You are on page 1of 33

This article was downloaded by: [University of Connecticut]

On: 12 October 2014, At: 07:28


Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Combustion Science and Technology


Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/gcst20

Hydrogen- and C 1–C 3 Hydrocarbon-


Nitrous Oxide Kinetics in Freely
Propagating and Burner-Stabilized
Flames, Shock Tubes, and Flow Reactors
a a a
O. A. Powell , P. Papas & C. B. Dreyer
a
Division of Engineering , Colorado School of Mines , Golden,
Colorado, USA
Published online: 03 Mar 2010.

To cite this article: O. A. Powell , P. Papas & C. B. Dreyer (2010) Hydrogen- and C 1–C 3 Hydrocarbon-
Nitrous Oxide Kinetics in Freely Propagating and Burner-Stabilized Flames, Shock Tubes, and Flow
Reactors, Combustion Science and Technology, 182:3, 252-283, DOI: 10.1080/00102200903357724

To link to this article: http://dx.doi.org/10.1080/00102200903357724

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
Combust. Sci. and Tech., 182: 252–283, 2010
Copyright # Taylor & Francis Group, LLC
ISSN: 0010-2202 print=1563-521X online
DOI: 10.1080/00102200903357724

HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS


OXIDE KINETICS IN FREELY PROPAGATING AND
BURNER-STABILIZED FLAMES, SHOCK TUBES, AND
FLOW REACTORS

O. A. Powell, P. Papas, and C. B. Dreyer


Division of Engineering, Colorado School of Mines, Golden, Colorado, USA
Downloaded by [University of Connecticut] at 07:29 12 October 2014

A detailed, chemical mechanism has been compiled for modeling the combustion of C1–C3
hydrocarbon-nitrous oxide mixtures. The compiled, chemical mechanism has been compared
and validated against available data in the literature, including flame structure measurements
of hydrogen-nitrous oxide and ammonia-nitrous oxide flames, flow reactor data on
hydrogen-nitrous oxide and moist, nitrous oxide decomposition mixtures, shock tube ignition
delay data on hydrogen- and methane-nitrous oxide mixtures, experimental flame speed data
on carbon monoxide-nitrous oxide mixtures, and experimental, laminar flame speed data for
hydrogen-, methane-, acetylene-, and propane-nitrous oxide mixtures. Feature sensitivity
studies and reaction path analyses have been employed to elucidate important pathways to
the overall reaction rate for these systems, and include the nitrous oxide decomposition step
(N2O(þM) Ð N2 þ O(þM)) and nitrous oxide reactions with H atoms (N2O þ H Ð
N2 þ OH and N2O þ H Ð NH þ NO). Modifications to the compiled mechanism were con-
sidered on the basis of the most sensitive reactions with the highest apparent rate constant
uncertainty, to closely predict the various experimental data considered.

Keywords: Chemical kinetics; Nitrous oxide mixture

INTRODUCTION
Studies on the reaction characteristics of fuel-nitrous oxide mixtures are
motivated by recent interest in using nitrous oxide as an oxidant in propulsion
systems (Frolik, 2003), the generation of flammable mixtures of hydrogen-nitrous
oxide in nuclear waste storage facilities from radiolysis and chemical reactions (Pfahl
et al., 2000), the use of nitrous oxide in propane-fueled vitiation systems for supersonic
testing facilities, the combustion of nitramine propellants, and understanding the for-
mation of nitrogen oxides in combustion systems. Despite these practical interests,
there have been very few studies on fuel-nitrous oxide reaction systems, and still fewer
studies have been undertaken to measure the laminar burning velocities of fuel-nitrous
oxide mixtures over a wide range of conditions. Accordingly, the authors recently have

Received 21 May 2009; revised 12 September 2009; accepted 22 September 2009.


The views expressed in this article are those of the authors and do not reflect the official policy or
position of the United States Air Force, Department of Defense, or the U.S. Government.
Address correspondence to P. Papas, Division of Engineering, 1610 Illinois Street, Colorado School
of Mines, Golden, CO 80401, USA; Tel.: (303) 273-3655; E-mail: ppapas@mines.edu

252
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 253

undertaken to measure the laminar burning velocity of hydrogen-, methane-,


acetylene-, and propane-nitrous oxide flames (Powell et al., 2009). These data will be
important to the development of comprehensive chemical kinetic models for fuel-
nitrous oxide systems and other combustion systems involving nitrogen oxidation.
Following our recent publication (Powell et al., 2009), work has been underta-
ken to refine the compiled chemical mechanism, previously reported in Powell et al.
(2009), for validation against the limited available data in the literature, including
flow reactor studies of H2=N2O=H2O=N2 (Allen et al., 1998), flame speeds studies
of CO=N2O with added hydrogen (Linteris et al., 2000), flame structure studies of
NH3-doped H2=N2O=Ar flames (Venizelos & Sausa, 1998), and ignition delay studies
of CH4=N2O=Ar mixtures (Drummond, 1969; Soloukhin, 1971). A summary table of
the experimental data used in developing the chemical mechanism is shown in
Table 1. This paper will highlight the results of this development, which has followed
a hierarchical approach, starting from the most basic N2O decomposition and H2=
Downloaded by [University of Connecticut] at 07:29 12 October 2014

N2O chemical sub-models and building in complexity to the NxOy=CO sub-model


and models involving higher hydrocarbon fuels. It should be emphasized that the
validation and development of a comprehensive fuel-nitrous oxide mechanism is
by no means complete, and the presentation here aims to highlight important areas
for future research. The goal of the current work is to compare our experimental
flame data and the available literature data with predictions using a modified version
of a recently compiled chemical mechanism (Powell et al., 2009). Based on sensitivity
analysis as well as a review of rate constant determinations in the literature of impor-
tant reactions identified by sensitivity and reaction flux analysis, rate modifications
to this compiled, chemical mechanism are examined to improve its capacity to
predict available experimental data.

Table 1 Various experimental data on fuel systems oxidized with nitrous


oxide

Experiment Reference

Flow reactor studies


N2O decomposition Allen et al. (1995)
H2=N2O=H2O=N2 Allen et al. (1998)
Flame speed studies
CO=N2O Linteris et al. (2000)
CH4=air Bosschaart and de Goey (2004)
C3H8=air Vagelopoulos et al. (1994)
H2=N2O=N2 Powell et al. (2009)
CH4=N2O=N2 Powell et al. (2009)
C2H2=N2O=N2 Powell et al. (2009)
C3H8=N2O=N2 Powell et al. (2009)
Flame structure studies
Neat and NH3-doped
H2=N2O=Ar Venizelos and Sausa (1998)
NH3=N2O=Ar Venizelos and Sausa (2002)
Induction time studies
H2=N2O=Ar Hidaka et al. (1985b)
CH4=N2O=Ar Drummond (1969)
Soloukhin (1971)
254 O. A. POWELL ET AL.

REACTION MECHANISM
Although chemical mechanisms (e.g., Mendiara & Glarborg, 2009; Tian et al.,
2009) have been developed for modeling nitrogen oxide chemistry in hydrocarbon
reactions systems, the authors are unaware of an extensive effort to specifically
validate a chemical mechanism against the available data involving direct oxidation
with nitrous oxide (e.g., flow reactor, flame structure studies, and shock tube ignition
delay data). Consequently, the authors have undertaken to compile a hydrocarbon-
nitrous oxide mechanism, to be discussed below, from existing published mechan-
isms appearing in the literature. It is recognized that this compiled scheme is neither
a comprehensive nor validated model, and thus should be viewed as a starting
mechanism here.
The compiled mechanism discussed here was assembled by grafting the updated
H2=O2 sub-mechanism from Li et al. (2004) as well as the NxHy=CO=H2=O2 sub-
Downloaded by [University of Connecticut] at 07:29 12 October 2014

mechanism (excluding its H2=O2 chemistry) originally developed by Allen et al.


(1997) (see also Table C8 in Glassman & Yetter, 2008) onto the C1–C3 hydrocarbon
mechanism of Qin et al. (2000) (excluding its H2=O2 chemistry). The thermodynamic
and transport data for the species derived from the mechanism of Qin et al. (2000)
were taken from this same reference, the thermodynamic data for the H2=O2 mech-
anism species were taken from Li et al. (2004), and the other thermodynamic data
needed for the NxHy=CO=H2=O2 mechanism species were taken from Mueller et al.
(2000). The corresponding transport data required for the latter two sub-mechanisms
were taken from Smith et al. (n.d.). The missing transport data (namely, N, HONO,
NO3, and HNO3) not found in these databases were taken from Konnov (2000).
The original C1–C3 mechanism of Qin et al. (2000) was optimized for different
experimental data, including laminar flame speeds for CH4 and C3H8 with air as the
oxidizer at 1 atm at / ¼ 0.98, 1.43 and / ¼ 0.8, 1.2, respectively. The original NxHy=
CO=H2=O2 chemical mechanism (Allen et al., 1995, 1998) was validated using flow
reactor data over wide ranges of pressure (0.5–14.0 atm) and temperature (750–
1100 K). The H2=O2 sub-mechanism found in the NxOy=CO=H2=O2 mechanism
developed by Allen et al. (1997) was recently updated by the same group (Li et al.,
2004) and was extensively validated against laminar premixed flame, shock tube
ignition delay, and flow reactor data over a wide range in conditions (/ ¼ 0.25–
5.0, 298–3000 K, 0.3–87 atm). As our compiled mechanism incorporates this updated
comprehensive H2=O2 model and other changes, model predictions against the orig-
inal flow reactor data of Allen et al. (1995, 1998), as well as the hydrocarbon-air
flame speed data reported by Qin et al. (2000), will require re-examination.
The following modifications were made to complete this first compiled
mechanism (Powell et al., 2009):

1. two reactions (N2O þ H Ð NNH þ O and NH þ NO Ð NNH þ O; Dean &


Bozzelli, 2000) were added based on their recommendation to include them in
high-temperature processes;
2. the rate expression for N2O þ H Ð NH þ NO from Dean and Bozzelli (2000) was
utilized;
3. the reaction NH þ H2O Ð HNO þ H2 from GRIMech 3.0 (Smith et al., n.d.) was
added; and
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 255

4. the reaction NH3 þ HO2 Ð NH2 þ H2O2 was added from the DeNOx mechanism
of Glarborg et al. (1994).

