DSpace at VNU: Experimental study on cellular instabilities in hydrocarbon hydrogen carbon monoxide-air premixed flames

DSpace at VNU: Experimental study on cellular instabilities in hydrocarbon hydrogen carbon monoxide-air premixed flames...

Experimental study on cellular instabilities in hydrocarbon/

a

Faculty of Civil Engineering, Ho Chi Minh City University of Technology, 268 Ly Thuong Kiet Street, District 10, Ho Chi Minh City, Viet Nam

bSchool of Mechanical Engineering, Pukyong National University, San 100, Yongdang-dong, Nam-gu, Busan 608-739, Republic of Korea

cEnvironment & Energy Research Division, Korea Institute of Machinery and Materials, 171 Jang-dong, Yuseong-gu,

Daejeon 305-343, Republic of Korea

a r t i c l e i n f o

Article history:

Received 10 August 2010

Received in revised form

12 February 2011

Accepted 15 February 2011

Available online 16 March 2011

Keywords:

Cell formation

Diffusional-thermal instability

Hydrocarbon

Hydrodynamic instability

Premixed flame

a b s t r a c t

To investigate cell formation in methane (or propane)/hydrogen/carbon monoxideeair premixed flames, the outward propagation and development of surface cellular instabil-ities of centrally ignited spherical premixed flames were experimentally studied in

a constant pressure combustion chamber at room temperature and elevated pressures Additionally, unstretched laminar burning velocities and Markstein lengths of the mixtures were obtained by analyzing high-speed schlieren images In this study, hydrodynamic and diffusional-thermal instabilities were evaluated to examine their effects on flame insta-bilities The experimentally-measured unstretched laminar burning velocities were compared to numerical predictions using the PREMIX code with a H2/CO/C1eC4 mecha-nism, USC Mech II, from Wang et al.[22] The results indicate a significant increase in the unstretched laminar burning velocities with hydrogen enrichment and a decrease with the addition of hydrocarbons, whereas the opposite effects for Markstein lengths were observed Furthermore, effective Lewis numbers of premixed flames with methane addi-tion decreased for all of the cases; meanwhile, effective Lewis numbers with propane addition increase for lean and stoichiometric conditions and increase for rich and stoi-chiometric cases for hydrogen-enriched flames With the addition of propane, the propensity for cell formation significantly diminishes, whereas cellular instabilities for hydrogen-enriched flames are promoted However, similar behavior of cellularity was obtained with the addition of methane, which indicates that methane is not a candidate for suppressing cell formation in methane/hydrogen/carbon monoxideeair premixed flames Copyrightª 2011, Hydrogen Energy Publications, LLC Published by Elsevier Ltd All rights

reserved

1 Introduction

The significance of global climate change and the depletion of

existing fossil fuels have led to the identification of

replace-ment fuels In this sense, hydrocarbons such as methane and

propane are considered to be an attractive potential fuel in spark-ignition engines[1,2] However, one of the problems is the release of carbon dioxide products if hydrocarbons are used as an alternative fuel In recent years, hydrogen has been widely used due to its advantages such as a high burning

* Corresponding author Tel.:þ82 51 629 6140; fax: þ82 51 629 6126

E-mail address:jeongpark@pknu.ac.kr(J Park)

A v a i l a b l e a t w w w s c i e n c e d i r e c t c o m

j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / h e

0360-3199/$e see front matter Copyright ª 2011, Hydrogen Energy Publications, LLC Published by Elsevier Ltd All rights reserved doi:10.1016/j.ijhydene.2011.02.085

