Iran. JJ.

C
Chem. C
Chem. Eng. Vol. 29, No. 4, 2010
Archive of SID

CFD Study on Hydrogen-Air Premixed Combustion

in a Micro Scale Chamber

Shabanian, Sayed Reza; Rahimi, Masoud*+; Khoshhal, Abbas

CFD Research Center, Chemical Engineering Department, Razi University, Kermanshah, I.R. IRAN

Ammar Abdulaziz Alsairafi

Faculty of Mechanical Engineering, College of Engineering and Petroleum, Kuwait University, KUWAIT

ABSTRACT: This paper reports a CFD modeling study to investigate the hydrogen-air mixture combustion
in a micro scale chamber. Nine species with nineteen reversible reactions were considered in the
premixed combustion model. The effect of operational and geometrical conditions including;
combustor size, wall conductivity, reactant flow rates and hydrogen feed splitting on combustion
stability and outlet gas temperature were investigated. The results show that the wall thermal
conductivity has a significant effect on the combustion especially at smaller chamber size with high
ratio of chamber surface area to its volume. In addition, the results reveal that high heat loss from
chamber wall, small chamber and high input rate may cause flame quenching. Moreover,
the modeling results indicate that a stable combustion in a micro combustor can be achieved at
an optimum operational condition.

KEY WORDS: Micro combustion, Hydrogen-air, CFD, Flame temperature, Heat transfer.

INTRODUCTION
In the recent years, there is a significant increase density of hydrogen-air systems is significantly higher
in the number of consumer, industrial, and military applications than that of batteries (approximately 140 MJ kg1 compared
utilizing portable electronic equipment, such as laptops, with 0.5 MJ kg1 for lithium-ion battery). The combustion
cellular telephones, and personal data assistants. All of based micro power device, such as micro gas turbine engine,
these devices require a portable power source, which, is one of the most promising designs due to its high power
most commonly, is provided by batteries. However, density. Micro combustor is one of the main components
for many of these applications, the relatively low energy density of micro power system uses hydrogen and hydrocarbon
of traditional batteries imposes burdensome weight and as fuels. Considerable experimental and numerical studies
power limitations on system design. Power generation have been carried out to investigate the effect of various
utilizing hydrocarbons or hydrogen-air systems offers geometrical and operational conditions on the flame stability
a promising alternative to traditional batteries. The energy limits and performance of micro combustors [1- 4].
* To whom correspondence should be addressed.
+ E-mail: masoudrahimi@yahoo.com
1021-9986/10/4/161 12/$/3.20

161

www.SID.ir
 Iran.
a . J.
J. C
Chem.
e .CChem.
e Eng. Shabanian S.R. et al. Vol. 29, No. 4, 2010
Archive of SID

An experimental investigation was carried out on monoxide was produced. The turbulent mixing of a
premixed combustion in cylindrical micro combustors passive scalar in confined multiple jet flows in a micro
(with a backward-facing step) by Li et al. [5]. They combustor was investigated by Cho et al. [11]. The authors
studied the effect of wall temperature and radiation heat flux performed their experiments for a micro combustor
on the performance of cylindrical micro combustors. having a circular disk baffle plate with a fuel injection
Their results showed that as the combustor length nozzle in the center and oxidant injection holes allocated
increases, the range of flow conditions for successful annularly. Their results revealed that the turbulent mixing
ignition becomes smaller. It was also shown that the is promoted by complex flow fields caused by the jet
backward-facing step employed in the combustor design flows and large vortical flow regions. An experimental
is effective in stabilizing the flame position. study performed by Zhou et al. [12] to improve the micro
In continuation of this research, they studied the porous combustion stability using an electrical heating.
media combustion of premixed hydrogenair in a planar The authors reported that when electrical heating power
micro combustor with a channel width of 1 mm [6]. of 1.05 and 4.70 W are used, the equivalence ratio ranges
The effects of flow conditions and position of the porous of stable combustion are extended from 0.362 - 6.52 to
media on both the wall temperature distribution and 0.178 - 7.66 and 0.126 - 9.43, respectively. A new micro
the emitter efficiency are investigated. Their experimental combustor configuration for a micro fuel cell reformer
results indicate that the wall temperature increases with integrated with a micro evaporator was studied
increasing mixture velocity, and higher emitter efficiency experimentally and numerically by Kim & Kwon [13].
is achieved for mixtures with U 0.8. They could extend the flame stability limits of micro
Cao & Xu [7] investigated the thermal performance combustor.
of a micro combustor for micro gas turbine system. They In addition to above experimentally research in this field,
found that the wall temperature of the micro combustor many theoretical works have been carried out by
and the temperature of the exhaust gas from combustor researchers. Raimondeau et al. [14] simulated the
reach to maximum values at excess air ratio of 0.9. methane-air flame propagation in micro tubes by solving
The effect of wall thickness of micro combustor on the the parabolic governing equations. In their model,
performance of micro Thermo Photo Voltaic (TPV) the discontinuity boundary conditions of temperature and
power generators was experimentally investigated by concentration as well as the destruction of radicals at
Wenming et al. [8], in their experiments, three kinds of the gassolid interface were employed. They found that
micro cylindrical combustors made from SiC with different close to entrance the heat loss and radical wall quenching
wall thickness were tested. The performance of the micro are key parameters in controlling the flame propagation
TPV system was measured for various flow rates and in these types of micro devices. In another investigation,
hydrogen/air ratios. Their results showed that when a numerical study of hydrogen-air combustion in micro
the inner diameter of the micro combustor is constant, channels was carried out by Leach et al. [15]. They
with the decrease of wall thickness, both maximum electrical concluded that heat exchange through the structure
power output and short-circuit current increase considerably. of the micro combustor can lead to broadening
In another study, the suppression of radical heat loss of the reaction zone (flame thickness).
to the combustor walls and flame quenching effects was Jejurkar & Mishra [16] discuss the flame stabilization
experimentally investigated by Miesse et al. [9]. In their mechanism for the annular micro combustor with the help
work, methane-oxygen diffusion flames were successfully of a computational model. In their study, thermal
stabilized between parallel plates having gaps of less than performance of the micro combustor is analyzed for
1 mm. Combustion of methane-oxygen in three types of variable inlet mass flow rates and different wall thermal
micro tubes with the outer diameter of 3 mm was conductivities. Their results for different inlet mass flow
experimentally studied by Junwei & Beijing [10]. rates and wall thermal conductivities revealed that there
The obtained results showed that when equivalence ratio are useful operating points in micro combustor for
was lower than unity, the methane was not completely different applications such as micropropulsion, heating
oxidized and a great amount of hydrogen and carbon and portable electric power generation.

