annotated visualization of the wick’s preparation process is presented in Fig. (2.5).
17
Testing Apparatus
1.33 E-1 m 7.62 E-2 m
1.27 E-1 m
2.54 E-2 m
6.35 E-2 m
?6.35 E-3 m
1.84 E-2 m 3.81 E-2 m 2.54 E-2 m
3.66 E-2 m
2.54 E-1 m
Figure 2.4: Burner Plan
Once ignited and allowed to burn undisturbed, the burner produced a nearly two
dimensional laminar line-flame through the gap in the glass panes. The height
of flame ranged from 2×10−2 m to 1.00×10−1 m, depending on the fuel type and
elapsed time in the flame’s evolution. A hexane flame at approximately 20 sec
after ignition is shown from the the coronal and sagittal planes in Fig.’s (2.6a) and
(2.6b) respectively.
18
Testing Apparatus
2.4 Testing Enclosure
The testing enclosure created a space where an open flame could burn safely
while simultaneously reducing the effects of errant air flows on the flame. The
enclosure was built on a steel bread board measuring 7.62×10−1 m × 7.62×10−1
m × 6.35×10−2 m. The surface of the board had a grid of screw holes spaced on
2.54×10−2 m squares that could accommodate 1/4− 20 threading. Erected on the
corners of the board were vertical metal supports measuring 6.35×10−1 m high.
The tops of the vertical supports were connected with horizontal supports, creating
a rectangular enclosure measuring 7.62×10−1 m × 7.62×10−1 m × 6.35×10−1 m.
A fine steel mesh screen with 1.00×10−6 m2 holes was then placed over the faces of
the enclosure, with only an opening for the tube left in the coverage. A schematic
of the enclosure is shown in Fig. (2.7), and pictures taken of the enclosure during
routine cleaning are shown in Fig.’s (2.8a) and (2.8b).
The burner was placed within the enclosure so that the flame’s base would be on
the center-line axis of the tube. As shown in Fig. (2.7), the top of the flame holder
was coplanar with the transverse plane of the tube, and the holder was positioned
so that the flame was 9×10−2 m away from the tube opening. A picture of the
burner in testing position is shown in Fig. (2.9a). The excess space in the testing
enclosure behind the burner was used to house the DAQ, some of which is shown
during calibration in Fig. (2.9b).
19
Testing Apparatus
1.2
7 E-
1 m
1.78 E-1 m
(a) All components from the previous
test were removed, and the burner
was allowed to cool to room
temperature.
1.0
8 E-
1 m
1.78 E-1 m
(b) A piece of backing insulation was
placed on the base plate.
5.0
2 E-
2 m
3.0
0 E-
2 m
7.6
2 E-
2 m
5.0
0 E-
2 m
(c) A second piece of insulation
containing a foil lined center cut-out
was placed on top of the first.
3.0
0 E-
2 m
5.0
0 E-
2 m
(d) The wick was placed in the foil
lined cutout, and 3.5 mL of fuel was
poured along its center-line axis.
5.0
8 E-
2 m
1.52 E-1 m
5.0
0 E-
3 m
(e) Two pieces of 3.18× 10−3 m thick
borosilicate place were placed over
the wick, leaving a 5× 10−3 m gap
for fuel to escape.
7.6
2 E-
2 m
1.2
7 E-
1 m
(f) Finally, the lid was placed over
the glass panels and secured to the
base plate with screws (not shown).
Figure 2.5: Wick preparation and burner assembly sequence
20
Testing Apparatus
(a) View of the flame sheet in the coronal
plane
(b) View of the flame sheet in the sagital
plane
Figure 2.6: Views of the flame sheet in the coronal and sagital planes.
1.91 E-1 m
8.89 E-2 m
7.62 E-2 m
6.99 E-1 m
6.35 E-1 m
7.62 E-1 m
Figure 2.7: Plan of Testing Enclosure
21
Testing Apparatus
(a) View of testing enclosure in the coro-
nal plane
(b) View of testing enclosure in the sagi-
tal plane
Figure 2.8: Views of testing enclosure
(a) Flame holder in position (b) DAQ duing calibration
Figure 2.9: Views of the flame holder and DAQ equipment in the testing
enclosure
22
Chapter 3
Data Acquisition
3.1 Overview
Data acquisition was limited to methods that did not interact directly with the
flame. Of particular interest were the acoustic conditions the flame experienced at
the point of extinction. These conditions included the acoustic pressure and the
speed of air movement around the flame. Other metrics of interest included the
flame’s burning rate and it’s movement during an acoustic cycle. The equipment
used to obtain this data is described in the following sections.
23
Data Acquisition
3.2 Acoustic Pressure
Measurements of acoustic pressure were made using a 1.27×10−2 m diameter
integrated constant current power (ICCP) microphone, manufactured by the BSWA
Technology Co., Model # MPA 231. The gauge was connected by BNC cable to a
signal conditioner, which was then connected to a Tektronix Digitial Oscilloscope,
Model # TDS 2004B. Using the oscilloscope, an rms voltage from the pressure
gauge was obtained, which was then converted into an rms pressure reading. The
conversion from voltage to pressure is described in Sec. (3.2.1). Product data
sheets for the pressure gauge, signal conditioner, and oscilloscope are included in
Appendix (B.5), (B.6), and (B.8) respectively.
3.2.1 Microphone Calibration
Accompanying the pressure gauge was the BSWA Calibrator, Model # CA111, and
its data sheet is included in Appendix (B.7). A calibration of the pressure gauge
was preformed prior to every testing session, and each testing session generally
lasted 4 to 6 hours. The calibration unit produced a 1000 Hz tone at 94 dB and
114 dB. Decibel reading are calculated from acoustic pressure as:
dB = 20 log
(
Prms
Pref
)
(3.1)
24
Data Acquisition
where Prms is the rms pressure of a complete acoustic cycle, and the reference
pressure (Pref ) is commonly taken as 20µPa [37]. Rearranging Eq. (3.1) yields:
Prms = Pref · 10 dB/20 (3.2)
Using Eq. (3.2), it was found that 94 dB corresponds to 1 Parms, and 114 dB
corresponded to 10 Parms. To convert from a voltage reading to a pressure reading,
a calibration constant (k) was desired such that:
Prms = kVrms (3.3)
Using the form presented in Eq. (3.3), a specific value of k for each calibration
pressure was calculated as:
1 Parms = k1V1,rms (3.4)
10 Parms = k2V2,rms (3.5)
An average calibration constant was then calculated as:
k =
1
2
(
1 Parms
V1,rms
+
10 Parms
V2,rms
)
(3.6)
25
Data Acquisition
Using the results for k from Eq. (3.6), an rms pressure was calculated from the
oscilloscope voltage using Eq. (3.3). The average value (µ) and standard deviation
(σ) of all calibration constants found during testing were:
µk = 0.0218
Pa
mV
, σk = 0.0004
Pa
mV
3.3 Anemometry
Anemometry readings were made using an Extec Hot Wire Anemometer, Model
# 407123. The measurements were made in units of ft/min, which provided the
highest resolution, and then converted to m/s. The product data sheet for the
anemometer is included in Appendix (B.9).
3.4 High Speed Video
A Phantom high speed camera, Model # V461, was used to capture high speed
videography. Videos were shot at either 400 or 1000 frames per second (fps) and
at a resolution of 640× 480. The product data sheet for the camera is included in
Appendix (B.10).
26
Data Acquisition
3.5 Mass Readings
Time-resolved mass readings were made using a Mettler Toledo Precision Balance,
Model # MS4002S. A RS232 - USB cable was used to connect the balance to a
computer, and a MATLAB script was written which read the data and stored it
in a text file. In order to use the balance, the burner’s support rods, shown in
Fig. (2.7), were removed. A separate stand was fabricated which provided the
burner with stability while on the balance; a picture of the stand is shown in
Fig. (3.1). When the balance was in use, it was situated in the enclosure so that
the burner still occupied the same position described in Sec. (2.4). The product
data sheet for the balance is included in Appendix (B.11).
Figure 3.1: Burner Balance Stand
27
Chapter 4
System Characterization
4.1 Overview
A detailed study of the testing apparatus was conducted prior to experimentation
with a flame. The specific objectives of the study were to understand the system’s
harmonics, the acoustic pressures generated, and nature of the acoustically induced
air movements.
4.2 System Harmonics
The first study conducted was meant to determine if the system showed resonance
at frequencies consistent with theory. The study began by calculating the resonant
frequencies for the 1st, 2nd, and 3rd harmonic modes of the tube. While the
28
System Characterization
tube’s length was 3.05 m, it was treated as being 2.44 m since the speaker was
mounted 0.61 m from the entrance. Pressure measurements were then taken at
regular intervals down the length of the tube at the calculated resonant frequencies,
and at frequencies within ±10 Hz of where resonance was expected. Using the
profiles generated, the approximate true frequencies of resonance were identified and
compared with those of theory. The calculated frequencies, measured frequencies,
and comparison of the results are presented in Sec.’s (4.2.1), (4.2.1), and (4.2.3)
respectively.
4.2.1 Predicted Resonant Frequencies
The speed of sound (cs) is related to the wavelength (λ) and frequency (ω) by the
relation:
cs = λω (4.1)
For a closed-open tube system, the length of resonance (LRn) for the n
th harmonic
is related to λ by the correlation:
LRn =
(2n− 1)λ
4
(4.2)
where LRn is equal to the the tube’s actual length (Lo) plus a correction factor
(e) [37]. The frequency of resonance can then be related to the tube’s length by:
29
System Characterization
Lo + e =
(2n− 1)λ
4
(4.3)
From experimentation, it is known that e = 0.3D, where D is the tube diameter
[37]. Substituting Eq. (4.1) into Eq. (4.3) and rearranging, the frequencies of
resonance for the nth harmonic can be calculated as:
ωRn =
(2n− 1)cs
4(Lo + 0.3D)
(4.4)
Since the speaker was mounted 0.61 m inside the 3.05 m long tube, the system
was treated as being 2.44 m long. Using cs = 343 m/s, Lo = 2.44 m and D = 0.254
m in Eq. (4.4), the frequencies for the 1st, 2nd, and 3rd harmonics were calculated
and are presented in Tab. (4.1).
Table 4.1: Calculated resonant frequencies of
the tube for the 1st, 2nd, and 3rd harmonics.
Mode n Calculated ωRn (Hz)
1 34.1
2 102.3
3 170.5
4.2.2 Observed Resonant Frequencies
To determine the approximate frequencies where resonance truly occured, a series
of frequencies was tested for each mode within a window centered on ωRn . The size
of the window was ± 10 Hz and divided into 2.5 Hz increments. For each frequency
30
System Characterization
tested, center-line pressure measurements were made at regular intervals down
the length of the tube. Profiles were then generated from these measurements by
using a spline fit in Matlab. For each frequency tested, measurements were made
at speaker powers of 25 W and 100 W. Representative profiles from the 1st, 2nd,
and 3rd harmonics are shown in Fig.’s (4.1), (4.2), and (4.3) respectively.
Distance From Speaker Face (m)
0 0.5 1 1.5 2
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
20
40
60
80
100
120
140
160
34.1 Hz
42.5 Hz
47.5 Hz
(a) 1st harmonic, 25 W.
Distance From Speaker Face (m)
0 0.5 1 1.5 2
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
20
40
60
80
100
120
140
160
34.1 Hz
42.5 Hz
47.5 Hz
(b) 1st harmonic, 100 W.
Figure 4.1: Representative pressure profiles from the 1st harmonic on the
interior of the tube. The profiles in Fig. (4.1a) were produced at a speaker
power of 25 W, and the profiles in Fig. (4.1b) at 100 W.
Distance From Speaker Face (m)
0 0.5 1 1.5 2
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
20
40
60
80
100
120
140
160
102.3 Hz
110 Hz
115 Hz
(a) 2nd harmonic, 25 W.
Distance From Speaker Face (m)
0 0.5 1 1.5 2
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
20
40
60
80
100
120
140
160 102.3 Hz
110 Hz
115 Hz
(b) 2nd harmonic, 100 W.
Figure 4.2: Representative pressure profiles from the 2nd harmonic on the
interior of the tube. The profiles in Fig. (4.2a) were produced at a speaker
power of 25 W, and the profiles in Fig. (4.2b) at 100 W.
31
System Characterization
Distance From Speaker Face (m)
0 0.5 1 1.5 2
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
20
40
60
80
100
120
140
160
170.5 Hz
177.5 Hz
180 Hz
(a) 3rd harmonic, 25 W.
Distance From Speaker Face (m)
0 0.5 1 1.5 2
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
20
40
60
80
100
120
140
160
170.5 Hz
177.5 Hz
180 Hz
(b) 3rd harmonic, 100 W.
Figure 4.3: Representative pressure profiles from the 3rd harmonic on the
interior of the tube. The profiles in Fig. (4.3a) were produced at a speaker
power of 25 W, and the profiles in Fig. (4.3b) at 100 W.
The approximate frequencies of true resonance were identified by visually inspecting
the family of profiles for each harmonic, and then identifying the profile that had
peak amplitudes. Results for the 2nd and 3rd harmonic were confirmed by noting
that peak amplitudes occured at the same frequency for both 25 W and 100
W. Intense vibrations of the testing apparatus occured though during the 1st
harmonic tests at 100 W. As seen in Fig. (4.1b), these vibrations made accurate
pressure measurements difficult and obscured the results. The approximate resonant
frequency for the first harmonic was, therefore, estimated from the 25 W profile
alone. The predicted and observed resonant frequencies for the first three modes
are presented in Tab. (4.2).
Table 4.2: Predicted and observed resonant frequencies of the
tube for the 1st, 2nd, and 3rd harmonics.
Mode n Predicted ωRn (Hz) Observed ωRn (Hz)
1 34.1 42.5
2 102.3 97.5
3 170.5 177.5
32
System Characterization
4.2.3 Comparison of Predicted and Observed Resonance
Rearranging Eq. (4.4), the length of a tube that resonates on the nth harmonic for
a given frequency can be calculated as:
Lo =
(2n− 1)cs
4ω
− 0.3D (4.5)
Using the values from Tab. (4.2) in Eq. (4.5), the observed frequencies of resonance
for the 1st, 2nd, and 3rd harmonics were found to correspond with tube lengths
of 1.94 m, 2.56 m, and 2.33 m respectively. On average then, the tube behaved
acoustically as having a length of 2.28 m. When compared to the actual effective
length of 2.44 m, it can be seen that the two are in relatively close agreement. It
was, therefore, concluded that the tube was behaving acoustically in a manner
consistent with theory.
4.3 Pressure Scaling
The next study of the system was designed to characterize the acoustic pressures
outside the tube in the region that the flame would inhabit. For every frequency
tested during the procedure described in Sec. (4.2), pressure measurements were
also taken at varying distances from the tube’s end along the projected center-line
axis of the tube. Representative profiles from the 1st, 2nd, and 3rd harmonics are
shown in Fig.’s (4.4), (4.5) and (4.6) respectively.
33
System Characterization
Distance From Tube Opening (m)
0 0.05 0.1 0.15
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
5
10
15
20
25
34.1 Hz
42.5 Hz
47.5 Hz
(a) 1st harmonic, 25 W.
Distance From Tube Opening (m)
0 0.05 0.1 0.15
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
5
10
15
20
25
34.1 Hz
42.5 Hz
47.5 Hz
(b) 1st harmonic, 100 W.
Figure 4.4: Representative pressure profiles from the 1st harmonic on the
exterior of the tube. The profiles in Fig. (4.4a) were produced at a speaker
power of 25 W, and the profiles in Fig. (4.4b) at 100 W.
Distance From Tube Opening (m)
0 0.05 0.1 0.15
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
5
10
15
20
25
102.3 Hz
110 Hz
115 Hz
(a) 2nd harmonic, 25 W.
Distance From Tube Opening (m)
0 0.05 0.1 0.15
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
5
10
15
20
25
102.3 Hz
110 Hz
115 Hz
(b) 2nd harmonic, 100 W.
Figure 4.5: Representative pressure profiles from the 2nd harmonic on the
exterior of the tube. The profiles in Fig. (4.2a) were produced at a speaker
power of 25 W, and the profiles in Fig. (4.2b) at 100 W.
The pressure measurements from each profile were then scaled by the pressure
at the tube opening. It was found that for each frequency tested, the scaled
pressures at 25 W and 100 W showed pointwise convergence. In other words,
for each frequency there was a proportional decay in the acoustic pressure as
the distance increased from the tube opening. When examined in aggregate, the
proportional decay profiles for all frequencies tested showed an overall linear trend.
34
System Characterization
Distance From Tube Opening (m)
0 0.05 0.1 0.15
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
5
10
15
20
25
170.5 Hz
177.5 Hz
180 Hz
(a) 3rd harmonic, 25 W.
Distance From Tube Opening (m)
0 0.05 0.1 0.15
A
co
us
tic
P
re
ss
ur
e
(P
a)
0
5
10
15
20
25
170.5 Hz
177.5 Hz
180 Hz
(b) 3rd harmonic, 100 W.
Figure 4.6: Representative pressure profiles from the 3rd harmonic on the
exterior of the tube. The profiles in Fig. (4.6a) were produced at a speaker
power of 25 W, and the profiles in Fig. (4.6b) at 100 W.
Representative samples of the scaled pressure profiles, along with the aggregate
line of fit, and coefficient of determination are shown in Fig. (4.7). Using these
results, the center-line acoustic pressure at distance x from the tube opening could
be estimated from the acoustic pressure at the tube opening (Po) by the relation:
P (x) = Po (−3.96x+ 0.92) (4.6)
where the units of x are m, and the units of it’s coefficient are m−1.
35
System Characterization
Distance From Tube Opening (m)
0 0.05 0.1 0.15
Sc
al
ed
 P
re
ss
ur
e
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
30.0 Hz, 25 W
34.1 Hz, 25 W
42.5 Hz, 25 W
97.5 Hz, 25 W
102.3 Hz, 25 W
110 Hz, 25 W
165 Hz, 25 W
170.5 Hz, 25 W
177.5 Hz, 25 W
y=-3.96x+0.92
R2=0.85
100 W
100 W
100 W
100 W
100 W
100 W
100 W
100 W
100 W
Figure 4.7: Representative samples of scaled exterior acoustic
pressures and the aggregate line of fit for all scaled pressures. Each
exterior pressure profile measured was scaled by the pressure at the
tube opening (Po) for that particular frequency and speaker power.
All scaled profiles showed pointwise convergence for each frequency
tested. When examined in aggregate, the decay profiles showed an
overall linear trend, and the line of fit was calculated from the
aggregate data set.
36
System Characterization
4.4 Acoustic Pressure Profiles
The next study was designed to determine how the burner’s presence would affect
the sound waves emanating from the tube. To investigate this, a series of acoustic
pressure profiles were made with and without the burner in place. The results were
then compared to quantify the aggregate effect of the burner’s presence.
4.4.1 Profiles Without Burner
The profiles without the burner were made by measuring the acoustic pressure in
the plane of the tube opening at varying distance from the opening. The acoustics
used for testing were generated at 40 Hz and 25 W of power to the speaker. The
pressure measurements were taken on a 0.05 m grid and at distances from 0 m
to 0.15 m in 0.05 m intervals. The pressures in all profiles were then normalized
by the center-line pressure at the tube opening. Finally, Matlab was used to
interpolate the pressures at interstitial points, creating normalized pressure profiles.
The individual profiles are shown in Fig. (4.8), and a schematic of the profiles in
relation to the tube is shown in Fig. (4.9).
4.4.2 Profiles with Burner
The pressure profiles with the burner in place were generated using the same basic
methodology described in Sec. (4.4.1). The presences of the burner, however, made
37
System Characterization
z (m)
-0.05
-0.15
0.05
0.15
-0.1
0.1
0
-0.15
0
-0.1
y (m)
0.2
-0.05
0
0.4
0.05
P/P
o
0.1
0.6
0.15
0.8
1
Sc
al
ed
 A
co
us
tic
 P
re
ss
ur
e
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
(a) x = 0 m
z (m)
-0.05
-0.15
0.15
0.05
-0.1
0.1
0
-0.15
0
-0.1
y (m)
0.2
-0.05
0
0.4
0.05
P/P
o
0.1
0.6
0.15
0.8
1
Sc
al
ed
 A
co
us
tic
 P
re
ss
ur
e
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
(b) x = 0.05 m
z (m)
-0.15
-0.05
0.05
0.15
-0.1
0.1
0
-0.15
0
-0.1
y (m)
0.2
-0.05
0
0.4
0.05
P/P
o
0.1
0.6
0.15
0.8
1
Sc
al
ed
 A
co
us
tic
 P
re
ss
ur
e
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
(c) x = 0.10 m
z (m)
-0.05
-0.15
0.15
0.05
-0.1
0.1
0
-0.15
0
-0.1
y (m)
0.2
-0.05
0
0.4
0.05
P/P
o
0.1
0.6
0.15
0.8
1
Sc
al
ed
 A
co
us
tic
 P
re
ss
ur
e
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
(d) x = 0.15 m
Figure 4.8: Normalized pressure profiles at varying distances from the tube
opening without the burner in place. Pressure measurements were taken at
points on a 0.05 m grid, which is indicated by the circles. The pressures were
then normalized by the center-line pressure at x = 0, and a surface was
interpolated from the data points. For convenience, the origin of the coordinate
system used was placed at the center of the tube opening.
it impossible to obtain measurements at all the points previously evaluated. In
addition, the only region of pertinent interest was where the flame would reside.