velocity and cleanly emitted products[3,4] However, hydrogen

mixtures cause cells on the flame surface to occur earlier,

which can induce turbulence in the unburned mixture and

cause a rapid increase in the flame propagation velocity, which

can cause a gas explosion In addition, a mixture of hydrogen

and carbon monoxide (i.e., syngas), which can form through

the gasification process of a variety of resources such as coal,

biomass, organic wastes, and refinery residuals [5], is also

a potential fuel Therefore, the combustion characteristics of

premixed flames combined with the use of hydrocarbons,

hydrogen, and carbon monoxide as fuels have been

continu-ously studied

In premixed flames, in addition to the laminar burning

velocity, a corrugated flame front due to the formation of

cellular instabilities is an interesting consideration Three

effects are related to the cellularity of premixed flames In this

study, the cellular instabilities of hydrocarbon/hydrogen/

carbon monoxideeair premixed flames were identified and

evaluated with respect to hydrodynamic and

diffusional-thermal instabilities Whereas body-force effects were not

significant and could be neglected because the laminar

burning velocities of the flames mentioned in this study are

large enough such that the flames overcome the impact of the

body-force factor[4] In the early stages of flame development,

the flame instabilities are primarily influenced by a

diffusive-thermal factor However, as the flame develops and the flame

radius increases, the hydrodynamic factor becomes dominant

[6] Initially, cellular instabilities are suppressed by the strong

curvature associated with a small flame radius However, as

the flame expands and flame stretch decreases, a state is

reached in which the cell development can no longer be

suppressed, and, consequently, cells will appear almost

instantaneously over the entire flame surface, i.e., the onset of

cellular instabilities, which is represented by the critical

radius, R

In response to the interest in controlling the unstable behavior of cellular flames, numerous studies have been con-ducted regarding cell formation in hydrogeneair and hydrogen/ hydrocarboneair flames [6e10] In our previous studies, the effects of hydrocarbon additions and dilutions to the hydrody-namic and diffusive-thermal factors of the cellular instabilities

in the syngaseair premixed flames were analyzed and dis-cussed[11,12] Additionally, an understanding of the formation

of cellular instabilities in hydrocarbon/hydrogen/carbon mon-oxideeair flames is not sufficient and should be discussed further Therefore, this study focuses on the laminar burning velocities, Markstein lengths, behavior of cell formations, and transition to cellularity of methane/hydrogen/carbon mon-oxideeair and propane/hydrogen/carbon monmon-oxideeair pre-mixed flames by enriching 40%, 60%, and 80% (by volume) of hydrogen (H2) and adding 5%, 10%, 15%, and 20% (by volume) of methane (CH4) and propane (C3H8) to the fuel blends at room temperature and elevated pressures for overall equivalence ratios of 0.8, 1.0, and 1.2 using the centrally ignited, outwardly propagating spherical flames method This method yields highly accurate results for both laminar flame speeds and cellular flame instabilities and can easily account for high initial pressures and high initial temperatures[6,7,13,14]

Control circuit

Charge amplifier

Pressure transducer Ignition system

Knife edge

Pin hole Halogen lamp

Spherical concave mirror

1.000

Digital display Pressure

transmitter

Bypass

Vacuum pump

Data acquisition

H2

CO CH4 or

C3H8 Air

Fig 1e Schematic representation of the experimental setup

Table 1e Effective volumetric fraction of hydrocarbon in the fuel blend,a

CH4 C3H8 CH4 C3H8 CH4 C3H8

2 Experimental and computational details

The experiments were conducted in a stainless steel,

cylin-drical constant-volume chamber with an inside diameter of

200 mm and a length of 220 mm Visual access was provided by

two 100-mm diameter, 40-mm thickness quartz windows

mounted opposite of each other on both flat sides of the

chamber Two tungsten electrodes with a diameter of 0.5 mm

were linked to a high voltage source (up to 10 kV) to ignite the

combustible mixture at the center of the chamber The

elec-trodes were movable, and thus the spark gap was manually

adjustable The spark gap varied from 0.7 mm to 2.0 mm; larger

gaps were used to ignite flames with small laminar burning

velocities that required relatively large ignition energies The

reactant mixtures were prepared within the chamber by

add-ing individual component gases at correspondadd-ing partial

pressures using an absolute pressure transmitter to reach the

desired initial pressure, Pu A period of 15 min was used to

ensure complete mixing and quiescent conditions Once the

spark was ignited, a flame kernel formed at the center of the

chamber, propagated outward spherically, and quenched when it touched the walls of the chamber The propagating spherical flame was imaged using schlieren photography with

