Fuel 89 (2010) 2211–2216

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

The effect of temperature on the adiabatic laminar burning velocities of CH4air

and H2air flames
A.A. Konnov *
Division of Combustion Physics, Dept. of Physics, Lund University, P.O. Box 118, S-221 00 Lund, Sweden

a r t i c l e i n f o a b s t r a c t

Article history: The effect of temperature on the adiabatic laminar burning velocities of CH4 + air and H2 + air flames was
Received 7 October 2009 analyzed. Available measurements were interpreted using correlation SL = SL0 (T/T0)a. Particular attention
Received in revised form 20 November 2009 was paid to the variation of the power exponent a with equivalence ratio at fixed (atmospheric) pressure.
Accepted 24 November 2009
Experimental data and proposed empirical expressions for a as a function of equivalence ratio were sum-
Available online 10 December 2009
marized. They were compared with predictions of detailed kinetic models in methane + air and hydro-
gen + air flames. Unexpected non-monotonic behavior of a was found in rich methane + air flames.
Keywords:
Modeling results are further examined using sensitivity analysis to elucidate the reason of particular
Methane
Hydrogen
dependences of the power exponent a on equivalence ratio.
Burning velocity Ó 2009 Elsevier Ltd. All rights reserved.
Temperature dependence

1. Introduction was used by Babkin and Kozachenko [3]. Coefficients C, a, D and E in

The adiabatic laminar burning velocity is a fundamental param-
eter of the combustible mixture, which depends on the stoichiom-
etric ratio, pressure and temperature. The laminar burning velocity
at standard conditions, i.e. atmospheric pressure and initial tem-
perature of 298 K is invaluable for characterization of combustion
properties of the given fuel, for understanding of the underlying
chemistry, validation of models, etc. However, in general applica-
tions, from domestic appliances to engines and gas turbines, pres-
sure and initial temperature of the mixture are often higher than
standard ones. It is therefore important to quantify the effects of
pressure and initial temperature on the adiabatic laminar burning
velocity of practical fuels such as methane.
Investigations of the laminar burning velocity of methane and
hydrogen flames are numerous and have quite long history. Varia-
tion of the laminar burning velocity, SL, with temperature was de-
scribed using different correlations. For instance, experiments
performed in 1950s and 1960s [1,2] were often approximated by
equations in the form
SL ¼ A þ BT a ;
where T is initial temperature of the mixture. It was early realized
that power exponent coefficient, a, varies with pressure and equiv-
alence ratio. Therefore complex equation in the form
SL ¼ CT a PDþET
* Tel.: +46 46 2229757.
E-mail address: Alexander.Konnov@forbrf.lth.se
Eq. (2) were found different for different equivalence ratios and
pressure ranges. The most popular correlation describing the effect
of initial temperature on the adiabatic laminar burning velocity
SL ¼ SL0 ðT=T 0 Þa ; ð3Þ
where T0 is the reference temperature and SL0 is the burning
velocity at this temperature, was used in this or equivalent form
since 1950s, e.g. [5,4]. Pressure dependence of the power expo-
nent coefficient, a, in Eq. (3) was sometimes neglected or aver-
aged, e.g. [6], yet its variation with equivalence ratio was
studied both in lean and rich flames. Accurate quantification of
the effect of temperature on the adiabatic laminar burning veloc-
ities over wide range of compositions is important from both fun-
damental and applied points of view. Burning velocity, SL, and its
variation with temperature in the form of Eq. (3) is often used for
calibration of turbulent burning velocities, e.g. [7,8]. This equation
is essential part of safety evaluation in relation to accidental
flame propagation in hydrogen + air + steam mixtures at nuclear
ð1Þ
power plants, e.g. [9].The goal of the present work was investiga-
tion of the variation of the power exponent a with equivalence
ratio at fixed (atmospheric) pressure. In the following available
experimental data and proposed empirical expressions for a are
summarized. They are compared with predictions of detailed ki-
ð2Þ netic models in methane + air and hydrogen + air flames. Model-
ing results are further examined using sensitivity analysis to
elucidate the rationale of particular dependences of the power
exponent a on equivalence ratio.