Therefore, profiles were only generated at x = 0.05 m and x = 0.10 m, and in the
region above the burner. For reference, a third profile was also generated at the
38
System Characterization
x=0 m
x=0.05 m
x=0.10 m
x=0.15 m
Figure 4.9: Schematic of the normalized pressure profiles presented in
Fig. (4.8) in relation to the tube.
actual flame position of x = 0.09 m. The profiles with the burner at x = 0.05 m
and x = 0.10 m are shown in Fig.’s (4.10a) and (4.10b) respectively; the pressure
profile at the flame position in shown in Fig. (4.10c).
4.4.3 Comparison of Profiles
A visual comparison of the pressure profile segments presented in Fig.’s (4.10a)
and (4.10b) to the full profiles presented in Fig.’s (4.8b) and (4.8c) indicated that
the presence of the burner had a small, but noticeable effect. In order to quantify
this effect, a direct comparison between comparable data points was made. For
each distance from the tube opening, an average ratio of the scaled pressures with
the burner (P̂H) to without the burner (P̂O) was calculated as:
µP̂H/P̂O =
1
ninj
nj∑
j=1
ni∑
i=1
PHij
POij
(4.7)
39
System Characterization
z (m)
-0.15
-0.05
0.15
0.05
-0.1
0.1
0
-0.15
0
-0.1
y (m)
0.2
-0.05
0
0.4
0.05
P/P
o
0.1
0.6
0.15
0.8
1
Sc
al
ed
 A
co
us
tic
 P
re
ss
ur
e
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
(a) x = 0.05 m
z (m)
-0.15
-0.05
0.05
0.15
-0.1
0.1
0
-0.15
0
-0.1
y (m)
0.2
-0.05
0
0.4
0.05
P/P
o
0.1
0.6
0.15
0.8
1
Sc
al
ed
 A
co
us
tic
 P
re
ss
ur
e
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
(b) x = 0.10 m
z (m)
-0.05
-0.15
0.05
0.15
-0.1
0.1
0
-0.15
0
-0.1
y (m)
0.2
-0.05
0
0.4
0.05
P/P
o
0.1
0.6
0.15
0.8
1
Sc
al
ed
 A
co
us
tic
 P
re
ss
ur
e
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
(c) Flame Position x = 0.09 m
Figure 4.10: Normalized pressure profiles at varying distances from the tube
opening with the burner in place. Pressure measurements were taken at points
on a 0.05 m grid, which is indicated by the circles. The pressures were then
normalized by the center-line pressure in Fig. (4.8a), and a surface was
interpolated from the data points. For convenience, the origin of the coordinate
system used was placed at the center of the tube opening.
From these averages an aggregate ratio was calculated, the results of which are
presented in Tab. (4.3). It was found that the acoustic pressures with the burner
in place were 1.22× greater than those without the burner in place. This increase
was most likely due to the acoustic waves reflecting off the burner’s surface.
40
System Characterization
Table 4.3: Ratio of normalized acoustic pressures measured
with the burner in place to corresponding pressures without
the burner.
Stat x = 0.05 m x = 0.10 m Aggregate
µP̂H/P̂O 1.20 1.25 1.22
σP̂H/P̂O 0.07 0.06 0.07
σ/µ 0.06 0.05 0.06
4.5 Acoustically Induced Flows
The initiation of acoustics induced air flows within the testing enclosure. When fog
was introduced into the enclosure, it also showed that there was a net flow to this air
movement moving away from the tube opening. To quantify the magnitude of the
air movement, measurements of flows were made using the the hot-wire anemometer.
As seen in Fig. (4.11), the magnitude of these flows showed sensitivity to both
acoustic frequency and acoustic pressure. Since these flows were being induced
by an oscillatory mechanism, it was not entirely clear whether the anemometer
was measuring the speed of the net flow moving away from the tube opening, or
an rms air speed of the oscillatory air movement. To elucidate the meaning of
these measurements, a Particle Image Velocimetry (PIV) study of the flow was
preformed.
4.5.1 PIV Analysis of Acoustic Flow
Visualization of the flow was achieved through the use of a fog machine, which
produced a cloud of atomized ethylene glycol droplets. The nozzle of the fog
41
System Characterization
Acoustic Pressure (Pa)
0 5 10 15 20 25 30 35 40 45
Ai
r S
pe
ed
 (m
·s-
1 )
0
0.2
0.4
0.6
0.8
1
1.2
30 Hz
35 Hz
40 Hz
45 Hz
50 Hz
Figure 4.11: Anemometer measurements of the acoustically induced
air flows, with burner in place, at differing acoustic pressures and
frequencies. Measurements were taken at x = 0.09 m from the tube
opening, at approximately 0.02 m above the burner’s surface.
machine was placed in the open end of the tube and fog was produced until it filled
the tube. The open end of the tube was then sealed with a cylindrical foam plug
for several minutes, giving time for the fog inside the tube to stagnate and excess
fog to dissipate. Just prior to testing, the plug was removed and the speaker was
activated. The acoustic excitations caused the fog in the tube to migrate into the
enclosure, where videography of the process was obtained.
Illumination for the videography was created through the use of a back-lighting
technique, which enhanced the visibility of the fog and yielded a sharper image.
The back-lighting was created by mounting a T-5 fluorescent tube lamp, encased
42
System Characterization
in a plastic diffuser, to the outside of the testing enclosure. The lamp was situated
so that the camera’s view of it was obscured by the burner. This had the effect of
creating the desired illumination without overexposing the image.
The video was recorded using the Phantom high speed camera which was situated
2.9 meters away from the burner. The camera used a 50 mm dual aperture lens,
with the outer aperture open all the way and the inner aperture set to F-2. The focal
point of the image was the burner’s center, and the depth of field was approximately
±0.05 m. The focal region of the image was, therefore, a 0.10 m wide region above
the burner, centered on the longitudinal axis of the burner. The video itself was
recorded at 400 frames per second, with auto exposure at a resolution of 640 ×
480 pixels.
Testing was done at 40 Hz and at acoustic pressures of 15 Pa, 37 Pa, and 54 Pa,
as measured at the tube opening. Each test was allowed to evolve until there was
an accumulation of fog around the enclosure that effectively obscured it from the
camera; this process generally took about two minutes. Measurements of air speed
and acoustic pressure were then taken at regular intervals from 0 to 1.5×10−2 m
from the tube opening and at a height of 2×10−2 m above the plane of the burner.
The PIV analysis was preformed on representative clips from the videos recorded
using the PIVLab application in Matlab. The clips consisted of 32 sequential
frames, which were loaded into PIVLab using A-B B-C sequencing. The Region
of Interest (ROI) measured 256 × 112 pixels, and is indicated by the rectangular
43
System Characterization
boundaries seen in the frames of Fig. (4.12). The distance scale for the frame was
calibrated by the width of the burner, which measured 0.13 m. The time between
each frame was 2.5 ms, which was determined by the recording rate of 400 fps.
For each frame, PIVLab calculated a u and v velocity component on a 90 point
mesh within the ROI. A representative sample of the PIV results from the 15 Pa
trial is shown in Fig. (4.12). Graphs of the calculated u and v components from
the 15 Pa trail, at various distances from the tube opening and 2×10−2 m above
the holder are shown in Fig.’s (4.13a) and (4.13b) respectively. Values of urms and
vrms were then calculated for each spatial point over the temporal domain. Finally,
the magnitude of the composite rms flow velocity was calculated as:
Urms =
√
u2rms + v
2
rms (4.8)
Specific values of Urms were obtained at the points in the PIV mesh that most
closely corresponded to the physical locations of the anemometer measurements.
When the PIV determined values of Urms are plotted along side the anemometer
measured air speeds, as shown in Fig. (4.14a), it was noted that two were in
generally good agreement. When the values of Urms were plotted against the
corresponding anemometer measured air speeds, as shown in Fig. (4.14b), it was
found that the line of best fit was linear, had a slope of approximately equal to
one, a y-intercept approximately equal to Zero, and a coefficient of determination
approximately equal to one. From this, it was concluded that the air speeds
44
System Characterization
(a) +0 ms (b) +5 ms (c) +10 ms
(d) +15 ms (e) +20 ms (f) +25 ms
Figure 4.12: Sample PIV analysis at 40 Hz and 15 Pa acoustic pressure, as
measured at the tube opening. The edge of the tube can be seen on the right
edge of the frame. Each vector shown in the ROI is the composite of a
calculated u and v component.
Time (ms)
0 10 20 30 40 50 60 70 80
Ve
lo
ci
ty
 (m
/s)
-1.5
-1
-0.5
0
0.5
1
1.5
0 m
0.08 m (3 in)
0.15 m (6 in)
(a) u component
Time (ms)
0 10 20 30 40 50 60 70 80
Ve
lo
ci
ty
 (m
/s)
-1.5
-1
-0.5
0
0.5
1
1.5
0 m
0.08 m (3 in)
0.15 m (6 in)
(b) v component
Figure 4.13: PIV calculated velocity components at 40 Hz and 17 Pa, as
measured at the tube opening. The u components are shown in Fig. (4.13a),
and the v components are shown in Fig. (4.13b). The values were taken from
the PIV mesh points that most closely corresponded to the indicated distance,
as measured from the opening of the tube, and at 2×10−2 m above the burner’s
surface.
45
System Characterization
measured by the anemometer were, in fact, the rms speeds of the oscillatory air
movement.
Distance From Tube Opening (m)
0.02 0.04 0.06 0.08 0.1 0.12 0.14
Ac
ou
st
ic
 B
ul
k 
Ai
r S
pe
ed
 (m
·s-
1 )
0
0.2
0.4
0.6
0.8
1
1.2
1.4
15 Pa, Anm
15 Pa, PIV
37 Pa, Anm
37 Pa, PIV
54 Pa, Anm
54 Pa, PIV
(a) Comparison of Urms and anemome-
ter measurements
Anemometer Measured Air Speed (m·s-1)
0 0.2 0.4 0.6 0.8 1 1.2 1.4
PI
V 
R
M
S 
Ai
r S
pe
ed
 (m
·s-
1 )
0
0.2
0.4
0.6
0.8
1
1.2
1.4
15 Pa
37 Pa
54 Pa
y = 1.07x+0.005
R2 = 0.97
(b) Plot of Urms vs. anemometer mea-
surements
Figure 4.14: Comparison of PIV determined values of Urms and the
corresponding anemometer measurements. Fig. (4.14a) shows the two
side-by-side, at various distances and acoustic pressures. Fig. (4.14b) shows the
values of Urms plotted against the corresponding anemometer measurements,
and the line of best fit.
46
Chapter 5
Free Burn Characterization
5.1 Overview
A study of the line-flame without acoustic perturbations was conducted to serve
as a point of reference for future tests. For each fuel used there were three
burns conducted, during which time videography and mass readings were obtained
simultaneously. From this data, values of mass loss rate (mlr or m˙), flame heights
(Lf), and flame width (Wf) were calculated. Since all three quantities were
parameterized by time, a direct comparison between the data points was made.
47
Free Burn Characterization
5.2 Mass Loss Rate
Values of m˙ were calculated from discrete mass readings taken at regular intervals
during a burn. To obtain this data, the burner was placed on the balance without
fuel and prepared for a test. The balance was then “Zeroed” and 3.5 mL of fuel
was added. Data acquisition was then initiated, the fuel was ignited, and the flame
was allowed to evolve without interference. During the evolution, mass readings
were captured digitally once every second. The procedure was repeated three times
for each fuel tested.
Using the data obtained, discrete values of m˙ for each burn were calculated as:
m˙i =
mi −mi+1
ti+1 − ti (5.1)
where the ordering of the numerator and denominator were reversed to achieve a
positive mlr. Values of m˙ were smoothed using a running average with a kernel
of ± 5 s, which accounted for no more than 5% of the data points in any given
set. The data points were averaged together and smoothed again using the same
methodology. The entire process was repeated for each fuel species, and the results
are shown in Fig. (5.1).
48
Free Burn Characterization
Time (s)
0 50 100 150 200 250 300 350 400 450
M
as
s
Lo
ss
R
at
e
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Pentane
Hexane
Heptane
Octane
Figure 5.1: Free burn mlr profiles for each fuel tested. Each profile is the
composite average of three individual profiles, and has been smoothed using
a running average with a kernel of ± 5 s.
49
Free Burn Characterization
5.3 Flame Height
Data for the flame height calculations was obtained simultaneously with the data
for m˙. For each burn, videography was obtained using a JVC Handycam situated
2.44 m (8 ft) from the flame. The metering for the videography was determined
with a pre-test burn using pentane. During the most luminous portion of the
flame’s evolution, the camera’s automatic metering set the aperture to F-5.6 with
an exposure time of 1/400 s. The metering mode for the camera was then set
to manual and these values were used for all tests. This was done so that pixel
intensities were all scaled the same, allowing for a direct comparison of pixels
between frames. The videos were then edited using Windows Live Movie Maker,
where the image was converted into a gray-scale, and extraneous footage from
before and after the burn was removed. A sample sequence of still images from a
hexane burn is shown in Fig. (5.2).
As seen in Fig. (5.3), the flame’s shape showed considerable variation over short time
intervals, and a flame height could not be reliably estimated by visual inspection
alone. To overcome this, the flame’s height had to be estimated for each frame and
smoothed over the temporal domain to find an average flame height profile.
To achieve this, the video was loaded into Matlab, and the indices of the pixels
that formed the ROI shown in Fig. (5.3) were identified. The pixel intensities
of every frame were evaluated, and those contained within the ROI were stored
using the data structure shown in Tab. (5.1). A matrix, defined by Eq. (5.2), was
50
Free Burn Characterization
(a) +30 s (b) +60 s (c) +90 s (d) +120 s
(e) +150 s (f) +180 s (g) +210 s (h) +240 s
Figure 5.2: Sample sequence of still images from footage of a hexane free burn. The
camera was situated 2.44 m away, and the video was shot at 30 fps with an aperture of
F-5.6 and an exposure time of 1/400 s.
(a) +60 s (b) +60.5 s (c) +61 s (d) +61.5 s
Figure 5.3: Sample sequence of still images from footage of a hexane free burn. The
flame shape showed considerable variability over short time intervals. The ROI is
indicated by the red rectangles.
then created whose columns were the average pixel intensities across the ROI in
the horizontal direction for each frame. The width of the ROI had a noticeable
effect on the values of these averages. If the width was increased, the value of the
averages was seen to decrease. This is consistent with the fact that by increasing
the width, more pixels with lower intensities were included. The ROI was therefore
limited to just the region above the wick.
51
Free Burn Characterization
Table 5.1: Data structure for flame height calculations. The
pixel intensities within the ROI are indexed spatially and
temporally.
Frame 1 Frame 2 · · · Frame k
I111 I121 · · · I1j1 I112 I122 · · · I1j2 I11k I12k · · · I1jk
I211 I221 · · · I2j1 I212 I222 · · · I2j2 I21k I22k · · · I2jk
...
...
. . .
...
...
...
. . .
...
...
...
. . .
...
Ii11 Ii21 · · · Iij1 Ii12 Ii22 · · · Iij2 Ii1k Ii2k · · · Iijk
I =
1
Nj
Nj∑
j=1
Iijk =

I11 I12 · · · I1k
I21 I12 · · · I2k
...
...
. . .
...
I i1 I i2 · · · I ik

(5.2)
I
∗
=
1
2∆F
k+∆F∑
k−∆F
I ik (5.3)
The values of I were then smoothed over k using a running average with a kernel
of ± 150 frames, which corresponded with ± 5 s of video footage. This created
a second matrix, I
∗
, defined in Eq. (5.3). Each column of I
∗
is the temporally-
smoothed luminous intensity profile of the flame at the time corresponding to the
frame. A representative sample of smoothed profiles from a hexane burn is shown
in Fig. (5.4).
To determine the flame height (Lf ), each column of I
∗
was first normalized by its
maximum value to create a matrix I˜∗, samples of which are shown in Fig. (5.5a).
This was done since the maximum luminosity of the flame changed over time. A
52
Free Burn Characterization
Pixel Intensity
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
H
ei
gh
t (m
)
0
0.01
0.02
0.03
0.04
0.05
0.06
49 s
97 s
146 s
195 s
Figure 5.4: Hexane flame luminous intensity profiles. Each profile
represents the luminous intensity of the flame over the ROI at the time
indicated. The intensities at each height were smoothed over ± 150
frames.
cutoff intensity (I∗◦ ) was then arbitrarily chosen, and a Matlab script was written
which went down each column of I˜∗ to find the first data point where I˜∗ik ≥ I∗◦ .
From the indices of this point, a time and value of Lf were extracted. The results
were then compared to the the video of the flame to check for consistency. To
determine the best value of I∗◦ , this process was repeated multiple times using
varying values of I∗◦ and sample video clips from each fuel. It was found that
I∗◦ = 0.70 yielded predictions of Lf which were most consistent with the visual
observations. Finally, using I∗◦ = 0.70, the process was applied to each of the three
videos for each fuel. As shown in Fig. (5.5b), the results were then averaged to
53
Free Burn Characterization
produce a composite flame height curves for each fuel. The composite flame height
curves for all fuels tested is shown in Fig. (5.6).
Normalized Pixel Intensity
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
H
ei
gh
t (m
)
0
0.01
0.02
0.03
0.04
0.05
0.06
49 s
97 s
146 s
195 s
(a) Normalized hexane flame intensity
profiles.
Time (s)
0 50 100 150 200 250
H
ei
gh
t (m
)
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
Burn 1
Burn 2
Burn 3
Avg
(b) Individual and composite hexane
flame height profiles.
Figure 5.5: Flame height determination method. The flame intensity profiles
were normalized by the maximum intensity in each frame, as shown in Fig.
(5.5a). A cutoff intensity of 0.7 was then used to determine Lf at the time
associated with the frame. The individual and composite average results are
shown in Fig. (5.5b).
54
Free Burn Characterization
Time (s)
0 50 100 150 200 250 300 350 400 450
Fl
am
e
H
ei
gh
t
(m
)
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
Pentane
Hexane
Heptane
Octane
Figure 5.6: Profiles of Lf for the alkane fuels tested. Each profile is
the composite of three individual profiles which have been averaged
together and smoothed. The initial dip in the octane profile is due to
the slow growth of the flame after the removal of the ignition lighter.
55
Free Burn Characterization
5.4 Flame Width
Due to reflections off the surface of the burner and visual obstructions in the
image frame, the process used to determine Lf could not be used to determine
Wf . Estimates of Wf were obtained instead from dimensional measurements of the
video image itself. To do so, a paper ruler was made by photocopying a Staedtler
brand metric ruler. The paper ruler was then taped to a computer screen cued to
play the video of a burn. The ruler was placed so that it was parallel to the burner
and and just below where the flame emanated from the burner slot. The left and
right edges of the flame were noted every five seconds, from which screen widths
were determined. A scaling factor was found by taking the ratio of the burner’s
true length to the measured screen length. Using this ratio, the values of Wf could
be estimated from the image measurements.
Since the sampling rate for Wf was significantly slower than that used for m˙ and
Lf , there was an insufficient amount of data to reliably determine the uncertainty
for each point. Therefore, instead of generating unique profiles for each burn, the
data points from all three burns using a particular fuel were plotted en masse. The
data points were then fitted with straight-line segments, from which a the value of
Wf could be estimated at any given time. The results, including the overall value
of the coefficient of determination, are shown in Fig. (5.7).
56
Free Burn Characterization
Time (s)
0 50 100 150 200 250 300 350 400 450 500
Fl
am
e
W
id
th
(m
)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Observed Width
Piece-Wise Fit (n=5)
R2 = 0:78
(a) Pentane flame width.
