ww.sciencedirect.com

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 1

Available online at w

journal homepage: www.elsevier .com/locate/he

Effect of N2/CO2 dilution on laminar burningvelocity of H2eair mixtures at high temperatures

Santosh K. Paidi a, Amrutha Bhavaraju b, Mohammad Akramb,Sudarshan Kumar b,*aDepartment of Mechanical Engineering, Indian Institute of Technology Bombay, Powai Mumbai 400 076, IndiabDepartment of Aerospace Engineering, Indian Institute of Technology Bombay, Powai Mumbai 400 076, India

a r t i c l e i n f o

Article history:

Received 28 June 2013

Received in revised form

29 July 2013

Accepted 4 August 2013

Available online 8 September 2013

Keywords:

Laminar burning velocity

Flame stabilization

Diverging channels

Diluted hydrogen flames

* Corresponding author. Tel.: þ91 22 2576 71E-mail addresses: sudar@aero.iitb.ac.in, s

0360-3199/$ e see front matter Copyright ªhttp://dx.doi.org/10.1016/j.ijhydene.2013.08.0

a b s t r a c t

The laminar burning velocities of H2eair mixtures diluted with N2 or CO2 gas at high

temperatures were obtained from planar flames observed in externally heated diverging

channels. Experiments were conducted for an equivalence ratio range of 0.8e1.3 and

temperature range of 350e600 K with various dilution rates. In addition, computational

predictions for burning velocities and their comparison with experimental results and

detailed flame structures have been presented. Sensitivity analysis was carried out to

identify important reactions and their contribution to the laminar burning velocity. The

computational predictions are in reasonably good agreement with the present experi-

mental data (especially for N2 dilution case). The burning velocity maxima was observed

for slightly rich mixtures and this maxima was found to shift to higher equivalence ratios

(V) with a decrease in the dilution. The effect of CO2 dilution was more profound than N2

dilution in reducing the burning velocity of mixtures at higher temperatures.

Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights

reserved.

1. Introduction combustible fueleair mixture and provides important informa-

The alarming issues of fossil-fuels depletion and increasing

concern for protection of environment calls for a shift from hy-

drocarbon fuels to cleaner alternatives, such as hydrogen fuel

based economy. Considering the slow developments in fuel cell

technology, the implementation of hydrogen or fuels blended

with hydrogen in gas turbine combustors and reciprocating type

internal combustionengines seems tobea feasible solution [1,2].

Thecharacteristicsof acombustionprocessaremainlygoverned

by laminar burning velocity of the fueleair mixture and it is

defined as the velocity with which a combustion wave propa-

gates into theunburned fueleairmixture inadirectionnormal to

the surface of combustion wave. Laminar burning velocity is

an important thermochemical property that characterizes a

24; fax: þ91 22 2572 2602.udar4@gmail.com (S. Kum2013, Hydrogen Energy P24

tion about reactivity, diffusivity and exothermicity of the given

fuels and their blends. The laminar burning velocity is also a

function of percentage of fuel in a given fueleair mixture, also

known as mixture equivalence ratio (V). Mixture equivalence

ratio is defined as the ratio of actual fueleairmass ratio to that of

stoichiometric fueleairmass ratio. It further indicates about the

availability of oxidizer in a mixture to completely burn the

available fuel in themixture (leanmixture,V< 1, stoichiometric

mixture, V ¼ 1 and rich mixture, V > 1). It is an important

property and helps in the development of new combustion sys-

tems, their flame stability, flame safety and emission control [3].

Laminar burning velocities of hydrogen, syngas (mixture of

H2 and CO), hydrogen enriched hydrocarbons and diluted

hydrogenmixtures have been reported in the literature [3e20].

ar).ublications, LLC. Published by Elsevier Ltd. All rights reserved.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 1 13813

Various methods have been widely used in the literature such

as constant-volume combustion bomb (expanding spherical

flame) method [3e12] using high-speed Schlieren photog-

raphy, heat-flux method [13,19], counter-flow stagnation

method [14,20] and burner-stabilized flame technique [16,21].

Qin et al. [15] have used a particle tracking velocimetry (PTV)

based technique and carried out image processing for burner-

stabilized flames to obtain the laminar burning velocities of

hydrogeneair mixtures. In these methods, the measured data

of burning velocities has been either extrapolated to zero heat

loss [13,19] or zero strain-rate conditions [3e12] to obtain

laminar burning velocity of a given fueleair mixture.