162

www.SID.ir
 Iran.
a . J.
J. C
Chem.
e .CChem.
e Eng. CFD Study on Hydrogen-Air Premixed Combustion ... Vol. 29, No. 4, 2010
Archive of SID

In another research, the effect of different wall important in macro combustion, should be considered
material on the performances of catalytic micro in micro combustion modeling.
combustor investigated by Zhou et al [17]. Their CFD Due to progresses in computer hardware and software
computations depicted that the thermal conductivity and and consequent increase of the calculation speed, the
inlet flow rates influenced the mode of combustion. Computational Fluid Dynamics (CFD) modeling
Heterogeneous micro combustion prevailed at higher wall technique would be a powerful and effective tool
thermal conductivity and low flow rates. for understanding the complex hydrodynamics in many
Kaisare & Vlachos [18] also tried to find suitable industrial processes. Since measuring instrument causes
reactor length, reactor opening size and wall thickness disturbance to the micro flame, thus CFD modeling
for homogeneous combustion in parallel plate narrow channels. can be a suitable way to study this type of process [23]
They solved the simplified one-dimensional governing with vanishing interference of experimental sensors in
equations and used single-step reaction mechanism flow hydrodynamics.
for this purpose. The flammability in terms of blowing-off A two-dimensional CFD study was carried out by
and flashing-back were recognized at different conditions. Benedetto et al. [24] to investigate the opportunity of
Li et al. [19] developed a simple 1D flame model setting up a novel scheme of micro combustors.
In that study, the combustor was divided into two parts.
to analyze the heat transport occurring in the cylindrical
In the first part, the walls of the micro combustor were
micro combustors. They extend the model of Li et al. [20,21]
coated with catalyst, while in the second part the catalyst
to quantify two important ratios, namely, the heat loss
was absent and only homogenous combustion could
ratio (the ratio of the heat loss from the flame to the heat
taken placed. Numerical results showed that this hybrid
generation in the reaction zone) and the heat recirculation
micro combustor allows operating at high inlet gas
ratio (the ratio of the heat recirculation through the solid
velocities and thus high input powers, without
structure to the upstream to the heat loss from the flame).
encountering blow-out and maintaining almost complete
As an improvement over Li et al. model [20,21], the one-step
fuel conversions. In continuation of their research [25],
global reaction mechanisms were included.
Benedetto et al. studied the effect of the cross-sectional
The effects of various parameters such as the
geometry on the ignition-extinction behavior of catalytic
combustor size, fuel property, fuelair equivalence ratio
micro combustors. For this purpose, two different types of
and unburned mixture temperature on the heat loss ratio
circular and square cross-section were investigated.
and the heat recirculation ratio were investigated. Their Stability maps were built and compared with lean propane-
results indicated that these parameters have significant air combustion at different inlet gas velocities. Results
effects on the two ratios, and therefore should be demonstrated that at low inlet velocities the square
carefully managed in order to achieve efficient and stable cross-section channel has more resistance to extinction than
combustion. A numerical investigation was performed by the cylindrical one.
Chen et al. [22] to study the effects of catalytic walls on In most of the theoretical studies dealing with micro
hydrogen-air flames in a micro tube. Their results showed combustors, two-dimensional CFD models have been
that the combustor channel wall and the fuel mixture adopted [26-28]. Two-dimensional models neglect
characteristics can play a significant role in micro the circumferential changes of velocity, temperature,
channel combustion. species concentration as well as heat and mass fluxes.
In the above mentioned modeling studies, regarding At the small-scales, these variations are important. Therefore,
to this fact that the combustion chamber characteristic ignoring of these variations may causes incorrect theoretical
length and gas flow path are larger than the molecular predictions for micro scale combustion systems.
mean free path, the fluid continuity is established and Regarding to above studies in the literature, in the
Navier-Stokes equations and no-slip wall condition present work it has been tried to predict the place of
are employed in micro combustion such as conventional flame front as well as the conditions that the flame can be
combustion in macro systems. However, some stable. Moreover, the effect of splitting of the fuel
parameters such as heat wall conductivity that are not injection on overall combustion performance is another