Time (s)
0 50 100 150 200 250 300 350 400 450 500
Fl
am
e
W
id
th
(c
m
)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Observed Width
Piece-Wise Fit (n=6)
R2 = 0:72
(b) Hexane flame width.
Time (s)
0 50 100 150 200 250 300 350 400 450 500
Fl
am
e
W
id
th
(c
m
)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Observed Width
Piece-Wise Fit (n=5)
R2 = 0:79
(c) Heptane flame width.
Time (s)
0 50 100 150 200 250 300 350 400 450 500
Fl
am
e
W
id
th
(m
)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Observed Width
Piece-Wise Fit (n=5)
R2 = 0:79
(d) Octane flame width.
Figure 5.7: Flame width determination. Visual observations of the flame’s width
were obtained every five seconds in each video. The results from all three videos
for each fuel were then plotted en masse and fitted with straight line segments.
57
Free Burn Characterization
5.5 Flame Height Scaling
When juxtaposed next to each other, there is an obvious qualitative relationship
between the profiles of m˙ and Lf for each fuel. As seen in Fig. (5.8), each profile
has peaks and points of inflection at approximately the same time. However, since
the burner used in this study was sui generis, existing flame height correlations
were found to be inadequate predictors of the observed flame heights.
Time (s)
0 50 100 150 200 250 300 350 400 450
M
as
s 
Lo
ss
 R
at
e 
(kg
·s-
1 )
#10-4
0
0.5
1
1.5
2
2.5
Pentane
Hexane
Heptane
Octane
(a) MLR Profiles
Time (s)
0 50 100 150 200 250 300 350 400 450
Fl
am
e 
H
ei
gh
t (m
)
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
Pentane
Hexane
Heptane
Octane
(b) Flame Height Profiles
Figure 5.8: Side-by-side comparison of mlr and flame height profiles for fuels
tested. The profiles show peaks and points of inflection at comparable times.
It was still desirable, though, to examine the link between the m˙ and Lf and Wf .
Since these quantities were all parametrized by time, a direct comparison on a
temporal basis could be made. To do so, an mlr per unit-width of flame (m˙′) was
calculated as:
m˙′ = m˙/Wf (5.4)
58
Free Burn Characterization
A heat release rate per unit flame-width (Q˙′) was then calculated as:
Q˙′ = ∆hcm˙′ (5.5)
where the heat of combustion per unit mass (∆hc) for each fuel is discussed in
Sec. (7.3.1). Values of Lf were then plotted against Q˙
′, as shown in Fig. (5.9).
_Q0 (W/m) #104
0 2 4 6 8 10 12
L
f
(m
)
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
Pentane
Hexane
Heptane
Octane
Fit
R2 = 0:80
Figure 5.9: Flame height vs. heat release rate per unit flame width.
Values of Lf below 0.01 m were centered about a constant value.
Values of Lf above 0.01 m showed a linear relationship with Q˙
′.
Close examination of Fig. (5.9) showed that values of Lf < 0.01 m were clustered
about a constant value of Q˙′, and this point appeared to constitute a minimum
heat flux below which a flame could not exist. On average, this value was found to
be 3.73× 104 W/m, with a standard deviation 0.95× 104 W/m. The fact that a
59
Free Burn Characterization
minimum value of Q˙′ exists is consistent with stagnant layer theory, which requires
a minimum heat release rate to sustain a flame [17].
For values of Lf ≥ 0.01 m, there appeared to be a linear relationship between Lf
and Q˙′. The line of fit for this section was calculated to be:
Lf =
(
3.89× 10−7 m2
W
)
Q˙′ − 4.5× 10−3 m (5.6)
This linearity is consistent with the Burke-Schumann analysis for laminar diffusion
flame height [38].
60
Chapter 6
Burning Rate in an Acoustic
Field
6.1 Overview
A study of the flame’s burning rate while experiencing acoustic perturbations was
conducted to further elucidate the effects of acoustics on the flames produced. Only
hexane fuel was used for this study, since the study required a large number of tests
and it was the only fuel for which there was an adequate supply. To conduct the
tests, the vertical rods used to support the burner were removed and the Mettler
Toledo balance was placed in front of the tube opening, as shown in Fig. (6.1). The
burner was situated on the balance so that it still occupied the same position shown
in Fig. (2.7). Samples were placed in the holder, ignited, and allowed to burn
61
Burning Rate
freely for 10 seconds. The speaker was then activated and the flame was allowed to
burn under acoustic excitation until it self-extinguished. After self-extinguishment,
measurements of the acoustic pressure and rms air speed were taken at the flame
position approximately 0.02 m above the flame holder.
Burner
Support
Balance
Figure 6.1: Burner placed on balance for mlr experiments.
The burner was situated so that it occupied the same
position indicated in Fig. (2.7).
Tests were conducted as 30, 35, 40, and 45 Hz. For each frequency used, a
series of tests were done at increasing acoustic pressure. The pressures tested for
each frequency were limited to those below which acoustic extinction occured, a
phenomenon described in Chap. (7). After the acoustic experiments, the speaker
was removed and replaced with a fan. This facilitated tests using forced flows and
served as a basis for comparison. At each acoustic pressure and fan speed tested,
three trials were performed, and a composite profile was generated using the same
methodology described in Sec. (5.2). Since the fuel used for these experiments
62
Burning Rate
came from a different stock than that used for the free-burn characterizations
described in Chap. (5), additional free-burn tests were performed. On average, 3.5
mL of the stock used for these experiments burned for 11 s less than that used in
the previous experiments.
6.2 MLR Profiles
Representative mlr profiles for each frequency tested are presented in Fig. (6.2).
For each profile, the rms acoustic pressure and rms air speed measured at the
flame position are shown in the legend. Each graph also contains the composite
free-burn profile, which is indicated by the black dashed line.
With the exception of the profiles at 45 Hz, the acoustics had an appreciable effect
on the growth phase of the flame. At low pressures and air speeds, the growth
phase of the flame was inhibited by the acoustics. As the pressure and air speed
increased though, the growth phase was enhanced. This can be seen by noting that
as the acoustic pressure and air speed increased, the peak values of m˙ increased
and the time to achieve peak m˙ decreased. After peak m˙ was achieved though, all
profiles tended to converge towards the free-burn profile. This indicated that the
decay phase of the flame was insensitive to the acoustic perturbations.
Representative profiles of m˙ using a fan driven-flow are presented in Fig. (6.3).
The fan had a minimum operating power of 1.5 W, meaning the slowest flows
63
Burning Rate
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
4.63 Pa, 0 m/s
8.11 Pa, 0.19 m/s
13.67 Pa, 0.45 m/s
(a) 30 Hz
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
4.29 Pa, 0.02 m/s
11.24 Pa, 0.29 m/s
14.96 Pa, 0.39 m/s
19.45 Pa, 0.66 m/s
(b) 35 Hz
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
3.36 Pa, 0.02 m/s
11.54 Pa, 0.16 m/s
17.85 Pa, 0.40 m/s
24.31 Pa, 0.55 m/s
(c) 40 Hz
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
2.60 Pa, 0 m/s
10.56 Pa, 0.22 m/s
18.19 Pa, 0.43 m/s
27.31 Pa, 0.66 m/s
(d) 45 Hz
Figure 6.2: Hexane mlr profiles at varying frequencies and acoustic pressures.
Each profile is the composite of three individual tests. The acoustics tended to
inhibit flame growth at low pressures and air speeds, and enhanced flame
growth as the pressure and air speed increased. All profiles though converged
towards the free-burn profile after peak m˙ was achieved.
that could be tested were 0.24 m/s. The fastest flows that could be tested were
limited by the speed at which extinction would occur, which was found to be 0.74
m/s. This was significantly less than the fan-driven extinction flow for a hexane
flame discussed in Chap. (7), which was found to be 1.58 m/s. The discrepancy
is attributed to the presence of the mass balance in the flow path, which was not
present in extinction experiments and which significantly affected air flows.
64
Burning Rate
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
Fan 0.24 m/s
Fan 0.45 m/s
Fan 0.64 m/s
Fan 0.70 m/s
Figure 6.3: Hexane mlr profiles at varying fan driven flows.
Each profile is the composite of three individual tests. As the
fan driven flows increased the growth phase of the flame was
inhibited. All profiles though converged towards the free-burn
profile after peak m˙ was achieved.
Although the trends were not nearly as dramatic in the fan-driven profiles as they
were in the acoustic profiles, certain patterns were discernible. As previously noted,
increasing the acoustic air speed enhanced the growth phase of the flame. In
contrast though, increasing the fan-driven air speed inhibited the flames’ growth.
This can be seen in Fig. (6.3) by noting the rightward shit of the profiles as the flow
speed increases. As with the acoustic profiles though, once peak m˙ was achieved,
the profiles converged towards the free-burn profile.
65
Burning Rate
6.3 MLR Profile Comparisons
To further understand the effects of the acoustics on burning rate, comparisons of
m˙ were made by examining profiles under comparable conditions. First, profiles
at different frequencies but comparable acoustic pressures were examined. Then,
profiles at different frequencies but comparable air speeds were examined. Included
in this examination were profiles made with a fan-driven flow whose magnitude
was comparable. Finally, average mlr’s were calculated and then compared on the
basis of acoustic pressure and rms air speed.
6.3.1 Comparable Acoustic Pressures
Graphs of mlr profiles at comparable acoustic pressures are presented in Fig. (6.4);
the error bars have been omitted for clarity. At constant acoustic pressure, no link
could be found between frequency and peak m˙, or frequency and time to achieve
peak m˙. More obvious, though, were the effects of increasing acoustic pressure.
As the acoustic pressure increased, both the rate of flame growth and magnitude
of peak m˙ increased. Since the magnitude of the rms air speed increased with
acoustic pressure, this observation is consistent with the trends noted in Sec. (6.2).
66
Burning Rate
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
30 Hz, 4.63 Pa, 0 m/s
35 Hz, 4.29 Pa, 0.02 m/s
45 Hz, 4.31 Pa, 0.03 m/s
(a) 4.41 Pa
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
30 Hz, 11.56 Pa, 0.41 m/s
35 Hz, 11.24 Pa, 0.29 m/s
40 Hz, 11.54 Pa, 0.16 m/s
(b) 11.45 Pa
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
35 Hz, 19.45 Pa, 0.66 m/s
40 Hz, 21.11 Pa, 0.47 m/s
45 Hz, 19.84 Pa, 0.49 m/s
(c) 20.14 Pa
Figure 6.4: Mlr profiles at varying frequencies and comparable acoustic
pressures. The pressures noted in the subtitles are the average acoustic
pressures of the profiles presented.
6.3.2 Comparable Air Speeds
Graphs of mlr profiles at comparable rms air speeds are presented in Fig. (6.5).
Included in these graphs are profiles with a fan-driven flow at roughly the same
magnitude. Among the acoustically-perturbed flames, no discernible pattern could
be identified with respect to changes in frequency at a constant air speed. All
acoustically perturbed flames, though, did exhibit faster growth and peak values
of m˙ than the corresponding fan profile. Perhaps most probative is the consistency
67
Burning Rate
of the acoustically-perturbed profiles. With the exception of the 35 Hz profile in
Fig. (6.5b), all acoustic profiles were nearly identical at comparable rms air speeds.
This suggests that for the acoustic profiles, rates of growth were most strongly
influenced by the magnitude of oscillatory air movement. The acoustic pressure
and frequency only influenced mlr insofar as they contributed to the rms air speed.
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
30 Hz, 13.67 Pa, 0.45m/s
40 Hz, 21.12 Pa, 0.47 m/s
45 Hz, 18.19 Pa, 0.43 m/s
Fan 0.45 m/s
(a) 0.45 m/s
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
35 Hz, 16.68 Pa, 0.53 m/s
40 Hz, 24.31 Pa, 0.55 m/s
45 Hz, 23.01 Pa, 0.56 m/s
Fan 0.58 m/s
(b) 0.58 m/s
Time (s)
0 50 100 150 200 250
M
LR
(k
g/
s)
#10-4
0
0.5
1
1.5
2
2.5 Free Burn
35 Hz, 19.45 Pa, 0.66 m/s
45 Hz, 27.31 Pa, 0.66 m/s
Fan 0.64 m/s
(c) 0.64 m/s
Figure 6.5: Profiles of m˙ at varying frequencies and comparable air speeds. The
air speeds noted in the subtitles are the average acoustic air speeds of the
profiles presented. The mlr profile from a fan driven flow at a comparable speed
is also presented.
68
Burning Rate
6.3.3 Average Mass Loss Rates
Values of average mlr (m˜) were calculated by taking the ratio of total fuel lost to
total burning time. These values were compared on the basis of acoustic pressure
(P ) and rms air speed (UA); the results are presented in Fig.’s (6.6a) and (6.6b)
respectively. Included in Fig. (6.6b) are values of m˜ for the flames perturbed by
the fan-driven flows.
Examining Fig’s (6.6a) and (6.6b), it can be seen that there is an obvious positive
correlation between m˜ with both P and UA. Interestingly though, values of m˜ for
the fan-driven experiments were roughly constant, with an average of 0.86× 10−4
kg/s. To determine which parameter m˜ was more closely associated with in the
acoustic experiments, two statistical tests were performed. The first test calculated
the coefficient of correlation (r) between the aggregate data set of m˜ and the argued
parameters of P and UA. The second test was to fit linear curves to the aggregate
data set of m˜ and then compare the coefficients of determination for the same
argued parameters.
For m˜ vs. P , it was found that r = 0.92 and R2 = 0.83. For m˜ vs. UA though, it
was found that r = 0.96 and R2 = 0.91; the line of fit for this data is shown in Fig.
(6.6c). The results of these two tests indicate that m˜ was more strongly influenced
by UA than by P . They also bolster the supposition made in Sec. (6.3.2), that
acoustic pressure only influenced the burning rate insofar as it contributed to the
oscillatory movement of air.
69
Burning Rate
Although not perfectly analogous, the link between the speed of oscillatory air
movement and burning rate is consistent with the observations of other authors.
In studies of flame spread in opposed flows, both Fernandez-Pello and De Ris et
al showed that increased flow speed enhanced burning rate. This was due to the
flames being forced closer to the fuel surface, which enhanced heat transfer into
the fuel bed [39, 40]. It is reasonable to conclude that a similar phenomenon was
occurring due to the oscillatory air movement over the fuel, and this conclusion is
consistent with the visually observed behavior of the flame during testing. Using
the same analogy, the linearity of the trend is also consistent with the work of
Hu et al, who studied the burning rate of various sized gasoline pool fires in cross-
flows. The linearity observed bu Hu arose from the same mechanism described by
Fernandez-Pello and De Ris, where the flame was forced closer to the fuel’s surface
with increasing air speed [41].
70
Burning Rate
Acoustic Pressure(Pa)
0 5 10 15 20 25 30
M
LR
(k
g/
s)
#10-4
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
30 Hz
35 Hz
40 Hz
45 Hz
(a) Average mlr vs. acoustic pressure.
Air Speed (m/s)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
M
LR
(k
g/
s)
#10-4
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
30 Hz
35 Hz
40 Hz
45 Hz
Fan
(b) Average mlr vs. rms air speed.
Air Speed (m/s)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
M
LR
(k
g/
s)
#10-4
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
30 Hz
35 Hz
40 Hz
45 Hz
em = (1:04Urms + 0:82)# 10!4
R2 = 0:91
(c) Average mlr vs. rms air speed with fitted
curve.
Figure 6.6: Profiles for average mlr. Fig. (6.6a) shows m˜ plotted against P ,
while Fig. (6.6b) shows m˜ plotted against the UA. Included in Fig. (6.6b), are
the values of m˜ for the fan driven experiments plotted against the bulk air
speed of the fan flow. Fig. (6.6c) shows the values of m˜ for the acoustic
experiments with the line of fit.
71
Chapter 7
Acoustic Extinction
7.1 Experimental Summary
Flames produced using the fuels n-pentane, n-hexane, n-heptane, n-octane, and
JP-8 aviation fuel were subjected to acoustic perturbations at varying frequencies
and acoustic pressures to determine extinction criteria. For comparison, fan-driven
flows were also created in the testing enclosure and conditions were measured.
7.1.1 Acoustic Extinction Results
The conditions required to cause an acoustic extinction of a particular fuel at
a particular frequency (ω) were determined by finding the lowest speaker power
that could cause three consecutive extinction events within 10 seconds of speaker
72
Acoustic Extinction
activation. For each test, the flame was allowed to burn unperturbed until it
reached a height of approximately 0.02 m. The speaker was then activated to
determine if acoustically driven extinction could be achieved. Immediately after
the flame was extinguished, the acoustic pressure (PA) and rms acoustic air speed
(UA) were measured. At extinction, the reported acoustic pressure (PAext) and rms
air speed of the acoustically induced flow (UAext) were calculated as the average of
the three individual trials. The results, including uncertainties, are summarized in
Tab. (7.1); the uncertainties are discussed in Appendix (A.1).
Table 7.1: Acoustic Flame Extinction Test Results
Fuel ω (Hz) PAext (Pa) UAext ( m · s−1)
Pentane 30 16.2 ±0.1 0.71 ±0.02
35 22.2 ±0.2 0.86 ±0.03
40 35.5 ±0.2 0.95 ±0.03
Hexane 30 14.6 ±0.1 0.65 ±0.02
35 19.5 ±0.1 0.75 ±0.02
40 27.0 ±0.2 0.68 ±0.02
45 28.4 ±0.2 0.86 ±0.03
Heptane 30 13.7 ±0.1 0.58 ±0.02
35 15.9 ±0.1 0.60 ±0.02
40 26.6 ±0.2 0.72 ±0.02
45 25.5 ±0.2 0.72 ±0.02
50 29.9 ±0.2 0.74 ±0.02
Octane 30 14.7 ±0.1 0.60 ±0.02
35 16.6 ±0.1 0.58 ±0.02
40 25.3 ±0.2 0.60 ±0.02
45 22.2 ±0.2 0.64 ±0.02
JP-8 30 16.1 ±0.1 0.57 ±0.02
35 24.3 ±0.1 0.55 ±0.02
40 23.2 ±0.2 0.59 ±0.02
45 20.6 ±0.1 0.66 ±0.03
Graphs of the data from Tab. (7.1) are shown in Fig.’s (7.1a) and (7.1b). The
uncertainties in PAext were not included in Fig. (7.1a) since they were smaller than
73
Acoustic Extinction
the data markers used. For each graph, the values of µ, σ, and the Coefficient of
Variation (CV=σ/µ), are shown in the legend.
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
P A
(P
a)
0
5
10
15
20
25
30
35
40
45
50
Pentane
Hexane
Heptane
Octane
JP-8
   7 =22
   < =5.8
</7 =0.27
(a) Average acoustic pressure (PA)
measured at extinction for
differing fuels at differing
frequencies
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
U
A
(m
/s
)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
Pentane
Hexane
Heptane
Octane
JP-8
   7 =0.68
   < =0.11
</7 =0.16
(b) Average acoustic rms air
speed (UA) measured at extinction
for differing fuels at differing
frequencies
Figure 7.1: Average acoustic extinction conditions for fuels tested. Each data
point is the average of the measured conditions at the lowest speaker power
which could cause three consecutive extinctions. The uncertainties in PA were
omitted from Fig. (7.1a) since they were smaller than the data markers used.
The curve for each fuel in Fig.’s (7.1a) and (7.1b) can be thought of as a boundary
delineating conditions where the flame can and cannot exist. As illustrated with
the heptane data in Fig.’s (7.2a) and (7.2b), at conditions below the boundary
the flame can continue to burn. At conditions above the boundary though, flame
extinction occurs.
74
Acoustic Extinction
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
P A
(P
a)
0
5
10
15
20
25
30
35
40
45
50
Heptane
Extinction
Flame
(a) Heptane extinction boundary
defined by PA
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
U
A
(m
/s
)
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
Heptane
Extinction
Flame
(b) Heptane extinction boundary
defined by UA
Figure 7.2: Heptane extinction boundaries defined by PA and UA. At acoustic
pressures below the curve in Fig. (7.2a) the flame continues to burn; at
pressures above the curve, flame extinction occurs. Likewise, the flame
continues to burn at acoustic flows below the curve in Fig. (7.2b); at flows
above the curve, flame extinction occurs.