Recently, Natrajan et al. [16] have studied the effect of CO2

dilution on laminar burning velocity of lean syngasmixtures for

a wide range of fuel compositions, temperatures and pressures

using Bunsen flame approach and wall stagnation flame

approach. This study was limited to lean and stoichiometric

fueleairmixtures. Prathap et al. [17] and Kishore et al. [18] have

studied the effect of dilution with N2 and CO2 on equimolar

mixtureofH2eCOgases. Prathapet al. [17] observeda significant

decrease in laminar burning velocity ofH2eCOmixtureswithN2

dilution anda shift in theburningvelocity peak fromvery rich to

slightly richmixtures (V¼ 1.4 for60%N2dilution fromV¼2.0 for

no dilution). Kishore et al. [18] observed an increase in burning

velocity on increasing the hydrogen content and compared the

experimentally observed results to the Davis mechanism [22]

predictions. In most of these studies, the laminar burning ve-

locities at low temperatures (near ambient) have been reported

and very limited data exists for flames at high temperatures.

Since the operating temperatures inmany combustion systems

of practical importance such as gas turbines and IC engines are

relatively much higher, it is important to measure the burning

velocities at higher mixture temperatures.

Present work reports the effect of N2 or CO2 dilution as indi-

vidual components on laminar burning velocities of H2eair

mixtures at highmixture temperatures using the planar flames

appearing in high aspect-ratio diverging mesoscale channels

suggested by Akram et al. [23e26]. In this technique, planar

flamepropagationmodeappears for a rangeofmixtureflowrate

and equivalence ratio conditions in these diverging channels

[23e26]. Thismethodproducesapparatus independent andnear

adiabatic burning velocities [26]. The burning velocities were

measured for a range of equivalence ratios (0.8 � V � 1.3) and

percentage dilutions at high temperature (350e600 K).

The experimental data has been compared to the existing

kinetic models for H2eair mixtures with N2 and CO2 dilution.

The effect of dilution with these gases is expected to help

understand the role of specific heat and molecular weight of

different diluents on laminar burning velocity of H2eair mix-

tures. The key chemical reactions affecting the combustion

characteristics and burning velocities of diluted H2eair mix-

tures have been discussed.

2. Methods and materials

2.1. Details of the experimental setup

In the present work, an externally preheated diverging chan-

nel based technique has been used. This technique has been

proposed and validated by Akram et al. for various fueleair

mixtures [23e26]. Details of the experimental setup are shown

in Fig. 1. A high aspect-ratiomesoscale diverging channel with

a rectangular cross-section has been used. The inlet dimen-

sions and divergence angle of the channel are 25 mm � 2 mm

and 10� respectively. The properties of quartz such as low

thermal conductivity, high heat capacity, low thermal

expansion and transparency make it a suitable material for

flame visualization. The channel was externally preheated

with liquefied petroleum gaseair (LPGeair) mixture burnt at

the top of a porous burner. External preheating of the channel

helps compensate for the heat loss from the propagating

flame to the walls and stabilize the flame in the channel. The

H2eair mixture diluted with N2/CO2 gases was supplied at the

inlet of the channel at ambient conditions (Tu ¼ 300 K and

1.0 atm). The flow rates of these gases were controlled and

monitored using mass flow controllers connected to a per-

sonal computer through a command module. These mass

flow controllers were accurate within �1.5% of the full scale.

The channel wall temperatures were measured using K-type

thermocouples of diameter 0.5 mm. The motion of the ther-

mocouples was controlled using a precise traverse with a

minimum resolution of 0.25 mm. The measured wall tem-

peratures were accurate within �5 K of the actual value.

Akram and co-workers [17,18] have shown that the tempera-

ture varies linearly in the longitudinal direction and remains

almost uniform in the transverse direction except near the

exit plane for all mixture flow rate conditions.

2.2. Measurement of laminar burning velocity

The planar flames obtained in the diverging channel were

considered for different equivalence ratios, percentages of

dilution and external heating rates. The flame positions were

recorded using a photographic camera and the velocity of

propagating flame was obtained by applying the mass con-

servation equation for the diluted fueleairmixture supplied at

the channel inlet and mixture entering the planar flame front

as given below.

rinlet � Ainlet � Uinlet ¼ rf � Af � Su (1)

Here r is the density of the mixture, Af flame area, Sulaminar burning velocity and Uinlet is the mixture velocity at

channel inlet. Since, rf/rinlet¼ Tu,o/Tu (using ideal gas equation),

this equation can be rewritten in the following form.