163

www.SID.ir
 Iran. JJ. C
Chem. C
Chem. Eng. Shabanian S.R. et al. Vol. 29, No. 4, 2010
Archive of SID

main issue in this study. The hydrogen-air mixture is Table 1: The grid independency study results.
chosen for three main reasons: (1) the reaction No. of control Predicted adiabatic Difference
mechanism is well known; (2) the reaction mechanism is volume flame temperature percentage

simpler compared to other hydrocarbon fuels; (3) the 148 563 2505.5
hydrogen-air flame is more practicable to be sustained 203 975 2450.8 2.183197 %
in micro scale combustors. For this purpose, a cylindrical
322 495 2420.4 1.240411 %
combustor with a fixed diameter/length ratio was modeled
500 027 2408.0 0.512312 %
using three dimensional CFD modeling technique.
855358 2407.6 0.016611 %
THEORY
The micro combustion modeling involves the solution CFD MODELING
of the laminar flows with heat transfer, species transport A three-dimensional CFD modeling of the reactive
and chemical reactions. Equations for continuity, gas flow in the micro combustor was performed using the
momentum, energy and species were solved by Fluent commercial CFD code, Fluent 6.2. [30]. The CFD
software, as are listed in Eqs. (1) to (4) [29]. In addition, modeling was carried out to find the behavior of
a User-Defined Subroutine (UDF) has been used combustion of hydrogen-air mixture in a cylindrical
to calculate the reaction rates. micro chamber. Moreover, the effect of various boundary
( ) =0
u j
(1)
conditions on flame characteristics has been investigated.
x j
Geometry and mesh generation
(
u j u i ) = p + ij + g (2)
In the present study, the micro combustor was created
i
x j x i x j and meshed using GAMBITTM. The inlet duct diameter
and length were one-half and one-fourth of the reactor's
(
u j h )=
kf
T N
h i J ij + (3) diameter, respectively. The modeled domain which
x j x j x j i =1 consists of two cylindrical chambers was meshed into
p almost 500000 tetrahedral cells. This mesh layout
uj + Sh + Srad
x j was found by examination of different cell sizes as
no further significant change was found for finer cells in
(
u j Yi ) = Jij + R (4) predicted adiabatic flame temperature. This temperature
i
x j x j has been calculated by modeling of hydrogen-air combustion
Where is the fluid density (kg/m3), u is velocity (m/s), in the studied combustor under adiabatic condition.
p is the static pressure (Pa), is the laminar stress tensor (Pa), The results of this grid consistency study for the combustion
g is the gravitational body force per unit volume (N/m3). chamber is presented in Table 1. As it is illustrated in this
h, kf and T are the sensible enthalpy (kJ/kg), table, the values of calculated temperature using grids 4
thermal conductivity (W/mK) and temperature (K) and 5 are quite close. In the other words, no significant
of fluid mixture, respectively. Moreover, hi is the sensible changes were observed in the predicted temperature for
enthalpy of species i (kJ/kg), Jij is the diffusion flux of grid 5 when it is compared with that of grid 4. Therefore,
species i in the j direction (kg/m2s), which was Grid 4 setup was chosen due to the lower required
the combination of full multi component diffusion and computation time. Fig. 1 shows the meshed configuration
thermal diffusion. In addition, N is the total number of of the inlet tube and the combustion chamber.
gas species, Sh is the source of energy caused by chemical
reaction (kW/m3) and Srad is the source of radiation Initial and boundary conditions
obtained from the solution of discrete ordinates radiation Uniform velocity inlets were employed as boundary
model equations. Finally, Yi is mass fraction of species i, conditions at the chamber inlet. In addition, the outlet
Ri is the net rate of production of species i by chemical boundary condition pressure was defined for the combustor
reaction (kg/m3s). outlet side. Different hydrogen-air mixture mass flow rates

164

www.SID.ir
 Iran. JJ. C
Chem. C
Chem. Eng. CFD Study on Hydrogen-Air Premixed Combustion ... Vol. 29, No. 4, 2010
Archive of SID

Table 2: Gas phase reaction mechanism for hydrogen-air combustion [31,32].