7.1.2 Fan-Driven Extinction Results
The conditions required to cause a fan-driven extinction of a particular fuel were
determined by finding the lowest fan power that could cause three consecutive
extinction events within 10 seconds of fan activation. The reported fan-driven
bulk air speeds at extinction (UFext) were calculated as the average of the three
individual trials. The results, including uncertainties, are summarized in Tab. (7.2);
the uncertainties are discussed in Appendix (A.1).
A graph of the data from Tab. (7.2), including the uncertainties, is shown in
Fig. (7.3). The data points have been plotted against the molar mass for each fuel
so that the error bars are visible.
75
Acoustic Extinction
Table 7.2: Fan-Driven Flame Extinction Test Results
Fuel UF ( m · s−1)
Pentane 1.55 ±0.03
Hexane 1.58 ±0.03
Heptane 1.89 ±0.04
Octane 1.96 ±0.04
JP-8 2.03 ±0.08
Molar Mass (kg/mol)
0.06 0.08 0.1 0.12 0.14 0.16 0.18
U
F
(m
/s
)
1.4
1.5
1.6
1.7
1.8
1.9
2
2.1
2.2
Pentane
Hexane
Heptane
Octane
JP-8
   7 =1.8
   < =0.2
</7 =0.11
Figure 7.3: Average fan driven bulk air speeds (UF )
measured for extinction of differing fuels. Each data
point is the average of the measured conditions at
the lowest speaker power which could cause three
consecutive extinctions.
7.2 Proposed Acoustic Extinction Theory
7.2.1 Comparison of Results
When the values of UFext are considered with the those of UAext , two observations
become apparent:
76
Acoustic Extinction
1. The average value of UFext (UFext = 1.8 m/s) is approximately 2.5× the
average value of UAext (UAext = 0.7 m/s).
2. The general “ordering” of the fuels between the two scenarios is reversed. For
the fan-driven extinctions, values of UFext were seen to increase with the fuel’s
molar mass, while for the acoustic extinctions, values of UAext were seen to
decrease. This is clearly shown in Fig. (7.4), which compares values of UFext
for the different fuels to those of UAext at 35 Hz. The data here is plotted
against the heat of combustion per unit mole (∆Hc) for each fuel, which
scales with the fuel’s molar mass [42]. This is done to facilitate a comparison
of extinction strain rates, which is discussed in Sec. (7.2.2).
7.2.2 Discussion of Results
Using a “Damko¨hler” analysis, it has been shown that flame stretch can be a primary
cause of flame extinction [18]. Flame stretch itself is caused by hydrodynamic
strain in the flow of the oxidizer and fuel, and by localized changes in flame speed
due to variations in temperature and species concentrations [43]. Except for cases
where the turbulent length scales of the flow are smaller than the flame thickness,
it is the hydrodynamic effects that have the strongest influence on flame stretch in
laminar flames [5].
Examining the fan-flow visualization in Fig. (7.5), it can be seen that a boundary
layer formed over the burner shortyly after the fan was activated. The fan-driven
77
Acoustic Extinction
"Hc (MJ/mol)
3 3.5 4 4.5 5 5.5 6 6.5 7
U
(m
/s
)
0
0.5
1
1.5
2
UF
UA at 35 Hz
Pentane
Hexane
Heptane
Octane
JP-8
Figure 7.4: Comparison of acoustically induced and fan driven air
flows at extinction. The values of UAext at 35 Hz and UFext are plotted
against ∆Hc for all fuels tested. Values of UFext are seen to increases
with ∆Hc while values of UAext are seen to decrease.
extinction experiments in this study are, therefore, similar to flames in a forced
flow over a stagnant fuel film, which were studied by Emmons. In his analysis,
Emmons showed that the flame existed in the boundary layer of the flow, and that
it separated the region of cooler air from the fuel bed. As the free-stream velocity
of the flow increased, the strain rate of the flow increased, the boundary layer
thickness decreased, and the flame was forced closer to the fuel. Extinction occured
when the free-stream velocity and corresponding strain rate became so great that
chemical kinetics could not compete with the mixing rate of the reactants, and
the value of Da dropped below a critical value. However, so long as combustion
78
Acoustic Extinction
chemistry was occurring, the flame existed within the boundary layer of the flow
[44, 45].
(a) +400 ms (b) +500 ms (c) +600 ms
Figure 7.5: Visualization of the fan-driven flow over the burner with UF = 1.6
m/s. The times noted are from when the fan was activated. A boundary layer
can clearly be seen forming over the burner as the flow develops.
Won et al. showed that the extinction strain rate (aE) for diffusion flames fueled
by large hydrocarbons scales as:
aE ∝ [DFYF,−∞∆Hc] [Kinetic Term] (7.1)
where DF is the diffusivity of the fuel into air, and YF,−∞ is the fuel-mass fraction
on the fuel side of the reaction zone [9]. For the combustion of an alkane though,
the number of moles of the heavier species in the mixture (e.g. CO2 and N2) all
scale approximately with the number of carbon atoms in the fuel. In addition, the
fuel’s molar mass also scales roughly with the number of carbon atoms. It can
be expected, therefore, that YF was roughly constant for all the fuels tested. In
addition, values of DF and the kinetic term should all be approximately the same
when compared among the fuels. It can be concluded then that as values of ∆Hc
79
Acoustic Extinction
for the fuels tested increased, the speed of the flow required to cause extinction
should also have increased.
Examining the fan data points in Fig. (7.4), it is seen that UFext does increase
with ∆Hc, as expected. In contrast, values of UAext are seen to decrease with ∆Hc.
When this observation is coupled with the fact that UAext is significantly less than
UFext , it suggests that flame stretch was not the cause of extinction in the acoustic
experiments, and that an alternate mechanism needs to be found.
7.2.3 Heuristic Framework
Reexamining the plots of PAext in Fig. (7.1a) and UAext in Fig. (7.1b), there is
a positive correlation between ω with PAext and UAext . The amount of scatter
though in PAext (CV=0.29) is nearly twice the amount seen UAext (CV=0.15). This
would suggest that the acoustic extinctions were more closely linked to UA than
PA. Since it is assumed that flame stretch was not the cause of extinction, it might
be reasonable to conclude that convective cooling of the wick was a cause.
To understand how this process might work, a simplified model of the flame is
utilized: fuel enters the flame region, reacts with the oxidizer, releases heat, a
portion of that energy is fed back into the fuel source, which drives more fuel into
the flame region. In this model, the propensity of the fuel to maintain this cycle
is best described by the Spalding B Number, and a large disruption to this cycle
would cause flame extinction.
80
Acoustic Extinction
Examining the image sequence presented in Fig. (7.6), it can be seen that as the
acoustic wave propagates over the burner, it temporarily displaces the flame from
the region above the wick. In contrast to the fan-driven flows, a boundary layer
never has time to fully form in the acoustic scenario. This means that the flame is
not confined to the region directly above the fuel, and that the fuel bed can be
exposed directly to cool air. It can be assumed that during the flame’s displacement
period, not only is the amount of heat feedback to the wick inhibited, but that
the exposed wick also experiences convective cooling from the acoustic flow. From
stagnant layer theory it is know that as the heat flux from the flame into the fuel
bed decreases, so too will the fuel’s mass loss rate (m˙′′) [17]. Furthermore, according
to the fire point theory described by Rashbash, there is a critical mass flux (m˙′′cr)
for any given fuel below which total flame extinction will occur [46, 47]. It would
be reasonable to conclude then that convective cooling of the fuel bed during the
flame’s displacement was creating conditions where m˙′′ < m˙′′cr. Assuming this is
true, a local Nusselt number (Nuξ) could be used to characterize the magnitude of
this cooling.
7.2.4 Proposed Extinction Criterion
In the model proposed, the B Number characterizes the fuel’s ability to maintain
the flame-fuel cycle, and the Nusselt Number characterizes the amount of disruption
to this cycle. A ratio of these two non-dimensional numbers might then constitute
a criterion by which acoustic extinctions can be predicted. For consistency with
81
Acoustic Extinction
(a) +0 ms (b) +4 ms (c) +8 ms (d) +12 ms
(e) +16 ms (f) +20 ms (g) +24 ms (h) +28 ms
Figure 7.6: Hexane Flame at 35 Hz, just prior to extinction. The sequence
shows the flame through one acoustic period (T = 0.0286 s). The flame begins
over the fuel region, which is indicated by the red marker, and is displaced as
the acoustic wave propagates over the flame holder. From Frames (7.6d) to
(7.6f), the flame is completely displaced from the fuel region. In Frame (7.6g) a
flame reemerges over the fuel bed, and in Frame (7.6h) the process starts over.
the boundary maps shown in Fig. (7.2), it would be useful to structure this ratio so
that larger values correspond with flame extinction, while smaller values correspond
with continued burning. It is, therefore, proposed that the ratio of Nuξ to B at
the point acoustic extinction was achieved is a constant, and that it that forms a
boundary delineating conditions where the flame can and cannot exist. This ratio
shall be called ΘA, as shown in Eq. (7.2).
ΘA =
Nuξ
B
(7.2)
Expressions for the terms in Eq. (7.2) shall be developed in the following sections.
82
Acoustic Extinction
7.3 B - Mass Transfer Number
In a study of ethanol and heptane pool fires of various areas with varying crosswinds,
Hu et al found that the ratio of radiation absorbed by the fuel to the heat needed
for vaporization (χa) decreased as the fuel area decreased and the crosswind
increased [48]. For their smallest heptane pool fire, which had an area of 10
cm2, they found χa = 0.26 with no crosswind and χa = 0.10 with a cross flow
of 0.7 m/s. In the experiments for this study the burner area was 2.5 cm2 and
UA = 0.68 m/s. Although the setup and testing conditions for this study were not
perfectly analogous to Hu’s, it can be assumed — based on their results and general
trends seen in their data — that the flames in this study were driven primarily by
convective heat transfer. With this assumption, Quintiere gives the B number as
[17]:
B =
YO2,∞(∆hc/r)− cp,air(Tb − T∞)
L
(7.3)
where it is assumed that YO2,∞ = 0.233 and T∞ = 298 K.
7.3.1 Fuel Specific Parameters
For the alkanes tested and certain gases of interest, data on the following parameters
was obtained from NIST Chemistry Web-Book [49], and the values are reproduced
in Tab. (7.3):
83
Acoustic Extinction
• Molar Mass (M)
• Liquid Specific Heat at 300 K (cp, liq)
• Liquid Heat of Formation (∆h◦f, liq)
• Gaseous Heat of Formation (∆h◦f, g)
• Heat of Vaporization (hv)
• Boiling Temperature at 1 atm (Tb)
Since JP-8 is distilled from crude oil it is not a “pure” substance, and its properties
are not as well defined. The values needed were, therefore, amalgamated from a
multitude of sources and are listed below. The results, where appropriate, are also
presented in Tab.’s (7.3) and (7.5).
• The molecular formula of JP-8 was approximated as C11H21 [50, 51], from
which the molar mass was estimated to be M = 1.53× 10−1 kg/mol.
• NIST gives the vapor rising temperature of JP-8 to be 182.8◦C, and this is
considered to be the initial boiling temperature [52]. The boiling temperature
was therefore estimated as Tb = 456 K.
• The specific heat of liquid JP-8 can be approximated from its temperature
as [52]:
cp(T ) = (2.193± 0.0055) + (3.996± 0.0011)× 10−3(T − 363.15) (7.4)
84
Acoustic Extinction
Using Eq. 7.4, the specific heat of JP-8 was evaluated at 1
2
(Tb +T∞) = 378 K,
and found to be cp,liq = 2252 J/(kg·K).
• Using data from the Defense Technical Information Center [53], the heat of
vaporization for JP-8 at T = 456 K was estimated to be hv = (2.85 ± 0.25)×
105 J/kg.
Table 7.3: Selected Properties of Fuels and Gases
Species M cp, liq ∆h
◦
f, liq at 298 K ∆h
◦
f, g hv Tb
×10−2 ×106 ×106 ×105
kg·mol−1 J·(kg·K)−1 J·kg−1 J·kg−1 J·kg−1 K
Pentane 7.215 2317 2.405 ±0.008 2.035 ±0.008 3.673 ±0.083 309.2 ±0.2
Hexane 8.612 2269 2.306 ±0.008 1.940 ±0.009 3.597 ±0.116 341.9 ±0.3
Heptane 10.02 2242 2.239 ±0.008 1.874 ±0.008 3.593 ±0.299 371.5 ±0.3
Octane 11.42 2230 2.191 ±0.016 1.824 ±0.006 3.589 ±0.350 398.7 ±0.5
JP-8 15.33 2252 N/A N/A 2.85± 0.25 456
CO2 4.401 N/A N/A 8.941 ±0.003 N/A N/A
H2O 1.802 N/A N/A 13.42 ±0.002 N/A N/A
O2 3.200 N/A N/A 0 N/A N/A
Using the data presented in Tab. (7.3), and assuming complete combustion in air
at stoichiometric conditions, as shown in Eq. (7.5), the mass ratio of oxygen to fuel
(r), latent heat of vaporization (L), and heat of combustion per unit mass (∆hc)
were calculated as follows, and the results are summarized in Tab. (7.5):
85
Acoustic Extinction
Cx Hy +
(
x+
y
4
)
(3.76 N2 + O2) −→ xCO2 + y
2
H2 O +
(
x+
y
4
)
3.76 N2 (7.5)
r =
(
x+ y
4
)
MO2
MCx Hy
(7.6)
L = hv + cp, liq(Tb − T∞) (7.7)
∆hc = x∆h
◦
f,CO2
MCO2
MCxHy
+
y
2
∆h◦f,H2O
MH2O
MCxHy
−∆h◦f,CxHy (7.8)
It was assumed that T∞ = 298 K in Eq. (7.7) and that water remained in a gaseous
state in Eq. (7.8). For JP-8, the value of ∆hc is given by multiple sources as
∆hc = 42.8× 106 J/kg [51, 53–55] The uncertainties in ∆hc (S∆hc) are discussed
in Sec. (A.2)
7.3.2 Air Properties
Air properties were interpolated from data presented by Turns in An Introduction
to Combustion [19]. For each fuel tested, the heat capacity of air was evaluated at
1
2
(Tb + T∞), while the kinematic viscosity (ν) and Prandtl (Pr) number of air were
evaluated at Tb. The results are presented in Tab. (7.4).
86
Acoustic Extinction
Table 7.4: Selected Properties of Air for Different Fuels
Fuel Tb cp, air ν Pr
×10−5
K J·(kg·K)−1 m2·s−1
Pentane 309.2 1007 1.68 0.706
Hexane 341.9 1008 2.01 0.701
Heptane 371.5 1008 2.33 0.696
Octane 398.7 1009 2.63 0.690
JP-8 456 1012 3.16 0.686
7.3.3 B Number Calculation
Using the data in Tab.’s (7.3), and (7.4) and the formulas in Eq.’s (7.6) through
(7.8), the values of the B-Mass Transfer Number were calculated and the results
are summarized in Tab. (7.5). The calculated values of B agreed reasonably well
with values presented in other sources [17, 53, 56].
Table 7.5: Calculated Fuel Properties
Fuel r L× 105 ∆hc × 106 B
J·kg−1 J·kg−1
Pentane 3.55 3.933 45.35 ±0.01 7.54
Hexane 3.53 4.593 45.10 ±0.01 6.39
Heptane 3.51 5.241 44.92 ±0.01 5.54
Octane 3.50 5.835 44.79 ±0.01 4.93
JP-8 3.40 6.21 42.80 ±0.18 4.49
87
Acoustic Extinction
7.4 Nusselt Number Correlation
7.4.1 Reynolds Number
Nusselt number correlations are based on the assumption that convective heat
transfer in a boundary layer scales with momentum transfer [57, 58]. Therefore, a
Reynolds Number (Re) had to be found, which in turn required a characteristic
length (`) to be defined. Since it was established that UA was an rms flow velocity
in Sec. (4.5.1), ` was defined as the rms displacement distance of a particle in an
acoustic cycle, and calculated as:
` =
UA
ω
(7.9)
Using Eq. (7.9), the Reynolds number was then calculated as:
ReA =
UA`
ν
=
U2A
νω
(7.10)
7.4.2 Mathematical Formulation
Nusselt number correlations are not well studied for oscillating flows [59]. It was
still necessary to characterize the ratio of convective to conductive heat transfer at
the wick’s surface, and it was assumed that a functional correlation existed which
88
Acoustic Extinction
linked a local Nusselt number at the flame position (Nuξ) with the Reynolds and
Prandtl numbers:
Nuξ = f( Re,Pr) (7.11)
Using the analogy of a non-oscillating flow over a flat plate , it was assumed Eq.
(7.11) took the form [57, 58]:
Nuξ = cRe
γ Prδ (7.12)
where c, γ, and δ are determined empirically.
7.4.3 Modified Nusselt Number
The values of the Prandtl numbers to be used in Eq. (7.12) were taken from Tab.
(7.4) and were practically constant for all tests (Pr = 0.70, σPr = 0.7 × 10−2).
Since c is also a constant, Eq. (7.12) was rewritten as:
Nuξ = C Re
γ
A (7.13)
where C = cPrδ. If Nu′ is defined as Nu′ = Nuξ/C, then Eq. (7.13) can be further
simplified as:
89
Acoustic Extinction
Nu′ = ReγA (7.14)
Finally, substituting Eq. (7.14) into Eq. (7.2) yields:
Θ′A =
ReγA
B
(7.15)
7.4.4 Exponent of Best Fit
The value of γ in Eq. (7.15) was chosen to optimize the results of Θ′A. To find
this value, a MATLAB script was written that calculated the values of Θ′A and
the corresponding CV’s of the data for −3 ≤ γ ≤ 3. The value of γ where the CV
was minimized would then indicate the exponent in Eq. (7.15) that produced the
least amount of scatter in values of Θ′A. The results from this script are shown in
Fig. (7.7), and the minimum occured at γ = 1/3. It should be noted though that
most Nusselt Number correlations for forced convection over a plate scale by Re1/2
[57, 58].
90
Acoustic Extinction
.
-3 -2 -1 0 1 2 3
<=
7
0
0.5
1
1.5
Calculated CV
Figure 7.7: Coefficients of Variation for Θ′A
with varying values of x. The minimum
occurs at γ = 1/3, where CVΘ′A=0.043.
7.5 Calculated Values of Θ′
7.5.1 Acoustic Results
Using the results from Sec. (7.4.4), Eq. (7.15) becomes:
Θ′A =
Re
1/3
A
B
=
U
2/3
A
(νω)1/3B
(7.16)
Applying the model described by Eq. (7.16) to the acoustic extinction data set
yields highly consistent results with Θ′Aext = 1.4, σΘ′A = 0.059, and CVΘ′A=0.043.
The results, including uncertainties, are shown in Fig. (7.8); the uncertainties are
discussed in Appendix (A.2.2). When the scatter in Θ′A is compared to that of
UAext (CVUA = 0.15) and PA (CVPA = 0.29), it becomes clear that Θ
′
Aext
is a much
91
Acoustic Extinction
more consistent descriptor of conditions at extinction than either UAext or PAext .
Furthermore, since Θ′Aext was calculated from conditions created by the lowest
speaker power that could consistently cause extinction, it is proposed that Θ
′
Aext
constitutes a critical value, below which the flame continues to burn and above
which total flame extinction occurs.
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
#
A
0
0
0.5
1
1.5
2
2.5
3
Pentane
Hexane
Heptane
Octane
JP-8
   7 =1.4
   < =0.061
</7 =0.043
Figure 7.8: Calculated values for Θ′Aext , based on conditions at the lowest
speaker power that could consistently cause total flame extinction. The
model shows significantly less scatter than either UAext or PAext .
7.5.2 Hypo and Hyper Critical Values
To test whether Θ
′
Aext constitutes a critical value for predicting extinction, values of
Θ′ had to be calculated for conditions above and below the minimum speaker power
92
Acoustic Extinction
required to cause flame extinction. Within the total data set, there existed several
instances where sufficient data was collected (i.e., 3 or more trials) at speaker powers
other than the minimum power required to cause extinction. From these data
points, values of Θ′A could be calculated using Eq. (7.16). The extinction results
from these trials, along with the calculated values of Θ′A and the uncertainties are
presented in Tab. (7.6).