Su ¼ Uinlet ��Ainlet=Af

�� ðTu=Tu;oÞ (2)

Here, Af is determined from the position of the stabilized

flame in the channel. With all other parameters known and

prior knowledge of temperature variation along the length of

the channel helps determining Tu. Substituting the values of

various parameters in Eq. (2) helps in determining the burning

velocity and its variation with mixture temperature as shown

in Fig. 2. The experiments are carried out for a range of

mixture velocities and equivalence ratios. For a given equiv-

alence ratio, the laminar burning velocity at standard condi-

tions (Su,o) and the temperature exponent (a) can be

determined by fitting a power law correlation to experimental

data as shown in Fig. 2. The laminar burning velocity at

Fig. 1 e Schematic diagram of the apparatus.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 113814

standard conditions (Su,o) and temperature exponent (a) can

be related to each other. The burning velocity at higher tem-

peratures can be obtained from the values of Su,o and a and

their dependence onmixture temperature as shown by Eq. (3).

Su ¼ Su;o � ðTu=Tu; oÞa (3)

2.3. Kinetic models

The laminar burning velocities of fueleair mixtures are pre-

dicted with various detailed mechanisms and the results are

compared to the present experimental results. San Diego

mechanism [27] and Li et al. [28] mechanismswere used for N2

dilution case. The predictions of both these reaction mecha-

nisms are very close to each other. Therefore, Li et al. [28]

mechanism has been employed in this work for the further

comparison of experimental data with the predictions, flame

structure studies and sensitivity analysis. Li et al. [28] mech-

anism and Davis H2eCO mechanism [22] were used for CO2

Fig. 2 e Laminar burning velocity H2eCO2eair mixture at

b [ 70% and V [ 1.0 and 1.2. The dotted line shows the

power law fit applied to the present data.

dilution cases. The Davis H2eCOmechanism [22] was used for

the case of CO2 dilution because some part of CO2 is expected

to dissociate back to CO within the stabilized flame front due

to high temperature. A brief summary of the various details

related to these mechanisms is given in Table 1. PREMIX [29]

code has been used to compute the laminar burning veloc-

ities for various mixture temperatures and equivalence ratio

conditions. Multicomponent diffusion and thermal diffusion

(Soret effect) were taken into account. The solution for the

burning velocity was seen to converge for the grid parameters

GRAD ¼ 0.05 and CURV ¼ 0.05.

3. Results and discussion

Laminar burning velocities were measured for various

mixture equivalence ratios (0.8 � V � 1.3), mixture tempera-

tures (350e600 K) and percentage dilution of H2eair mixture

with CO2 or N2 gas. The dilution of fueleair mixture with a

diluent, b is defined as the molar or volume percentage of

diluent in fuelediluent mixture (without air). Air or availabil-

ity of N2 in air is not considered in the present calculation of

dilution percentage.

Table 1 e Summary of kinetic mechanisms used in thepresent study.

Mechanism No. ofspecies

No. ofreactions

Diluent caseapplied for

Li et al. mechanism [28] 13 22 N2, CO2

San Diego mechanism [27] 28 52 N2

Davis H2eCO mechanism [22] 14 34 CO2

Percentage dilution; b ¼ No: of moles of diluent ðN2 or CO2ÞTotal no: of moles ½fuelþ diluent� in the mixture

� 100 (4)

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 1 13815

The effect of initial temperature, Tu, equivalence ratio,

V, and dilution, b on the laminar burning velocity of

diluted H2eair mixtures has been reported in the following

sections.

3.1. Planar flames

High aspect-ratio of a diverging channel provides a uniform

velocity and temperature distribution in the transverse di-

rectionwhich helps in the formation of planar flames. A direct

photograph of a planar flame stabilized in a diverging channel

for stoichiometric H2eair mixtures with 70% dilution (with

CO2 gas) at an inlet velocity of 1.0 m/s is shown in Fig. 3. The

stabilized planar flames are very thin which confirm that they

are planar in both transverse and depth directions. The heat

loss from these planar flames is compensated through

external heating and thermal feedback to upstream mixture.

Hence, the effect of heat loss on the reduction of burning ve-

locities is observed to be less than 5% [24,30].