This first, compiled mechanism, designated the Previous mechanism, should be


considered as a starting mechanism and contained 85 species and 587 reactions
(Powell et al., 2009). It should be noted that this starting mechanism does not
contain carbon-nitrogen species or their corresponding chemical interactions; their
inclusion will be considered using the mechanism developed by Tian et al. (2009)
and Mendiara and Glarborg (2009), designated the Glarborg mechanism, which
includes 97 species and 779 reactions (Mendiara & Glarborg, 2009).
For the purposes of this work, Table 2 summarizes the rate changes to the first
compiled mechanism. The reasons behind the selection of these particular reactions
and the rate changes made will be offered in the following, subsequent descriptions:
Downloaded by [University of Connecticut] at 07:29 12 October 2014

modifications have been made according to important reactions, characterized by


relatively large rate constant uncertainties, identified by sensitivity and reaction flux
analysis; and consideration of rate constant changes according to published rate
determinations within their reported uncertainty.
For the current mechanism used in this work, the following rate constant
changes to the NxOy=CO=H2=O2 chemistry used in Powell et al. (2009) were
adopted:

1. rate constants for NO þ H Ð N þ OH, NO þ O Ð N þ O2, NH þ NO Ð NNH þ O


from Baulch et al. (2005);
2. rate constant for NO þ N Ð N2 þ O from Mick et al. (1993);
3. rate constant for NO þ H(þM) Ð HNO(þM) from Riley et al. (2003);
4. rate constant for NH þ NO Ð N2 þ OH from Dean and Bozzelli (2000);
5. rate constant for NH2 þ O Ð HNO þ H from Venizelos and Sausa (2002);
6. rate constant for NH3 þ OH Ð NH2 þ H2O from Cohen and Westberg (1991);
7. rate constant for NH þ NO Ð N2O þ H from Dean and Bozzelli (2000);
8. rate constant for CO þ N2O Ð CO2 þ N2 from Milks and Matula (1973); and
9. rate constant for N2O þ H Ð N2 þ OH determined from a data refit (see
Powell et al., 2009), yielding k ¼ 2.53  1010 exp(4550 K=RT) þ 5.00 
1014 exp(18100 K=RT). Figure 1 shows Arrhenius plots of the new rate constant
compared with literature data.

The mechanism containing all the changes above will be designated the Current
mechanism in this paper.
Considering reactions 5, 6, and 10 (NH þ NO ! Products, see Table 2), several
authors have examined the branching ratio for k5=(k5 þ k6 þ k10). At 3000 K, the
current mechanism gives 0.12, while Dean and Bozzelli (2000) predicted 0.09 and
Miller and Melius (1992) predicted 0.29, compared to 0.19  0.10 as measured by
Mertens et al. (1991). At room temperature, the current model gives 0.10, while Dean
and Bozzelli (2000) predicted 0.09 and Miller and Melius (1992) predicted 0.19
compared to 0.15  0.05 as measured by Lillich et al. (1994). Furthermore, the
Current mechanism yields a ratio of k10=(k5 þ k6 þ k10) ¼ 0.90, which is within the
range of 0.65–0.91 as reported by Baulch et al. (2005).
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Table 2 Modifications to the chemical mechanism used by Powell et al. (2009)

Previous mechanism, Powell et al. (2009) Current mechanism

# Reaction A n Ea A n Ea Reference

1 NO þ H Ð N þ OH 1.69E þ 14 0.00 4.88E þ 4 3.44E þ 14 0.00 4.95E þ 4 Baulch et al. (2005)


2 NO þ O Ð N þ O2 1.81E þ 09 1.00 3.88E þ 4 6.87E þ 08 1.13 3.82E þ 4 Baulch et al. (2005)
3 NO þ N Ð N2 þ O 3.27E þ 12 0.30 0.00 2.00E þ 13 0.00 0.00 Mick et al. (1993)
4 NO þ H(þM) Ð HNO(þM) 8.96E þ 19 1.32 7.35E þ 2 5.00E þ 19 1.32 7.35E þ 2 Riley et al. (2003)
5 NH þ NO Ð N2 þ OH 2.16E þ 13 0.23 0.00 1.40E þ 17 1.49 1.31E þ 3 Dean and Bozzelli (2000)
6 NH þ NO Ð NNH þ O 1.70E þ 14 0.20 1.22E þ 4 9.00E þ 14 0.00 2.09E þ 4 Baulch et al. (2005)
7 N2O þ H Ð N2 þ OH 2.93E þ 10 0.00 4.57E þ 3 2.53E þ 10 0.00 4.55E þ 3
5.82E þ 14 0.00 1.79E þ 4 5.00E þ 14 0.00 1.81E þ 4 see texta

256
8 NH2 þ O Ð HNO þ H 6.63E þ 14 0.50 0.00 6.90E þ 13 0.00 0.00 Venizelos and Sausa (2002)
9 NH3 þ OH Ð NH2 þ H2O 2.04E þ 06 2.04 5.66E þ 2 2.51E þ 07 1.60 9.54E þ 2 Cohen and Westberg (1991)
10 NH þ NO Ð N2O þ H see text 3.80E þ 18 1.65 1.43E þ 4 Dean and Bozzelli (2000)
11 CO þ N2O Ð CO2 þ N2 5.01E þ 13 0.00 4.40E þ 4 2.09E þ 11 0.00 1.73E þ 4 Milks and Matula (1973)
12 NH þ H2O Ð HNO þ H2 2.00E þ 13 0.00 1.39E þ 4 removed Smith et al. (n.d.)
13 O þ CH3 Ð H þ H2 þ CO 3.37E þ 13 0.00 0.00 1.69E þ 13 0.00 0.00 Baulch et al. (2005)
14 OH þ CH3 Ð CH2 þ H2O 6.44E þ 17 1.34 1.42E þ 3 1.08E þ 16 0.91 5.46E þ 2 Baulch et al. (2005)
15 C2H4 þ H(þM) Ð C2H5(þM)y 6.00E þ 41 7.62 6.97E þ 3 2.00E þ 39 6.64 5.77E þ 3 Miller and Klippenstein (2004)
16 C2H3 þ H(þM) Ð H2 þ C2H2 3.03E þ 13 0.00 0.00 1.00E þ 13 0.00 0.00 Tanzawa and Gardiner, Jr. (1980)
a
Rate represented as sum of two Arrhenius expressions.
y
Low-pressure limit rate.
Specific reaction rate constants have the form k ¼ ATn exp (Ea=RT). Units are cm3, cal, mol, and sec, and Reference refers to reference for the Current mechanism.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 257
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 1 Arrhenius plots for N2O þ H Ð N2 þ OH comparing the new rate to literature data: (a) shows a
broad temperature range; (b) shows high temperature range indicated by dotted box in (a); (- - -) current
work; (—) Baulch et al. (2005) (reported as sum of two product channels N2 þ OH and NH þ NO); (- -&- -)
Dean et al. (1980); (- - - -) Dean et al. (1978); (- -~- -) Henrici and Bauer (1969); (- -- -) Hidaka et al.
4
(1985a); (þ) Marshall et al. (1989); (.) Arthur et al. (1997).

NUMERICAL METHODOLOGY
Numerical results for flame speed, flow reactor species profiles, and induction
time were obtained using the CANTERA-based programs (Goodwin, 2003). For
the determination of laminar flame speeds, the freely propagating, adiabatic one-
dimensional flame code provided with CANTERA was utilized. This code solves
the steady-state species and energy equations using a combined time-integration, New-
ton iteration technique. Soret effects were not included, and the mixture-average trans-
port descriptions were used. The computational differences arising from using the
multicomponent transport formulation, which is computationally more expensive,
in lieu of the mixture-average transport description was not fully quantified for all
conditions examined here. Computations using the multicomponent transport descrip-
tion, however, were conducted for the peak flame speed conditions for nitrogen-
diluted H2=N2O, CH4=N2O, and C2H2=N2O flames, later shown in Figures 15, 18,
and 20, respectively. For the conditions corresponding to the peak flame speed for
these systems, the peak laminar flame speeds computed using the multicomponent
transport description and the Current mechanism for the H2=N2O, CH4=N2O, and
C2H2=N2O flame systems decreased by 1.51%, increased by 1.35%, and decreased
by 1.36%, respectively. These computational changes are within the range of experi-
mental uncertainties reported on these same graphs. For all the flame speed determina-
tions reported here, grid independence was verified using more than 500 grid points.
For the simulation of flow reactor and shock tube experiments, a constant-
pressure, zero-dimensional model provided with CANTERA was utilized. This pro-
gram predicts homogeneous gas-phase kinetics by integrating the time-dependent
energy and species conservation equations. Numerical results for flame structure
profiles were obtained using the premixed, burner stabilized code provided with
258 O. A. POWELL ET AL.

CHEMKIN-PRO1 (CHEMKIN-PRO1, 2008). As with the CANTERA-based


simulations, Soret effects were not included, and mixture-average transport descrip-
tions were utilized.

RESULTS AND DISCUSSION


In previous work (Powell et al., 2009), the authors reported experimental
measurements of the adiabatic, laminar flame speeds for hydrogen-, methane-,
acetylene-, and propane-nitrous oxide. In this recent study, the importance of
increasing the rate constant for the reaction 7, N2O þ H Ð N2 þ OH, was discussed
to match flame speed data for H2=N2O=N2 flames. While maintaining the need to
increase this rate, the Current mechanism was modified and compared to various
experimental data sets available in the literature. Comparison with these experi-
mental data sets formed the basis for the changes to the mechanism shown in
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Table 2. These changes were necessary refinements for predicting the different
experimental data considered, and are discussed below in the subsequent sections.