Reactions AK(m,kmol,s) K EK,J/kmol

1. O2+H=OH+O 5.1E+13 -0.82 6.91E+7

2. H2+O=OH+H 1.80E+7 1 3.70E+7

3. H2+OH=H2O+H 1.20E+6 1.30 1.52E+7

4. OH+OH=H2O+O 6.00E+5 1.30 0

5. H2+O2=OH+OH 1.70E+10 0 2.0E+8

6. H+OH+M=H2O+M 7.50E+17 -2.60 0

7. O2+M=O+O+M 1.90E+8 0.5 4.001E+8

8. H2+M=H+H+M 2.20E+9 0.5 3.877E+8

9. H+O2+M=HO2+M 2.10E+12 -1 0

10. H+O2+O2=HO2+O2 6.70E+13 -1.42 0

11. H+O2+N2= HO2+ N2 6.70E+13 -1.42 0

12. HO2+H=H2+O2 2.50E+10 0 2.90E+6

13. HO2+H=OH+OH 2.50E+11 0 7.90E+6

14. HO2+O=OH+O2 4.80E+10 0 4.20E+6

15. HO2+OH=H2O+O2 5.00E+10 0 4.20E+6

16. HO2+HO2=H2O2+O2 2.00E+9 0 0

17. H2O2+M=OH+OH+M 1.30E+14 0 1.905E+8

18. H2O2+H=H2+HO2 1.70E+9 0 1.57E+7

19. H2O2+OH=H2O+HO2 1.0E+10 0 7.50E+6

Rate constants are given in the form k=AKT K exp (-EK/RT).
Enhancement factors:H2O=20.0 , Enhancement factors:H2O=6.0, H=2.0, H2=3.0.
Enhancement factors : H2O=21.0, H2=3.3, O2=0.0, N2=0.0

Physical properties and chemical reactions

In the present study, fluid flow regime is assumed
L laminar, as it is, and the fluid density has been calculated
using the ideal gas law. The fluid mixture specific heat,
viscosity, and thermal conductivity have been calculated
2D
using mass fraction weighted average of species properties.
The specific heat of each species has been calculated using
Hydrogen/air
a piecewise polynomial of temperature. In order to model
inlet
the gas-phase reaction, a detailed hydrogen-air oxidation
mechanism, which listed in Table 2, has been employed.
Fig. 1: The modeled micro combustor. This reaction mechanism includes 9 species and 19
elementary reactions.
were defined at the inlet side. No-slip boundary condition
was imposed at the combustion chamber wall. Combustion Solution strategy and convergence criterion
was initiated by introducing a small high temperature The whole modeling was carried out in steady state
zone within the reaction zone to simulate ignition. mode of calculations. The SIMPLE pressurevelocity
The initial mass fractions of all species were set equal to coupling algorithm, the standard pressure and the second
their respective inlet values. The influence of the order upwind discretization scheme for momentum were
gravitational force on the flow was not taken into account. employed in the model. In addition, the CFD simulation

165

www.SID.ir
 Iran.
a . J.
J. C
Chem.
e .CChem.
e Eng. Shabanian S.R. et al. Vol. 29, No. 4, 2010
Archive of SID

convergence is judged upon the residuals of all governing Table 3: A Comparison between the modeling results and
equations and the reported results related to residuals less experimental measurement (pressure=1 atm & =1).
than 1.0 106. A double precision, segregated solution Temperature/ Experimental measurement
Modeling results
composition (Glassman & Yetter, 2008)
solver is used to solve the governing equations.
H 0.0020 0.0021

RESULTS AND DISCUSSION H2 0.0150 0.0160

In this study, the performance of a micro combustor H2O 0.3230 0.3150
has been investigated using three dimensional CFD
N2 0.6440 0.6440
modeling. The performance of the combustor is evaluated
under various geometrical and operational conditions, OH 0.0070 0.0075
including: chamber size, wall heat conductivity, fuel-air O2 0.0050 0.0052
flow rate and splitting the hydrogen feed. These results
T(K) 2382 2408
may be used as design guidelines and recommendations
for the future development of micro combustion systems.
2500
Model Validation

Flame temperature (K)

2000
In order to validate the CFD model, a comparison
is done with the measured adiabatic flame temperatures
1500
and combustion product composition found by
Glassman & Yetter [33]. The comparison results are
1000
illustrated in Table 3. With trusting to the observed
reasonable agreement between the theoretical and experimental 500
results, the effects of various process operational conditions
on hydrogen-air combustion have been investigated. 0
0 0.001 0.002 0.003 0.004 0.005 0.006

Effect of feed flow rate Mass flow rate (g/s)

In this section, the effect of feed flow rate on Fig. 2: The flame temperature versus gas mixture mass flow rate
operational stability has been investigated. For this for the micro combustor.
purpose, a combustor with an inlet diameter of 0.42 mm
has been modeled. The feed fuel/air equivalence ratio of results confirm that the feed flow rate has an optimum
0.75, the wall convection coefficient of 200 W/m2.K and range which is from 0.001 to 0.003 g/s at the above
ambient temperature of 300 K are considered as boundary mentioned operation condition. These CFD predicted
conditions in the model. The geometry of studied results can be explained by a fact that at low flow rate,
chamber can be found in Fig. 1. The predicted temperatures due to low rate of generated heat in comparison with
at the flame core, which is place of maximum temperature, the rate of heat loss, the combustion process can not be
at various feed flow rate are shown in Fig. 2. self sustained and the flame will be quenched. Moreover,
The results show that the flame temperature increased with increase of input flow rate the flame can be more
sharply with increasing of feed flow rate at low flow rates stable as more heat releases from the reaction. However,
due to the fact that the increase of inlet velocity with increase of the feed flow rate more than this optimum
can establish more heat generation per unit time. Consequently, values, the effective reactant residence time decreases
the results reveal that the temperature reaches to and the reactant have less time for reaction. On the other hand,
a maximum value about 2000 K at feed mass flow rate of at higher gas flow rate the convection heat transfer
0.001 g/s. No future considerable change in temperature coefficient of product gas increases. Therefore, reduction
can be observed with the increase of the feed flow rate in flame temperature is expected. In the other words,
up to 0.003 g/s. However, at higher feed flow rate a stable combustion can be happened in almost
the predicted temperature reveal a decreasing trend. These 0.001 to 0.003g/s feed mass flow rate, the other two sides,