3.2. Effect of mixture temperature on laminar burningvelocity

The laminar burning velocity has been observed to increase

with an increase in the initial mixture temperature of the

fuelediluenteair mixture for all conditions. This can be seen

from Fig. 2, which shows the variation of the laminar burning

velocity for H2eair mixtures with CO2 as diluent and b ¼ 70% at

V ¼ 1.0 and 1.2. The power law fit gives the temperature

exponent, a and laminar burning velocity Su,owhich can beused

Fig. 3 e Direct flame photograph of b [ 70% CO2 dilution

flame at V [ 1.0 and Uinlet [ 1.0 m/s.

to obtain laminar burning velocity at higher mixture tempera-

tures for given conditions of mixture equivalence ratio and

dilution using Eq. (2). The predicted values of a for various cases

of dilution were found by plotting the Su at various tempera-

tures with temperature ratio Tu/Tu,o and fitting a power law

curve at each equivalence ratio. The effect of mixture equiva-

lence ratio and percentage dilution on laminar burning velocity

was studied with both N2 and CO2 as diluents and obtained

experimental data was compared with the predictions.

3.3. H2eair mixtures diluted with N2 diluent gas

Inthepresentstudy, theexperimentswithN2asdiluentgaswere

performedwith b¼ 70% and 80%. For higher dilution cases, (e.g.

b¼ 90%), the visibility of the flameswas very poor and for lower

dilution rates, the flames were not stable within the diverging

portion of the channel for most cases due to very high burning

velocity of mixtures at highermixture temperatures.

3.3.1. Effect of the mixture equivalence ratio on temperatureexponentThe temperature exponent, a is a strong function of mixture

equivalence ratio of the mixture at a particular diluent per-

centage, b. For each diluent condition, the variation of a with

mixture equivalence ratio, V at different values of b has been

studied and compared with the predictions of various chem-

ical kinetics models. Various results for N2 diluted H2eair

flames are shown in Fig. 4.

Fig. 4a shows the variation of the predicted and measured

temperature exponent for two different dilution rates of

b ¼ 70% and 80% N2. Second degree polynomials were fitted to

the experimental data and predictions to explain the trend.

The temperature exponent decreases with an increase in the

equivalence ratio for b ¼ 70% case. For the case of b ¼ 80%, a

attains a minimum value for slightly richer mixtures

(1.1 < F < 1.2) and increases again. For lower dilution, a

minimum value perhaps exists for very rich mixtures. Li et al.

[28] mechanism predictions for the temperature exponent a

were observed to be in good agreement with the experimental

data for N2 dilution case.

3.3.2. Effect of the mixture equivalence ratio on laminarburning velocityThe laminar burning velocity first increases and then de-

creases after attaining a maxima on the richer side for a

given dilution percentage. The equivalence ratio at which

the maxima occurs depends on the dilution percentage, b.

Fig. 4b shows the variation of laminar burning velocity at

ambient conditions for the dilution rates of b ¼ 70% and 80%

with N2 gas. Second degree polynomials were fitted to the

experimental data and predictions to explain the trend. The

laminar burning velocity was found to increase with an

Fig. 4 e (a) Variation of temperature exponent, a (b) mixture

burning velocity with equivalence ratio at different diluent

percentages, b for N2 dilution case at 300 K. The dashed

lines show the Li et al. mechanism [25] predictions (c)

Laminar burning velocity at high mixture temperatures for

b [ 80%.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 113816

increase in the equivalence ratio initially, attain a maxima at

V ¼ 1.3 and then decreases for higher equivalence ratios for

b ¼ 70%. For the case of higher dilution with b ¼ 80%, similar

trend was observed with a maxima at V ¼ 1.2. The maxima

appears to shift towards higher equivalence ratios with a

decrease in the dilution, in agreement with the existence of

maxima at V ¼ 1.7 for pure H2eair mixtures. The Li et al. [28]

mechanism predictions for the laminar burning velocities

are in good agreement with the experimental data at both

b ¼ 70% and b ¼ 80% case. The Li et al. [28] predictions lie

within �5% error bars of the experimental data for both

b ¼ 70% and 80% N2 dilution.

At higher mixture temperatures, the variation of the

laminar burning velocity with equivalence ratio remains the

same, though the magnitude increases with mixture tem-

perature for a given equivalence ratio and dilution. Fig. 4c

shows the variation of laminar burning velocity with equiva-

lence ratio for b ¼ 80% N2 dilution and high temperatures.

3.3.3. Effect of dilution on the temperature exponent andlaminar burning velocityThe temperature exponent, a was found to increase with an

increase in the percentage dilution, b at all equivalence ratios.

This variation can be observed in Fig. 4a. The laminar burning

velocity was found to increase with a decrease in the per-

centage of dilution b for all equivalence ratios. This variation

can be seen from Fig. 4b. The maximum for the laminar

burning velocity was found to shift to richer side with a

decrease in dilution as noted above.