0016-2361/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.fuel.2009.11.038
2212 A.A. Konnov / Fuel 89 (2010) 2211–2216

2. Modeling details
The CHEMKIN – II collection of codes [10–12], including trans-
port properties [13] from Sandia National Laboratories, was used
for the modeling of adiabatic premixed flames. Multi-component
diffusion and thermal diffusion options were taken into account.
Adaptive mesh parameters were GRAD = 0.1 and CURV = 0.5. Rela-
tive and absolute error criteria were RTOL = 1.E-5 and ATOL = 1.E-9
respectively; total number of grid points was typically 150–200.
Three detailed kinetic mechanisms were used for the modeling.
First, methane + air flames were modeled using the Konnov de-
tailed C/H/N/O reaction mechanism for the combustion of small
hydrocarbons [14]. The current version of the mechanism (Release
0.6) consists of 1230 reactions among 129 species. It differs from
the previous Release 0.5 only in the part of the prompt-NO sub-
mechanism. Therefore it retains predicting capabilities and perfor-
mance of the previous version in terms of laminar flame modeling.
Validation of this Release 0.5 of the Konnov mechanism was sum-
marized recently [14]. In the present work the modeling was used
mostly to compare experimentally observed trends with those pre-
dicted by the calculations. No attempts to modify the mechanism
in order to improve its performance and agreement with the
experiments were made.
To avoid any ambiguity, methane + air flames were also modeled
using the GRI-mech. 3.0 [15]. Calculated burning velocities with the
GRI-mech. are slightly higher than with the Konnov mechanism,
however, the shape of the dependences of the laminar burning
velocity as a function of equivalence ratio are very close. Finally,
hydrogen + air flames were modeled using extensively validated
mechanism of hydrogen combustion [16]. The adiabatic laminar
burning velocities of CH4 + air and H2 + air flames were calculated
and analyzed within 298–398 K range. Normalized sensitivity coef-
ficients in premixed flames have been calculated using the Premix
code. Modeling has been performed in the range of equivalence ra-
tios from 0.5 to 2 for methane flames and from 0.35 to 8.5 for hydro-
gen flames. These ranges are wider than corresponding flammability
limits at standard conditions; one should remember, however, that
increase of the initial temperature extends these limits. Very rich
and very lean mixtures could be formed locally in turbulent non-
premixed or partially premixed flames burning in a flamelet regime.
3. Results and discussion
3.1. Methane + air flames
Andrews and Bradley [2] reviewed experiments with methane
flames prior to 1972 and noted considerable scatter in both the
magnitude of the burning velocity and its rate of increase with tem-
perature. For stoichiometric flames they proposed correlation
SL = 10 + 0.000459T2.31 based on the data reviewed in [17], which
was then corrected to the form SL = 10 + 0.000371T2 [2]. Since then
it was generally accepted that the power exponent a for methane
flames is 2, while for other hydrocarbon + air mixtures it ranges be-
tween 1.5 and 2 [18]. Metghalchi and Keck [19,20] investigated the
effect of temperature on the burning velocities of several fuels using
Eq. (3) and proposed linear function of the power exponent a with
equivalence ratio independent of the fuel type: a = 2.18–0.8(/  1)
in the range of equivalence ratios 0.8–1.2. Dahoe and de Goey [21]
analyzed both Eqs. (1) and (3) and processed essentially the same
experimental data as did Andrews and Bradley [2]. They concluded
that mean value of a in Eq. (3) for stoichiometric methane + air
flames is 1.89 with apparent scattering from 1.5 to 2.2.
Experimental data not covered in these reviews [2,21] are sum-
marized in Table 1 and shown in Fig. 1. Also included are the values
of a obtained by Babkin and Kozachenko [3] in lean flames, since
majority of the measurements have been performed in rather lim-
ited range of equivalence ratios, typically from 0.8 to 1.2. Babkin
and Kozachenko [3] found that the power exponent a rapidly in-
creases from stoichiometric to lean methane mixtures. This obser-
vation was subsequently confirmed by Lauer and Leuckel [22].
Sharma et al. [6] demonstrated that the power exponent a in-
creases also from stoichiometric to rich methane mixtures. In the
range of equivalence ratios 0.8–1.2 they proposed pressure-aver-
aged expressions for a as 1.68//½ in the lean flames and 1.68 /½
in the rich flames. Existence of the minimum of a in the close to
stoichiometric mixtures was also confirmed by Gu et al. [23], by
Liao et al. [24], and by Hermanns et al. [25]. In other studies, how-
ever, it was overlooked. Elia et al. [26] obtained a fixed value of a
(not shown in Fig. 1) over the range of equivalence ratios 0.8–
1.2. Iijima and Takeno [27] observed modest increase of a in the
range of / from 0.8 to 1.3, while Bose et al. [28] and Takizawa et
al. [29] found modest decrease of a from lean to rich mixtures.
Other works shown in Fig. 1 and listed in Table 1 presented values
of a only for stoichiometric flames [30–34]. The only linear func-
tion of the power exponent a with equivalence ratio which dra-
matically diverges from all available measurements in rich
mixtures was derived by Stone et al. [35]. They deduced burning
velocities from pressure–time measurements made in near zero-
gravity conditions in a spherical constant volume chamber. Origi-
nal results have not been reported, instead correlation as a function
of initial temperature and pressure was derived [35].
Due to existence of the minimum of a in the close to stoichiom-
etric mixtures it is logical to use quadratic function to approximate
its variation with equivalence ratio. Liao et al. [24] investigated