Table 7.6: Analysis of Hypo and Hyper Critical Extinction Conditions
Fuel ω (Hz) UA (m·s−1) Extinction Θ′A SΘ′A
Pentane 30 0.61 No 1.20 0.04
Hexane 30 0.52 No 1.20 0.06
Hexane 35 0.69 No 1.38 0.05
Heptane 30 0.48 No 1.25 0.09
Heptane 35 0.53 No 1.26 0.08
Heptane 45 0.69 No 1.38 0.09
Octane 30 0.50 No 1.38 0.09
Pentane 30 0.77 Yes 1.41 0.05
Heptane 45 0.86 Yes 1.60 0.10
Heptane 50 0.78 Yes 1.45 0.09
JP-8 40 0.64 Yes 1.51 0.05
According to the theory presented, Θ
′
Aext constitutes a critical value below which
the flame can continue to burn and above which flame extinction occurs. When
the values of Θ′A from Tab. (7.6) are plotted, as shown in Fig. (7.9), it is seen that
Θ
′
Aext does in fact form a boundary between flammability and extinction. For those
trials where extinction did not occur and which are indicated by green markers,
values of Θ′A were less than Θ
′
Aext . For those trials where extinction did occur
above the minimum speaker power and which are indicated by red markers, Θ′A
93
Acoustic Extinction
! (Hz)
30 35 40 45
#
A
0
1.2
1.25
1.3
1.35
1.4
1.45
1.5
1.55
1.6
Pentane
Hexane
Heptane
Octane
JP-8
Flame
Extinction
#0Aext
Figure 7.9: Calculated values of Θ′A for hypo and hyper
critical data points. Trails where extinction was not achieved
are indicated by green markers and trials where extinction was
achieved are indicated by red markers. Values of Θ′A < Θ
′
Aext
correspond uniquely with continued burning, while values of
Θ′A > Θ
′
Aext correspond uniquely with extinction.
was greater than Θ
′
Aext . The errors bars in Θ
′
A have been omitted from Fig. (7.9)
for clarity.
7.5.3 Fan Driven Results
To test if the model for Θ′A was unique to the acoustically-driven extinction events,
the model was also applied to the fan-driven extinctions. In this analysis though, `
was taken to be the distance from windward edge of the flame holder to the center
of the fuel bed, which was 0.067 m. The Reynolds number was then calculated as:
94
Acoustic Extinction
ReF =
UF `
ν
(7.17)
The ratio of the modified Nusselt number to B number for the fan-driven flows
(Θ′F ) was calculated as:
Θ′F =
(UF `/ν)
1/3
B
(7.18)
The calculated values of Θ′F at extinction(Θ
′
Fext
), including the uncertainties, are
plotted in Fig. (7.10); the uncertainties are discussed in Appendix (A.2.3). In
contrast to the acoustic results presented in Sec. (7.5.1), the values of Θ′Fext showed
slightly increased scatter (CVΘ′F = 0.14) when compared to that seen in UFext
(CVUF =0.10). This indicates that Θ
′
F is a less consistent descriptor of conditions at
extinction, and that the proposed model does not explain the fan driven extinction
results.
Molar Mass (kg/mol)
0.06 0.08 0.1 0.12 0.14 0.16 0.18
#
F
0
0
0.5
1
1.5
2
2.5
3
3.5
4
Pentane
Hexane
Heptane
Octane
JP-8
   7 =3.1
   < =0.44
</7 =0.14
Figure 7.10: Calculated values of Θ′F for fan
driven extinction events. The model shows
slightly more scatter than UF .
95
Acoustic Extinction
7.6 Limitations and Considerations
7.6.1 System Limitations
The performance characteristics of the speaker used for this study posed significant
limitations on the range of frequencies and acoustic pressures that could be tested.
The speaker had a rated power of 250 W, and as seen in Fig. (7.11a), this operating
limit was quickly reached. In fact, the one test series shown in Fig. (7.11a) that
occured above 250 W caused irreversible damage to the speaker, which required
it to be replaced. The actual acoustic power per unit area (SWL′′) of the waves
produced was calculated using Eq. (7.19) [37], and those results are presented in
Fig. (7.11b).
SWL′′ = PAUA (7.19)
When the ratio of SWL′′ to speaker power is taken, as shown in Fig. (7.12), it
becomes apparent that as the frequency was increased, there was a “diminishing
return” on the power of the acoustics generated. If the speaker had been rated for
a higher power, acoustic extinction of the flames could likely have been achieved at
higher frequencies. This supposition is consistent with the findings of Whiteside,
who concluded that acoustic extinction could be achieved at any frequency, provided
that a critical acoustic velocity could be achieved [4].
96
Acoustic Extinction
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
Sp
ea
ke
r
Po
w
er
(W
)
0
50
100
150
200
250
300
350
400 Pentane
Hexane
Heptane
Octane
JP-8
Rated Power
(a) Speaker power required to
cause acoustic extinction. The
black line indicates the rated
power of the speaker.
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
SW
L0
0
(W
/m
2 )
0
5
10
15
20
25
30
35
40
45
50
Pentane
Hexane
Heptane
Octane
JP-8
(b) Sound Power Level per unit
area (SWL′′) of the acoustic waves
produced at the minimum speaker
power required to cause extinction.
Figure 7.11: Comparison of the power required by the speaker and power of the
acoustic wave generated at extinction.
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
SW
L0
0
/
Sp
ea
ke
r
Po
w
er
(m
!
2 )
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7 Pentane
Hexane
Heptane
Octane
JP-8
Figure 7.12: Ratio of SWL′′ to speaker
power. As the frequency was increased the
ratio was seen to decrease
7.6.2 Considerations of Applicability
When compared to other flame suppression techniques, the concept of acoustic
flame suppression has several unique advantages. Chief among these advantages
97
Acoustic Extinction
is that there is no flame suppressing agent which needs to be applied. For fire
protection scenarios where there are sensitive mechanical and electrical systems,
this feature is especially desirable. It would therefore be beneficial to consider
whether this technique could be applied to fires larger than the ones used for the
study. To do so, a heptane pool fire whose diameter (D) is 1 m was considered,
and the speaker size required to control this fire was be estimated.
According to the model developed, acoustic flame extinction requires the oscillatory
movement of air to be sufficiently fast that convective cooling prevents enough fuel
from being liberated to sustain the flame. Although heat transfer in a pool fire of
this size is dominated by radiation [17, 19, 60], for the purposes of scaling it was
assumed that convection was still the primary mode of heat transfer. Another key
consideration to explore though is the observation of McKinney and Dunn-Rankin,
who found that acoustic extinction of droplet flames required the flame to be
displaced by at least the droplet’s radius [23]. When the ratio of ` to the burner
width (wb) was taken for the extinction data in Tab. (7.1) – the results of which
are shown in Fig. (7.13) – its average was found to be 3.66.
With these considerations in mind, an a priori analysis would indicate that for a
pool fire, `/D ≥ 1. Taking this ratio at it’s minimum, it was assumed that ` ≈ D.
Using Eq. (7.9), the acoustic velocity required for extinction was calculated as:
UA = ωD (7.20)
98
Acoustic Extinction
! (Hz)
30 32 34 36 38 40 42 44 46 48 50
`=
w
b
0
1
2
3
4
5
6
7
8
9
10
Pentane
Hexane
Heptane
Octane
JP-8
7 = 3:66
< = 0:64
<=7 = 0:18
Figure 7.13: Ratio of rms displacement distance (`) to the
burner width (wb).
Reexamining Fig. (7.12), it can be seen that the lower the frequency, the more
efficiently the speaker produced acoustics. Assuming this trend is true for any
speaker, 30 Hz was used for the scaling since it is the lowest frequency known to
cause acoustic extinction. Evaluating Eq. (7.20) with ω = 30 Hz and D = 1 m, a
value of UA = 30 m/s was calculated. When these values are then used in Eq. (7.16)
along with the appropriate values for ν and B, it was found that ΘApool = 19.6.
Recalling the results from Sec.’s (7.5.1) and (7.5.2), and assuming that Θ
′
Aext = 1.4
is a universal extinction constant, it was concluded that an rms air speed of 30
m/s at a frequency of 30 Hz could extinguish a heptane pool fire 1 m in diameter.
To estimate the size the speaker though an estimate of the acoustic pressure was
required. Such an estimate was made by using the characteristic specific acoustic
99
Acoustic Extinction
impedance of air (Zo) at 298 K, which is given as 410 Pa·s·m−1 [37]. The acoustic
pressure was then calculated as:
PA = UAzo
= 1.23× 104 kPa (7.21)
Using Eq. (3.1), it was found that 1.23×104 kPa = 144 dB. Commercially available
sub-woofers generally produce acoustics no louder than 110 dB at 1 m [61]. It
is possible, though, to produce low frequency acoustics at higher pressures. In
1996 ARL studied the use of “weaponized” acoustics at 20 Hz and 165 dB [62].
Given that the calculated acoustic pressure to control the pool fire is 4× larger
than a “weaponized” acoustic pressure generated with specialty equipment, the
use of acoustics alone to suppress this fire would be highly impractical.
The fact that acoustics alone would not be suitable for controlling larger fires
should not necessarily preclude development of the technology. Even if flame
extinction can’t be achieved with acoustics alone, acoustics can certainly be used
to destabilize the flame and even force it to become detached. If the acoustics were
coupled with a known suppressing agent (e.g., water mist, dry-chem, CO2), the
combination of flame destabilization, increased mixing, and delivery of the agent
directly to the fuel bed could prove to be an efficient means of fire suppression.
100
Chapter 8
Conclusions
A burner and testing apparatus were constructed to study the interaction of acoustic
waves with a laminar flame produced with various liquid fuels. The fuels used in
the study were the alkanes n-pentane through n-octane and JP-8. The burner was
designed to produce a line-flame which approximated a flame sheet. To produce the
line-flame, fuel-laden wicks were placed under panes of borosilicate glass, leaving
only a 5.0×10−2 m × 5.0×10−3 m gap through which fuel could escape. The testing
apparatus itself consisted of a 0.25 m diameter, 3.05 m long PVC tube with the
burner placed at one end and a sub-woofer mounted inside the tube at the other
end.
Acoustics were produced at frequencies of 30-50 Hz and at acoustic pressures of
5-50 Pa. Measurements of the pressures inside the tube showed that the system
achieved resonance at frequencies consistent with theory. Measurements of the
101
Conclusions
pressures outside the tube showed a linearly proportional decay as distance from
the tube opening increased. Also of interest were the air movements within the
enclosure which were induced by the acoustics. Measurements of air movement
were made with a hot-wire anemometer, and PIV analysis showed that the quantity
being measured was the rms speed of the oscillatory air movement.
Prior to the acoustic testing with a flame, studies of the alkane flames’ unperturbed
burning rate, flame height, and flame width were conducted. To collect this
data, the burner was placed on a mass balance, the fuels were ignited, and the
flames were allowed to evolve while the process was simultaneously recorded via
videography. Mass loss rates were then evaluated from the temporally indexed
mass readings of the balance, and flame heights were evaluated from the temporally
and spatially indexed pixel intensities of the videography. Estimates of the flame’s
width were made from direct measurements of the flame’s width in the video image.
Using each fuel’s heat of combustion, a direct comparison was made between each
flame’s height and heat release rate per unit flame width. A linear relationship was
observed between these quantities, which is consistent with the Burke-Schumann
model.
To more thoroughly understand the effects of the acoustics on the flame, a study of
flame’s burning rate under acoustic excitation was preformed. Samples of hexane
were ignited and allowed to evolve until burnout under various combinations of
acoustic frequency and pressure, during which mass readings were recorded. Since
the acoustics induced air movements, which the flame was sensitive to, mlr profiles
102
Conclusions
using a fan at various air speeds were also generated for comparison. It was found
that the acoustic and fan flows only affected the flame during the growth phase.
Once peak mlr was achieved, all profiles converged towards the free burn profile. A
comparison of mlr profiles at differing frequencies and comparable acoustic pressures
revealed no obvious trends. A comparison of mlr profiles at differing frequencies
and comparable rms air speeds were found to be very similar. In addition, there
was a strong correlation between average mlr and rms air speed at all frequencies.
These observations indicated that it was the effects of the oscillatory air movement
which influenced the flame, and that frequency and acoustic pressure influenced
the flame’s burning rate only insofar as they influenced oscillatory movement of air.
It is believed that the observed increase in burning rate was due to the flame being
forced closer to the fuel surface, which enhanced heat transfer into the fuel bed.
Finally, a study of acoustically-driven extinction was carried out using the alkanes
and JP-8. Samples were ignited and then subjected to acoustic perturbations at
various frequencies and speaker powers until the minimum speaker power was found
that could cause three consecutive flame extinctions. The acoustic pressure and
rms air speeds were measured during each trial and, along with frequency, these
measurements were used to characterize the acoustic conditions. The minimum fan
driven flows required to cause extinction for each fuel was also evaluated. Analysis of
the data showed that the fan-driven air speed required to cause acoustic extinction
of each fuel increased with the fuel’s heat of combustion per unit mole, and that
this trend was consistent with extinction strain rate theory. Using acoustics though,
103
Conclusions
the rms air speed at extinction was seen to decrease, which indicated that flame
stretch was not the cause of extinction using acoustics.
High speed video showed that during acoustic excitation, the flame would become
detached, forced away from the fuel bed, and then returned and reattached until
the next cycle. It was theorized that during this displacement phase the fuel
bed experienced convective cooling, and that this eventually caused the fuel’s mlr
to fall below the critical amount needed to sustain the flame. This hypothesis
was consistent with that developed by other authors to explain the acoustically
driven extinction of droplet flames. To explore this hypothesis, the flame was
conceptualized in a simple model between flame and fuel source. In this model, the
Spalding B Number was used to characterize the interplay between the flame and
fuel, and a Nusselt Number was used to characterize the convective cooling of the
fuel bed. Mathematical expressions were then developed for each of the numbers
and they were evaluated using values reported in the literature and conditions
measured during experiment.
It was found that at the minimum speaker power required to cause extinction,
the ratio of the Nusselt Number to B Number was a constant for all fuels at all
frequencies tested. It was found that when this ratio was below the constant, the
flame continued to burn. If the ratio was greater than or equal to the constant, then
flame extinction occured. It was therefore asserted that this constant constituted a
boundary between regions of flammability and flame extinction.
104
Appendix A
Error Analysis
A.1 Data Uncertainties
Since the data points for each extinction event were only calculated from three trials,
there were an insufficient number of measurements to estimate the uncertainty
directly. To overcome this, uncertainties were estimated from a larger data set,
which included trials where either extinction was not achieved or extinction was
achieved at speaker powers higher than the minimum required.
The analysis began with a selection of data subsets that contained three or more
measurements of the system at homologous conditions (i.e., the same frequency and
speaker power). A total of 17 subsets (NS) were used, with each subset containing
105
Appendix A: Uncertainties and Error Propagation
three to five trials (Ni). Amongst all data subsets there were a total of 53 trials
(NT ), and the data was structured as shown in Tab. (A.1).
Table A.1: Uncertainty Analysis Data Structure
Data Subset Trial PA UA
1 1 PA11 UA11
2 PA12 UA12
3 PA13 UA13
2 1 PA21 UA21
2 PA22 UA22
3 PA23 UA23
4 PA24 UA24
...
...
...
...
i j PAij UAij
For each data subset, an average was calculated using Eq. (A.1). Each data point
within the subset was then normalized by the average of that subset using Eq.
(A.2).
X i =
1
Ni
Ni∑
j=1
Xij (A.1)
X̂ij =
Xij
Xi
(A.2)
An average of all normalized data points was then calculated by Eq. (A.3), and a
relative standard deviation of these normalized data points was calculated by Eq.
(A.4).
106
Appendix A: Uncertainties and Error Propagation
X˜ =
1
NT
NS∑
i=1
Ni∑
j=1
X̂ij (A.3)
σX̂ =
√√√√ 1
NT
NS∑
i=1
Ni∑
j=1
(
X̂ij − X˜
)2
(A.4)
The relative standard deviations of the acoustic pressure and rms air speed were
found to be σP̂A = 0.006 and σÛA = 0.035. The standard deviation of an extinction
data point from Tab. (7.1) was then estimated as:
σX = σX̂X (A.5)
On average, the individual extinction values of PA were 1.04 standard deviations
from the their subset average, and the individual extinction values of UA were 0.60
standard deviations from the their subset average. With estimates for the standard
deviations of each extinction data point, the uncertainties in the data from Tab.
(7.1) were calculated as:
SPA = 2σPA/
√
N (A.6)
SUA = 2σUA/
√
N (A.7)
107
Appendix A: Uncertainties and Error Propagation
where N = 3. For the fan driven flows, it was assumed that σÛF = σÛA , and the
uncertainties in the fan extinction data points were also estimated by Eq. (A.7).
In addition to the random uncertainties within the bulk air speed measurements
there was also systemic uncertainties from the anemometer, which are discussed
further in Sec. (A.2.1).
A.2 Fuel Parameter Uncertainties
Uncertainties in the values of ∆hc were calculated through an error propagation of
Eq. (7.8), as shown in Eq. (A.8).
S∆hc =
√∑( ∂hc
∂∆Hf,i
Si
)2
=
√(
x
MCO2
MCxHy
S∆H◦f,CO2
)2
+
(
y
2
MH2O
MCxHy
S∆H◦f,H2O
)2
+
(
−S∆H◦f,CxHy
)2
(A.8)
A.2.1 Model Uncertainties
Estimates of the uncertainties in Θ′A and Θ
′
F were made through a modified error
propagation of Eq.’s (7.16) and (7.18), respectively.
108
Appendix A: Uncertainties and Error Propagation
A.2.2 Uncertainties in Acoustic Analysis
In its fully expanded form for the acoustic analysis, Eq. (7.16) is:
Θ′A =
U
2/3
A [hfg + cp, liq(Tb − T∞)]
(νω)1/3 [YO2,∞(∆hc/r)− cp, air(Tb − T∞)]
(A.9)
It is assumed that the main sources of error in Eq. (A.9) came from uncertainties
in UA, hv, and ∆hc. The error in Θ
′
A was, therefore, estimated as:
SΘ′A =
√√√√√
(
∂Θ′A
∂UA︸ ︷︷ ︸
Term A1
SUA
)2
+
(
∂Θ′A
∂hfg︸ ︷︷ ︸
Term A2
Shfg
)2
+
(
∂Θ′A
∂∆hc︸ ︷︷ ︸
Term A3
S∆hc
)2
+
∂Θ′
∂UA︸︷︷︸
Term A1
SUA,Sys
(A.10)
The final element of Eq. (A.10) represents the systemic uncertainty in the anemome-
ter, which was given by the manufacturer as ± 3%. Each partial derivative indicated
in Eq. (A.10) is shown below:
109
Appendix A: Uncertainties and Error Propagation
Term A1:
∂Θ′A
∂UA
=
2Re
1/3
A
3UAB
(A.11)
Term A2:
∂Θ′A
∂hfg
=
Re
1/3
A
BL
(A.12)
Term A3:
∂Θ′A
∂∆hc
=
−YO2,∞Re1/3A
rB2L
(A.13)
A.2.3 Uncertainties in Fan Driven Analysis
In its fully expanded form for the fan driven analysis, Eq. (7.16) is:
Θ′F =
(UF `/ν)
1/3 [hfg + cp, liq(Tb − T∞)]
[YO2,∞(∆hc/r)− cp, air(Tb − T∞)]
(A.14)
It is assumed that the main sources of error in Eq. (A.14) came from uncertainties
in UF , hv, ∆hc, and `. The error in Θ
′
F was then estimated as:
SΘ′F =
√√√√√
(
∂Θ′F
∂UF︸ ︷︷ ︸
Term F1
SUF
)2
+
(
∂Θ′F
∂hfg︸ ︷︷ ︸
Term F2
Shfg
)2
+
(
∂Θ′F
∂∆hc︸ ︷︷ ︸
Term F3
S∆hc
)2
+
(
∂Θ′F
∂`︸ ︷︷ ︸
Term F4
S`
)2
+
∂Θ′
∂UF︸︷︷︸
Term F1
SUF,Sys (A.15)
110
Appendix A: Uncertainties and Error Propagation
Term F1:
∂Θ′F
∂UF
=
Re
1/3
F
3UFB
(A.16)
Term F2:
∂Θ′F
∂hfg
=
Re
1/3
F
BL
(A.17)
Term F3:
∂Θ′F
∂∆hc
=
−YO2,∞Re1/3F
rB2L
(A.18)
Term F4:
∂Θ′F
∂`
=
Re
1/3
F
3B`
(A.19)
111
Appendix B
Product Data Sheets
112
Agilent 33220A 20 MHz Function/Arbitrary 
Waveform Generator
Data Sheet
• Fully compliant to 
 LXI Class C specification
•  20 MHz Sine and 
  Square waveforms 
•  Pulse, Ramp, Triangle, 
  Noise, and DC waveforms
•  14-bit, 50 MSa/s, 64 k-point
  arbitrary waveforms
•  AM, FM, PM, FSK, and 
  PWM modulation types
•  Linear & logarithmic sweeps
  and burst operation
•  10 mVpp to 10 Vpp 
  amplitude range
•  Graph mode for visual 
  verification of signal settings 
•  Connect via USB, GPIB 
  and LAN
Uncompromising performance 
for functions and waveforms
The Agilent Technologies 33220A 
function/arbitrary waveform genera-
tor uses direct digital synthesis 
(DDS) techniques to create a stable, 
accurate output signal for clean, low 
distortion sine waves. It also gives 
you square waves with fast rise and 
fall times up to 20 MHz and linear 
ramp waves up to 200 kHz.