3.4. H2eair mixtures diluted with CO2 as diluent gas

3.4.1. Effect of the equivalence ratio on temperature exponentFig. 5a shows the variation of temperature exponent with

equivalence ratio for the dilution rates of b ¼ 60% and 70%

with CO2 as diluent gas. The second degree polynomial was

fitted to both experimental data and predictions using Li et al.

[28] and Davis H2eCOmechanism [22] as shown in Fig. 5a. The

temperature exponent, a has been observed to decrease with

an increase in equivalence ratio (for 0.8 � F � 1.3) for both

b ¼ 60% and 70% cases in the present study. Perhaps, there

could be a minima for richer mixtures with F � 1.3 in both the

cases. Although, the Li et al. [28] mechanism predictions are

not in complete agreement with the experimental data, they

are reasonably close to the experimental results. The pre-

dictions for b ¼ 70% case are relatively closer to experimental

data. The Davis H2eCO mechanism [22] completely over-

predicts the values of temperature exponent as compared to

the present experimental data.

3.4.2. Effect of the equivalence ratio on laminar burningvelocityThe variation of laminar burning velocity at ambient condi-

tions with equivalence ratio for the case of CO2 dilution with

b ¼ 60% and 70% is shown in Fig. 5b. The predictions from Li

et al. [28] mechanism are also shown in the figure for com-

parison with the experimental data. The laminar burning ve-

locity was initially found to increase with equivalence ratio,

attain amaxima at aroundV¼ 1.2 and then decrease at higher

mixture equivalence ratios for b ¼ 60% dilution case. A similar

trend was observed for b ¼ 70% case, but the maxima shifts to

V ¼ 1.1. The experimental data and predictions have been

fitted with a second degree polynomial and shown using

continuous lines in Fig. 5b. The shift in the maxima towards

increasingVwith a decrease in the dilution, b is similar to the

shift observed for N2 diluted mixtures. This shift in the

mixture equivalence ratio for maximum burning velocity is in

agreement with the fact that for pure H2eair mixtures, the

maxima occurs at around V ¼ 1.7. The laminar burning ve-

locities predicted using Li et al. mechanism [28] are relatively

close to the experimental data for b ¼ 70% and far away for

b ¼ 60% case. The predictions of Davis mechanism [22] are

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 1 13817

comparatively not in good agreement with the present

experimental data. For higher mixture temperatures, the

laminar burning velocity increases with equivalence ratio as

shown in Fig. 5c.

3.4.3. Effect of dilution on the temperature exponent andlaminar burning velocityThe temperature exponent, a increases with the dilution rates

for all equivalence ratios as shown in Fig. 5a. The laminar

burning velocity increases with a decrease in diluent per-

centage as shown in Fig. 5b. The variation of both temperature

exponent and laminar burning velocity with CO2 as diluent is

similar to that with N2 as diluent.

Fig. 5 e (a) Variation of temperature exponent, a (b) mixture

burning velocity with equivalence ratio at different diluent

percentages b for CO2 dilution. The dashed and dotted lines

show Li et al. mechanism predictions [25] and Davis

mechanism predictions [19] respectively. (c) Variation of

laminar burning velocity for high mixture temperatures at

b [ 70%.

3.5. Comparison of the effect of diluent on burningvelocity

The reduction in laminar burning velocity for CO2 dilution

case is higher as compared to N2 dilution case, as shown in

Fig. 6a. The higher heat capacity of CO2 along with the

chemical effect of CO2 dilution is perhaps responsible for

reduction in the burning velocity [30]. More details on the ef-

fect of dilution and diluent type are discussed in the next

section. A detailed comparison of the laminar burning velocity

data with dilution is shown in Fig. 6b. A straight line was fitted

to the experimental data at V ¼ 1.0 for both N2 and CO2 dilu-

tion cases. The predictions from Li et al. mechanism [28] are

shownwith a dashed line in the figure. It can be observed that

the Li et al.mechanism [28] predictions for theN2 dilution case

overlap with the experimental data indicating a good agree-

ment with experimental results. For the case of CO2 dilution,

the predictions of Li et al. mechanism [28] are closer to the

experimental data than Davis mechanism [22]. It can be

clearly seen from the slopes of straight lines that CO2 as a

diluent reduces laminar burning velocity by a larger extent

than N2.

3.6. Flame structure studies

To understand the effect of various parameters on laminar

burning velocity, detailed flame structure studies were con-

ducted using Li et al. mechanism [22] because the predictions

are in good agreement with the experimental data as reported

earlier.