166

www.SID.ir
 Iran. JJ. C
Chem. C
Chem. Eng. CFD Study on Hydrogen-Air Premixed Combustion ... Vol. 29, No. 4, 2010
Archive of SID

K 2400

2300

Flame temperature (K)

2200
a). mass flow rate= 0.0005g/s
2100

b). mass flow rate= 0.001g/s

2000

1900
D=0.12mm
c). mass flow rate= 0.002g/s 1800 D=0.22mm
D=0.42mm
1700
3.5 4.5 5.5 6.5 7.5
d). mass flow rate= 0.005g/s
Inlet velocity (m/s)

Fig. 3: The temperature distribution in an axial slice inside Fig. 4: Effects of combustor size on the flame temperature
the micro combustor for different flow rates. in various inlet velocities.

less than 0.001 and more than 0.003, which involves with in cases that the energy loss cannot be compensated by
a very high temperature gradient, an unstable flame the heat released from reaction at flame zone, which leads
is expected. The temperature distributions in an axial slice to the flame quenching. Therefore, in this part the effect
of micro combustor at four different mass flow rates of combustion chamber size on flame temperature has
are shown in Fig. 3. It can be seen from the figure that been studied. For this purpose, three combustors with the
the flame core (high temperature zone) shifts toward inlet diameters of 0.12, 0.22 and 0.42 mm have been
the downstream from the inlet as the mass flow rate modeled. The results have been analyzed for four inlet
increased. In addition, Fig. 3 illustrates the thermal velocities of 4, 5, 6 and 7m/s layouts. In addition, the heat
quenching of micro scale combustion at low and high transfer coefficient on the outer wall is kept at 50 W/m2.K.
feed flow rates as compared with normal flames formed The variations of flame temperature at various inlet velocities
at moderate feed flow rates. Therefore, at very low for three mentioned combustors are given in Fig. 4.
as well as very high feed flow rates due to lack of existence As can be seen in this figure, for each chamber size
of high temperature outlet gas and stable combustion, there is not a considerable difference in flame temperature
the system cannot work efficiently. at various employed feed velocities. However, there are
In summary, the above obtained results reveal that the significant differences among the predicted flame
flame temperature and position can be changed by temperatures for these three combustor sizes at each inlet
variation of inlet gas velocity. In other words, velocity. The results also show that the lowest
the maximum temperature shifts toward the exit by temperatures were established by the smallest chamber
increasing the gas velocity while extinction can be and the highest temperatures obtained using the biggest
occurred by lowering the inlet gas flow rate. This is chamber. The average difference in the predicted flame
in agreement with the results obtained by Yang et al. [3]. temperature between the smallest and biggest chamber is
They found that increasing the gas velocity has contrary almost 400 K, which is quite significant. This can be
effects on combustion. They explained these effects by explained by fact that as combustor size decreases, the
analyzing the flow residence time and released energy. ratio of heat transfer surface to chamber volume
A similar result was obtained by Li et al. [34] for increases. This can causes an increase in ratio of heat loss
methane-air combustion system. to heat generated and consequent decrease in flame
temperature. Therefore, it can be concluded that combustor
Effect of the combustor size size has significant effect on flame temperature. This is
It is well known that the effect of heat loss from in agreement with a study carried out by Li et al. [35].
combustion chamber in micro scale is more significant They reported that the flame temperature decreases
than the conventional combustion. This is more important by decreasing the combustor size for hydrogen-air systems.

167

www.SID.ir
 Iran.
a . J.
J. C
Chem.
e .CChem.
e Eng. Shabanian S.R. et al. Vol. 29, No. 4, 2010
Archive of SID

K
90
80

Water production rate

70
60

(kmole/m3.s)
a). K=150 W/m/K
50
40
K=150 W/m.K
30
b). K=15 W/m/K K=0.7 W/m.K
20
K=15 W/m.K
10
0
c). K=0.7 W/m/K 0 0.2 0.4 0.6 0.8 1
Dimentionless length (x/L)

Fig. 5: The temperature distribution in an axial slice inside Fig. 6: Water production rate distribution along the central
the micro combustor for different wall heat conductivity. axis of combustion chamber with different wall heat conductivity.