a 300-W halogen light source and a pair of 150-mm diameter spherical concave mirrors and was recorded using a high-speed digital camera (Phantom v7.2) that operated at 10,000 frames per second with a resolution of 512  384 pixels Simultaneously, the pressure inside of the chamber during combustion was measured using a water-cooled piezo-electric pressure transducer (Kistler 6061B) along with a charge amplifier (Kistler 5011B) and was transferred to a computer via

a data acquisition system (NI 9215A) After combustion, the chamber was vented to the laboratory exhaust system and purged using an air compressor to remove condensed water vapor prior to refilling for the next test The measurements were restricted to flames with radii larger than 6 mm and smaller than 30 mm The lower bound provided sufficient time for the removal of disturbances introduced by ignition as well

as minimized curvature and transient effects associated with the finite thickness of the flame, whereas the upper bound allowed the flames to avoid wall interference and limited the pressure increases during the measurement period to values less than 1.0% of the initial pressure[15] Details of the exper-imental setup are shown inFig 1

a

b

Fig 2e Experimental (points) and calculated (lines)

unstretched laminar burning velocities of various

hydrocarbon/hydrogen/carbon monoxideeair mixtures at

P [ 0.2 MPa and 4 [ 1.0

a

b

Fig 3e Markstein lengths of various hydrocarbon/ hydrogen/carbon monoxideeair mixtures at Pu[ 0.1 MPa and4 [ 0.8

In this study, the volumetric fraction of each fuel is given as

VHCþ VH 2þ VCO

(1) where i¼ CH4, C3H8, H2, and CO and VHC, VH 2and VCOare the

volumes of hydrocarbon (CH4or C3H8), hydrogen and carbon

monoxide in the fuel blends, respectively

To compare the effects of methane and propane to the

premixed flames, due to the heavier hydrocarbon consisting

of more carbon and hydrogen atoms and having more fuel

content, it will have a larger effect than smaller molecular

hydrocarbon for the same molar amount of addition

There-fore, it is useful to define one mixture using an effective

volumetric fraction of hydrocarbon in the fuel blend, which

can be expressed as

XHCVHCþVAirHCþ1XHC

2



VH 2þVAirH 2þ1XHC

2



VCOþVAirCO

(2) where VAireHC, VAireH 2, and VAireCOare the volumes of air

cor-responding to hydrocarbon, hydrogen and carbon monoxide,

respectively Corresponding values ofa of XHC¼ 5%, 10%, 15%,

and 20% of various hydrocarbon/hydrogen/carbon

mon-oxideeair mixtures are shown inTable 1

For a spherically expanding flame, the stretched flame velocity, Sn, which represents the flame propagation speed, is calculated from the instantaneous flame radius measured from the experiments using the following equation[16e18]

where R is the instantaneous radius of the flame in the schlieren photographs and t is time Therefore, Sn can be obtained directly from the images of the flame The flame stretch rate, K, is the Lagrangian time derivative of the loga-rithm of the area A of any infinitesimal element of the surface

[16,19,20]

K¼dðlnAÞ

dt ¼1 A

dA

where A is the surface area of the flame For a spherically outwardly expanding flame front, the flame stretch rate due to the combined effects of curvature and flame motion can be simplified as

K¼A1dAdt¼4pR128pRdRdt ¼2RdRdt¼2R n (5)

Fig 4e Effective Lewis numbers for various equivalence ratios in CH4/H2/COeair and C3H8/H2/COeair premixed flames

From Eqs.(3) and (5), the stretched flame velocity, Sn, and

the flame stretch rate, K, can be calculated The stretched

flame speed can be related to the flame stretch rate by the

linear relationship[2,16,17,20]

where S1is the unstretched flame speed and Lbis the burned

gas Markstein length that represents the influence of the

flame speed on the flame stretch rate The unstretched flame

speed, S1, can be obtained as the intercept value at K¼ 0, in the plot of Snagainst K, and the burned gas Markstein length, Lb, is the negative value of the slope of the SnK curve Due to S1

being known, the unstretched laminar burning velocity, which is defined as the unstretched upstream flame speed, can be determined from mass conservation as

S0

u¼ S1

r

b

ru



(7) Fig 5e Schlieren images of C3H8/H2/COeair flames for XC 3 H 8[0:15, 4 [ 1.2 at various initial pressures