Table 1
Experimental data and empirical expressions for the power exponent a at fixed pressure close to 1 atm in methane + air flames.

Temperature range (K) Pressure range (atm) Range of / Correlation a at / = 1 Year Ref.
323–473 1–70 0.6–1 1.98 1966 [3]
290–525 0.6–4 1 1.726 1978 [30,31]
300–600 1–8 0.8–1.2 1.68//½ / < 1 1.68 1981 [6]
1.68 /½ / > 1
296–500 0.4–10 1 2.0–2.4 1986 [32]
291–500 0.5–30 0.8–1.3 1.6 + 0.22(/  1) 1.6 1986 [27]
298–500 1–8 1 1.8 1988 [33]
300–520 1–6 0.8–1.3 2.5839 1993 [28]
298–673 1–20 1, 0.588 1.95 1995 [22]
293–454 0.5–10.4 0.6–1.4 1.42–1.98(/  1) 1.42 1998 [35]
300–400 1–10 0.8–1.2 1.612 2000 [23]
298–550 0.75–70 0.8–1.2 1.857 2001 [26]
300–400 0.5–1.5 0.6–1.4 7.98–12.15/ + 5.75/2 1.58 2004 [24]
280–330 0.8–1.1 0.7–1.3 1.88–0.095(/  1) 1.88 2005 [29]
298–498 1–5 1 1.653 2007 [34]
298–418 1 0.8–1.2 9.11–14.8/ + 7.279/2 1.589 2009 [25]
 A.A. Konnov / Fuel 89 (2010) 2211–2216 2213

inaccuracies in the measurements of the laminar burning velocity

3 itself. It was until Wu and Law [36] demonstrated experimentally
that the flame stretch due to flame front curvature and/or flow
divergence must be taken into account in the data processing.
New stretch-corrected measurements in methane + air flames at
Temperature power exponent, α