Comparison with Flow Reactor Data


The H2=O2 sub-mechanism found in the original NxOy=CO=H2=O2 mechanism
developed by Allen et al. (1997) was recently updated by the same group (Li et al.,
2004) and extensively validated against hydrogen-oxygen laminar premixed flames,
shock tube ignition delay, and flow reactor data in the literature. In order to develop
a robust mechanism following a hierarchal approach, the current compiled
mechanism incorporates this updated comprehensive H2=O2 model. Given the rate
constant changes in the updated mechanism of Li et al. (2004), it becomes necessary
to re-examine the original flow reactor data of Allen et al. (1995, 1998) in light of all
the rate constant revisions.
First, the flow reactor data for the decomposition of nitrous oxide with added
water from the study of Allen et al. (1995) were re-examined. Figure 2 shows the
N2O, O2, NO, and NO2 flow reactor profiles from the study of Allen et al. (1995)
compared to predictions using the current mechanism. Given the agreement, the
modifications incorporated in the current mechanism were judged to not affect the
capacity for reproducing the originally reported nitrous oxide decomposition flow
reactor results.
Following a hierarchical approach, the H2=N2O flow reactor studies of Allen
et al. (1998) were examined next. The comparison with the flow reactor data and
previously reported flame speed data (Powell et al., 2009) points to some important
kinetic steps that require careful examination. First, the high-pressure H2=N2O flow
reactor studies have demonstrated that NO þ H(þM) Ð HNO(þM) is important
with regard to both the overall reaction rate as well as NO production. Allen et al.
also found that the most sensitive reactions to the overall H2=N2O rate included
N2O(þM) Ð N2 þ O(þM), N2O þ H Ð N2 þ OH, and N2O þ H Ð NH þ NO, and
these findings are consistent with those of our previous flame speed studies (Powell
et al., 2009). The rate constant expressions reported in the literature for the nitrous
oxide decomposition reaction are quite consistent, and the rate utilized by Allen et al.
was derived in their earlier work (Allen et al., 1995). Consequently, this rate constant
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 259
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 2 Species profiles of N2O decomposition diluted with nitrogen at 10.5 atm and 1123 K. The initial
mole fractions of N2O and H2O were 1.00  102 and 5.57  104, respectively. (symbols) experimental
data from Allen et al. (1995); (—) predictions using the Current mechanism.

expression has remained unchanged given its low uncertainty. The two product
channels of N2O þ H, on the other hand, show a relatively higher degree of uncer-
tainty, and thus were considered for modification. First, the reverse direction for
N2O þ H Ð NH þ NO was considered (shown as reaction 10 in Table 2) in the cur-
rent work to facilitate matching experimental NO profiles. In examining the NO flow
reactor profiles as well as the NO=Ar flame structure profiles of Venizelos and Sausa
(1998), the pre-exponential factor for this reaction (NH þ NO Ð N2O þ H) was
slightly increased (21%) from the value recommended by Dean and Bozzelli
(2000); in addition, the rate constant for the reaction NO þ H(þM) Ð HNO(þM)
was increased to 5.00  1019, compared to the value of 3.10  1019 used by Allen
et al. (1998). The increased rate constant for this reaction is consistent with the
experimental value reported by Riley et al. (2003). Second, the rate constant for
the N2O þ H Ð N2 þ OH reaction has been found to be quite sensitive to the flame
speed data reported previously (Powell et al., 2009), and a new, double Arrhenius
expression was determined based on high-temperature rate data (e.g., Dean et al.,
1978) and the intermediate temperature data of Marshall et al. (1989) (see Powell
et al., 2009). This new rate constant expression is 2.53  1010 exp(4550 K=RT) þ
5.00  1014 exp(18,100 K=RT). Figure 1 shows Arrhenius plots of the new rate
constant compared with literature data.
Figure 3 shows the H2, H2O, N2O, NH3, and NO profiles using the Current
mechanism compared to the flow reactor data from Allen et al. (1998), and
Figure 4 shows the H2, N2O, and NO profiles with added NO using the Current
mechanism. These computed flow reactor profiles and others shown in Figures 3
and 4 were not shifted in time to match the experimental 50% fuel consumption
value. Referring to Figure 3a, the Current mechanism gives a slightly faster, overall
reaction rate compared to the flow reactor data from Allen et al. (1998). As shown in
Figure 3b, the mechanism over-predicts the peak NO and NH3 values by 63% and
260 O. A. POWELL ET AL.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 3 H2=N2O profiles diluted with N2 at 3 atm, 995 K. The initial mole fractions for H2 and N2O were
5.3  103 and 1.1  102, respectively. (symbols) experimental data from Allen et al. (1998); (lines) pre-
dictions using the Current mechanism.

32%, respectively, while roughly matching the overall time scale of the experiment.
Figure 4 shows a comparison of the results from Allen et al. (1998) for a H2=
N2O=N2 mixture with added NO with the Current mechanism. The predicted N2O
and H2O profiles are characterized by a slightly slower rate compared to the experi-
mental data, while the H2 profile is reproduced quite well. Figure 5 shows a compari-
son of the results from Allen et al. (1998) and the current model with varying
amounts of added NO. This graph shows the N2O mole fractions at 0.35 s residence

Figure 4 H2=N2O profiles diluted with N2 at 3 atm, 995 K with added NO. The initial mole fractions for
H2, N2O, and NO were 5.3  103, 1.06  102, and 1.5  104, respectively. (symbols) experimental data
from Allen et al. (1998); (—) predictions using the Current mechanism.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 261
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 5 N2O mole fractions at 0.35 s residence time, 3 atm, 995 K with various amounts of added NO. The
initial mole fractions for H2 and N2O were 5.3  103 and 1.06  102, respectively. (symbols) experi-
mental data from Allen et al. (1998); (—) prediction using the Current mechanism.

time for experiments where the system pressure and the initial temperature were
3 atm and 995 K, respectively. The peak difference between the experimental and
computational results is 15% with no added NO, while the differences are less than
5% at the other data points. Figure 6 shows a comparison of the results from Allen
et al. (1998) and the Current mechanism with added NH3. As discussed by Allen et al.
(1998), the NH3 profile stays relatively constant, indicating the production of NH3
during the reaction process, and the Current mechanism captures this trend. The
H2 profile is reproduced quite well, whereas the N2O and H2O profiles are character-
ized by slightly slower rates.

Comparison with Flame Structure Measurements


In this section, the flame structure measurements for both H2=N2O=Ar
(Venizelos & Sausa, 1998) and NH3=N2O=Ar (Venizelos & Sausa, 2002) flames will
be examined. In their earlier work, these authors used molecular beam mass
spectrometry (MB=MS) and laser-induced fluorescence (LIF) to measure profiles
for various species, including NO and O2 for stoichiometric neat and NH3-doped
H2=N2O=N2 flames at 30 torr.
They used thermocouple-measured temperatures (corrected for radiation
losses) to validate NH and OH LIF temperature measurements. Venizelos and Sausa
used these data to validate a previously assembled mechanism (Sausa et al., 1996). In
the latter work, they used similar techniques (MB=MS, LIF, and thermocouple
measurements) to measure and predict species profiles for a stoichiometric NH3=
N2O=Ar flame at 60 torr to facilitate the modification and validation of the mechanism
from their earlier work. In considering these flame systems, Allen et al. (1998)
262 O. A. POWELL ET AL.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 6 Species profiles at 3 atm, 995 K with added NH3. The initial mole fractions for H2, N2O, and NH3
were 5.3  103, 1.16  102, and 8.0  104, respectively. (symbols) experimental data from Allen et al.
(1998); (—) predictions using the Current mechanism.

discussed the importance of NxHy kinetics for their H2=N2O=N2 flow reactor
conditions. As they discussed, the rate constant for reaction 9, NH3 þ OH Ð NH2 þ
H2O is not well characterized; consequently, they reduced the pre-exponential term
for this reaction to control the speed of the reaction. Similarly, Venizelos and Sausa
(2002) lowered the pre-exponential term of this reaction in their mechanism to match
the species profiles of their NH3=N2O=Ar flames. In the current work, the authors
reduced the pre-exponential term for reaction 9 to the value reported by Cohen and
Westberg (1991) to 2.51  107. For comparison, at 1000 K, the rate constant
expression used by Allen et al. (1998) yielded 5.60  1011 cm3mol1s1, and the rate
expression used by Venizelos and Sausa (2002) yielded 1.03  1012 cm3mol1s1,
whereas the current rate expression yields 9.78  1011 cm3mol1 s1.
Venizelos and Sausa (1998) measured the flame structure for neat (i.e.,
undoped) and NH3-doped H2=N2O=Ar flames. Figures 7a–7c show the mole fraction
ratios NO=Ar, O2=Ar, and NH3=Ar, respectively, for these flames compared with
predictions using the Current mechanism. For modeling these burner-stabilized
flames, the experimentally measured temperature profile reported by Venizelos and
Sausa (1998) was specified as an input for the computation. The computational
NO=Ar and O2=Ar profiles fall within the reported experimental uncertainty, while
the NH3=Ar profile is slightly higher than the reported experimental uncertainty.
Overall, the agreement between the experimental and computational results is judged
to be quite reasonable.
Venizelos and Sausa (2002) also conducted flame structure measurements for
NH3=N2O=Ar flames, and Figures 8a and 8b show the mole fraction ratios NO=Ar
and O2=Ar for these flames compared with predictions using the Current mechanism.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 263
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 7 Species profiles for neat and NH3-doped H2=N2O flames diluted with Ar at 30 torr compared with
literature data. Reactant flow rates for H2, N2O, and Ar were 1.41, 1.43, and 1.06 slm, respectively, for the
neat flame and doped with 0.14 slm NH3: (a) XNO=XAr; (b) XO2=XAr; (c) XNH3=XAr for NH3-doped flame.
(symbols) from Venizelos and Sausa (1998); (.) molecular beam mass spectroscopy, () laser-induced
fluorescence; (—) prediction using the Current mechanism.