Fig 6e Schlieren images of various CH/H/COeair and CH /H/COeair premixed flames

whereruis the density of the unburned mixtures andrbis the

density of the burned products

The PREMIX code[21]was used to predict the unstretched

laminar burning velocity, which was then compared to

experimental data The chemical mechanism in this study

was the H2/CO/C1eC4model, USC Mech Version II, which was

developed by Wang et al.[22]and consists of 784 elementary

reactions with 111 species This model was chosen because it

includes all of the species required in this study

3 Factors that influence cellular instabilities

Hydrodynamic effects have the most significant influence on

the flame instability and are caused by the thermal expansion

ratio through the flame front, which is defined as the ratio of

the density of unburned gas (ru) to the density of burned gas

(rb) at two sides of the flame front[23] The flame thickness is

also an important parameter that influences hydrodynamic

instability; if the flame is thin, then it will reduce the influence

of curvature and enhance the baroclinic torque intensity,

which is dependent on the density gradient across the flame

and the transverse pressure gradient along the flame

Vr  VP=r2[24]

Diffusive-thermal effects are caused by the preferential

diffusion of mass versus heat and are represented by the

Lewis number, Le, which is a ratio of the heat diffusivity of the

mixture to the mass diffusivity of the limiting reactant relative

to the abundant inert[16,25] If the Lewis number of the flame

is smaller than, equal to or larger than the critical value, Leeff

(slightly less than unity), then the flame will be unstable,

neutral or stable regarding the diffusive-thermal effect,

respectively

In this study, three fuels (H2, CO and hydrocarbon (HC)) were

used in the mixture; thus, the fuel Lewis number of the

reactant is a weighted average of the Lewis numbers of the

three fuels[9e11], which is given as

LeF¼ 1 þqHCðLeHC 1Þ þ qH 2 LeH 2 1
þ qCOðLeCO 1Þ

where LeHC, LeH 2 and LeCOare the fuel Lewis numbers of the

hydrocarboneair mixture at fHC¼ ðXHC=XAÞ=ðXF=XAÞst,

hydro-geneair mixture at fH 2¼ ðXH 2=XAÞ=ðXF=XAÞst, and carbon

monoxideeair mixture at fCO¼ ðXCO=XAÞ=ðXF=XAÞst,

respec-tively XFand XAare the mole fractions of fuel and air in the

reactant mixture [9,11], q¼ qHCþ qH 2þ qCO is the total heat

release, where qj( j¼ HC, H2, CO) is the nondimensional heat

release associated with the consumption of species j, which is

defined as qj¼ QYj=cPTu, where Q is the heat of reaction, Yjis

the supply mass fraction of species j, cPis the specific heat of

the unburned gas and Tuis the unburned gas temperature[10]

The effective Lewis number is defined as the combination

of the fuel and oxidizer Lewis numbers[11,12,26]

Leeff¼ 1 þðLeE 1Þ þ ðLeD 1ÞA1

1þ A1

(9)

where LeEand LeD are the Lewis numbers of excessive and deficient reactants, respectively A1¼ 1 þ bðF  1Þ is

a measurement of the mixture strength, whereF is a ratio of the mass of excess-to-deficient reactants in the fresh mixture relative to their stoichiometric ratio (F ¼ 1/4 for 4  1 and F ¼ 4 for 4 > 1) and b ¼ EaðTad TuÞ=R0T2

ad is the Zeldovich number, where Tad the adiabatic flame temperature,

Ea¼ 2R0p½vlnðruS0

uÞ=vð1=TadÞ is the activation energy, R0is the universal gas constant Tadalong withruandrbin the previous section were assumed to be in equilibrium and were calculated using the EQUIL code[27]

The laminar flame thickness, lf, is a characteristic length scale that is used to evaluate the hydrodynamic instability and to normalize the critical radius to obtain the critical Peclet number for the onset of cellular instabilities, which will be discussed in further detail later In this study, the character-istic flame thickness is given by Law et al.[10]as

lf¼l=cP

ruS0 u

(10) wherel and cPare the thermal conductivity and specific heat

at 1200 K, respectively, which is an approximate average of the free stream and flame temperatures[10].s ¼ ru/rbis referred

to as the thermal expansion ratio

4 Results and discussion

Markstein lengths The laminar burning velocity is one of the key parameters in combustion research Thus, an accurate measurement of the