Fig. 7 shows the variation of flame temperature along axial

direction for stoichiometric mixtures at various dilution con-

ditions. The adiabatic flame temperature is higher for N2

diluted flames than CO2 diluted flames at both 300 K and 500 K

mixture temperatures. Similarly, due to high heat release

rates at higher temperatures, the flame temperatures are

higher for mixtures at 500 K than at 300 K for both N2 and CO2

dilution cases. It is to be noted that the concentration of

diluent was kept constant at 70% for this comparative study

shown in Fig. 7.

Figs. 8 and 9 show the variation of concentration of various

stable species (i.e. H2, CO2, O2 andH2O) and unstable species or

radicals (H, OH, O, H2O2 and HO2) for stoichiometric mixtures

at different conditions. The origin of the abscissa of these

plots is an arbitrary location in the channel. The flame tem-

perature plotted on the secondary axis of Fig. 8a gives an idea

about the location of flame. The scale used for different plots

are different for better readability of these plots.

The magnitudes of the mole fractions of radicals H2O2 and

HO2 have been multiplied by a large number to ensure their

visibility on the plot along with the other radicals as shown in

Fig. 8b. This also shows that the maximum concentration of

these radicals is orders of magnitude lower than that of the

other radicals such as H, OH and O which implies that these

radicals do not significantly contribute to the reaction rates.

The concentration of H radical is significantly higher than OH

radical concentration for all the flames studied in this section.

The effect of the diluent type can be clearly seen by

comparing the plots shown in Fig. 8a and c which show the

flame structures for stoichiometric mixtures with b ¼ 70% for

Fig. 6 e (a) Comparison of laminar burning velocities at

b [ 70% with N2 and CO2 diluents at 300 K. (b) Variation of

laminar burning velocity at V [ 1.0 and T [ 300 K with

percentage dilution, b. Dashed and dotted lines show Li

et al. mechanism predictions [25] and Davis mechanism

predictions [19] respectively.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 113818

N2 and CO2 diluted mixtures at 300 K temperature. The H and

OH radical concentration peaks can be observed to be higher

for the N2 dilution case. The concentration peak of H radical is

about 5 times higher for N2 dilution case than CO2 dilution

case. According to Kwon and Faeth [8], the reduction in

Fig. 7 e Variation of predicted flame temperature along

axial direction for stoichiometric mixtures at various

dilutions and initial temperatures.

concentration peak of H radicals is expected to cause a

reduction in laminar burning velocity of flames involving

hydrogen and oxygen as reactants. This reduction in H con-

centration peak explains the higher reduction in laminar

burning velocity with CO2 dilution as compared to N2 dilution

case.

The effect of percentage dilution, b can be observed by

comparing the plots shown in Figs. 8a and 9a which show the

flame structure of N2 diluted flames at F ¼ 1 for ambient

conditions with b ¼ 70% and 80% respectively. The H and OH

radical peaks are about 3 times higher for b ¼ 70% than for

b ¼ 80% which is in agreement with higher laminar burning

velocity at lower dilutions. It can be seen that with increasing

dilution, themole fractions of various intermediate and stable

species decrease.

The comparison of the plots in Figs. 8d and 9d shows that

the H and OH radical concentrations at 500 K are significantly

higher than those at 300 K for b ¼ 70% CO2 dilution. It can be

inferred from this comparison that the concentration of H

and OH radicals significantly affects the laminar burning

velocity.

3.7. Sensitivity analysis

Since the Li et al. [28] mechanism predictions are close to the

experimental data (especially for N2 dilution), sensitivity

analysis was performed for stoichiometric mixtures at 300 K

for b ¼ 70% and 80% and at 500 K for b ¼ 70%. The sensitivity

analysis helps in identifying the important reactions affecting

the laminar burning velocity of various fueleair mixtures

considered in the present work. The normalized sensitivity

coefficients of seven most important reactions are shown in

Fig. 10. It can be seen that the chain branching reactions R1, R2

and R6 and the chain propagation reaction R3 which produces

H and OH radicals, have positive sensitivity coefficients and

hence increase laminar burning velocity. The remaining re-

actions are chain terminating reactions and have negative

sensitivities because they consume radicals resulting in a

reduction of laminar burning velocity. The magnitude of

negative coefficients, especially that of reaction R4 involving

the third body species M (and their chaperon efficiency for N2

and CO2 species) becomes significant with an increase in the

dilution from b ¼ 70e80% which explains the reduction in

laminar burning velocity. On comparing the sensitivity co-

efficients for b¼ 70% at 300 K and 500 K, it can be observed that

the positive coefficients are almost the same in both the cases

but the negative coefficient of the reaction R4 at 500 K is

significantly lower than that at 300 K mixture temperature,

thus explaining the increase in the laminar burning velocity

for higher mixture temperatures.