While the Current mechanism predicted slightly higher NO=Ar values over all spatial
positions, the values are within the reported uncertainty. Additionally, the Current
mechanism predicted slightly higher values for O2=Ar.

Comparison with Laminar Flame Speed Studies


CO/N2O flames. Linteris et al. (2000) measured the laminar flame speed for
bone-dry CO=N2O flames and CO=N2O flames with added amounts of hydrogen.
The background water levels for their experiments were in the range of 5–15 ppm.
264 O. A. POWELL ET AL.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 8 Species profiles for NH3=N2O flames diluted with Ar compared at 60 torr with literature data.
Reactant flow rates for NH3, N2O, and Ar were 1.03, 1.59, and 0.54 slm, respectively: (a) XNO=XAr; (b)
XO2=XAr. (symbols) experimental data from Venizelos and Sausa (2002); (.) molecular beam mass spec-
troscopy, () laser-induced fluorescence; (—) prediction using the Current mechanism.

These authors examined the effects of small background contaminate levels, and
their computations showed that background hydrogen levels of 10 ppm only changed
the burning velocity by 1% compared to the bone-dry case. Following their work, the
Current mechanism included the rate constant for CO2þ N2O Ð CO2 þ N2 from
Milks and Matula (1973). Their rate expression for the exchange reaction, CO þ
N2O Ð CO2 þ N2, is higher than that recommended by Allen et al. (1997); conse-
quently, more research is required to isolate the appropriate rate expression. By
diluting bone-dry CO=N2O flames with nitrogen, Linteris et al. clearly demonstrated
that the rate constant from Milks and Matula for reaction 11, with its associated
activation energy, was most appropriate for predicting flame speeds for this system.
Figure 9 shows the laminar burning velocity for CO=N2O flames with added H2
using the Current mechanism compared to the flame speed data of Linteris et al.
(2000). For the current, computational results, the initial temperature for the reac-
tants was assumed to be 298 K. Linteris et al. showed a similar underprediction of
their experiments and computational results at higher added H2 levels. Figure 9 also
shows the predicted flame speeds using the Glarborg mechanism (Mendiara &
Glarborg, 2009; Tian et al., 2009). The lower flame speed results using the Glarborg
mechanism (a reduction of 66% at 0% H2) is primarily attributed to a difference in
the rate constant expression utilized for the direct exchange reaction, CO þ
N2O Ð CO2 þ N2. The rate constant used in the Current mechanism for this reaction
is 62% higher at 2000 K than that used in the Glarborg mechanism. Linteris et al.
demonstrated that this exchange reaction and CO þ OH Ð CO2 þ H are the most
sensitive reactions across the range of added H2 shown in Figure 9. The rate constant
for this latter reaction in the Current mechanism is only 6% lower (at 2000 K) than
the rate used in the Glarborg mechanism, further substantiating that the direct
exchange reaction is the major cause for the reduced flame speed predictions using
the Glarborg mechanism.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 265
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 9 Burning velocity of stoichiometric CO-N2O flames at 1 atm with varying values of added H2. The
initial temperature was assumed to be 298 K. (symbols) experimental data from Linteris et al. (2000); (—)
predictions using the Previous mechanism (Powell et al., 2009); () predictions using the Current mech-
anism; (- - -) predictions using the Glarborg mechanism (Mendiara & Glarborg, 2009; Tian et al., 2009).

CH4/air flames. Given our rate constant changes to the hydrogen-oxygen


chemistry inherent in the hydrocarbon mechanism of Qin et al. (2000), it became
prudent to re-evaluate the hydrocarbon-air flame speed predictions reported in that
original work. In this section, flame speed predictions for the CH4=air will be exam-
ined, while predictions for C3H8=air systems will be examined in the next section.
Figure 10 shows unstretched, adiabatic flame speeds for CH4=air mixtures using
the Previous and Current mechanisms compared to flame speed data reported in
the literature, including data from Vagelopoulos et al. (1994) originally used by
Qin et al. (2000) for comparison purposes. For stoichiometric and fuel-rich con-
ditions, the computational results using the Previous mechanism are in agreement
with the experimental results shown; however, the mechanism overpredicts the
experimental results for fuel-lean conditions. At /  1, Vagelopoulos et al. (1994)
reported a flame speed of 38.7 cm=s, while the Previous mechanism slightly overpre-
dicted this value, yielding 40.2 cm=s. For comparison purposes, the Glarborg mech-
anism gives a value of 36.3 cm=s. The recent experimental work by Bosschaart and
de Goey (2004) reported that the flame speed for stoichiometric CH4=air flames
approaches a value of 36 cm=s. Also, these authors report a typical uncertainty of
0.5 cm=s around / ¼ 1.0, increasing to 1.0 cm=s at equivalences ratios with smaller
heat release (i.e., / < 0.7 and / > 1.3). In comparing the Previous mechanism predic-
tions with the results from Bosschaart and de Goey (2004), the model predictions are
in reasonable agreement under fuel-rich conditions, but overpredict experimental
data under near-stoichiometric and fuel-lean conditions.
To improve the agreement with experimental data under fuel-lean conditions,
the authors undertook flame speed sensitivity analyses to highlight sensitive reactions
whose rate constant definitions also show a relatively high degree of uncertainty.
266 O. A. POWELL ET AL.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 10 Adiabatic, laminar flame speed, SL0 , for CH4=air mixtures at 1 atm. () experimental data from
Egolfopoulos et al. (1989); (.) experimental data from Vagelopoulos et al. (1994); (~) experimental data
from Gu et al. (2000); (&) experimental data from Bosschaart and de Goey (2004) (taken at 0.99 atm); (—)
predictions using the Previous mechanism; () predictions using the Current mechanism (see Table 2).

The feature sensitivity analysis, CHEMKIN-PRO1 (2008; also see Kee et al., 1995)
on the laminar burning rate (laminar flame speed multiplied by the unburned mixture
density) identifies main reaction paths by determining the influence of all reaction rate
perturbations on the laminar mass burning rate. Figure 11 shows the most sensitive
reactions (magnitude > j0.02j) for CH4-air mixtures for three equivalences ratios.
Based on the sensitivity=uncertainty analyses, four rate constant revisions were made
to the Previous mechanism, reported in Table 2 as reactions 13–16. First, Baulch et al.
(2005) report that the branching ratios for reaction 13, O þ CH3 Ð H þ H2 þ CO, and
O þ CH3 Ð H þ CH2O are 80% and 20%, respectively, and are roughly independent
of temperature. The pre-exponential factor of the former rate was lowered by half, so
that these branching ratios are 77% and 23% in the Current mechanism. For compari-
son purposes, these branching ratios are 82% and 18% in the Glarborg mechanism.
Second, the rate for reaction 14, OH þCH3 Ð CH2 þH2 O, was lowered to the value
recommended by Baulch et al. (2005). The combination of these first two revisions
lowered the flame speed under fuel-lean CH4-air conditions, thereby improving the
agreement with the experimental data. Finally, two additional revisions were made
to improve predictions under fuel-rich conditions:

1. the rate constant for reaction 15, C2H4 þ H(þM) Ð C2H5(þM), was incorporated
from the Glarborg mechanism; and
2. the rate constant for reaction 16, C2H3 þ H Ð H2 þ C2H2, was lowered to that
reported by Tanzawa and Gardiner, Jr. (1980).

The rate constant from Tanzawa and Gardiner, Jr. (1980) is within the reliability
reported by Baulch et al. (2005). Table 3 shows the ratio of the modified rate
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 267
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 11 Flame speed sensitivity for the most sensitive reactions (magnitude > j0.02j) using the
Previous mechanism for CH4=air flames: (13) O þ CH3 Ð H þ H2 þ CO; (14) OH þ CH3 Ð CH2 þ H2O;
(15) H þ C2H4(þM) Ð C2H5(þM); (17) H þ O2 Ð O þ OH; (18) H þ OH þ M Ð H2O þ M; (19) H þ O2
(þM) Ð HO2(þM); (20) CO þ OH Ð CO2 þ H; (21) HCO þ M Ð H þ CO þ M; (22) H þ CH3(þM) Ð
CH4(þM); (23) H þ CH4 Ð CH3 þ H2; (24) HO2 þ CH3 Ð OH þ CH3O; (25) OH þ CH4 Ð CH3 þ H2O;
(26) O þ CH3 Ð H þ CH2O; (27) CH3 þ CH3(þM) Ð C2H6(þM).

Table 3 Rate constant ratios of the Current mechanism to the Previous mechanism
from Powell et al. (2009) at 1500 K and 2500 K

# Reaction 1500 K 2500 K

1 NO þ H Ð N þ OH 1.61 1.77
2 NO þ O Ð N þ O2 1.20 1.18
3 NO þ N Ð N2 þ O 0.68 0.58
4 NO þ H(þM) Ð HNO(þM) 0.56 0.56
5 NH þ NO Ð N2 þ OH 0.42 0.26
6 NH þ NO Ð NNH þ O 1.15 4.20
7 N2O þ H Ð N2 þ OH 0.79 0.82
8 NH2 þ O Ð HNO þ H 4.03 5.20
9 NH3 þ OH Ð NH2 þ H2O 0.43 0.36
10 NH þ NO Ð N2O þ H (new rate)
11 CO þ N2O Ð CO2 þ N2 32 0.90
12 NH þ H2O Ð HNO þ H2 (removed)
13 O þ CH3 Ð H þ H2 þ CO 0.50 0.50
14 OH þ CH3 Ð CH2 þ H2O 0.52 0.58
15 C2H4 þ H(þM) Ð C2H5(þM)a 0.16 0.11
16 C2H3 þ H(þM) Ð H2 þ C2H2 0.33 0.33
a
Low pressure limit rate.
268 O. A. POWELL ET AL.

constants to those used in the Previous mechanism by Powell et al. (2009) at


temperatures of 1500 K and 2500 K.
The CH4=air predictions using these modifications are shown in Figure 10 and
are designated as Current mechanism. The four revisions effectively improved the
agreement with experimental data under fuel-lean conditions. Specifically, the
Current mechanism gives a flame speed value of 35.6 cm=s at / ¼ 1.0, which is within
the experimental uncertainty reported by Bosschaart and de Goey (2004). The
Current mechanism slightly underpredicts (6%) the peak experimental flame speed
value of 37.3 cm=s at / ¼ 1.1 reported by Bosschaart and de Goey.