Fig 7e Comparison of the suppression of cellular instabilities of H2/COeair premixed flames with CH4and

C H additions at similara

unstretched laminar burning velocity is necessary to assess

combustion theories and for the validation of numerical

models [1,13,28] Fig 2 compares experimentally-measured

and predicted unstretched laminar burning velocities of

various hydrocarbon/hydrogen/carbon monoxideeair

mixtu-res at Pu¼ 0.2 MPa and 4 ¼ 1.0; the two results are in good

agreement The unstretched laminar burning velocities

increase significantly along with an increase in the amount of

H2, and this increasing tendency is stronger at high percentages

of hydrogen concentration On the other hand, the unstretched

laminar burning velocities decrease along with hydrocarbon

addition to the fuel blend The primary reason for this decrease

is thermal effects due to an increase in the heat release and

thus an increase in the adiabatic flame temperature for

hydrogen enrichment, whereas the opposite tendency was

observed for hydrocarbon addition[29,30]

As previously mentioned, the negative slope of the linear

relationship between Snand K is defined as the burned gas

Markstein length, Lb If Lb< 0, then the flame speed increases

along with an increase in the flame stretch rate In this case, if

any protuberance occurs on the flame front, then the flame

speed increases, which increases the instability of the flame

On the other hand, for the case of a positive Lb, the flame front

instabilities will be restricted, and the flame will be stable

[2,20].Fig 3shows that the Markstein length decreases along

with an increase in hydrogen enrichment, whereas it

increases along with an increase in hydrocarbon addition

This effect indicates that the flame instability becomes more susceptible to an increase in the hydrogen fraction, and the flame front becomes more stable with the addition of hydro-carbons in the reactant mixture

Flame stability is another important characteristic of pre-mixed flames As previously mentioned, two instabilities of premixed flame were observed in this study: diffusional-thermal instability and hydrodynamic instability The effec-tive Lewis numbers, Leeff, which represent the influence of diffusive-thermal effects on premixed flames, are analyzed

Fig 4shows the effective Lewis numbers of CH4/H2/COeair and C3H8/H2/COeair flames for different equivalence ratios This figure indicates that the effective Lewis numbers of all of the mixtures are larger than unity; therefore, the diffusional-thermal instability can be sufficiently suppressed As the content of H2concentration increases in the fuel blends, as shown inFig 4a and b, the effective Lewis numbers increase for rich and stoichiometric mixtures and decrease for lean mixtures Conversely, the effective Lewis numbers of pre-mixed flames with propane addition increase for lean and stoichiometric mixtures and decrease for rich mixtures, as shown inFig 4d This effect is due to opposite tendencies of the effective Lewis number of hydrogeneair flames and pro-paneeair flames[8,31] This indicates that a modulation of the

Fig 8e Experimentally-measured critical radii for the onset of cellular instabilities for various equivalence ratios in CH4/H2/