Pulse generation
The 33220A can generate 
variable-edge-time pulses up to 
5 MHz. With variable period, pulse 
width, and amplitude the 33220A is 
ideally suited to a wide variety of 
applications requiring a flexible pulse 
signal.
Custom waveform generation
Use the 33220A to generate complex 
custom waveforms.  With 14-bit reso-
lution, and a sampling rate of 50 
MSa/s, the 33220A gives you the flex-
ibility to create the waveforms you 
need. It also lets you store up to four 
waveforms in nonvolatile memory.
The Agilent IntuiLink arbitrary wave-
form software allows you to easily 
create, edit, and download complex 
waveforms using the waveform edi-
tor. Or you can capture a waveform 
using IntuiLink for Oscilloscope and 
send it to the 33220A for output. To 
find out more about IntuiLink, visit 
www.agilent.com/find/intuilink.
Appendix B: Product Data Sheets
113
Measurement Characteristics
Easy-to-use functionality
Front-panel operation of the 33220A 
is straight-forward and user friendly. 
You can access all major functions 
with a single key or two. The knob or 
numeric keypad can be used to 
adjust frequency, amplitude, offset, 
and other parameters. You can even 
enter voltage values directly in Vpp, 
Vrms, dBm, or as high and low lev-
els. Timing parameters can be 
entered in Hertz (Hz) or seconds.
Internal AM, FM, PM, FSK, 
and PWM modulation make it easy 
to modulate waveforms without the 
need for a separate modulation 
source. Linear and logarithmic 
sweeps are also built in, with sweep 
rates selectable from 1 ms to 500 s. 
Burst mode operation allows for a 
user-selected number of cycles per 
period of time. GPIB, LAN, and 
USB interfaces are all standard, plus 
you get full programmability using 
SCPI commands.
External frequency reference 
(Option 001)
The 33220A external frequency ref-
erence lets you synchronize to an 
external 10 MHz clock, to another 
33220A, or to an Agilent 33250A. 
Phase adjustments can be made 
from the front panel or via a com-
puter interface, allowing precise 
phase calibration and adjustment.
Waveforms
Standard  Sine, Square, Ramp,   
  Triangle, Pulse, 
  Noise, DC
Built-in arbitrary Exponential rise,   
  Exponential fall, 
  Negative ramp,   
  Sin(x)/x, Cardiac
Waveforms Characteristics 
Sine
Frequency Range 1 µHz to 20 MHz
Amplitude Flatness[1], [2] (relative to 1 kHz)
 < 100 kHz  0.1 dB 
 100 kHz to 5 MHz  0.15 dB 
 5 MHz to 20 MHz  0.3 dB  
Harmonic distortion[2], [3]
  < 1 VPP ≥ 1 VPP
 DC to 20 kHz  -70 dBc -70 dBc
 20 kHz to 100 kHz  -65 dBc -60 dBc
 100 kHz to 1 MHz  -50 dBc -45 dBc
 1 MHz to 20 MHz  -40 dBc -35 dBc
Total harmonic distortion[2], [3]
 DC to 20 kHz  0.04%
Spurious (non-harmonic)[2], [4]
 DC to 1 MHz  -70 dBc
 1 MHz to 20 MHz  -70 dBc + 6 dB/octave
Phase noise 
  (10 kHz offset) -115 dBc / Hz, typical 
Square
Frequency range 1 µHz to 20 MHz
Rise/Fall time  < 13 ns
Overshoot  < 2% 
Variable duty cycle 20% to 80% (to 10 MHz)
  40% to 60% (to 20 MHz)
Asymmetry (@ 50% duty) 
  1% of period + 5 ns
Jitter (RMS) 1 ns + 
  100 ppm of period
Ramp, Triangle
Frequency range 1 µHz to 200 kHz
Linearity  < 0.1% of peak output
Variable Symmetry 0.0% to 100.0%
Pulse
Frequency range 500 µHz to 5 MHz
Pulse width  20 ns minimum,
  (period ≤ 10s) 10 ns resolution
Variable edge time  < 13 ns to 100 ns 
Overshoot < 2% 
Jitter (RMS) 300 ps + 
  0.1 ppm of period
Noise
Bandwidth   9 MHz typical
Arbitrary
Frequency range 1 µHz to 6 MHz
Waveform length  2 to 64 k points
Amplitude resolution  14 bits (including sign)
Sample rate  50 MSa/s
Min. Rise/Fall Time 35 ns typical
Linearity  < 0.1% of peak output
Settling Time  < 250 ns to 0.5% 
  of final value
Jitter (RMS) 6 ns + 30 ppm
Non-volatile memory  four waveforms
Common Characteristics
Frequency  
Accuracy[5] ± (10 ppm + 3 pHz)
      in 90 days
  ± (20 ppm + 3 pHz) 
      in 1 year
Resolution   1 µHz
Amplitude
Range   10 mVPP to 
  10 VPP into 50 Ω
  20 mVPP to 20 VPP 
  into open circuit
Accuracy[1], [2] (at 1 kHz)  
  ± 1% of setting 
  ± 1 mVPP
Units  VPP, Vrms, dBm
Resolution  4 digits
DC Offset
Range (peak AC + DC) ± 5 V into 50 Ω
  ± 10 V into open circuit
Accuracy[1], [2]  ± 2% of offset setting
  ± 0.5% of amplitude 
  ± 2 mV
Resolution 4 digits
Main Output
Impedance 50 Ω typical
Isolation  42 Vpk maximum 
   to earth
Protection  Short-circuit protected, 
   overload automatically 
   disables main output
External Frequency Reference (Option 001)
Rear Panel Input
 Lock Range  10 MHz ± 500 Hz
 Level  100 mVPP to 5 VPP
 Impedance  1 kΩ typical, 
  AC coupled 
 Lock Time  < 2 seconds
Rear Panel Output
 Frequency  10 MHz
 Level  632 mVPP 
  (0 dBm), typical
 Impedance 50 Ω typical, 
  AC coupled2
Appendix B: Product Data Sheets
114
Phase Offset
 Range  + 360° to - 360°
 Resolution  0.001°
 Accuracy  20 ns
Modulation
AM
 Carrier waveforms Sine, Square, 
  Ramp, Arb
 Source  Internal/External
 Internal modulation  Sine, Square, Ramp, 
  Triangle, Noise, Arb 
  (2 mHz to 20 kHz)
 Depth  0.0% to 120.0%
FM
 Carrier waveforms Sine, Square, 
  Ramp, Arb
 Source  Internal/External
 Internal modulation  Sine, Square, Ramp, 
  Triangle, Noise, Arb 
  (2 mHz to 20 kHz)
 Deviation  DC to 10 MHz
PM
 Carrier waveforms Sine, Square, 
  Ramp, Arb
 Source  Internal/External
 Internal modulation  Sine, Square, Ramp, 
  Triangle, Noise, Arb 
  (2 mHz to 20 kHz)
 Deviation 0.0 to 360.0 degrees
PWM
 Carrier waveform Pulse
 Source  Internal/External
 Internal modulation Sine, Square, Ramp,   
  Triangle, Noise, Arb 
  (2 mHz to 20 kHz)
 Deviation 0% to 100% 
  of pulse width
FSK
 Carrier waveforms Sine, Square, 
  Ramp, Arb
 Source  Internal/External 
 Internal modulation 50% duty cycle   
  square (2 mHz 
  to 100 kHz)
External Modulation Input[6] 
 (for AM, FM, PM, PWM)
 Voltage range ± 5 V full scale 
 Input impedance 5 kΩ typical
 Bandwidth DC to 20 kHz
Sweep
 Waveforms Sine, Square, 
  Ramp, Arb
 Type  Linear or Logarithmic
 Direction  Up or Down
 Sweep time 1 ms to 500 s
 Trigger  Single, External, 
  or Internal
 Marker falling edge of sync   
  signal (programmable  
  frequency)
Burst[7]
Waveforms Sine, Square, Ramp,  
  Triangle, Pulse, 
  Noise, Arb
Type Counted (1 to 50,000 
  cycles), Infinite, Gated
Start/Stop Phase  -360° to +360°
Internal Period  1 µs to 500 s 
Gate Source  External trigger
Trigger source  Single, External 
  or Internal
Trigger Characteristics
Trigger input
 Input level TTL compatible
 Slope Rising or Falling,   
  selectable
 Pulse width > 100 ns
 Input impedance >10 kΩ, DC coupled
 Latency < 500 ns
 Jitter (rms) 6 ns (3.5 ns for pulse)
Trigger output
 Level TTL compatible 
  into ≥ 1 kΩ
 Pulse width > 400 ns
 Output Impedance 50 Ω, typical
 Maximum rate 1 MHz
 Fanout ≤ 4 Agilent 33220As
Programming Times (typical)
Configuration times
  USB LAN GPIB
 Function Change 111 ms 111 ms 111 ms
 Frequency Change 1.5 ms 2.7 ms 1.2 ms
 Amplitude Change 30 ms 30 ms 30 ms
 Select User Arb 124 ms 124 ms 123 ms
Arb Download Times 
(binary transfer) USB LAN GPIB
 64 k points 96.9 ms 191.7 ms 336.5 ms
 16 k points 24.5 ms 48.4 ms 80.7 ms
 4 k points 7.3 ms 14.6 ms 19.8 ms
General
Power Supply  CAT II
  100 - 240 V @   
  50/60 Hz (-5%, +10%)
  100 - 120 V @ 400 Hz  
  (±10%)
Power Consumption  50 VA max
Operating Environment  IEC 61010
  Pollution Degree 2
  Indoor Location
Operating Temperature  0°C to 55°C
Operating Humidity  5% to 80% RH, 
  non-condensing
Operating Altitude Up to 3000 meters
Storage Temperature  -30°C to 70°C
State Storage Memory  Power off state 
  automatically saved.  
  Four user-configurable  
  stored states
Interface  USB, GPIB, and 
  LAN standard
Language  SCPI - 1993, 
  IEEE-488.2
Dimensions (W x H x D)
 Bench top  261.1 mm x 103.8 mm 
  x 303.2mm
 Rack mount  212.8mm x 88.3mm 
  x 272.3mm
Weight  3.4 kg (7.5 lbs)
Safety Designed to  UL-1244, CSA 1010,   
  EN61010
EMC Tested to  MIL-461C, EN55011,  
  EN50082-1
Vibration and Shock  MIL-T-28800, Type III,  
  Class 5
Acoustic Noise  30 dBa
Warm-up Time  1 hour
Warranty  1 year standard
Footnotes
1. Add 1/10th of output amplitude and offset 
 spec per °C for operation outside the range 
 of 18°C to 28°C
2. Autorange enabled
3. DC offset set to 0 V
4. Spurious output at low amplitude is 
 –75 dBm typical
5. Add 1 ppm/°C average for operation 
 outside the range of 18°C to 28°C 
6. FSK uses trigger input (1 MHz maximum)
7. Sine and square waveforms above 
 6 MHz are allowed only with 
 an “infinite” burst count
3
Measurement Characteristics (Continued)
Appendix B: Product Data Sheets
115
SPECIFICATIONS
DIAMETER: 8" (200MM) 
SENSITIVITY (2.83V @ 1M): 91dB
POWER HANDLING: 250W      (1000W )PEAKRMS
FREQUENCY RESPONSE: 30HZ ~ 400HZ
NOMINAL IMPEDANCE: 4 OHMS
VOICE COIL DIAMETER: 2.0" (51.0MM) 
DIMENSIONS:
THIELE-SMALL PARAMETERS
VOICE COIL DC RESISTANCE: REVC (OHMS) . . . . . . 3.30
VOICE COIL INDUCTANCE @ 1 KHZ: LEVC (MH) . . . . . . . 2.31
DRIVER RADIATING AREA: SD (IN2) . . . . . . . . 33.17
SD (M2) . . . . . . . 214.00
MOTOR FORCE FACTOR: BL (TM) . . . . . . . . 12.56
COMPLIANCE VOLUME: VAS (FT3) . . . . . . . . 0.54
VAS (LITERS) . . . . . 15.23
SUSPENSION COMPLIANCE: CMS (µM/N) . . . . 232.70
MOVING MASS, AIR LOAD: M MS(GRAMS). . . . 124.20
MOVING MASS, DIAPHRAGM: MMD (GRAMS) . . . 122.40
FREE-AIR RESONANCE: FS (HZ) . . . . . . . . 29.60
MECHANICAL Q: QMS . . . . . . . . . . . 6.10
ELECTRICAL Q: QES. . . . . . . . . . . 0.483
TOTAL Q: QTS . . . . . . . . . . . . 0.45
MAGNETIC-GAP HEIGHT: HAG (IN) . . . . . . . . .0.315
HAG (MM) . . . . . . . . 8.00
VOICE COIL HEIGHT: HVC (IN) .. . . . . . . . . 1.34
HVC (MM) . . . . . . . . . . . 34
MAXIMUM EXCURSION: XMAX (IN) . . . . . . . 0.513
XMAX (MM) . . . . . . 13.00 
SEALED BOX VOLUME 
(INCLUDES DRIVER DISPLACEMENT)
SEALED ENCLOSURE
FREQUENCY RESPONSE @ 2.83V
SEALED ENCLOSURE 
CONE EXCURSION @ 250W
mounting depth
5-3/16" (131mm)
cutout diameter
7-5/16" (186mm)
outer diameter
8-7/16" (214mm)
VBOX = 0.35 ft3 (9.91 liters)
Reference 860w
(side view)
Harman Consumer Group, Inc.,  250 Crossways Park Drive,  Woodbury, NY 11797,  USA www.infinitysystems.com
20  Hz 30 40 50 60 70 80 90 100 200 300
dBSPL
60
65
70
75
80
85
90
95
100
105
110
Out-of-Car
In-Car
12 dB LP @ 100Hz
20  Hz 30 40 50 60 70 80 90 100 200 300
 M
5u
10u
20u
50u
100u
200u
500u
1m
2m
5m
10m
20m
30m
Reference 860w 8" Woofer – Technical Data
Appendix B: Product Data Sheets
116
20  Hz 30 40 50 60 70 80 90 100 200 300
dBSPL
60
65
70
75
80
85
90
95
100
105
110
VENTED BOX VOLUME 
(INCLUDES DRIVER/PORT DISPLACEMENTS)
VENTED ENCLOSURE
FREQUENCY RESPONSE @ 2.83V
VENTED ENCLOSURE 
CONE EXCURSION @ 250W
BAND-PASS BOX VOLUME
(INCLUDES DRIVER/PORT DISPLACEMENTS)
BAND-PASS ENCLOSURE
FREQUENCY RESPONSE @ 2.83V
BAND-PASS ENCLOSURE
CONE EXCURSION @ 250W
Reference 860w
(side view)
Port
fB = 40.0Hz
diameter = 2-1/2"
                   64mm
length = 8-13/16"
               224mm
VBOX  = 0.6 ft3 (16.99 liters) VREAR = 0.54 ft3 (15.29 liters) VFRONT = 0.5 ft3
              (14.16 liters)
Reference 860w
(side view)
diam. = 3"
          77mm
Port
fB= 55Hz
length =
 6-7/8"
175mm
Out-of-Car
Harman Consumer Group, Inc.,  250 Crossways Park Drive,  Woodbury, NY 11797,  USA www.infinitysystems.com
20  Hz 30 40 50 60 70 80 90 100 200 300
dBSPL
60
65
70
75
80
85
90
95
100
105
110
20  Hz 30 40 50 60 70 80 90 100 200 300
 
M
5u
10u
20u
50u
100u
200u
500u
1m
2m
5m
10m
20m
30m
20  Hz 30 40 50 60 70 80 90 100 200 300
 M
5u
10u
20u
50u
100u
200u
500u
1m
2m
5m
10m
20m
30m
In-Car
12 dB LP @ 100Hz
In-Car
12 dB LP @ 100Hz
Out-of-Car
Reference 860w 8" Woofer – Technical Data
Appendix B: Product Data Sheets
117
2 5 0 7  W a r r e n  S t r e e t ,  E l k h a r t ,  I N  4 6 5 1 6  U S A   |   5 7 4 . 2 9 5 . 9 4 9 5   |   w w w . A E T e c h r o n . c o m  
 
 
  8102 Datasheet Information subject to change. 12/21/12 
 
 
8102 SPECIFICATION SHEET 
The AE Techron 8102 power supply 
amplifier features an advanced switch-
mode design that results in low noise 
and distortion, and high power density. 
It is configured as a single-channel, 
DC-coupled, controlled voltage 
amplifier ideal for reactive loads. The 
8102 can provide up to 16 Arms or 
235 Vrms continuous output. It offers a 
continuous, full-power frequency 
bandwidth of DC to 5 kHz.  
The 8102 also features an integrated 
switching power supply that reduces 
weight and allows the unit to fit a 
standard 2U rack space. The 8102 
operates from single-phase, 120-volt 
AC mains (230 VAC version 
available).  
The 8102 amplifier is built and tested 
to the most stringent quality standards 
for long life and outstanding 
performance. The AE Techron brand 
is known throughout the world for its 
robust precision amplifiers as well as 
its product service and support. 
Performance 
Specifications are for units driven into 
an 8-ohm load, (20 times voltage gain) 
and operating from 120 VAC, unless 
otherwise specified.  
“Standard 1 kHz Power" refers to 
maximum average power in watts at 1 
kHz with 0.1% THD. 
Frequency Response: 
±3 dB from DC to 5 kHz at 1 watt 
Signal to Noise Ratio:  
< 105 dB (ref. rated power, DC to 5 
kHz, A-weighted).  
  
Features 
x Up to 16 Arms and 235 Vrms continuous output. 
x Full-power frequency bandwidth of DC – 5 kHz. 
x Compact design; only 2U of rack space and 27 lbs. 
x Switching power supply for reduced weight. 
x Installs easily into a standard 19-inch rack or stands alone for 
bench top operation. 
x Built-in protection circuitry safely provides for sustained high-
power output, with protection against input overloads, improper 
output connection (including shorts and improper loads), and 
excessive temperature, voltage or current. 
x Operates from single-phase, 120-volt AC mains, (230 VAC 
version available). 
Appendix B: Product Data Sheets
118
  8102 Datasheet Information subject to change. 12/21/12 
 
Total Harmonic Distortion (THD):  
<0.35% at full rated power, from DC 
to 5 kHz.  
I.M. Distortion:  
<0.35% at 60 Hz and 7 kHz at 4:1, 
from -40 dB to full rated power. 
DC Output Offset:  
< 15 mV 
Input Impedance (nominally balanced, 
nominally unbalanced):  
10 k ohms, 5 k ohms. 
Maximum Input Voltage: 
± 10 V balanced or unbalanced 
Common Mode Rejection (CMR) 
(20Hz to 1kHz, typical):  
50 dB 
Load Impedance: 
2 – 62 ohm 
Gain Control (when enabled, switch 
selectable): 
Voltage gain adjustable from 20 to 0  
or from 63 to 0 
Front Panel Controls and Indicators 
Fault Indicator:  
Red LEDs, flash when the amplifier 
output has stopped operating. Usually 
this means that the amplifier must be 
serviced. 
Thermal Indicator:  
Red LEDs, illuminate when the 
amplifier has shut down, or is very 
near shutting down, due to thermal 
stress or overload. 
Ready Indicator:  
Green LEDs, illuminate when the 
amplifier is initialized and ready to 
produce output.  
Input Signal Indicator:  
Green LEDs, illuminate when the amplifier’s input signal is above 
–40 dBu (8 mVrms). 
Output Signal Indicator, -20 dB:  
Green LEDs, illuminate when the amplifier’s output signal is within 
20 dB of clipping. 
Output Signal Indicator, –10 dB:  
Green LEDs, illuminate when the amplifier’s output signal is within 
10 dB of clipping. 