elevated temperatures are quite limited; it is therefore interesting

2
1
0.4 0.8 1.2 1.6
Equivalence ratio
Fig. 1. Power exponent coefficient, a, in Eq. (3) for methane + air flames. Crosses:
[25], open diamonds: [34], solid diamonds: [22], open circles: [28], solid circle:
[30,31], open squares: [23], solid triangles: [3], open triangles: [6], long-dash line:
[27], dash-dot line: [35], dash-double dot line: [24], dash-triple dot line: [29], solid
line: modeling using the Konnov mech., short-dash line: modeling using the GRI-
mech. 3.0.
flames of natural gas with air in the range of / from 0.6 to 1.4 and
proposed quadratic function for a, which is in satisfactory agree-
ment with the measurements in methane + air flames as well
[22,25,34,6]. Hermanns et al. [25] determined the power exponent
a for methane + air and also for hydrogen enriched methane + air
flames. These measurements show moderate dependence of a on
the hydrogen content up to 20 vol% of H2 in the fuel. Quadratic
function for a was derived for the range of / from 0.8 to 1.2 [25].
Significant scattering of the available experimental data and
empirical expressions for the power exponent a is clearly due to
to analyze the effect of temperature on the adiabatic laminar burn-
ing velocities using detailed kinetic models over extended range of
equivalence ratios not yet accessed experimentally.
Contemporary measurements of laminar burning velocities are
always accompanied by numerical modeling. For example, Ogami
and Kobayashi [37] compared their experimental data at high tem-
peratures and high pressures with predictions of the GRI-mech. 3.0
and found good agreement at atmospheric pressure at room and at
higher temperatures. However, processing of the modeling results
to derive the power exponent a is very rare. Mishra [38] compared
experiments of Andrews and Bradley [2], Hill and Hung [33], Stone
et al. [35] and Garford and Rallis [30] in stoichiometric mixtures
with his modeling and concluded that most of them are in overall
satisfactory agreement. Derived from the modeling power temper-
ature coefficient was a = 1.575. Kobayashi et al. [8] noticed signif-
2.0
icant scattering of the existing data on pressure and temperature
exponents and used GRI-mech. 2.11 to derive the following values
of a: 1.9, 2.0, and 2.6 at / = 0.9, 0.8, and 0.6, respectively.
In the present work adiabatic laminar burning velocities of
methane + air have been calculated in the range of equivalence ra-
tios from 0.5 to 2. From these the power exponents a were derived.
Calculated dependences using the Konnov mechanism and the
GRI-mech. 3.0 are shown in Fig. 1. Both models are in good agree-
ment with the measurements in methane + air flames in the range
of / from 0.8 to 1.2 [22,25,34,6]. Furthermore, they qualitatively
reproduce rapid increase of a from stoichiometric to very lean
methane mixtures. Unexpected non-monotonic behavior of a was
found in rich methane + air flames. Both dependences peak at /
= 1.4, at higher equivalence ratios they level off approaching a = 2.
This variation of the power exponents a clearly indicates signif-
icant changes in the combustion chemistry that is modification of
the relative role of the elementary reactions. To elucidate the rea-
son of this change, modeling results were examined using sensitiv-
ity analysis. Normalized sensitivity coefficients with respect to

Reactions
0.6 10
11
35
36
0.4
38
45
52
53
Sensitivity

0.2
74
87
97
0.0 98
99
119
158
-0.2 166
167
168
284
-0.4
294
0.4 0.8 1.2 1.6 2.0
Equivalence ratio

Fig. 2. Normalized sensitivity coefficients for the laminar burning velocity of methane + air flames.
2214 A.A. Konnov / Fuel 89 (2010) 2211–2216