COeair and C H/H/COeair premixed flames

diffusional-thermal instability is obtained for rich and

stoi-chiometric flames of the amount of hydrogen increases,

whereas this modulation is attained for lean and

stoichio-metric mixtures for the propane addition cases.Fig 4c shows

that the effective Lewis numbers in the premixed flames with

methane addition always slightly decrease along with an

increase in the methane fraction; thus, the addition of

methane may not be effective to diminish the

diffusional-thermal instability for syngaseair flames

The effects of initial pressure to destabilize the flame front

of the mixtures are observed by analyzing schlieren images of

the premixed flames by adding 15% propane by volume, as

shown inFig 5 Three important parameters (Leeff,s, lf) are

tabulated at the bottom of the figure For Pu¼ 0.1 MPa, the

flame front remains smooth; however, the cells form earlier,

and the size of the cells is smaller at higher Pu As the initial

pressure increases, the diffusive-thermal effect does not

affect the destabilization of the flame front because the

effective Lewis number maintains the same amount of

pres-sure change This effect can be attributed to the

hydrody-namic instability, which is related to the thermal expansion

ratio, s, and the flame thickness, lf The augmentation of

cellular instabilities along with an increase in the initial

pressure results from an enhancement of the hydrodynamic

instability due to a substantial decrease in the flame

thick-ness, whereas the thermal expansion ratio maintains nearly

the same value

The flame instability characteristics of hydrocarbon/ hydrogen/carbon monoxideeair flames with hydrogen enrichment as well as methane and propane additions are shown inFig 6 For hydrogen enrichment,Fig 6a and b indi-cates that the flame front destabilizations are promoted due to

an enhancement in both the diffusional-thermal instability due to the decrease in the effective Lewis number and the hydrodynamic instability caused by the decrease in the flame thickness and the small changes in the thermal expansion ratio For the methane addition,Fig 6c indicates that there are

no differences in the sequences of the flame front surfaces between the premixed flames with and without methane addition As the amount of methane addition in the fuel blend increases, the thermal expansion ratio maintains a nearly constant value, and the flame thickness slightly increases; thus, they only slightly affect the hydrodynamic effect However, the effective Lewis number slightly decreases to promote the diffusive-thermal effect The two effects combine and cause the premixed flames with methane addition to exhibit flame surfaces similar to the original flame Schlieren images of the premixed flames with propane added to the reactant mixtures are shown inFig 6d, which indicate that the propensity of stabilization tends to be progressively promoted For propane addition, the thermal expansion ratio increases, and the flame thickness also increases Thus, the net effect of the two factors related to hydrodynamic insta-bility is negligible The remaining parameter that affects the

Fig 9e Experimentally-measured critical Peclet numbers for various equivalence ratios in CH4/H2/COeair and C3H8/H2/