Clip Indicator:  
Red LEDs, illuminate when the amplifier’s output signal reaches 
the onset of audible clipping. The Clip Indicators also will 
illuminate during Thermal Level Control (TLC) limiting or when the 
input compressor/limiter is protecting the amplifier from input 
overload. 
Cooling Vents:  
Front-to-rear forced airflow. 
Power Indicator:  
Blue LED indicates AC power has been applied and is within the 
safe operating range of the power supply. The LED will flash when 
the AC line voltage is approximately 15% above or 25% below the 
nominal rated value. 
Data Indicator:  
Feature not implemented. 
Bridge Indicator:  
Illuminates when the amplifier is receiving AC power. 
Power Switch:  
Push-on / push-off switch. 
Back Panel Controls and Connectors 
Power Cord Connector:  
Standard 15 amp IEC inlet. A circuit breaker located near the IEC 
power inlet protects the amplifier from excessive AC current draw. 
Reset Switch:  
Resets the circuit breaker that protects the power supply. 
Ventilation Grille:  
Air flow is front to back. Do not block the ventilation grilles. 
  
Appendix B: Product Data Sheets
119
Appendix B: Product Data Sheets
120
Appendix B: Product Data Sheets
121
14 15
Model
Photos
Diameter
Standards (IEC61672)
Microphone
Optimized
Preamplifier
Frequency Response (Hz)
Open-circuit Sensitivity (mV/Pa) (±2dB)
Output Impedance (Ω)
Dynamic Range (dBA)
Inherent Noise (dBA)
Operating Temperature (°C)
Operating Humidity (RH)
Temperature Coefficient (dB/°C)
Humidity Coefficient (dB/%RH)
Pressure Coefficient (250 Hz) (dB/kPa)
Length (mm)
Input Connector
Corresponding Model with TEDS
1/4″
Class I
Integrated
Free Field
Integrated
20 ~ 20k
12.5
< 110
35 ~ 130
< 35
-10 ~ 50
0 ~ 95%
23°C(15 ~ 35 °C): < ±0.3 dB; 0 ~ 40 °C:
 < ±1.5 dB-10 ~ 50 °C: < ±3.0 dB; 
with 1000Hz, at reference 
temperature 23 °C
20% ~ 90% RH: < ±0.8 dB
with 1000Hz, at reference 
temperature 23 °C, 
Humidity 50% RH
-0.06
24
SMB
--
1/2″
Class I
MP201
Free Field
MA231(TEDS optional)
20 ~ 20k
45
< 50
16 ~ 134
< 16
-30 ~ 80
0 ~ 95%
0.005
0.003
-0.004
91
BNC
MPA261
1/2″
Class I
MP231
Free Field
MA231(TEDS optional)
20 ~ 20k
40
< 50
17 ~ 136
< 17
-30 ~ 80
0 ~ 95%
0.005
0.003
-0.004
91
BNC
MPA271
MPA201 MPA231 MPA215 MPA401MPA416* MPA436* MPA418
MPA Series Microphones
MPA201 / MPA231 / MPA215 / MPA416 / MPA436 / MPA418 / MPA401
*The MPA416 & MPA436 are the most suitable models for array uses. The frequency responses meet the IEC 61672 Class 1 requirements.
1/2″
Class  II
MP215
Free Field
MA231(TEDS optional)
20 ~ 12.5k
40
< 110
23 ~ 135
< 23
-20 ~ 80
0 ~ 95%
<± 0.3 dB (0 ~ 40 °C) 
with 250Hz, at reference 
temperature 23 °C
0.007
-0.03
91
BNC
MPA265
1/4″
Class II
MP418
Free Field
MA418
20 ~ 16k
10
< 110
36 ~ 135
< 36
0 ~ 40
0 ~ 98%
<± 0.6 (0 ~ 40 °C)
 at reference 
temperature 23 °C
0.015
-0.06
64
SMB
--
1/4″
Class I
MP401
Free Field
MA401
20 ~ 70k
5
< 110
35 ~ 155
< 35
-20 ~ 80
0 ~ 98%
-0.009
0.003
-0.007
67
SMB
--
1/4″
Class I
Integrated
Free Field
Integrated
20 ~ 20k
50
< 110
29 ~ 127
< 29
-10 ~ 50
0 ~ 95%
15 ~ 35 °C: < ±0.3 dB; 0 ~ 40 °C: 
< ±1.5 dB; -10 ~ 50 °C: < ±3.0 dB;
with 1000Hz, at reference 
temperature 23 °C
20% ~ 90% RH: < ±0.8 dB
with 1000Hz, at reference 
temperature 23 °C, 
Humidity 50% RH
-0.06
61
SMB
MPA466
Appendix B: Product Data Sheets
122
18 19
Model
Photos
Number of Input Channels
Connector of Input
Number of Output Channels
Connector of Output
Frequency Response (Hz)
Gain
Polarization Voltage
Output Power for Preamplifier
Power Supply
Filter
Operating Temperature (°C)
Operating Humidity (RH)
Dimension (mm)
Weight (g)
2
BNC
2
BNC
5 ~ 200k
--
0 V
4 mA
1 × 9 V Battery or 220 V
--
-10 ~ 50
0 ~ 95%
113 × 70 × 45
160
1
BNC
1
BNC
5 ~ 200k
× 1 , × 10
0 V
4 mA
220 V
--
-10 ~ 50
0 ~ 95%
113 × 70 × 45
160
4
BNC
4
BNC
5 ~ 200k
× 0.1, × 1 , × 10
0 V
4 mA
220 V
--
-10 ~ 50
0 ~ 95%
310 × 250 × 65
1500
1
7-pin LEMO
1
BNC
1 ~ 1M
--
200 V or 0 V
28 V
220 V
--
-10 ~ 50
0 ~ 95%
113 × 70 × 45
190
2
7-pin LEMO
2
BNC
5 ~ 200k
-20 dB, 0 dB, 20 dB, 40dB
200 V or 0 V
28 V or 120 V
220 V
Lin, AW, HP, Dir
-10 ~ 50
0 ~ 95%
260 × 60 × 150
1130
MC102 MC141 MC104 MC711 MC722
Microphone Conditioning Units
MC102 / MC141 / MC104 / MC711 / MC722
 BSWA
Appendix B: Product Data Sheets
123
25
Sound Calibrators
 CA111 / CA114 / CA115
SPECIFICATIONS
Sound Calibrators
CA111/CA114/CA115 is small sound source 
for calibrating measurement microphones, sound 
level meters, and other sound measurement 
equipments. The calibrator can be used on 1/2-inch 
and 1/4-inch microphones with adaptor. 
CA111 conforms to IEC 60942:2003 Class 1, ANSI 
S1.40-1984 and GB/T 15173-1994.
CA114/115 conforms to IEC 60942:2003 Class 2 
standards.
APPLICATIONS
Calibration of measurement microphones, sound 
level meters, and other sound measurement 
equipments.
Checking the linearity of equipments.
FEATURES
Conforms to IEC60942:2003 Class 1/Class 2, ANSI 
S1.40-1984, and GB/T 15173-1994.
1 kHz calibration frequency for all weighting 
networks.
CA111: Dual 94 & 114 dB sound pressure level 
outputs.
CA114: 94 dB sound pressure level outputs.
CA115: 114 dB sound pressure level outputs.
Calibration accuracy ± 0.3 dB. 
Designed with highly stable level and frequency.
CA111 for two-keypad operation and CA114/115 for 
one-keypad operation.
Fits 1/2” microphones and 1/4” microphones with 
adaptor.
Powered by 2×AAA battery and automatic power off 
to conserve battery life.
Model
Standard
Sound Pressure Level
Frequency
Microphone Diameter
Harmonic Distortion
Equivalent Free-field Level
Equivalent Random Incidence Level
Reference Conditions
Environmental Conditions
Power Supply
Dimension(mm)
Weight
CA111
IEC60942:2003 Class 1, ANSI S1.40-1984, 
GB/T 15173-1994
94.0 dB ±0.3 dB and 114.0 dB ±0.3 dB
CA114/CA115
IEC60942:2003 Class 2, ANSI 
S1.40-1984, GB/T 15173-1994
94.0dB/114.0 dB ±0.3 dB
1000Hz ±0.5%
According to IEC61094-4: 1/2” & 1/4”
<2% Stabilization Time: <10 s
-0.2 dB for 1/2” Microphones
+0.0 dB for 1/2”, 1/4”
Ambient Temperature: 25°C (77°F)  / Ambient Pressure: 101.3 kPa  
/ Humidity: 55% RH / Effective Load Volume: 250 mm3
Temperature: -10°C-50°C (14°F -122°F)
Pressure: 65 kPa to 108 kPa
Humidity: 10 to 90%RH (non-condensing)
Batteries: 1.5 V LR6 (AA battery) × 2
Lifetime: Typically 40 hours with alkaline batteries at 25°C (77°F)
48 × 70 × 70mm
180 g, including batteries
Appendix B: Product Data Sheets
124
Digital Storage Oscilloscopes
TDS2000C Series Datasheet
The TDS2000C Digital Storage Oscilloscope Series provides you with
affordable performance in a compact design. Packed with standard
features - including USB connectivity, 16 automated measurements, limit
testing, data logging, and context-sensitive help - the TDS2000C Series
oscilloscopes help you get more done in less time.
Key performance specifications
200 MHz, 100 MHz, 70 MHz, 50 MHz bandwidth models
2- and 4-channel models
Up to 2 GS/s sample rate on all channels
2.5k point record length on all channels
Advanced triggers including pulse width trigger and line-selectable
video trigger
Key features
16 automated measurements and FFT analysis for simplified waveform
analysis
Built-in waveform limit testing
Automated, extended data logging feature
Autoset and signal auto-ranging
Built-in context-sensitive help
Probe check wizard
11-language user interface
144 mm (5.7 inch) active TFT color display
Small footprint and lightweight - only 124 mm (4.9 inches) deep and
2 kg (4.4 lb)
USB 2.0 host port on the front panel for quick and easy data storage
USB 2.0 device port on the rear panel for easy connection to a PC or
for direct printing to a PictBridge® -compatible printer
Includes National Instrument's LabVIEW SignalExpress™ TE Limited
Edition and Tektronix OpenChoice® Software for connecting to your
bench
Lifetime warranty. Limitations apply. For terms and conditions, visit
www.tektronix.com/lifetimewarranty
Digital precision for accurate measurements
With up to 200 MHz bandwidth and 2 GS/s maximum sample rate, no other
digital storage oscilloscope offers as much bandwidth and sample rate for
the price. Tektronix proprietary sampling technology provides real-time
sampling with a minimum of 10X oversampling on all channels, all the time
to accurately capture your signals. Sampling performance is not reduced
when using multiple channels.
Critical tools for troubleshooting your device
Advanced triggers - rising/falling edge, pulse width, and video - help you
quickly isolate your signals of interest. Once you've captured a signal,
advanced math capabilities and automated measurements can speed your
analysis. Quickly perform an FFT or add, subtract, or multiply waveforms.
Sixteen automated measurements quickly and reliably calculate important
signal characteristics such as frequency or rise time, while the built-in Limit
Test function enables you to easily identify problems in your signal.
Quickly and easily capture waveforms with advanced triggering.
www.tektronix.com    1
Appendix B: Product Data Sheets
125
Specifications
All specifications apply to all models unless noted otherwise.
Overview
TDS2001C TDS2002C TDS2004C TDS2012C TDS2014C TDS2022C TDS2024C
Display (QVGA LCD) TFT on all models
Bandwidth 50 MHz 70 MHz 70 MHz 100 MHz 100 MHz 200 MHz 200 MHz
Channels 2 2 4 2 4 2 4 
External trigger input Included on all models
Sample rate on each channel 500 MS/s 1.0 GS/s 1.0 GS/s 2.0 GS//s 2.0 GS/s 2.0 GS/s 2.0 GS/s
Vertical system
Record length 2.5k points at all time bases on all models
Vertical resolution 8 bits
Vertical sensitivity 2 mV to 5 V/div on all models with calibrated fine adjustment
DC vertical accuracy ±3% on all models
Vertical zoom Vertically expand or compress a live or stopped waveform
Maximum input voltage 300 VRMS CAT II; derated at 20 dB/decade above 100 kHz to 13 Vp-pAC at 3 MHz
Position range 2 mV to 200 mV/div +2 V;
>200 mV to 5 V/div +50 V
Bandwidth limit 20 MHz for all models
Input impedance 1 MΩ in parallel with 20 pF
Input coupling AC, DC, GND on all models
Horizontal system
Time base accuracy 50 ppm
Horizontal zoom Horizontally expand or compress a live or stopped waveform
TDS2000C Digital Storage Oscilloscopes
www.tektronix.com    5
Appendix B: Product Data Sheets
126
Trigger system
Trigger modes Auto, Normal, Single Sequence
Trigger types
Edge (rising/falling) Conventional level-driven trigger. Positive or negative slope on any channel. Coupling selections: AC, DC, Noise Reject, HF
Reject, LF Reject
Video Trigger on all lines or individual lines, odd/even or all fields from composite video, or broadcast standards (NTSC, PAL, SECAM)
Pulse width (or glitch) Trigger on a pulse width less than, greater than, equal to, or not equal to, a selectable time limit ranging from 33 ns to 10 s
Trigger source
2-channel models CH1, CH2, Ext, Ext/5, AC Line
4-channel models CH1, CH2, CH3, CH4, Ext, Ext/5, AC Line
Trigger view Displays the trigger signal while the Trigger View button is depressed
Trigger signal frequency readout Provides a frequency readout of the trigger source
Acquisition system
Acquisition modes
Peak detect High-frequency and random glitch capture. Captures glitches as narrow as 12 ns (typical) at all time base settings from 5 µs/div to
50 s/div
Sample Sample data only
Average Waveform averaged, selectable: 4, 16, 64, 128 
Single sequence Use the Single Sequence button to capture a single triggered acquisition sequence
Roll mode At acquisition time base settings of >100 ms/div
Waveform measurements
Automatic waveform
measurements
Period, Frequency, +Width, -Width, Rise Time, Fall Time, Max, Min, Peak-to-Peak, Mean, RMS, Cycle RMS, Cursor RMS, Duty
Cycle, Phase, Delay
Cursors
Types Amplitude and time
Measurements ΔT, 1/ΔT (frequency), ΔV
Waveform math
Operators Add, Subtract, Multiply, FFT
Sources
2-channel models CH1 - CH2, CH2 - CH1, CH1 + CH2, CH1 x CH2
4-channel models CH1 - CH2, CH2 - CH1, CH3 - CH4, CH4 - CH3, CH1 + CH2, CH3 + CH4, CH1 x CH2, CH3 x CH4
FFT Windows: Hanning, Flat Top, Rectangular
2,048 sample points
Datasheet
6   www.tektronix.com
Appendix B: Product Data Sheets
127
Specifications subject to change without notice.
Copyright © 2008-2011 Extech Instruments Corporation. All rights reserved including the right of reproduction in whole or in part in any form.
www.extech.com 9/19/11 - R1
Telescoping probe is designed to fit into small openings
And measures airflow down to 40ft/min (0.2m/s)
Heavy Duty Hot Wire Thermo-Anemometer
Ordering Information:
407123 ........................Heavy Duty Hot Wire Thermo-Anemometer
407123-NIST ................Heavy Duty Hot Wire Thermo-Anemometer w/ Calibration Traceable to NIST.
407001 ........................Data Acquisition Software and Serial Cable
407001-USB ................USB Adaptor for 407001
380340 ........................Battery Operated Datalogger
153117 ........................117V AC Adaptor
153220 ........................220V AC Adaptor
Features:
• Telescoping probe is ideal for measuring in
HVAC ducts and other small vents;
extends up to 7ft (2.1m) maximum length with cable
• Super large 1.4" (36mm) dual LCD display
• MAX/MIN, Data Hold
• Optional Data Acquisition software (407001)
and Datalogger (380340);
• Complete with telescoping probe with cable,
six AAA batteries and protective holster.
NOTE: AC Adaptor not available for this model.
Specifications Resolution Basic Accuracy
Air Velocity
0.2 to 20m/s 0.1m/s ±3%
40 to 3940ft/min 10ft/min
0.5 to 45MPH 0.1MPH
1.0 to 31knots: 0.1knots
0.7 to 72km/h 0.1km/h
Temperature & Windchill
0 to 50°C 0.1° ±0.8°C
32 to 122°F 0.1° ±1.5°F
Dimensions 7 x 2.9 x 1.3" (178 x 74 x 33mm)
Weight 17oz (482g)
Optional Data Acquisition Software Optional Battery Operated Datalogger
Appendix B: Product Data Sheets
128
1when it’s too fast to see, and too important not to.®
DATA SHEET
For the most current version visit www.visionresearch.com
Subject to change Rev September 2013
Key Benefits:
WHEN IT’S TOO FAST TO SEE, AND TOO IMPORTANT NOT TO®
The Phantom v641 is the second generation v640 camera. It smaller and lighter 
than its predecessor and has a number of new convenience features requested 
by users. 
The v641 provides a 4 megapixel sensor and greater than 6 gigapixels/second 
throughput. That means full-resolution frame rates of 1450 frames-per-second 
(fps), and 1920 x 1080 HD-resolution frame rates of 2560 fps. The minimum 
frame rate is 10 fps.
Take the wide view with our custom-designed 2560 x 1600 pixel CMOS 
sensor. The aspect ratio of the v641 allows you to keep moving targets in-frame 
longer and see more of the event you are recording. 
Key Features:
10-1450 frames-per-second (fps) at full resolution. 
Maximum FPS: 219,000 @ 256 x 8
2560 x 1600 CMOS sensor
Minimum Exposure (shutter speed): 1 μs
High-resolution timing system: Better than 20 ns resolution
Extreme Dynamic Range (EDR): two different exposures 
within a single frame
Internal Shutter Mechanism: hands-free/remote current 
session reference (CSR)
Memory Segmentation: Up to 63 segments
Non-volatile, hot-swappable Phantom CineMag 
memory magazines (128 GB, 256 GB & 512 GB)
CineMag to CineStation®
Range Data input
Built-in Memory: 8 GB, 16 GB, 32 GB
Breakthrough Sensitivity: ISO (ISO-12232 SAT Method) 
Mono: 16,000 T and 6400 D  
Color: 1600 T and 1600 D  
QE 60% peak; NEP 0.011 fJ
Pixel Bit-depth: 12-bit
Gb Ethernet, 10 Gb Ethernet with optional CineStream 
X2SR module
Image-Based Auto-Trigger
Burst Mode
IRIG & SMPTE Time Code
Genlock
2560 x 1600 resolution
10-1450 fps at full resolution
Breakthrough sensitivity
Phantom CineMag® compatible
v641
Phantom v641
Shown with optional CineMag interface 
and On-Camera Controls
Appendix B: Product Data Sheets
129
2DATA SHEET
v641
when it’s too fast to see, and too important not to.®
Shutter speeds down to 1 microsecond and a global electronic shutter allow 
for crisp, sharp images with little or no image blur or motion artifacts.
With a peak quantum efficiency (QE) of 60% – greatly improved over current 
sensor designs – and a significant reduction in readout noise, along with the 
addition of microlens technology, the v641’s four megapixel resolution can 
be used to full advantage at speeds that normally called for large-pixel, lower 
resolution cameras.
That makes the v641 ideal for applications where high sensitivity and high 
resolution are needed. Coupled with a 1.4 microseconds straddle time the v641 
is ideal for PIV applications, for example.
Each camera supports 12-bit pixel depth. Smaller bit-depth gives you more 
recording time and smaller files. Greater bit-depth gives you more gray levels and 
finer detail. With the greater latitude of 12 bits, you can pull more detail out of the 
image.
The v641’s high-resolution timing system yields a timing resolution of 
better than 20 nanoseconds. Frame rate, frame synchronization and exposure 
accuracy are all improved over previous generations of high-speed cameras. 
And, an external frame synchronization signal is available via a dedicated BNC 
for easier cabling and increased signal integrity. A GenLock input is available for 
synchronizing the playback of recorded cines to other video gear.
Of course, the v641 offers our unique Extreme Dynamic Range (EDR) feature 
giving you the ability to get two different exposures within a single frame. 
And, with auto exposure, the camera adjusts to changing lighting conditions 
automatically.
There is an internal shutter for cutting off all light to the sensor when doing a 
session-specific black reference (CSR). You now can do remote CSRs through 
software control without the need to manually cover the lens! 
The v641 comes standard with 8 GB of high-speed dynamic RAM, but you can 
order 16 GB or 32 GB versions. Our segmented memory allows you to divide 
this into up to 63 segments so you can take multiple shots back-to-back without 
the need to download data from the camera.