4. Conclusions

The present work reports the laminar burning velocities of

H2eair mixtures diluted with N2 and CO2 as diluent gases at

higher mixture temperatures using planar flames formed in

externally preheated mesoscale diverging channels. The

following conclusions can be drawn from the present work.

Fig. 8 e (a, b) Predicted flame structure of stoichiometric mixture with b[ 70% N2 dilution at 300 K using Li et al. mechanism

[25]. (c, d) Predictedflame structure of stoichiometricmixturewith b[ 70%CO2 dilution at 300Kusing Li et al.mechanism [25].

Fig. 9 e (a, b) Predicted flame structure of stoichiometric mixture with b [ 80% N2 dilution at 300 K using Li et al. mechanism

[25]. (c, d) Predicted flame structure of stoichiometricmixture with b[ 70% CO2 dilution at 500 K using Li et al. mechanism [25].

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 1 13819

Fig. 10 e Normalized sensitivity coefficients of the most dominant reactions computed using Li et al. mechanism [25] for

stoichiometric mixtures with N2 as diluent at different conditions.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 113820

1. The variation of laminar burning velocity of H2eair mix-

tures diluted with N2 and CO2 as diluent gases has been

studied experimentally and compared with predictions of

various detailed reaction mechanisms at higher mixture

temperatures.

2. The temperature exponent has been observed to decrease

for leaner mixtures with an increase in equivalence ratio

for almost all cases and increase after attaining a minima

for slightly richer mixtures (V ¼ 1.1e1.3).

3. Laminar burning velocity was found to be maximum for

slightly richer mixtures (V ¼ 1.1e1.3).

4. Laminar burning velocity increases with a decrease in the

percentage of dilution for both N2 and CO2 dilution cases.

Temperature exponent increases with an increase in the

dilution.

5. It was also noted that the effect of CO2 as diluent was more

profound than N2 gas due to higher specific heat of CO2.

The data at high temperatures will be very useful in the

design of various high temperature combustion devices.

Nomenclature

Ainlet cross-sectional area of the channel inlet, m2

Af cross-sectional area of the flame, m2

Su,o laminar burning velocity at the reference initial

temperature of 300 K, m/s

Su laminar burning velocity at temperature Tu, m/s

Tu,o initial temperature of the unburned mixture, K

Tu temperature of the unburned mixture, K

Uinlet inlet mixture velocity, m/s

Uairflow velocity of the cold air flow, m/s

a temperature exponent

b molarpercentageofdiluent in fuelediluentmixture,%

r density

F mixture equivalence ratio

r e f e r e n c e s

[1] White CM, Steeper RR, Lutz AE. The hydrogen-fueled internalcombustion engine: a technical review. Int J HydrogenEnergy 2006;31:1292e305.

[2] Blarigan PV, Keller JO. A hydrogen fuelled internalcombustion engine designed for single speed/poweroperation. Int J Hydrogen Energy 1998;23:603e9.

[3] Milton BE, Keck JC. Laminar burning velocities instoichiometric hydrogen and hydrogenehydrocarbon gasmixtures. Combust Flame 1984;58:13e22.

[4] Ilbas M, Crayford AP, Yilmaz I, Bowen PJ, Syred N. Laminar-burning velocities of hydrogeneair andhydrogenemethaneeair mixtures: an experimental study.Int J Hydrogen Energy 2006;31:1768e79.

[5] Wang J, Huang Z, Kobayashi H, Ogami Y. Laminar burningvelocities and flame characteristics of COeH2eCO2eO2

mixtures. Int J Hydrogen Energy 2012;37:19158e67.[6] Marley SK, Roberts WL. Measurements of laminar burning

velocity and Markstein number using high-speedchemiluminescence imaging.Combust Flame2005;141:473e7.

[7] Dowdy DR, Smith DB, Taylor SC. The use of expandingspherical flames to determine burning velocities and stretcheffects in hydrogen/air mixtures. Proc Combust Instit1991;23:325e32.

[8] Kwon OC, Faeth GM. Flame/stretch interactions of premixedhydrogen-fueled flames: measurements and predictions.Combust Flame 2001;124:590e610.