C3H8/air flames. Figure 12 shows unstretched, adiabatic flame speeds for


C3H8=air mixtures using the Current mechanism compared to various experimental
flame speed data from the literature. Similar to the CH4=air results, the Previous
mechanism gives higher flame speed predictions under fuel-lean conditions com-
Downloaded by [University of Connecticut] at 07:29 12 October 2014

pared to the Current mechanism, whereas both mechanisms give similar predictions
under fuel-rich conditions. The experimental values from Vagelopoulos et al. (1994)
and Bosschaart and de Goey (2004) are quite similar across the entire equivalences
ratio range shown in Figure 12, and thus are taken collectively as a reference for the
discussions here. For fuel-lean and near-stoichiometric conditions, the Previous
mechanism overpredicts these experimental data by about 7–10%, with larger
differences occurring under very lean conditions (about 25% at / ¼ 0.6). Figure 13
shows the most sensitive reactions to the C3H8=air laminar flame speed at three
different equivalence ratios. This figure shows the relative sensitivity of the four revi-
sions reported previously in Table 2 and incorporated in the Current mechanism.
Similar to the CH4-air system, the predictions using the Current mechanism are in
reasonable agreement with the experimental data across the entire range of
equivalence ratio.

Figure 12 Adiabatic, laminar flame speed, SL0 , for C3H8=air mixtures at 1 atm. () experimental data from
Law et al. (1988); (.) experimental data from Vagelopoulos et al. (1994); (&) experimental data from
Bosschaart and de Goey (2004); (—) predictions using the Previous mechanism; () predictions using
the current mechanism (see Table 2).
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 269
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 13 Flame speed sensitivity for the most sensitive reactions (magnitude > j0.02j) using the Previous
mechanism for C3H8=air flames: (14) OH þ CH3 Ð CH2 þ H2O; (17) H þ O2 Ð O þ OH; (18) H þ OH þ
M Ð H2O þ M; (19) H þ O2(þM) Ð HO2(þM); (20) CO þ OH Ð CO2 þ H; (21) HCO þ M Ð H þ CO þ M;
(22) H þ CH3(þM) Ð CH4(þM); (24) HO2 þ CH3 Ð OH þ CH3O; (27) C3H6 þ H(þM) Ð iC3H7(þM); (28)
HO2 þ OH Ð H2O þ O2; (30) HCO þ H Ð CO þ H2; (31) C3H6 þ OH Ð aC3H5 þ H2O; (32) C3H8 þ H Ð
H2 þ iC3H7; (33) H þ C2H3 Ð H2 þ C2H2.

H2/N2O flames. In order to differentiate the unique, non-chain branching


characteristics of combustion systems oxidized by nitrous oxide from the chain
branching characteristics of systems oxidized by air, it becomes informative to
explore a reaction path analysis of a hydrogen-nitrous oxide flame. Figure 14 shows
the reaction pathways for a H2=N2O, / ¼ 1.24 flame diluted with nitrogen such that
N2O=(N2O þ N2) ¼ 0.27. This analysis was conducted at the location in the flame
corresponding to a flame temperature that is 50% of the overall temperature rise
from ambient. The values shown adjacent to each path represent the rate of progress
for a given reaction. To determine the actual rate of progress, one needs to multiply
three terms: the scale factor located at the bottom of the figure, the listed factor for a
given channel, and the percentage for a given reaction. The result of this operation
will be the rate of progress for that reaction in units of kmol m3 s1. For example,
referring to Figure 14, the rate of progress for the reaction H þ H2O Ð H2 þ OH is
19  0.164  67%, yielding 2.088 kmol m3 s1.
From the reaction path diagram shown in Figure 14, the following character-
istics become apparent. First, the H radical is primarily attacked by nitrous oxide via
the chain propagating reaction N2O þ H Ð N2 þ OH. Unlike fuel-oxygen flames,
chain branching via H þ O2 Ð OH þ H is relatively unimportant for nitrous oxide
flames. As reported in previous work (Allen et al., 1998; Powell et al., 2009), the
N2O þ H pathways are quite important to the overall rate of reaction and thus flame
speed. Reactions that compete with the H, OH, and O chain carriers significantly
270 O. A. POWELL ET AL.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 14 Reactions pathways for H2=N2O flame at 1 atm, / ¼ 1.24, and 50% of the temperature rise using
the Current mechanism; rate of progress units, kmol m3 s1. Percentages that do not sum to 100% are due
to rounding error.

impact the overall reaction rate; thus, the N2O þ H Ð NH þ NO reaction channel,
although characterized by a relatively small branching ratio, can lead to significant
inhibition of the system. Second, the vast majority of the attack on molecular hydro-
gen comes about from reaction with the OH radical, H2 þ OH Ð H2O þ H, and a
very small fraction via attack by the O radical. Third, to accurately predict H2=
N2O combustion characteristics, particularly under fuel-rich conditions, it becomes
important to include the chemistry involving N2Hx=NHx species (Allen et al.,
1998; Powell et al., 2009). Under the conditions of their flow reactor experiments,
Allen et al. (1998) found that the inclusion of N2Hx=NHx chemistry significantly
reduced the predicted overall reaction rate, and thereby was necessary to bring their
measured species profiles in agreement with their model predictions. Consistent with
this finding, our previous work (Powell et al., 2009) demonstrated that the reaction
steps involving these species act to inhibit the overall H2=N2O reaction rate or flame
speed under the fuel-rich conditions shown in Figure 14.
Figure 15 shows unstretched, adiabatic laminar flame speeds for H2=N2O=N2
mixtures using the Current mechanism compared to the experimental flame speed
data reported by Powell et al. (2009). These flames were diluted with nitrogen such
that N2O=(N2O þ N2) ¼ 0.27 on a volumetric basis. The adiabatic laminar flame speed
values, SL0 , shown in Figure 15—as well as Figures 18, 20, and 21, discussed later—
were measured using the flat-flame burner technique, where SL0 was determined from
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 271
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 15 Adiabatic, unstretched, laminar flame speed, SL0 , for H2=N2O mixtures at 0.8 atm diluted with
N2 (N2O=(N2O þ N2) ¼ 0.27) compared with literature values. (.) experimental data from Powell et al.
(2009); (—) predictions using the Previous mechanism from Powell et al. (2009); () predictions using
the Current mechanism; (- - -) predictions using the Glarborg mechanism (Mendiara & Glarborg, 2009;
Tian et al., 2009).

extrapolation to the zero heat loss condition (Powell et al., 2009). In the
near-stoichiometric and fuel-rich regions, the Current mechanism shows improved
performance over the Previous mechanism (Powell et al., 2009). For comparison pur-
poses, Figure 15 also shows the computational results using the Glarborg mechanism.
In the fuel-lean region, the results from all three mechanisms are virtually identical
and somewhat underpredict the experimental results from Powell et al. (2009). In
the near-stoichiometric region (up to / ¼ 1.41), the Current mechanism gives values
closer to the experimental results. Above / ¼ 1.41, all three mechanisms somewhat
overpredict the experimental results with the Current mechanism yielding the lowest
overprediction.
To illustrate the important reaction paths for the hydrogen-nitrous oxide flame
systems, it becomes interesting to explore the mechanistic differences between the
Current and Glarborg mechanisms. The Glarborg mechanism, which was based in
part on the recent CH4=NH3=O2 flame studies of Tian et al. (2009) and the CH4=
NH3=O2 flow reactor studies of Mendiara and Glarborg (2009), was developed using
both reaction path analysis of the fuel depletion as well as sensitivity analyses of the
conversion of NO. Using both these chemical mechanisms, feature sensitivity
analyses to the laminar flame speed under the conditions at / ¼ 1.24 reported in
Figure 15 were undertaken. Consistent with our previous work (Powell et al.,
2009), the four most sensitive reactions to flame speed at / ¼ 1.24 using the Current
mechanism (in order of decreasing sensitivity) are N2O þ H Ð N2 þ OH,
N2O(þM) Ð N2 þ O(þM), N2O þ H Ð NH þ NO, and N2O þ H Ð NNH þ O. These
same four reactions also appear as the most sensitive for the Glarborg mechanism
that incorporates similar rate expressions as the Current mechanism (within
50%) for the first three reactions. For the N2O þ H Ð NNH þ O reaction, however,
272 O. A. POWELL ET AL.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 16 Reaction pathways for a CH4=N2O=N2 flame (N2O=(N2O þ N2) ¼ 0.42 at 0.81 atm, / ¼ 0.78,
and 50% of the temperature rise using the Current mechanism; rate of progress units, kmol m3 s1.
Percentages that do not sum to 100% are due to rounding error.