You are able to record directly to our Phantom CineMag non-volatile, 
hot-swappable memory magazines. They mount on the CineMag compatible 
version of the camera. Continuously record full-resolution cines into 
Phantom v641
provides a  
4 megapixel 
sensor and 
greater than  
6 gigapixels/ 
second 
throughput.
Appendix B: Product Data Sheets
130
3non-volatile memory at up to 195 fps (360 fps for 1920 x 1080). That’s about 
4.5 minutes of continuous recording into the 256 GB CineMag or 9 minutes
into the 512 GB CineMag.
Or, record at higher speeds into camera RAM, then manually or automatically 
move your cine to the CineMag. If you need to take multiple shots back-to-
back, you don’t have to wait for a time-consuming download of camera memory 
over Ethernet. Instead, just upload the camera memory to a CineMag at about  
800 megapixels/second, then take your next shot! 
With CineMag storage you get maximum data protection and an ideal storage 
medium for secure environments.
Move the CineMag from the camera to a CineStation connected to a PC and 
view, edit, and save your cines using the Phantom Software supplied with the 
camera. 
Keep them in their original cine raw format, or convert them to TIFF, QuickTime, 
AVI, or a number of other formats. Move the files from the CineStation to a disk 
or tape deck via 10 Gb Ethernet, dual HD-SDI, or Component Video outputs.
When used on a tracking mount, elevation and azimuth data can be transferred 
to the camera and associated with image frames through our unique Range 
Data input.
View your recordings immediately. There are two Versatile Dual HD-SDI 
ports that can be used in one of four different modes: 2 identical 4:2:2 outputs; 
1 dual HD-SDI 4:4:4 output; independent 4:2:2 outputs where one is live and 
one is playback; or 4:4:4 playback on the dual HD-SDI while you have a live 
image on the component viewfinder. Yes, a component video viewfinder port has 
been added so any viewfinder compatible with our Phantom HD camera can now 
be used with the v641.
The v641 is controlled by the feature-rich Phantom Software. If you’ve used any 
Phantom camera before, you will know how to run the v641. As an option, you 
can add On-Camera Controls (OCC) to get full control of the camera without 
the need to connect to a PC. We also provide a full-featured Remote Control 
Unit (RCU) for wired or wireless control.
The v641 comes in two base models, either with or without a CineMag interface. 
An optical low-pass filter is available as an option.
H V FPS*
256 8 219,200
256 64 90,200
256 128 53,900
256 256 29,800
512 384 16,200
512 512 12,300
640 480 10,700
800 600 7,370
1280 720 5,350
1280 800 4,820
1280 1024 3,780
1920 1080 2,560
2048 1024 2,700
2048 1600 1,730
2560 1600 1,450
*Typical results
ISO SAT
Mono Color
ISO SAT T ISO SAT D ISO SAT T ISO SAT D
16,000 6400 1600 1600
Appendix B: Product Data Sheets
131
MS-TS Precision Balances
Engineered for Reliable Performance 
The robust construction of the MS-TS precision  
balances makes them perfect for heavier tasks in 
the laboratory or out on the factory floor. The unique 
MonoBloc weighing cell delivers the accuracy you need 
and is fully protected against accidental overload.
Delivering consistently reliable results, even in harsh 
environments, these balances will also deliver a fast 
return on your investment.
The 7” extra-large color TFT touchscreen display is 
operable through cotton, silicon and rubber gloves. 
An intuitive user interface and 18mm high digits bring 
comfort to your daily tasks. 
Metal housing ensures  
long balance lifetime
The full die-cast aluminum housing not 
only protects the weighing cell from 
environmental influences and impacts,  
it is also resistant to harsh chemicals, 
including acetone.
Proven weighing cell  
delivers reliable results
Our renowned MonoBloc weighing  
cell, with proFACT automatic internal 
adjustment, delivers consistently reliable 
results. Built-in overload protection 
ensures a long balance lifetime. 
MinWeigh function assists  
dosing process
During weighing-in, the weight value 
remains red until the net sample weight 
is above the pre-programmed minimum 
value. It is clear to see when process 
tolerances have been met.
Built-in function simplifies  
balance leveling
The built-in LevelControl function issues 
a warning when the balance is not level 
and provides onscreen guidance to help 
you level the balance correctly within 
seconds. 
Trusted Results at Your Fingertips
Worry-free Weighing, 0.01 – 0.1 g
M
S-
TS
 P
re
ci
si
on
Appendix B: Product Data Sheets
132
MS-TS Precision Balances, 0.01 – 0.1 g
Technical Specifications MS1602TS MS3002TS MS4002TS MS4002TSDR MS6002TS MS6002TSDR MS12002TS MS8001TS
Limit values    
Maximum capacity 1620 g 3200 g 4200 g 4200 g 6200 g 6200 g 12200 g 8200 g
Maximum capacity, fine range – – – 820 g – 1220 g –   –
Readability 0.01 g 0.01 g 0.01 g 0.1 g 0.01 g 0.1 g 0.01 g 0.1 g
Readability, fine range – – – 0.01 g – 0.01 g   –   –
Repeatability 0.01 g 0.01 g 0.01 g 0.06 g 0.01 g 0.06 g 0.01 g 0.1 g 
Repeatability, fine range – – – 0.01 g – 0.01 g  –  –
Linearity deviation 0.02 g 0.02 g 0.02 g 0.08 g 0.02 g 0.08 g 0.025 g 0.2 g
Typical values
Repeatability 0.007 g 0.007 g 0.007 g 0.05 g 0.007 g 0.05 g 0.007 g 0.07 g
Repeatability, fine range – – – 0.007 g – 0.007 g  –  –
Linearity deviation 0.006 g 0.006 g 0.006 g 0.06 g 0.006 g 0.06 g 0.008 g 0.06 g
Sensitivity offset (test weight)
0.018 g  
(1600 g)
0.018 g  
(3000 g)
0.024 g  
(4000 g)
0.024 g  
(4000 g)
0.036 g 
(6000 g)
0.036 g  
(6000 g)
0.048 g  
(12000 g)
0.24 g  
(8000 g)
USP minimum sample weight  
(5% load, k=2, U=0.10%)
14 g 14 g 14 g 14 g 14 g 14 g 14 g 120 g
Minimum sample weight  
(5% load, k=2, U=1%)
1.4 g 1.4 g 1.4 g 1.4 g 1.4 g 1.4 g 1.4 g 12 g
Settling time 1.5 s 1.5 s 1.5 s 1.5 s 1.5 s 1.5 s 1.5 s 1 s
Dimensions       
Weighing pan size, W×D (mm) 170 × 200 170 × 200 170 × 200 170 × 200 170 × 200 170 × 200 170 × 200 190 × 226
For more informationwww.mt.com/GWP
GWP®
 Good Weighing Practice™
Mettler-Toledo AG 
Laboratory Weighing 
CH-8606 Greifensee, Switzerland 
Tel. +41 44 944 22 11 
Fax +41 44 944 30 60 
Subject to technical changes 
© 01/2015 Mettler-Toledo AG 
30248772
Global MarCom Switzerland 
www.mt.com/ms-precision
Embedded applications 
Weighing, statistics for all appli-
cations, check weighing, totaling, 
piece counting, formulation, percent 
weighing, factor weighing, dynamic 
weighing, density, dosing and PC 
Direct (for easy data transfer).
Features
Ac
cu
ra
te
  
Re
su
lts
MonoBloc weighing cell
Strong overload protection
proFACT inernal adjustment
MinWeigh warning function
Full metal housing
Ef
fic
ie
nt
  
O
pe
ra
tio
n
7” Extra-large color TFT  
touchscreen
18mm high digits
LevelControl function
Statistical data analysis
Easy cleaning
LevelLock
Q
ua
lit
y 
As
su
ra
nc
e ISO-Log
Sample ID input
Passcode protection
Se
am
le
ss
  
Pr
oc
es
s 3 interfaces: USB device, 
USB host, RS232
Bluetooth option
PC Direct application
Accessories
The large draft shield protects your 
balance from air currents. Get faster 
and more accurate results.
External draft shield
Wirelessly sends data between the 
balance and a PC, tablet or printer. 
No additional software needed.
Bluetooth adaptors
Self-adhesive sheet protects your 
balance and absorbs minor spills. 
Peel off and replace as required.
Pan protectors
Fast, high quality printouts on 
paper, self-adhesive labels and 
continuous self-adhesive paper 
(including barcodes).
P-5x thermal printers
For further information on  
acces sories, please visit 
www.mt.com/lab-accessories
Appendix B: Product Data Sheets
133
Bibliography
[1] P.J. DiNenno and G.M. Taylor. Halon and halon replacement agents and
systems. In A.E. Cote, editor, Fire Protection Handbook, pages 17.93–17.121.
National Fire Protection Association, Quincy, MA, twentieth edition, 2008.
[2] S. McCormick, M. Clauson, and H. Cross. Us army ground vehicle crew
compartment halon replacement program. In Halon Options Technical Working
Conference, pages 229–236, May 2000.
[3] R.M. Gagnon. Design of Special Hazard and Fire Alarm Systems. Thomson
Delmar Learning, Clifton Park, NY, second edition, 2008.
[4] G.M. Whiteside. Instant flame suppression. Technical report, United States
Department of Defense, 2013.
[5] F.A. Williams. Combustion Theory. Benjamin/Cummings Publishing Com-
pany, Menlon Park, CA, 2 edition, 1985.
[6] J.G. Quintiere and A.S. Rangwala. A theory for flame extinction based on
flame temperature. Fire and Materials, 28:387–402, 2004.
[7] Y. Jin and B.D. Shaw. Computational modeling of n-heptane droplet combus-
tion in air–diluent environments under reduced-gravity. International Journal
of Heat and Mass Transfer, 53(25–26):5782–5791, 2010.
134
Bibliography
[8] H.Z. Yu and J.S. Newman. Theories of fire extinguishment. In A.E. Cote,
editor, Fire Protection Handbook, pages 2.80–2.92. National Fire Protection
Association, Quincy, MA, twentieth edition, 2008.
[9] S.H. Won, S. Dooley, F. L. Dryer, and Yiguang Ju. A radical index for the
determination of the chemical kinetic contribution to diffusion flame extinction
of large hydrocarbon fuels. Combustion and Flame, 159(4):541–551, 2012.
[10] F. Kreith, S.A. Berger, S.W. Churchill, J.P. Tullis, F.M. White, A.T. McDon-
ald, A. Kumar, J.C. Chen, T.F. Irvine, M. Capobianchi, F.E. Kennedy, E.R.
Booser, D.F. Wilcock, R.F. Boehm, R.D. Reitz, S.A. Sherif, and B. Bhushan.
Fluid mechanics. In F. Kreith, editor, Mechanical Engineering Handbook. CRC
Press, Boca Raton, FL, 1999.
[11] K.K. Kuo and R. Acharya. Fundamentals of Turbulent and Multiphase Com-
bustion. John Wiley & Sons, Hoboken, NJ, 2012.
[12] K.W. Jenkins, M. Klein, N. Chakraborty, and R.S. Cant. Effects of strain
rate and curvature on the propagation of a spherical flame kernel in the
thin-reaction-zones regime. Combustion and Flame, 145:415–434, 2006.
[13] S.M. Candel and T.J. Poinsot. Flame stretch and the balance equation for
the flame area. Combustion Science and Technology, 70(1–3):1–15, 1990.
[14] A. Lin˜a´n. The asymptotic structure of counterflow diffusion flames for large
activation energies. Acta Astronautica, 1:1007–1039, 1974.
[15] B. Karlovitz, D.W. Denniston, D.H. Knapschaefer, and F.E. Wells. Fourth sym-
posium (international) on combustion studies on turbulent flames. Symposium
(International) on Combustion, 4(1):613–620, 1953. ISSN 0082-0784.
[16] V.R. Katta, T.R. Meyer, M.S. Brown, J.R. Gord, and W.M. Roquemore.
Extinction criterion for unsteady, opposing-jet diffusion flames. Combustion
and Flame, 137(1–2):198–221, 2004. ISSN 0010-2180.
135
Bibliography
[17] J.G. Quintiere. Fundamentals of Fire Phenomena. John Wiley & Sons, West
Sussex, England, 2006.
[18] V.R. Lecoustre, P. Narayanan, H.R. Baum, and A. Trouve. Local extinction
of diffusion flames in fires. Proceedings of the Tenth International Symposium
on Fire Safety Science, pages 583–596, 2011.
[19] S.R. Turns. An Introduction to Combustion. McGraw-Hill, New York, NY, 3
edition, 2007.
[20] S. McAllister, J.Y. Chen, and A.C. Fernandez-Pello. Fundamentals of Com-
bustion Processes. Springer, New York, NY, 2011.
[21] J.S. Kim and F.A. Williams. Contribution of strained diffusion flames to
acoustic pressure response. Combustion and Flame, 98:279–299, 1994.
[22] W.C. Strahle. A theoretical study of droplet burning: Transient and periodic
solutions. Technical report, National Aeronautics and Space Administration,
1963.
[23] D.J. McKinney and D. Dunn-Rankin. Acoustically driven extinction in a
droplet stream flame. Combustion Science and Technology, 161:24–48, 2000.
[24] C.I. Sevilla-Esparza, J. L. Wegener, S. Teshome, J. I. Rodriguez, O. I. Smith,
and A. R. Karagozian. Droplet combustion in the presence of acoustic excita-
tion. Combustion and Flame, 161:1604–1619, 2014.
[25] S.K. Kim, H.S Choi, H.J. Kim, Y.S. Ko, and C.H. Sohn. Finite element analysis
for acoustic characteristics of combustion stabilization devices. Aerospace
Science and Technology, 42:229 – 240, 2015. ISSN 1270-9638.
[26] J. O’Connor, V. Acharya, and T. Lieuwen. Transverse combustion instabili-
ties: Acoustic, fluid mechanic, and flame processes. Progress in Energy and
Combustion Science, 49:1–39, 2015. ISSN 0360-1285.
136
Bibliography
[27] A. Miglani, S. Basu, and R. Kumar. Suppression of instabilities in burning
droplets using preferential acoustic perturbations. Combustion and Flame,
161(12):3181–3190, 2014. ISSN 0010-2180.
[28] H.Y. Wang, J.K Bechtold, and C.K. Law. Forced oscillations in diffusion
flames near diffusive-thermal resonance. International Journal of Heat and
Mass Transfer, 51:630–639, 2008.
[29] S. Kartheekeyan and S.R. Chakvarthy. An experimental investigation of an
acoustically excited laminar premixed flame. Combustion and Flame, 146:
513–529, 2006.
[30] Q. Wang, H.W. Huang, H.J. Tang, M. Zhu, and Y. Zhang. Nonlinear response
of buoyant diffusion flame under acoustic excitation. Fuel, 103:364–372, 2013.
[31] Y. Hardalupas and A. Selbach. Imposed oscillations and non-premixed flames.
Progress in Energy and Combustion Science, 28:75–104, 2002.
[32] K. Deng, Z. Shen, M. Wang, Y. Hu, and Y.J. Zhong. Acoustic excitation
effect on {NO} reduction in a laminar methane-air flame. Energy Procedia,
61:2890–2893, 2014. ISSN 1876–6102. International Conference on Applied
Energy, {ICAE2014}.
[33] L.W. Chen, Q. Wang, and Y. Zhang. Flow characteristics of diffusion flame
under non-resonant acoustic excitation. Experimental Thermal and Fluid
Science, 45:227–233, 2013.
[34] Li-Wei Chen. Application of piv measurement techniques to study the charac-
teristics of flame–acoustic wave interactions. Flow Measurement and Instru-
mentation, 45:308–317, 2015. ISSN 0955-5986.
[35] F. Nicoud and T. Poinsot. Thermoacoustic instabilities: Should the rayleigh
criterion be extended to include entropy changes? Combustion and Flame,
142:153–159, 205.
137
Bibliography
[36] S. Nair. Acoustic characterization of flame blowout phenomenon. PhD thesis,
Georgia Institute of Technology, School of Aerospace Technology, May 2006.
[37] W.W. Seto. Schaum’s Outline of Theory and Problems of Acoustics. McGraw-
Hill, New York, NY, 1st edition, 1971.
[38] C.K. Law. Combustion Physics. Cambridge University Press, New York, NY,
2006.
[39] A.C. Fernandez-Pello, S.R. Ray, and I. Glassman. Eighteenth symposium
(international) on combustion flame spread in an opposed forced flow: the effect
of ambient oxygen concentration. Symposium (International) on Combustion,
18(1):579 – 589, 1981. ISSN 0082-0784.
[40] J.N. De Ris. Spread of a laminar diffusion flame. Symposium (International)
on Combustion, 12(1):241 – 252, 1969. ISSN 0082-0784.
[41] L. Hu, S. Liu, Y. Xu, and D. Li. A wind tunnel experimental study on burning
rate enhancement behavior of gasoline pool fires by cross air flow. Combustion
and Flame, 158(3):586 – 591, 2011. ISSN 0010-2180.
[42] M.A. Fox and J. K. Whitesell. Core Organic Chemistry. Jones and Bartlett,
Sudbury, MA, 1997.
[43] M. Matalon. Lecture 8: Flame stretch and lewis number effects. https://www.
princeton.edu/cefrc/Files/201120Lecture20Notes/Matalon/Lecture-8.pdf,
2011.
[44] H.W. Emmons. The film combustion of liquid fuel. Z. angew. Math. Mech.,
36:60–71, 1956.
[45] V. Raghavan, A.S. Rangwala, and J.L. Torero. Laminar flame propagation on
a horizontal fuel surface: Verification of classical emmons sslution. Combustion
Theory and Modelling, 13(1):121–141, 2009.
138
Bibliography
[46] D.J. Rashbash. A flame extinction criterion for fire spread. Combustion and
Flame, 26:411–412, 1976.
[47] A. Tewarson. Generation of heat and chemical compounds in fires. In SFPE
Handbook. National Fire Protection Association, Quincy, MA, third edition,
2002.
[48] L. Hu, S. Liu, and L. Wu. Flame radiation feedback to fuel surface in medium
ethanol and heptane pool fires with cross air flow. Combustion and Flame,
160(2):295–306, 2013.
[49] NIST chemistry webbook. http://webbook.nist.gov/chemistry/. Accessed:
2015-07-12.
[50] R. Arnold and W.E. Anderson. Droplet burning jp-8/silica gels. 48th AIAA
Aerospace Sciences Meeting, 48(421), 2010.
[51] T. Edwards, B. Harrison, and L. Maurice. Properties and usage of air force
fuel: Jp-8. In 39th Aerospace Sciences Meeting and Exhibit, AIAA, 2001,
volume 39 of Aerospace Sciences Meeting. American Institute of Aeronautics
and Astronautics, August 2001.
[52] T.J. Bruno, M. Huber, A. Laesecke, E. Lemmon, M. McLinden, S.L. Outcalt,
R. Perkins, B.L. Smith, and J.A. Widegren. Thermodynamic, transport,
and chemical properties of “reference” JP-8. Technical Report NISTIR 6659,
National Institute of Standards and Technology, July 2010.
[53] Defense Technical Information Center. DTIC-132106 (unclassified): Aviation
fuel properties. Technical report, United States Department of Defense, 1983.
[54] Exxon Mobile. World jet fuel specification. Technical report, 2005.
[55] MIL-DTL-83133E Detail specification; turbine fuels, aviation, kerosene types,
nato f-34 (jp-8), nato f-35, and jp-8+100. Technical report, United States
Department of Defense, 2005.
139
Bibliography
[56] D.B. Spalding. The combustion of liquid fuels. Proc. 4th Internation Sympo-
sium on Combustion, pages 847–864, 1953.
[57] F.P. Incropera, David P. Dewitt, and et al. Fundamentals of Heat and Mass
Transfer. John Wiley and Sons, Danvers, MA, 6 edition, 2007.
[58] A.F. Miles. Basic Heat and Mass Transfer. Irwin, Inc., Chicago, IL, 1995.
[59] J.F. Brady. Nusselt numbers of laminar, oscillating flows in stacks and
regenerators with pores of arbitrary cross-sectional geometry. Journal of the
Acoustical Society of America, 133(4):2004–2013, 2013.
[60] B. Karlsson and J. Quintiere. Enclosure Fire Dynamics. CRC Press, Boca
Raton, FL, 2000.
[61] J. Borwick. Loudspeaker and Headphone Handbook. Focal Press, Oxford, third
edition, 2001.
[62] J. Volcler. Extremely Loud: Sound as a Weapon. The New Press, New York,
NY, 2013.
140