Effect of the wall heat conductivity to the other regions. This can help the preheating of reactant
In micro combustor systems, heat transfer from the gas as well as improving the ignition and flame stability.
flame zone to the incoming gases is quite important in order As far as water is the product of hydrogen-air
to have a stable flame. A system with maximum axial and combustion, more water production rate along the rector
minimum crosswise wall heat conduction is ideal for has been used as a criterion to show the quality of
micro combustors, as in this case the heat loss reaction progress. Therefore, the water production has
is minimized while the heat recirculation is maximized [35]. been used to judge about combustion performance.
In this section, the effect of heat conduction within In order to illustrate to opposite roles of wall
chamber wall on combustion characteristics in a theoretical conductivity, water production rate along the combustor
micro combustion chamber has been studied. For this for three wall conductivities are studied. Fig. 6 reveals
purpose the wall thickness is selected to be 0.1mm and that the water production rate for k=15 W/m.K is more
three kinds of material with different thermal conductivity than those of thermal conductivities of 0.7 and 150.
of 0.7, 15 and 150 W/m/K are used in the modeling. In the other words, the figure implies that the water production
A convective heat transfer coefficient of 40W/m2/K from rate has higher value at a moderate value of wall
chamber wall to environment and an inlet velocity of conductivity. It can be explained by the fact that higher
1m/s have been chosen. Fig. 5 shows temperature wall conductivity reduces the flame temperature. In addition,
contours in an axial slice of the combustor with three this has an advantage of more efficient inlet gas preheating.
different thermal conductivities. It can be clearly see As a complementary study, some quantitative predicted
in Fig. 5 that as the wall thermal conductivity decreased, temperatures from micro combustors with different wall
the maximum temperature in the flame core increased. thermal conductivity are shown in Table 4. As can be seen
This can be argued that as the heat conductivity of wall in this table, as the heat conductivity of combustor wall
decreases, the heat loss to the ambient decreases. This causes increased, the exit gas temperature increased and
the ratio of heat generation to heat loss in the combustion both the flame temperature and the average temperature
chamber increases, which stabilized the combustion of combustor decreased.
inside the chamber. In addition, Fig. 5 illustrates a more A similar results were previously reported by
uniform temperature distribution and higher exit gas Norton & Vlachos [1,2]. They investigated the effect of
temperature has been established as the wall heat wall heat conductivity in micro combustion systems with
conductivity increased. This can be attributed to that the propane-air as well as methane-air mixtures. They studied
axial heat conduction in the chamber wall, which provides the positive and negative effects of wall heat conductivity
a way for the heat transfer from the combustion region on flame stability and found that a moderate value of

168

www.SID.ir
 Iran. JJ. C
Chem. C
Chem. Eng. CFD Study on Hydrogen-Air Premixed Combustion ... Vol. 29, No. 4, 2010
Archive of SID

Table 4: Predicted temperatures at various wall heat conductivity.

Wall heat conductivity (W/m/K) Flame temperature (K) Exit gas temperature (K) Average combustor temperature

400 1769 530 509

200 1776 527 570

15 1785 507 584

0.7 1836 361 608

0.2 1861 331 616

heat conductivity is suitable for having a stable flame 2500

with low heat loss. 2250

2000

Effect of splitting the hydrogen feed 1750

T em peratu re(K)
As was illustrated before, the heat of combustion from 1500
Three -
1250
reaction between hydrogen and oxygen caused high Two -
1000 One -
temperature zone close to the inlet nozzle. This usually
750
causes undesirable effects such as considerable heat losses
500
from chamber, radical destructions and temperature
250
increases more than the allowable temperature of
0
combustor's wall. In order to overcome this problem and 0 1 2 3 4 5 6 7
Length(mm )
reach to a more uniform temperature distribution,
the hydrogen has been entered into the chamber in two
and three portions with aim of minimizing the formation Fig. 7: The temperature inside the combustor with different
of hot entrance zone. The total hydrogen flow rate was set layout.
at a fixed value for all cases. For the cases of two and
three injection stages, the total of hydrogen divided in In this layout, the temperature profile is more uniform
two and three equal portions and injected to reactor with and the outlet temperature is 140 K higher than the first
an angle of 90 .In all cases a chamber inlet diameter of layout. Therefore, it can be concluded that with more
0.42mm, convection heat transfer coefficient of 50W/m2.K, distributing of hydrogen feed it is possible to reach
inlet velocity of 6m/s and the ambient temperature of a more efficient system.
300 K were chosen. The temperature profile along the In summary, the above mentioned results reveal that
chamber at different layout of fuel injection is shown in the distribution of feed along the combustor can be
Fig. 7. As can be seen in one stage layout, which the a good strategy for controlling the hotspot appeared at the
whole of fuel is entered via the inlet nozzle, the temperature entrance of the reactor as well as having more uniform
rises to a maximum temperature, 2340 K, sharply. temperature profile along the combustor. This is also
Consequently, due to heat loss from the wall, the temperature can be quite important for preventing wall damage as well as
of outlet hot gas is 1310K. In second layout, which the safety of a combustor. This is in agreement with the
hydrogen is divided into two parts (the first half initially results obtained by Barrio et al. [36]. They found that
with the feed and the second half in the middle of the split of feed along a catalytic reactor allowed a noticeable
combustor), the temperature profile inside the combustor decrease of the maximum temperature in the reactor.
is completely different. The figure shows that the maximum
temperature in two stage layout is almost 250 K lower than CONCLUSIONS
that of first layout. In addition, the predicted results show Here, a study on hydrogenair premixed flame in
that the outlet temperature is 120 K higher than that of first micro combustors with different physical and boundary
layout. Finally, in the third layout there are three temperature conditions has been performed. In order to investigate
peaks with lower values compared two other layout. the effects of the combustor size, three micro cylindrical