the rate constant used in the Glarborg mechanism is about 17 times faster than the
rate in the Current mechanism at flame temperatures for this system.
There are other important differences between the Current and Glarborg
mechanisms. First, the Glarborg mechanism incorporates a different set of reactions
involving NHx=NxHy interactions, and fewer pathways involving N2Hx. Specifically,
the Glarborg mechanism incorporates different pathways involving the reactions of
N2H3 with radicals (e.g., O, H, HO2, and NH2) and, of particular importance, does
not contain the reactions NH þ NH2 þ M ! N2H3 þ M and NH3 þ NH2 ! N2H3 þ
H2. These latter two pathways have been found to be particularly important in
inhibiting the hydrogen-nitrous oxide system under fuel-rich flame conditions
(Powell et al., 2009), as well as for the flow reactor experimental conditions reported
by Allen et al. (1998). Second, the Glarborg mechanism (cf., Tian et al., 2009)
includes nitrogenous intermediates or CN species such as HCN, HNCO, and
CH2NH, which have been shown to form in the NH3=CH4 flames that they exam-
ined. Although these species will not impact the H2=N2O flames discussed here, their
impact on hydrocarbon-nitrous oxide flame predictions will be addressed in later
sections of this paper. Despite the mechanistic differences between the Current
and Glarborg mechanisms, however, the predicted flame speed results shown in
Figure 15 for H2=N2O flames using both of these mechanisms are quite similar.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 273

CH4/N2O flames. As an introduction to the discussion of hydrocarbon-


nitrous oxide combustion systems, a reaction path analysis for a CH4=N2O flame will
be contrasted with that of a CH4-air flame. Figure 16 shows a reaction path diagram
for a CH4=N2O=N2 flame at / ¼ 0.78 and diluted with nitrogen such that N2O=
(N2O þ N2) ¼ 0.42, whereas Figure 17 shows the reaction pathways for a CH4=air
flame at 1 atm and / ¼ 0.76. The peak flame temperature for the CH4-air and
CH4=N2O=N2 flame are 1940 K and 2290 K, respectively. Both reaction path
analyses were conducted at a location within the flame corresponding to 50% of
the temperature rise. In comparing these two figures, the following characteristics
are revealed:

. The production of the methyl radical (CH3) for both the nitrous oxide and air
flames occurs primarily from attack by H and OH radicals on molecular methane.
. The conversion of the methyl radical (CH3) to CH2O occurs primarily from attack
Downloaded by [University of Connecticut] at 07:29 12 October 2014

by O2, as well as O and OH radicals for both flames. In contrast to CH4-air


flames, however, the conversion of methyl radicals via CH3 þ O2 Ð CH2O þ OH
in CH4=N2O flames will be relatively unimportant due to low level of O2 in these
systems.
. In CH4-air flames, the production of the methoxy radical (CH3O) occurs primarily
from reaction of the methyl radical with HO2. The production of the hydroperoxy
radical HO2 via H þ O2 þ M Ð HO2 þ M, however, is diminished in importance in

Figure 17 Reaction pathways for CH4=air flame at 1 atm, / ¼ 0.76, and 50% of the temperature rise using
the Current mechanism; rate of progress units, kmol m3 s1. Percentages that do not sum to 100% are due
to rounding error.
274 O. A. POWELL ET AL.

the nitrous oxide system due to the lack of O2. Consequently, the conversion of
CH3 to CH3O from attack by the hydroperoxy radical, HO2, is relatively
unimportant for the N2O flame considered.
. The conversion of CH2O to the HCO radical occurs primarily from attack by H,
OH, and O radicals in CH4-air flames, whereas this conversion occurs primarily
from attack by H, OH, and CH3 in CH4=N2O flames.
. The conversion of the HCO radical to carbon monoxide, CO, occurs primarily
from collisions with third bodies (HCO þ M), as well as attack by O2 and H radi-
cals for CH4-air flames. In CH4=N2O flames, this conversion occurs primarily
from collisions with third bodies, as well as attack from NO.
. Finally, the oxidation of carbon monoxide, CO, to carbon dioxide, CO2, occurs
via the main reaction CO þ OH Ð CO2 þ H for hydrocarbon-air flames. In
hydrocarbon-nitrous oxide flames, however, oxidation of CO can also occur via
the direct exchange reaction CO þ N2O Ð CO2 þ N2. For the CH4=N2O flame
Downloaded by [University of Connecticut] at 07:29 12 October 2014

considered in Figure 16, about 79% of the CO is oxidized to CO2 directly by N2O.

Figure 18 shows laminar, unstretched, adiabatic flame speeds for CH4=N2O=N2


mixtures (N2O=(N2O þ N2) ¼ 0.42) using the Current mechanism compared to the
experimental flame speed data of Powell et al. (2009). The largest difference between
the Current results and the experimental data is 20% and occurs at / ¼ 0.60. The
peak, experimental flame speed value of 25.1 cm=s occurs at / ¼ 1.24, and the differ-
ence with the predicted value using the Current mechanism is 13%. Compared to our

Figure 18 Adiabatic, unstretched, laminar flame speed, SL0 , for CH4=N2O mixtures at 0.8 atm diluted with
N2 (N2O=(N2O þ N2) ¼ 0.42) compared with literature values. (.) experimental data from Powell et al.
(2009); (—) predictions using the Previous mechanism from Powell et al. (2009); () predictions using
the Current mechanism; (- - -) predictions using the Glarborg mechanism (Mendiara & Glarborg, 2009;
Tian et al., 2009); (– - –) predictions using the Glarborg mechanism (Mendiara & Glarborg, 2009; Tian
et al., 2009) with CN chemistry removed.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 275

Previous mechanism (Powell et al., 2009), the Current mechanism shows a closer
reproduction of the experimental results across all equivalence ratios.
For comparison purposes, Figure 18 also shows the results using the Glarborg
mechanism. This mechanism reproduces well the experimental data from Powell et al.
(2009). To account for the differences in the flame speed predictions using the differ-
ent mechanisms, Figure 19 shows a flame speed sensitivity analyses for CH4=N2O=N2
flames diluted with nitrogen (N2O=(N2O þ N2) ¼ 0.42) for fuel-lean, stoichiometric,
and fuel-rich conditions. For the stoichiometric condition (/ ¼ 1.0), the four most
sensitive reactions using the Current mechanism are (in order of decreasing sensi-
tivity): N2O(þM) Ð N2 þ O(þM), N2O þ H Ð N2 þ OH, NH þ NO Ð N2O þ H
(negative sensitivity), and CO þ OH Ð CO2 þ H. As discussed in our previous work
(Powell et al., 2009), the three most sensitive reactions for this CH4=N2O flame have
also been shown to be the most sensitive for hydrogen- and other hydrocarbon-
nitrous oxide flames. Thus, the most sensitive reactions with respect to the laminar
Downloaded by [University of Connecticut] at 07:29 12 October 2014

flame speed are within the hydrogen-nitrous oxide sub-mechanism for these flame
systems. A similar flame speed sensitivity analysis, not shown here, was conducted
using the Glarborg mechanism, and also revealed the same first three reactions
reported above to be the most sensitive. The different rate constant used for these
three reactions within the Current and Glarborg mechanisms partially accounts
for the differences in predictions shown in Figure 18. Although there are numerous
mechanistic differences between these two mechanisms, one characteristic worth
exploring further is the influence of reactions involving carbon-nitrogen (CN) species

Figure 19 Flame speed sensitivity for the most sensitive reactions (magnitude > j0.02j) using the Previous
mechanism for CH4=N2O flames diluted with N2 (N2O=(N2O þ N2) ¼ 0.42). (5) N2O þ H Ð N2 þ OH; (10)
NH þ NO Ð N2O þ H; (11) CO þ N2O Ð CO2 þ N2; (20) CO þ OH Ð CO2 þ H; (34) H2 þ OH Ð
H2O þ H; (35) CH2 þ O2 Ð H þ H þ CO2; (36) NH þ OH Ð HNO þ H; (37) N2O(þM) Ð N2 þ O(þM);
(38) N2O þ O Ð NO þ NO.
276 O. A. POWELL ET AL.

on the flame speed predictions for the CH4=N2O flames examined. The Glarborg
mechanism includes CN species and their associated chemical interactions, whereas
the Current mechanism does not. To examine the importance of this chemistry on
laminar flame speed predictions, Figure 18 also shows model predictions using the
Glarborg mechanism without the CN chemistry inherent in this same mechanism.
The Glarborg mechanism with its CN chemistry removed showed an increase in
flame speed predictions in the near stoichiometric and fuel-rich conditions. For
the conditions examined here, it appears that the exclusion of carbon-nitrogen chem-
istry, as in our Current mechanism, does not significantly impact flame speed predic-
tions. In the next section, the influence of carbon-nitrogen chemistry on C2H2=N2O
flames will be explored further; however, more research will be necessary to fully
understand the importance of this chemistry with regard to flame speed predictions,
in particular under undiluted flame conditions.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

C2H2/N2O flames. Figure 20 shows unstretched, adiabatic, laminar flame


speeds for C2H2=N2O=N2 mixtures (N2O=(N2O þ N2) ¼ 0.42) using the Current
mechanism compared to the flame speed data of Powell et al. (2009). With the
exception of / ¼ 0.60, the results using the Current mechanism reproduce, within
experimental uncertainty, the experimental data reported by Powell et al. (2009).
For comparative purposes, the flame speed predictions using the Glarborg
mechanism are also shown in Figure 20. Under fuel-rich conditions, the results using
the Glarborg mechanism somewhat overpredict the experimental results.