COeair premixed flames

suppression of cellular instabilities is the effective Lewis

number In this case, the effective Lewis number significantly

increases along with an increase in the propane

concentra-tion; therefore, it will affect the flame front instabilities due to

the suppression of the diffusive-thermal effect

To examine the effects of methane and propane for

suppression of cellular instabilities of H2/COeair flames,Fig 7

shows a comparison of schlieren images of the flame surfaces

in H2/COeair flames with methane and propane additions for

4 ¼ 0.8 and Pu¼ 0.2 MPa at similar a As shown in this figure,

similar behavior of the flame front instabilities is observed for

the H2/COeair and CH4/H2/COeair flames Furthermore, the

cells form for the methane addition case, whereas wrinkles do

not form for the C3H8/H2/COeair flame, and only a few large

cracks are observed The thermal expansion ratio and the

flame thickness of H2/COeair flames with methane and

propane additions are larger compared to those of the H2/

COeair flame Therefore, the net effect of hydrodynamic

instability is negligible However, for the propane addition

case, the effect of diffusional-thermal instability diminishes

due to a significant increase in the effective Lewis number As

a result, a combination of the two effects causes the H2/

COeair flame with propane addition to affect the cellular

instabilities compared to the CH4/H2/COeair flame Therefore,

the H2/COeair flame can be suppressed by adding propane to

the fuel blend

In this study, the onset condition of flame instabilities was

obtained from the plot of the stretched flame velocity versus

the flame stretch rate (Eq.(6)) and is represented by the critical

radius, Rcr In this case, the critical radius was detected at the

moment when the burning velocity increased significantly

and lost the linear relationship between Sn and K [9] The

critical radius was also determined from the sequence of

flame images in which the cells appeared spontaneously and

uniformly over the entire flame surface Both methods had

similar results.Fig 8shows the critical radius at the onset of

instabilities of the CH4/H2/COeair and C3H8/H2/COeair flames

for different equivalence ratios For hydrogen enrichment, as

shown inFig 8a and b, the critical radii significantly decrease,

which indicates that cells form at earlier stages, whereas the

critical radius of the CH4/H2/COeair flames is slightly larger, as

shown inFig 8c For propane addition,Fig 8d shows that the

critical radius significantly increases As a result, the larger

critical radius of the C3H8/H2/COeair flame compared to that

of the CH4/H2/COeair flame is in good agreement with the

result of the comparison of the suppression for cellular

instabilities between methane and propane additions to the

H2/COeair flames shown inFig 7 This indicates that the onset

of instabilities will be delayed if the propane concentration

increases, whereas a similar situation occurs in the methane

addition case The influence of diffusive-thermal effects on

the results of flame instabilities is also shown inFig 8d The

critical radii of lean flames are smaller or larger than those of

rich flames at small or large percentages of propane addition,

respectively, because of an increase or decrease in the

effec-tive Lewis numbers in lean/rich H2/COeair flames with

propane addition, as shown inFig 4d

Fig 9 shows the experimentally-measured critical Peclet number, Pecr, which is defined as the critical radius normal-ized by the flame thickness, Pecr¼ Rcr/lf, for various equiva-lence ratios of CH4/H2/COeair and C3H8/H2/COeair premixed flames As shown inFig 9a and b, the critical Peclet number significantly decreases along with hydrogen enrichments, which indicates that the flame front instabilities appear at smaller radii.Fig 9c indicates that the critical Peclet number maintains almost the same values for CH4/H2/COeair flames Meanwhile, the critical Peclet number increases significantly along with an increase in propane in the fuel blends, as shown

inFig 9d, which indicates that the stabilizing effect becomes stronger if propane is added to the H2/COeair mixtures For a given reactant mixture and overall equivalence ratio, the initial chamber pressure gradually increases in incre-ments of 0.01 MPa The chamber pressure at which the flame loses its stability at a critical radius around 23e25 mm is defined as the critical initial pressure, Pcr [11,12] Fig 10a shows variations in critical initial pressures for 4 ¼ 1.2 of hydrogen enrichment flames, and the critical initial pressures

of H2/COeair flames with methane and propane additions are compared in Fig 10b The critical initial pressure decreases fairly linearly with hydrogen enrichment, which indicates that cells appear earlier and results in a smaller Pcr.Fig 10b

a

b

Fig 10e (a) Critical initial pressures of CH4/H2/COeair and

C3H8/H2/COeair with hydrogen enrichment at 4 [ 1.2 and (b) Comparison of critical initial pressures of H2/COeair flames with methane and propane additions for4 [ 1.2

shows that the critical initial pressure of the flame with

methane addition increases slightly, whereas the critical

initial pressure in the H2/COeair flame with propane addition

increases significantly This indicates that, for the initiation of

cellular instabilities at a similar critical radius, the C3H8/H2/

COeair flame requires a higher initial pressure of the reactant

mixture compared to the CH4/H2/COeair flame

5 Conclusions

An experimental study was conducted regarding the flame

characteristics of CH4/H2/COeair and C3H8/H2/COeair

pre-mixed flames The following conclusions can be made:

(1) The unstretched laminar burning velocities increase along

with an increase in the hydrogen fraction and decrease

along with the addition of hydrocarbons Conversely, the

Markstein length decreases along with an increase in

hydrogen enrichment, whereas it increases along with an

increase in hydrocarbon addition, which indicates that the

flame becomes stable or unstable along with an increase in

the percentage of hydrocarbon or hydrogen, respectively

(2) The modulation of diffusional-thermal instability is

obtained for rich and stoichiometric mixtures of hydrogen

enrichment flames as well as lean and stoichiometric

mixtures of C3H8/H2/COeair flames due to an increase in

the effective Lewis number Meanwhile, the addition of

methane did not diminish the diffusional-thermal

insta-bility due to a decrease in the effective Lewis number

(3) For an increase in the initial pressure, the onset of cellular

instabilities is obtained at an earlier stage, and the flame is

more unstable because of the enhancement of

hydrody-namic instability due to a significant decrease in the flame

thickness

(4) The three parameters, Rcr, Pecr, Pcr, decrease for hydrogen

enrichment, significantly increase for propane addition

and maintain almost the same values for methane

addi-tion For hydrogen enrichment, this indicates that the

flame will be more unstable and cellular instabilities will

appear earlier However, the opposite tendency occurs such

that cellular instabilities will be suppressed with propane

addition Meanwhile, the H2/COeair flames could not be

diminished for methane addition to the reactant mixtures

Acknowledgements

This research was supported by Basic Science Research

Program through the National Research Foundation of Korea

(NRF) funded by the Ministry of Education, Science and

Technology (2010-0021916)