169

www.SID.ir
 Iran.
a . J.
J. C
Chem.
e .CChem.
e Eng. Shabanian S.R. et al. Vol. 29, No. 4, 2010
Archive of SID

combustors with inlet diameters of 0.42, 0.22 and 0.12 mm u Velocity vector, m s-1
have been modeled. The numerical results indicate that u, ui, uj Mean velocity components, m s-1
the flame temperature is lower for the smaller combustor. xi, xj Cartesian coordinate, m
This has been explained by the fact that as the
combustion chamber size decreases, the ratio of surface Greek symbols
area to volume increases. Consequently, heat loss from Equivalence ratio
the wall of chamber is increased, which affect on flame eff Stress tensor, Pa
temperature. In this study, it has been showed that
, T, eff Laminar, turbulent and effective
a moderated feed flow rate is required for a stable power
viscosity, Pa s
generation in micro combustors. When feed flow rate
Kinematic viscosity, m2/s
is too low, the performance of the combustor is limited
Density, kg m-3
by flame quenching in the chamber. However, at high
mass flow rates of fuel-air, the inlet gas mixture have not
Received : Jan. 1, 2010 ; Accepted : Jan. 17, 2011
enough time for reaction and flame may be blown out,
which also leads to low combustion efficiency.
REFERENCES
The results of this study reveal that the wall thermal
[1] Norton D.G., Vlachos D.G., Combustion
conductivity is another parameter which plays
Characteristics and Flame Stability at the
an important role in the performance of micro combustor.
Microscale: a CFD Study of Premixed Methane-Air
The results show that lower wall heat conductivity
Mixtures, Chem. Eng. Sci., 58, p. 4871 (2003).
can reduces the heat loss from the system and causes a stable
[2] Norton D.G., Vlachos D.G., A CFD Study of
combustion. On the contrary, the higher heat conductivity
Propane-Air Microflame Stability, Combust. Flame.,
can increases preheating of reactants and decreasing wall
138, p. 97 (2004).
thermal stress. From this investigation, it can be concluded
[3] Yang W.M., Chou S.K., Shu C., Li Z.W., Xue H.,
that the split of hydrogen feed along the combustor
Combustion in Micro Cylindrical Combustors with
allowed a noticeable decrease in maximum temperature
and Without a Backward Facing step, Appl. Therm.
of the combustion chamber. The hydrogen feed splitting
Eng., 22, p. 1777 (2002).
can causes a more uniform temperature to be established
[4] Karagiannidis S., Mantzaras J., Jackson G.,
in the chamber. Finally, it can be concluded that by
Boulouchos K., Hetero-homogeneous Combustion
CFD modeling of combustion of premixed hydrogen-air
and Stability Maps in Methane-Fueled Catalytic Micro
mixture in a micro scaled chamber, it is possible to design
Reactors, Proc. Combust. Inst., 31, p. 3309 (2007).
an efficient system without need to many expensive
[5] Li J., Chou S.K., Huang G., Yang W.M., Li Z.W.,
experimental analysis.
Study on Premixed Combustion in Cylindrical
Micro Combustors: Transient Flame Behavior and Wall
Acknowledgment Heat Flux, Exp. Therm. Fluid. Sci., 33, p. 764
The authors wish to express their thanks to the Renewable (2009).
Energy Organization of Iran, for supporting of this work. [6] Li J., Chou S.K., Li Z.W., Yang W.M., Experimental
Investigation of Porous Media Combustion in
Nomenclature a Planar Micro Combustor, Fuel, 89, p. 708 (2010).
Cp Specific heat capacity, kJ/kgK [7] Cao H.L., Xu J.L., Thermal Performance of a Micro
D Inlet diameter of combustor, m Combustor for Micro Gas Turbine System, Energy
Eij Linear deformation rate, s-1 Conv. Manag., 48, 1569 (2007).
H Heat transfer coefficient, W/m2K [8] Wenming Y., Siawkiang Ch., Chang Sh., Hong X.,
L Combustor length, m Zhiwang L., Effect of Wall Thickness of Micro
x Distance from the inlet in central axis, m Combustor on the Performance of Micro-
K Thermal conductivity, W/mK Thermophotovoltaic Power Generators, Sensor.
T Temperature, K Actuator. , 119, p. 441 (2005).

170

www.SID.ir
 Iran.
a . J.
J. C
Chem.
e .CChem.
e Eng. CFD Study on Hydrogen-Air Premixed Combustion ... Vol. 29, No. 4, 2010
Archive of SID