Figure 20 Adiabatic, unstretched, laminar flame speed, SL0 , for C2H2=N2O mixtures at 0.8 atm diluted with
N2 (N2O=(N2O þ N2) ¼ 0.42) compared with literature values. (.) experimental data from Powell
et al. (2009); (—) predictions using the Previous mechanism from Powell et al. (2009); (  ) predictions
using the Current mechanism (see Table 2); (- - -) predictions using the Glarborg mechanism (Mendiara
& Glarborg, 2009; Tian et al., 2009); (– - –) predictions using the Glarborg mechanism (Mendiara &
Glarborg, 2009; Tian et al., 2009) with CN chemistry removed.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 277

Figure 20 also shows the results from the Glarborg mechanism with its
associated CN chemistry removed. As with the CH4=N2O system discussed
previously, the flame speed determinations with the CN chemistry removed give rela-
tively higher flame speed values. The influence of the CN chemistry on the laminar
flame speed determinations can be expected to vary with the associated flame
temperature and thus the degree of dissociation. To explore the effect of flame tem-
perature, a series of computations were undertaken for C2H2=N2O flames at
/ ¼ 1.41, using the Glarborg mechanism with and without CN chemistry, for differ-
ent nitrogen dilution levels N2O=(N2O þ N2). For an undiluted flame (i.e., N2O=
(N2O þ N2) ¼ 100%) at / ¼ 1.41 and 1 atm, the unaltered Glarborg mechanism pre-
dicts a flame speed of 188.3 cm=s and a peak flame temperature of 3748 K. This flame
speed value can be compared to the experimental value of 170 cm=s reported by
Parker and Wolfhard (1952) at / ¼ 1.39 and our experimental value of
159  6.3 cm=s at / ¼ 1.41 and 0.81 atm based on an extrapolation (Powell et al.,
Downloaded by [University of Connecticut] at 07:29 12 October 2014

2009). Repeating this calculation with the CN chemistry removed from the Glarborg
mechanism leads to an increase in the flame speed by 6.5 cm=s to 195 cm=s. At a
dilution level of 42% (N2O=(N2O þ N2) ¼ 42%) and / ¼ 1.41, the flame speed
increased by 2.1 cm=s (from 52.4 cm=s to 54.5 cm=s) by removing the CN chemistry,
where the peak flame temperature was about 2840 K. This flame speed value can be
compared to our previously reported experimental value of 49  1.6 cm=s at 0.81 atm
(Powell et al., 2009). At a 22% dilution level and / ¼ 1.41, the flame speed increased
by only 0.22 cm=s (from 7.42 cm=s to 7.64 cm=s) by removing the CN chemistry,
where the peak flame temperature was about 2040 K. These results show that the
effect of CN chemistry on laminar flame speed determinations increases with increas-
ing flame temperature; consequently, inclusion of CN chemistry for nitrous oxide
flame speed determinations will be important for accurately modeling undiluted
conditions.

C3H8/N2O flames. Finally, Figure 21 shows flame speeds for C3H8=N2O=N2


mixtures (N2O=(N2O þ N2) ¼ 0.42) using the Current mechanism compared to the
experimental flame speed data of Powell et al. (2009). The largest difference between
the Current mechanism results and the experiment results is 17% and occurs at
/ ¼ 0.60. The difference at the peak, experimental flame speed value of 25.9 cm=s
(at / ¼ 1.10) is 6%. The predictions using the Current mechanism are comparable
to the previously reported predictions using the Previous mechanism (Powell et al.,
2009). Both mechanisms somewhat underpredict experimental flame speed determi-
nations across the entire range in equivalence ratio examined, pointing to the need
for further research on C3H8=N2O=N2 flames. Comparison with the Glarborg
mechanism was not possible, as this mechansim does not include C3 species.

Comparison with Ignition Delay Data


H2/N2O/Ar mixtures. Following the definition of Akbar et al. (2000) and
Kaneshige et al. (1997), the ignition delay or induction time was defined as the time
where the maximum rise in temperature occurred. Figure 22 shows the induction
times for H2=N2O=Ar mixtures using the Current mechanism compared to literature
data. Hidaka et al. (1985b) determined the induction time for these systems by
278 O. A. POWELL ET AL.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Figure 21 Adiabatic, unstretched, laminar flame speed, SL0 , for C3H8=N2O mixtures at 0.8 atm diluted with
N2 (N2O=(N2O þ N2) ¼ 0.42) compared with literature values. (.) experimental data from Powell et al.
(2009); (—) predictions using the Previous mechanism from Powell et al. (2009); (  ) predictions using
the Current mechanism.

Figure 22 Induction times for H2=N2O=Ar mixtures at 2 atm compared with literature values. 1%H2,
1%N2O: (.) experimental data from Hidaka et al. (1985b); (- - -) predictions using the Current mechanism.
1%H2, 2%N2O: (&) experimental data from Hidaka et al. (1985b); (  ) predictions using the Current
4
mechanism. 0.5%H2, 1%N2O: ( ) experimental data from Hidaka et al. (1985b); (—) predictions using
the Current mechanism.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 279

monitoring electronically energized OH* profiles behind reflected shock waves over
the temperature range 1400–2000 K. Their mechanism predicted somewhat longer
induction times compared to their experimental data. In their study, Hidaka et al.
(1985b) utilized the rate constant from Baulch et al. (1973) for the important reaction
N2O þ H Ð N2 þ OH, and this rate constant expression is appreciably slower than
the rate expression incorporated into the Current mechanism (the significance of this
reaction was discussed previously in this article). For example, at 2000 K, the rate
expression in the Current mechanism gives a value of 5.27  1012 cm3mol1sec1
compared to a value of 1.75  1012 cm3mol1sec1 using the rate expression from
Baulch et al. (1973). The faster rate constant for the important chain propagating
step, N2O þ H Ð N2 þ OH, in the Current mechanism leads to shorter calculated
induction times compared to the computations of Hidaka et al. (1985b). Overall,
the induction time computations using the Current mechanism reproduce the experi-
mental results from Hidaka et al. (1985b) quite well.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

CH4/N2O/Ar mixtures. Figure 23 shows the induction times for CH4=N2O=


Ar mixtures using the Current mechanism compared to literature data. Drummond
(1969) defined the induction time, or ignition delay time, as the time between passage
of a reflected shock and attainment of the maximum OH concentration after reaction
occurred. These authors detected CN through the use of spectrographic plates before
the reaction was completed; however, our induction time sensitivity studies, not
reported here, indicate that pathways involving CN species are not particularly

Figure 23 Induction times for CH4=N2O=Ar mixtures at various pressures compared with literature
values. 2%CH4, 8%N2O, 3.5 atm: (.) experimental data from Drummond (1969); (—) predictions using
4
the Current mechanism. 1%CH4, 4%N2O, 1 atm: ( ) experimental data from Soloukhin (1971); (- - -)
predictions using the Current mechanism. 4%CH4, 16%N2O, 3.5 atm: (þ) experimental data from
Drummond (1969); (— - —) predictions using the Current mechanism. 6%CH4, 24%N2O, 3.5 atm: (&)
experimental data from Drummond (1969); () predictions using the Current mechanism.
280 O. A. POWELL ET AL.

significant to the induction time of these systems. The computational results using
the Current mechanism, which does not include CN species, reproduce quite well
the experimental results of Drummond (1969) shown in Figure 23. Soloukhin
(1971) used CO2- and OH-emission records to determine the induction time in 1%
CH4=4% N2O systems diluted with Ar at temperatures of 1600–3200 K. Soloukhin
determined an overall activation energy of 29.5 kcal=mol from his experimental data.
As evidenced by the slope of the computational lines shown on Figure 23, the
computational results using the Current mechanism seem to capture the overall
activation energy of these systems quite well. Given the scatter in the experimental
data in Figure 23, the predictions using the Current mechanism for these CH4
mixtures are considered to be in reasonable agreement with the experimental results
shown.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

CONCLUSIONS
A detailed, chemical mechanism has been compiled and modified from
previously published mechanisms to model the combustion of C1–C3 hydrocarbon-
nitrous oxide mixtures. This chemical mechanism, which is still under development,
has been compared and validated against available data in the literature, including
the flow reactor studies of N2O decomposition (Allen et al., 1995), flow reactor
studies of H2=N2O=H2O=N2 mixtures (Allen et al., 1998), flame speeds studies of
CO=N2O with added hydrogen (Linteris et al., 2000), recent flame speed studies
for hydrogen-nitrous oxide and hydrocarbon-nitrous oxide mixtures (Powell et al.,
2009), flame structure studies of neat and NH3-doped H2=N2O flames (Venizelos
& Sausa, 1998), and NH3=N2O flames (Venizelos & Sausa, 2002), as well as shock
tube ignition delay studies of H2=N2O=Ar mixtures (Hidaka et al., 1985b) and
CH4=N2O=Ar mixtures (Drummond, 1969; Soloukhin, 1971). Although overall
predictions using the complied Current mechanism compare reasonably well with
these experimental data, further work is required to provide more extensive data
on fuel-nitrous oxide flames such as laminar flame speeds and flame structure
measurements to develop a more robust, comprehensive mechanism.

ACKNOWLEDGMENTS
Financial support from AFRL=RZA is gratefully acknowledged. The authors
also would like to thank Prof. Anthony Dean for his fruitful insights and suggestions
and Prof. P. Glarborg for kindly providing an electronic copy of the latest chemical
mechanism developed by his group.

REFERENCES
Akbar, R., Kaneshige, M., Schultz, E., and Shepherd, J. 2000. Detonations in H2–N2O–CH4–
NH3–O2–N2 mixtures. Technical report, Graduate Aeronautical Laboratories, California
Institute of Technology, Explosion Dynamics Laboratory Report FM97-3.
Allen, M.T., Yetter, R.A., and Dryer, F.L. 1995. The decomposition of nitrous oxide at
1.5 p 0.5 atm and 1103 T 1173 K. International Journal of Chemical Kinetics,
27(9), 883.
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 281

Allen, M.T., Yetter, R.A., and Dryer, F.L. 1997. High pressure studies of moist carbon
monoxide=nitrous oxide kinetics. Combustion and Flame, 109(3), 449.
Allen, M.T., Yetter, R.A., and Dryer, F.L. 1998. Hydrogen=nitrous oxide kinetics—
implications of the NxHy species. Combustion and Flame, 112(3), 302.
Arthur, N.L., Cooper, I.A., and Gershenzon, Y.M. 1997. Reaction of H atoms with N2O at
374–628 K. Journal of the Chemical Society, Faraday Transactions, 33, 3485.
Baulch, D.L., Bowman, C.T., Cobos, C.J., Cox, R.A., Just, T., Kerr, J.A., Pilling, M.J.,
Stocker, D., Troe, J., Tsang, W., Walker, R.W., and Warnatz, J. 2005. Evaluated kinetic
data for combustion modeling: Supplement II. Journal of Physical and Chemical
Reference Data, 34(3), 757.
Baulch, D.L., Drysdale, D.D., Horne, D.G., and Lloyd, A.C. 1973. Evaluated kinetic data for
high temperature reactions. Butterworths London, 1–3, 2.
Bosschaart, K.J., and de Goey, L.P.H. 2004. The laminar burning velocity of flames propagat-
ing in mixtures of hydrocarbons and air measured with the heat flux method. Combustion
and Flame, 136, 261.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

CHEMKIN-PRO1. 2008. Release 15082, Reaction Design, San Diego, Calif.