[9] Tse SD, Zhu DL, Law CK. Morphology and burning rates ofexpanding spherical flames in H2/O2/inert mixtures up to 60atmospheres. Proc Combust Instit 2000;28:1793e800.

[10] Lamoureux N, Djebayli-Chaumeix N, Paillard CE. Laminarflame velocity determination for H2eaireHeeCO2 mixturesusing the spherical bomb method. Exp Therm Fluid Sci2003;27:385e93.

[11] Qiao L, Kim CH, Faeth GM. Suppression effects of diluents onlaminar premixed hydrogen/oxygen/nitrogen flames.Combust Flame 2005;143:79e96.

[12] Hu E, Huang Z, He J, Miao H. Experimental and numericalstudy on laminar burning velocities and flame instabilities ofhydrogeneair mixtures at elevated pressures andtemperatures. Int J Hydrogen Energy 2009;34:8741e55.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 1 3 8 1 2e1 3 8 2 1 13821

[13] Hermanns RTE, Konnov AA, Bastiaans RJM, de Goey LPH.Laminar burning velocities of dilutedhydrogeneoxygenenitrogen mixtures. Energy Fuels2007;21:1977e81.

[14] Yu G, Law CK, Wu CK. Laminar flame speeds of hydrocarbonair mixtures with hydrogen addition. Combust Flame1986;63:339e47.

[15] Qin X, Kobayashi H, Niioka T. Laminar burning velocity ofhydrogeneair premixed flames at elevated pressure. ExpTherm Fluid Sci 2000;21:58e63.

[16] Natrajan J, Lieuwen T, Seitzman J. Laminar flame speeds ofH2/CO mixtures: effect of CO2 dilution, preheat temperature,and pressure. Combust Flame 2007;151:104e19.

[17] Prathap C, Ray A, Ravi MR. Investigation of nitrogen dilutioneffects on the laminar burning velocity and flame stability ofsyngas fuel at atmospheric condition. Combust Flame2007;155:145e60.

[18] Kishore VR, Ravi MR, Ray A. Adiabatic burning velocity andcellular flame characteristics of H2eCOeCO2eair mixtures.Combust Flame 2011;158:2149e64.

[19] Bosschaart KJ, de Goey LPH. Detailed analysis of the heat fluxmethod for measuring burning velocities. Combust Flame2003;132:170e80.

[20] Vagelopoulos CM, Egolfopoulos FN. Direct experimentaldetermination of laminar flame speeds. Proc Combust Instit1998;27:513e9.

[21] Pareja J, Burbano HJ, Amell A, Carvajal J. Laminar burningvelocities and flame stability analysis of hydrogen/airpremixed flames at low pressure. Int J Hydrogen Energy2011;36:6317e24.

[22] Wang H, You X, Joshi AV, Davis SG, Egolfopoulos F, Law CK.USC Mech version II. High-temperature combustion reactionmodel of H2/CO/C1eC4 compounds. http://ignis.usc.edu/USC_Mech_II.htm; May 2007.

[23] Akram M, Kumar S. Experimental studies on dynamics ofmethaneeair premixed flame in meso-scale divergingchannels. Combust Flame 2011;158:915e24.

[24] Akram M, Kumar S. Measurement of laminar burningvelocity of liquefied petroleum gaseair mixtures at elevatedtemperatures. Energy Fuels 2012;26:3267e74.

[25] Akram M, Kumar S. Experimental studies on the dynamics ofpremixed methane-air flames in various aspect ratiochannels. in: Proceedings of the 23rd ICDERS; Irvine, CA,USA, July 24�29, 2011.

[26] Akram M, Minaev S, Kumar S. Investigations on theformation of planar flames in mesoscale divergent channelsand prediction of burning velocity at high temperatures.Combust Sci Technol 2013;185:645e60.

[27] Chemicalekinetic mechanisms for combustion applications,http://web.eng.ucsd.edu/mae/groups/combustion/mechanism.html.

[28] Li J, Zhao Z, Kazakov A, Dryer FL. An updated comprehensivekinetic model of hydrogen combustion. Int J Chem Kinet2004;36:1e10.

[29] Kee RJ, Grear JF, Smooke MD, Miller JA, Meeks E. A FORTRANprogram for modeling steady laminar one-dimensionalpremixed flames. Report no. SAND85-8240; 1985.

[30] Akram M, Ratnakishore V, Kumar S. Laminar burningvelocity of propane/CO2/N2eair mixtures at elevatedtemperatures. Energy Fuels 2012;26:5509e18.