r e f e r e n c e s

[1] Metghalchi M, Keck JC Laminar burning velocity of

propaneeair mixtures at high temperature and pressure

Combust Flame 1980;38:143e54

[2] Gu XJ, Haq MZ, Lawes M, Woolley R Laminar burning velocity and Markstein lengths of methaneeair mixtures Combust Flame 2000;121:41e58

[3] Aung KT, Hassan MI, Faeth GM Flame stretch interactions of laminar premixed hydrogen/air flames at normal

temperature and pressure Combust Flame 1997;109:1e24 [4] Qiao L, Kim CH, Faeth GM Suppression effects of diluents on laminar premixed hydrogen/oxygen/nitrogen flames Combust Flame 2005;143:79e96

[5] Park J, Lee DH, Yoon SH, Vu TM, Yun JH, Keel SI Effects of Lewis number and preferential diffusion on flame characteristics in 80% H2/20% CO syngas counterflow diffusion flames diluted with He and Ar Int J Hydrogen Energy 2009;34:1578e84

[6] Law CK, Kwon OC Effects of hydrocarbon substitution on atmospheric hydrogeneair flame propagation Int J Hydrogen Energy 2004;29:867e79

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

[8] Tang C, He J, Huang Z, Jin C, Wang J, Wang X, et al Measurements of laminar burning velocities and Markstein lengths of propaneehydrogeneair mixtures at elevated pressures and temperatures Int J Hydrogen Energy 2008;33: 7274e85

[9] Hu E, Huang Z, He J, Zheng J, Miao H Measurements of laminar burning velocities and onset of cellular instabilities

of methaneehydrogeneair flames at elevated pressures and temperatures Int J Hydrogen Energy 2009;34:5574e84 [10] Law CK, Jomaas G, Bechtold JK Cellular instabilities of expanding hydrogen/propane spherical flames at elevated pressures: theory and experiment Proc Combust Inst 2005; 30:159e67

[11] Vu TM, Park J, Kwon OB, Kim JS Effects of hydrocarbon addition on cellular instabilities in expanding syngaseair spherical premixed flames Int J Hydrogen Energy 2009;34: 6961e9

[12] Vu TM, Park J, Kwon OB, Bae DS, Yun JH, Keel SI Effects of diluents on cellular instabilities in outwardly propagating spherical syngaseair premixed flames Int J Hydrogen Energy 2010;35:3868e80

[13] Dowdy DR, Smith DB, Taylor SC, Williams A The use of expanding spherical flames to determine burning velocities and stretch effects in hydrogen/air mixtures Proc Combust Inst 1990;23:325e32

[14] Hassan MI, Aung KT, Faeth GM Measured and predicted properties of laminar premixed methane/air flames at various pressures Combust Flame 1998;115:539e50 [15] Tseng LK, Ismail MA, Faeth GM Laminar burning velocities and Markstein numbers of hydrocarbon/air flames Combust Flame 1993;95:410e26

[16] Clavin P Dynamic behavior of premixed flame fronts in laminar and turbulent flows Prog Energy Combust Sci 1985; 11:1e59

[17] Bradley D, Gaskell PH, Gu XJ Burning velocities, Markstein lengths, and flame quenching for spherical methaneeair flames: a computational study Combust Flame 1996;104: 176e98

[18] Law CK, Sung CJ Structure, aerodynamic, and geometry of premixed flamelets Prog Energy Combust Sci 2000;26:459e505 [19] Williams FA Combustion theory 2nd ed Redwood City, CA: Addison-Wesley; 1985

[20] Bradley D, Hicks RA, Lawes M, Sheppard CGW, Woolley R The measurement of laminar burning velocities and Markstein numbers for iso-octaneeair and iso-octane en-heptaneeair mixtures at elevated temperatures and pressures in an explosion bomb Combust Flame 1998;115: 126e44