Cohen, N., and Westberg, K.R. 1991. Chemical kinetic data sheets for high-temperature
reactions, part II. Journal of Physical and Chemical Reference Data, 20(6), 1211.
Dean, A.M., and Bozzelli, J.W. 2000. Combustion chemistry of nitrogen. In Gardiner, W.C.,
Jr. (Ed.) Gas-Phase Combustion Chemistry. Springer, New York, p. 125.
Dean, A.M., Johnson, R.L., and Steiner, D.C. 1980. Shock-tube studies of formaldehyde
oxidation. Combustion and Flame, 37, 41.
Dean, A.M., Steiner, D.C., and Wang, E.E. 1978. A shock tube study of the H2=O2=CO=Ar
and H2=N2O=CO=Ar systems: Measurements of the rate constant for H þ N2O ¼ N2 þ
OH. Combustion and Flame, 32, 73.
Drummond, L. 1969. Shock-induced reactions of methane with nitrous and nitric oxides.
Bulletin of the Chemical Society of Japan, 42, 285.
Egolfopoulos, F.N., Cho, P., and Law, C.K. 1989. Laminar flame speeds of methane-air
mixtures under reduced and elevated pressures. Combustion and Flame, 76, 375.
Frolik, S.A. 2003. Hybrid rockets. Aerospace America, 41, 68.
Glarborg, P., Dam-Johansen, K., Miller, J.A., Kee, R.J., and Coltrin, M.E. 1994. Modeling
the thermal DeNOx process in flow reactors. Surface effects and nitrous oxide formation.
International Journal of Chemical Kinetics, 26(4), 421.
Glassman, I., and Yetter, R.A. 2008. Combustion, 4th ed., American Press, Boston, Mass.
Goodwin, D.G. 2003. An open source, extensible software suite for CVD process simulation.
Division of Engineering and Applied Science, California Institute of Technology.
Gu, X.J., Haq, M.Z., Lawes, M., and Woolley, R. 2000. Laminar burning velocity and
markstein lengths of methane-air mixtures. Combustion and Flame, 121, 41.
Henrici, H., and Bauer, S.H. 1969. Kinetics of the nitrous oxide-hydrogen reaction. Journal of
Chemical Physics, 50(3), 1333.
Hidaka, Y., Takuma, H., and Suga, M. 1985a. Shock-tube studies of N2O decomposition and
N2O-H2 reaction. Bulletin of the Chemical Society of Japan, 58, 2911.
Hidaka, Y., Takuma, H., and Suga, M. 1985b. Shock-tube study of the rate constant for excited
OH (2Rþ) formation in the N2O-H2 reaction. Journal of Physical Chemistry, 89(23), 4903.
Kaneshige, M., Akbar, R., and Shepherd, J.E. 1997. Hydrocarbon-air-nitrous oxide detona-
tions. Western States Section, Combustion Institute, Spring Meeting, April 14–15.
Kee, R.J., Rupley, F.M., Meeks, E., and Miller, J.A. 1995. Chemkin-III: A FORTRAN
chemical kinetics package for the analysis of gas-phase chemical kinetics. Sandia National
Laboratories Report SAND 96–8215.
Konnov, A.A. 2000. Detailed reaction mechanism for small hydrocarbons combustion. Release
0.5. Available at: http://homepages.vub.ac.be/~akonnov/ (accessed March 2007).
282 O. A. POWELL ET AL.

Law, C.K., Zhu, D.L., and Yu, G. 1988. Propagation and extinction of stretched premixed
flames. Proceedings of the Combustion Institute, 21, 1419.
Li, J., Zhao, Z., Kazakov, A., and Dryer, F.L. 2004. An updated comprehensive kinetic model
of hydrogen combustion. International Journal of Chemical Kinetics, 36, 566.
Lillich, H., Schuck, A., Volpp, H.-R., and Wolfrum, J. 1994. Kinetic studies of the reaction
NH(X3R) þ NO and NH(X3R) þ O at elevated temperatures. Proceedings of the
Combustion Institute, 25, 993.
Linteris, G.T., Rumminger, M.D., and Babushok, V.I. 2000. Premixed carbon monoxide-
nitrous oxide-hydrogen flames: Measured and calculated burning velocities with and
without Fe(CO)5. Combustion and Flame, 122, 58.
Marshall, P., Ko, T., and Fontijn, A. 1989. High-temperature photochemistry kinetics
studies of the reactions of H(12S) and D(12S) with N2O. Journal of Physical Chemistry,
93(5), 1922.
Mendiara, T., and Glarborg, P. 2009. Ammonia chemistry in oxy-fuel combustion of methane.
Combustion and Flame, 156, 1937.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Mertens, J.D., Chang, A.Y., Hanson, R.K., and Bowman, C.T. 1991. A shock tube study
of the reactions of NH with NO, O2 and O. International Journal of Chemical Kinetics,
23(2), 173.
Mick, H.J., Matsui, H., and Roth, P. 1993. High-temperature kinetics of Si atom oxidation by
NO based on Si, N, and O atom measurements. Journal of Physical Chemistry, 97, 6839.
Milks, D., and Matula, R.A. 1973. A single-phase shock-tube study of the reaction between
nitrous oxide and carbon monoxide. Proceedings of the Combustion Institute, 14, 84.
Miller, J.A., and Klippenstein, S.J. 2004. The H þ C2H2(þM) Ð C2H3(þM) and
H þ C2H2(þM) Ð C2H5(þM) reactions: Electronic structure, variational transition-state
theory, and solutions to a two-dimensional master equation. Physical Chemistry Chemical
Physics, 6(6), 1192.
Miller, J.A., and Melius, C.F. 1992. Kinetic and thermodynamic issues in the formation of
aromatic compounds in flames of aliphatic fuels. Combustion and Flame, 91(1), 21.
Mueller, M.A., Yetter, R.A., and Dryer, F.L. 2000. Kinetic modeling of the CO=H2O=O2=
NO=SO2 system: Implications for high-pressure fall-off in the SO2 þ O(þM) ¼SO3(þM)
reaction. International Journal of Chemical Kinetics, 32, 317. Available at: http://
www.princeton.edu/combust/database/other.html (accessed March 2007).
Parker, W.G., and Wolfhard, H.G. 1952. Some characteristics of flames supported by NO and
NO2. Proceedings of the Combustion Institute, 4, 420.
Pfahl, U.J., Ross, M.C., Shepherd, J.E., Pasamehmetoglu, K.O., and Unal, C. 2000. Flamm-
ability limits, ignition energy, and flame speeds in H2–CH4–NH3–N2O–O2–N2 mixtures.
Combustion and Flame, 123(1–2), 140.
Powell, O.A., Papas, P., and Dreyer, C. 2009. Laminar burning velocities for hydrogen-,
methane-, acetylene-, and propane-nitrous oxide flames. Combustion Science and
Technology, 181(7), 917.
Qin, A., Lissianski, V.V., Yang, H., Gardiner, W.C., Davis, S.G., and Wang, H. 2000. Com-
bustion chemistry of propane: A case study of detailed reaction mechanism optimization.
Proceedings of the Combustion Institute, 28, 1663.
Riley, P.S., Cosic, B., and Fontijn, A. 2003. The H þ NO recombination reaction over a wide
temperature range. International Journal of Chemical Kinetics, 35, 374.
Sausa, R.C., Singh, G., Lemire, G.W., and Anderson, W.R. 1996. Molecular beam mass
spectrometric and modeling studies of neat and NH3-doped low-pressure H2=N2O=Ar
flames: Formation and consumption of NO. Proceedings of the Combustion Institute,
26, 1043.
Smith, G.P., Golden, D.P., Frenklach, M., Moriarty, N.W., Eiteneer, B., Goldenberg, M.,
Bowman, C.T., Hanson, R.K., Song, S., Gardiner, W.C., Jr., Lissianski, V.V., and
HYDROGEN- AND C1–C3 HYDROCARBON-NITROUS OXIDE KINETICS 283

Qin, Z. GRI-Mech. Available at: http://www.me.berkeley.edu/gri-mech/ (accessed


March 2007).
Soloukhin, R. 1971. High-temperature oxidation of ammonia, carbon monoxide and methane
by nitrous oxide in shock waves. Proceedings of the Combustion Institute, 13, 121.
Tanzawa, T., and Gardiner, W.C., Jr. 1980. Thermal decomposition of ethylene. Combustion
and Flame, 39, 241.
Tian, Z., Li, Y., Zhang, L., Glarborg, P., and Qi, F. 2009. An experimental and kinetic
modeling study of premixed NH3=CH4=O2=Ar flames at low pressure. Combustion and
Flame, 156, 1413.
Vagelopoulos, C.M., Egolfopoulos, F.N., and Law, C.K. 1994. Further considerations on the
determination of laminar flame speeds with the counterflow twin-flame technique.
Proceedings of the Combustion Institute, 25, 1341.
Venizelos, D.T., and Sausa, R.C. 1998. Laser-induced fluorescence, mass spectrometric, and
modeling studies of neat and NH3-doped H2=N2O=Ar flames. Combustion and Flame,
115, 313.
Downloaded by [University of Connecticut] at 07:29 12 October 2014

Venizelos, D.T., and Sausa, R.C. 2002. Chemical kinetic modeling and mass spectrometric and
laser-induced fluorescence measurements of NH3=N2O flames. Technical report, Army
Research Laboratory Report ARL-TR-2782.

You might also like