laminar burning velocity of methane + air flames are shown in
Fig. 2. They have been calculated with the step of 0.1 in equiva-
lence ratio. The most sensitive reactions are listed in Table 2. The
sensitivity analysis has been applied to methane flames in many
instances both in lean and in rich flames, e.g. [15,39]. To facilitate
comparison, reaction numbering in Fig. 2 and in Table 2 corre-
sponds to the numbering in the GRI-mech. 3.0 [15]. Table 2 also in-
cludes the sign of the normalized sensitivity coefficient for each
reaction. In full agreement with the earlier results [15] reaction
H þ O2 ¼ O þ OH
has the highest positive sensitivity at all conditions. In stoichiome-
tric and lean mixtures reaction
OH þ CO ¼ H þ CO2
has the second ranking among reactions with positive sensitivity.
From moderately lean up to very rich mixtures reaction
H þ CH3 ðþMÞ ¼ CH4 ðþMÞ
is the most important among reactions with negative sensitivity. In
the very lean mixtures reaction
H þ O2 þ H2 O ¼ HO2 þ H2 O
shows up as the most important termination step.
Fig. 2 clearly shows dramatic change of the role of many reac-
tions when equivalence ratio approaches 1.4. Sensitivity of reac-
tion (52) peaks at / = 1.4, sensitivity of reaction
H þ C2 H4 ðþMÞ ¼ C2 H5 ðþMÞ
also has a local maximum here. Even more important, sensitivities
of several reactions change their sign at / = 1.4, see circled region
in Fig. 2. This is noted in Table 2 as ‘‘+/” if the sign is positive at
/ < 1.4 and negative at / > 1.4, and ‘‘/+” in the opposite case.
Reactions
O þ CH3 ¼ H þ CH2 O
HCO þ H2 O ¼ H þ CO þ H2 O
HCO þ M ¼ H þ CO þ M
O þ CH3 ¼ H þ H2 þ CO
accelerate flame propagation from lean to moderately rich mixtures
thanks to production of H atoms, which speed up chain-branching
reaction (38). Reactions
O þ CH4 ¼ OH þ CH3 ð11Þ
H þ O2 þ H2 O ¼ HO2 þ H2 O ð35Þ
H þ O2 þ N2 ¼ HO2 þ N2 ð36Þ
HCO þ O2 ¼ HO2 þ CO ð168Þ
ð38Þ
inhibit flame propagation through consumption of active O and H
atoms and competition with reactions (10), (166), (167), and
(284). In the mixtures with equivalence ratio higher than 1.4 the
role of these two groups of reactions switches. H atoms are not
ð99Þ
key chain-leading radicals anymore, while CH3 radicals keep their
importance. Thus competition for H atoms formation in the first
group and H atoms consumptions in the second group loses its sig-
nificance, and formation of CH3 radicals in reaction (11) and con-
ð52Þ
sumption in reaction (10) play major role.
This consideration could be further substantiated by the com-
parison of the maximum concentrations of the key radicals in
ð35Þ methane + air flames as shown in Fig. 3. Flame structure modeling
clearly shows that OH radicals are most important in lean and stoi-
chiometric flames, H atoms are still important in moderately rich
flames, however their concentration at / > 1.4 drops dramatically.
Super-equilibrium concentrations of methyl radicals are not as
high as of OH and H, yet they become major chain-carriers (since
[CH3] > [H]) at equivalence ratios slightly above 1.4. One can con-
ð74Þ clude that present analysis allows to clearly distinguish four re-
gimes of methane + air combustion: lean, stoichiometric, rich,
and ‘‘over-rich” at equivalence ratios above 1.4.
Unexpected non-monotonic behavior of a found in rich meth-
ane + air flames could be anticipated in rich flames of other hydro-
carbons since reactions discussed above are not fuel-specific and
are pertinent to all H/C/O fuels. This also implies that no anomalies
ð10Þ should be expected in hydrogen + air flames. Analysis of the effect
of temperature on the adiabatic laminar burning velocities of
H2 + air flames presented in the following has been performed to
ð166Þ
ensure this conclusion.
ð167Þ
0.008
ð284Þ

Table 2
0.006
The most sensitive reactions with respect to the laminar burning velocity of
methane + air flames.