[9] Miesse C., Masel R., Short M., Shannon M., [22] Chen G., Chen C., Wu C., Chao Y., Effects of
Experimental Observations of Methane-Oxygen Catalytic Walls on Hydrogen-Air Combustion Inside
Diffusion Flame Structure in a Sub-Millimeter a Micro Tube, Appl. Catal., 332, p. 89 (2007).
Microburner, Combust. Theory Model., 9, p.77 (2005). [23] Leach T.T., Cadou C.P., The Role of Structural Heat
[10] Junwei L., Beijing Z., Experimental Investigation on Exchange and Heat Loss in the Design of Efficient
Heat Loss and Combustion in Methane-Oxygen Silicon Micro Combustors, Proc. Combust. Inst., 30,
Micro Tube Combustor, Appl. Therm. Eng., 28, p. 2437 (2005).
p. 707 (2008). [24] Benedetto A.D., Sarli V.D., Russo G., A Novel
[11] Choi H.S., Park T.S., Suzuki K., Turbulent Mixing Catalytic-Homogenous Micro Combustor, Catal.
of a Passive Scalar in Confined Multiple Jet Flows Today, 147S, p. S156 (2009).
of a Micro Combustor, Int. J. Heat Mass Transfer, [25] Benedetto A.D., Sarli V.D., Russo G., Effect
51, p. 4276 (2008). of Geometry on the Thermal Behavior of
[12] Zhou J., Wang Y., Yang W., Liu J., Wang Z., Cen K., Catalytic Micro Combustors, Catal. Today, 155, p.
Improvement of Micro Combustion Stability 116 (2010).
Through Electrical Heating, Appl. Therm. Eng., 29, [26] Maruta K., Takeda K., Ahn J., Borer K., Sitzki L.,
p. 2373 (2009). Ronney P.D., Deutschmann O., Extinction Limits of
[13] Kim K.B., Kwon O.C., Studies on a Two-Staged Catalytic Combustion in Micro Channels, Proc.
Micro Combustor for a Micro Reformer Integrated Combust. Inst. 29, p. 957 (2002).
with a Micro Evaporator, J. Power Sourc., 182, [27] Norton D.G., Wetzel E.D., Vlachos D.G., Thermal
p. 609 (2008). Management in Catalytic Micro-Reactors, Ind. Eng.
[14] Raimondeau S., Norton D., Vlachos D.G., Masel R.I., Chem. Res., 45, p. 76 (2006).
Modeling of High Temperature Micro Burners, [28] Hua J., Wu M., Kumar K., Numerical Simulation of
Proc. Combust. Inst., 29, p. 901 (2002). the Combustion of Hydrogen-Air Mixture in
[15] Leach T.T., Cadou C.P., Jackson G.S., Effects of Micro-Scaled Chambers. Part I: Fundamental Study,
Structure Conduction and Heat Loss on Combustion Chem. Eng. Sci., 60, p. 3497 (2005).
in Micro Channels, Combust. Theory Model., 10, [29] Shi L., Bayless D.J., Prudich M., A Model of Steam
p. 85 (2006). Reforming of Iso-Octane: The Effect of Thermal
[16] Jejurkar S.Y., Mishra D.P., Numerical Characterization Boundary Conditions on Hydrogen Production and
of a Premixed Flame Based Annular Micro Combustor, Reactor Temperature, Int. J. hydrogen energy, 33,
Int. J. hydrogen energy, 35, p. 9755 (2010). p. 4577 (2008).
[17] Zhou J., Wang Y., Yang W., Liu J., Wang Z., Cen K., [30] Fluent 6.2: Fluent Inc., Lebanon, NH, USA,
Combustion of Hydrogen-Air in Catalytic Micro (2005).
Combustors Made of Different Material, Int. J. [31] Andrae J.C.G., Bjornbom P.H., Wall Effects of
hydrogen energy, 34, p. 3535 (2009). Laminar Hydrogen Flames Over Platinum and Inert
[18] Kaisare N.S., Vlachos D.G., Optimal Reactor Surfaces, American Inst. Chem. Eng. J., 46, p. 1454
Dimensions for Homogeneous Combustion in Small (2000).
Channels, Catal. Today, 120, p. 96 (2007). [32] Giovangigli V., Smooke M.D., Extinction
[19] Li J., Chou S. K., Li Z., Yang W., Development of of Strained Premixed Laminar Flames with
1D Model for the Analysis of Heat Transport in Complex Chemistry, Combust. Sci. Tech., 53, p. 23
Cylindrical Micro Combustors , Appl. Therm. Eng., (1987).
29, p. 1854 (2009). [33] Glassman I., Yetter R. A., "Combustion", 4th Ed,
[20] Li Z.W., Chou S.K., Shu C., Yang W.M., Predicting Chap 1, p. 30, Academic Press, California, (2008).
the Temperature of a Premixed Flame in a Micro [34] Li J., Chou S.K., Yang W.M., Li Z.W., A Numerical
Combustor, J. Appl. Physic., 96, p. 3524 (2004). Study on Premixed Micro Combustion of CH4-Air
[21] Li Z.W., Chou S.K., Shu C., Yang W.M., Entropy Mixture: Effects of Combustor Size, Geometry and
Generation During Micro Combustion, J. Appl. Boundary Conditions on Flame Temperature, Chem.
Physic., 97, p. 84914 (2005). Eng. J., 150, p. 213 (2009).

171

www.SID.ir
 Iran.
a . J.
J. C
Chem.
e .CChem.
e Eng. Shabanian S.R. et al. Vol. 29, No. 4, 2010
Archive of SID

[35] Li J., Chou S. K., Li Z. W., Yang W. M.,

A Comparative Study of H2-Air Premixed Flame in
Micro Combustors with Different Physical and
Boundary Conditions, Combust. Theory Model., 12,
p. 325 (2008).
[36] Barrio V.L., Schaub G., Rohde M., Rabe S., Vogel
F., Cambra J.F., Arias P.L., Gemez M.B., Reactor
Modeling to Simulate Catalytic Partial Oxidation
and Steam Reforming of Methane. Comparison of
Temperature Profiles and Strategies for Hot Spot
Minimization, Int. J. Hydrogen Energy, 32, p. 1421
(2007).

172

www.SID.ir