No. Reaction Sign

Mole fraction

10 O + CH3 = H + CH2O +/

11 O + CH4 = OH + CH3 /+ 0.004
35 H + O2 + H2O = HO2 + H2O /+
36 H + O2 + N2 = HO2 + N2 /+
38 H + O2 = O + OH +
45 H + HO2 = O2 + H2 
52 H + CH3 (+M) = CH4 (+M) 
53 H + CH4 = CH3 + H2  0.002
74 H + C2H4 (+M) = C2H5 (+M) +
87 OH + HO2 = O2 + H2O 
97 OH + CH3 = CH2 (S) + H2O +
98 OH + CH4 = CH3 + H2O 
99 OH + CO = H + CO2 +
119 HO2 + CH3 = OH + CH3O + 0.000
158 2CH3 (+M) = C2H6 (+M) 
0.4 0.8 1.2 1.6 2.0
166 HCO + H2O = H + CO + H2O +/
167 HCO + M = H + CO + M +/ Equivalence ratio
168 HCO + O2 = HO2 + CO /+
284 O + CH3 = H + H2 + CO +/ Fig. 3. Maximum concentrations of radicals in methane + air flames at standard
294 C2H3 + O2 = O + CH2CHO + conditions. Solid line: [OH], dashed line: [H], dash-dot line: [CH3], dash-double dot
line: [HO2]  10.
 A.A. Konnov / Fuel 89 (2010) 2211–2216 2215

Table 3
Experimental data and empirical expressions for the power exponent a at fixed pressure close to 1 atm in hydrogen + air flames.

Temperature range (K) Pressure range (atm) Range of / Correlation a at / = 1 Year Ref.
300–700 1 1–1.95 1.721 1957 [5]
296–523 1 0.52–4.4 1.571 + 0.3839  (0.42  x) (x < 0.42) 1.618 1983 [42]
1.571–0.2476  (0.42  x) (x > 0.42)
300–550 0.5–7 1 1.26 1984 [44]
291–500 0.5–25 0.5–4 1.54 + 0.026(/  1) 1.54 1986 [27]

3.2. Hydrogen + air flames
Measurements of the laminar burning velocity of hydrogen
flames are as numerous as those of methane. However, due to
practical reasons, air was considered most often as an oxidizer of
methane. Hydrogen combustion with air, however, is not fre-
quently implemented, and large body of experiments with hydro-
gen has been performed with oxidizers different from air.
Therefore, available data on the effect of temperature on the adia-
batic laminar burning velocities of H2 + air flames are very scarce.
Other oxidizers not covered in the present work vary from pure
oxygen, e.g. [40], to different combinations of oxygen, nitrogen
and steam, e.g. [41].
Available measurements of the power exponent coefficient a for
hydrogen + air flames are summarized in Table 3 and shown in
Fig. 4. It is a common practice of presenting experimental measure-
ments in hydrogen flames as a function of mole fraction of H2 in
the mixture, denoted here as x, not as a function of equivalence ra-
tio. This makes the graphs more ‘‘symmetric” because hydrogen
flames propagate up to very rich mixtures. Heimel [5] determined
the power exponent a in stoichiometric flames (x = 0.296) and in
the flames with the maximum burning velocity, which he found
at x = 0.45. Liu and MacFarlane [42] identified this maximum at
x = 0.42 and proposed linear correlations below and above this va-
lue with the minimum a = 1.571. Iijima and Takeno [27] covered
practically the same range of equivalence ratios and observed
modest increase of a without minimum.
Equivalence ratio
0.3 0.5 1 2
5
In the present work adiabatic laminar burning velocities of
hydrogen + air have been calculated in the range of equivalence ra-
tios from 0.35 to 8.5. From these the power exponents a were de-
rived as shown in Fig. 4. Modeling results confirm small variation
of a in the range of / from 1 to 3. In the very lean and very rich
mixtures, however, the temperature power exponent increases
dramatically. Sensitivity analysis reveals the reason of this behav-
ior. Normalized sensitivity coefficients for laminar burning velocity
of hydrogen + air flames calculated with the step of 0.03 in the
mole fraction x and shown in Fig. 5 agree with the previous calcu-
lations of Brown et al. [43] performed in the range of / from 0.7 to
2.8. Chain-branching reaction
H þ O2 ¼ OH þ O ðH16Þ
possesses dominant and rapidly increasing sensitivity towards rich
mixtures. Thus, any intensification of this reaction due to, e.g., tem-
perature rise will accelerate burning velocity. Reaction
H þ HO2 ¼ OH þ OH ðH20Þ
converts excessive H atoms in the very rich flames into OH, while
reaction
H þ HO2 ¼ H2 þ O2 ðH22Þ
serves as major terminating step. Inversely, in lean mixtures H
atoms produced from OH radicals in reaction
H2 þ OH ¼ H2 O þ H ðH17Þ
lead to acceleration of reaction (H16). The only reaction, which sen-
sitivity changes its sign close to the flammability limits is
4 6 8
2

Reactions

H1
Temperature power exponent, α

H9
4
H 15
H 16
1 H 17
H 20
Sensitivity

H 22
3
H 24

0
2

0.0 0.2 0.4 0.6 0.8 -1

Mole fraction of H2 0.0 0.2 0.4 0.6 0.8

Mole fraction of H2
Fig. 4. Power exponent coefficient, a, in Eq. (3) for hydrogen + air flames. Crosses:
[5], diamond: [44], dashed line: [42], dash-dot line: [27], solid line: present Fig. 5. Normalized sensitivity coefficients for the laminar burning velocity of
modeling. hydrogen + air flames.
2216 A.A. Konnov / Fuel 89 (2010) 2211–2216
H þ O2 ðþMÞ ¼ HO2 ðþMÞ:
Altogether, it is the high sensitivity of the key chain-branching and
chain-propagation reactions that leads to the rapid rise of the power
exponents a in the very lean and very rich flames of hydrogen.
4. Conclusions
The effect of temperature on the adiabatic laminar burning
velocities of CH4 + air and H2 + air flames was analyzed. Available
measurements were interpreted using correlation SL = SL0 (T/T0)a.
Experimental data and proposed empirical expressions for a as a
function of equivalence ratio of methane + air flames were first
summarized. They show significant scattering; only recent data
in rather narrow range of equivalence ratios (0.8–1.2) possess con-
sistency. Implementation of linear empirical expressions for the
temperature power exponent should be discouraged due to exis-
tence of the minimum of a in near-stoichiometric mixtures. Exper-
imental data were compared with predictions of detailed kinetic
models. Unexpected non-monotonic behavior of a was found in
rich methane + air flames. Modeling results were further examined
using sensitivity analysis to elucidate the reason of particular
dependences of the power exponent a on equivalence ratio. Pres-
ent analysis allowed to clearly distinguish four regimes of meth-
ane + air combustion: lean, stoichiometric, rich, and ‘‘over-rich”
at equivalence ratios above 1.4. The power exponent a for hydro-
gen + air flames does not show this peculiarity. However, it rapidly
increases in the very lean and very rich flames. This is explained by
the high sensitivity of the key chain-branching and chain-propaga-
tion reactions in hydrogen combustion.
References
[1] Dugger GL. Effect of initial mixture temperature on flame speed of methane–
air, propane–air, and ethylene–air mixtures. NACA Report 1061, Lewis Flight
Propulsion Lab.; 1952.
[2] Andrews GE, Bradley D. Combust Flame 1972;19:275–88.
[3] Babkin VS, Kozachenko LS. Combust Explo Shock Waves 1966;2:46–52.
[4] Dugger GL, Graab DD. Proc Combust Inst 1953;4:302–10.
[5] Heimel S. Effects of initial mixture temperature on burning velocity of
hydrogen–air mixtures with preheating and simulated preburning. NACA
Technical Note 4156, Lewis Flight Propulsion Lab.; 1957.
[6] Sharma SP, Agrawal DD, Gupta CP. Proc Combust Inst 1981;18:493–501.
[7] Koroll GW, Kumar RK, Bowles EM. Combust Flame 1993;94:330–40.
ðH9Þ [8] Kobayashi H, Kawahata T, Seyama K, Fujimari T, Kim J-S. Proc Combust Inst
2002;29:1793–800.
[9] Chan CK, Loesel Sitar JL, Beauvais R, Mayinger F. Nucl Eng Des 1996;166:
463–9.
[10] Kee RJ, Rupley FM, Miller JA. Sandia National Laboratories Report, SAND89-
8009; 1990.
[11] Kee RJ, Grcar JF, Smooke MD, Miller JA. Sandia National Laboratories Report,
SAND85-8240; 1990.
[12] Lutz AE, Kee RJ, Miller JA. Sandia National Laboratories Report, SAND87-8248;
1990.
[13] Kee RJ, Dixon-Lewis G, Warnatz J, Coltrin ME, Miller JA. Sandia National
Laboratories Report, SAND86-8246; 1990.
[14] Konnov AA. Combust Flame 2009;156:2093–105.
[15] Smith GP, Golden DM, Frenklach M, Moriarty NW, Eiteneer B, Goldenberg M,
et al. GRI-mech. 3.0. <http://www.me.berkeley.edu/gri_mech/>.
[16] Konnov AA. Combust Flame 2008;152:507–28.
[17] Andrews GE, Bradley D. Combust Flame 1972;18:133–53.
[18] Gülder ÖL. Correlations of laminar combustion data for alternative S.I. engine
fuels. SAE tech pap ser. SAE 841000; 1984.
[19] Metghalchi M, Keck JC. Combust Flame 1980;38:143–54.
[20] Metghalchi M, Keck JC. Combust Flame 1982;48:191–210.
[21] Dahoe AE, de Goey LPH. J Loss Prev Proc Ind 2003;16:457–78.
[22] Lauer G, Leuckel W. Arch Combust 1995;15:7–23.
[23] Gu XJ, Haq MZ, Lawes M, Wooley R. Combust Flame 2000;121:41–58.
[24] Liao SY, Jiang DM, Cheng Q. Fuel 2004;83:1247–50.
[25] Hermanns RTE, Konnov AA, Bastiaans RJM, de Goey LPH, Lucka K, Kohne H.
Fuel 2010;89:114–21.
[26] Elia M, Ulinski M, Metghalchi M. ASME Trans 2001;123:190–6.
[27] Iijima T, Takeno T. Combust Flame 1986;65:35–43.
[28] Bose PK, Sharma SP, Mitra S. Laminar burning velocity of methane–air mixture
in the presence of a diluents. In: Transport phenomena in thermal engineering.
Begell House, Inc.; 1993. p. 648–53.
[29] Takizawa K, Takahashi A, Tokuhashi K, Kondo S, Sekiya A. Combust Flame
2005;141:298–307.
[30] Garford AM, Rallis CJ. Combust Flame 1978;31:53–68.
[31] Rallis CJ, Garford AM. Prog Energy Combust Sci 1980;6:303–29.
[32] Mol’kov VV, Shamonin VG, Baratov AN. Flame propagation by a homogeneous
gaseous mixture at elevated pressures and temperatures. Combustion of
heterogeneous and gaseous mixtures. In: Proceedings of the eighth all-union
symposium on combustion and explosion. Chernogolovka: OIKhF AN SSSR;
1986. p. 11–4.
[33] Hill PG, Hung J. Combust Sci Technol 1988;60:7–30.
[34] Han P, Checkel MD, Fleck BA, Nowicki NL. Fuel 2007;86:585–96.
[35] Stone R, Clarke A, Beckwith P. Combust Flame 1998;114:546–55.
[36] Wu CK, Law CK. Proc Combust Inst 1984;20:1941–9.
[37] Ogami Y, Kobayashi H. JSME Int J 2005;48:603–9.
[38] Mishra DP. Fuel 2003;82:1471–5.
[39] Bui-Pham MN, Miller JA. Proc Combust Inst 1994;25:1309–15.
[40] Zitoun R, Deshaies B. Combust Flame 1997;109:427–35.
[41] Kusharin AYu, Popov OE, Agafonov GL. Khim Fiz 1995;14:179–89.
[42] Liu DDS, MacFarlane R. Combust Flame 1983;49:59–71.
[43] Brown MJ, Smith DB, Taylor SC. Combust Flame 1999;117:652–6.
[44] Milton BE, Keck JC. Combust Flame 1984;58:13–22.