Bulletin of the JSME Vol.10, No.

1, 2015

Journal of Thermal Science and Technology

A numerical study of propane-air combustion in meso-scale

tube combustors with concentric rings
Fudhail Abdul MUNIR* and Masato MIKAMI*
*Graduate School of Science and Engineering, Yamaguchi University
2-16-1 Tokiwadai, Ube, Yamaguchi 755-8611, Japan
E-mail: [email protected]

Received 8 January 2015

Abstract
Dwindling energy resources and strong demand for better power sources as compared to conventional batteries
have sparked research interest in micro power generation. The invention of state-of-the-art electronic devices
requires more energy capacity, shorter charging period and light in weight, characteristics of which batteries
lack. Therefore, in recent years micro power generation systems have been seen as a potential alternative to
batteries owing to the obvious advantages that it has. It is essential to fully understand the underlying factors
that affect the combustion stability in meso and micro-scale combustors. One of the popular methods to
examine these factors is by performing numerical simulations. This paper demonstrates an axisymmetric
two-dimensional steady state numerical simulation of propane-air combustion in meso-scale cylindrical tube
combustors with concentric rings. The inner diameter of the tube is set to 3.5 mm and the wall thickness is
specified to 0.7 mm. The concentric rings are placed between the unburned and burned gas region. The main
function of these rings is to act as a flame holder where a stable flame can be easily established. The wall
thermal conductivity in the unburned and burned gas region is varied from 1 W/m/K to 1000 W/m/K and the
results in terms of gas, inner wall, outer wall surface temperature distribution, the blowout limits and
combustion efficiency are analyzed and presented. In addition, the effect of the inlet velocity and the
equivalence ratio is also investigated. The results show that the inlet velocity and equivalence ratio have
significant impacts on the flame temperature, which in turn change the wall temperature distribution. Although
the wall thermal conductivity has minimal effect on the flame temperature, both inner and outer wall surface
temperature are greatly affected. Consequently, this variation of wall temperature contributes to the significant
changes on the blowout limits. It is also shown that the combustion efficiency is influenced by the wall thermal
conductivity of the combustors.

Key words: Micro power generation, Micro and meso-scale combustors, Numerical simulations, Blowout
limits

1. Introduction
Meso and micro-scale combustion phenomena are the key components in micro power generation. An
experimental study has shown that flame stabilization in a confined space or a narrow channel is achievable (Miesse, et
al., 2004). However, the difficulty of sustaining a stable flame for a micro combustor due to its high surface to volume
ratio is a major issue in micro combustion (Suenaga, et al., 2011; Maruta, 2011). This high ratio causes a large portion
of heat loss from the flame to the wall, which leads to thermal quenching (Kaisare, et al., 2012). In recent years,
numerical simulations have been utilized as prominent methods to study the combustion feasibility in meso and
micro-scale channels. By performing the verification, any newly designed of micro combustors can be thoroughly
studied and optimized (Kurdyumov, et al., 2009).

Paper No.15-00029
[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers
1

0123456789
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

As early as in 2000s, Computational Fluid Dynamics (CFD) has been utilized to numerically investigate the flame
propagation in micro-burners. Raimondeau et al. (2002) demonstrated methane (CH4) combustion in a two-dimensional
tubular micro combustor, which was modeled using a detailed multi-component transport and gas chemistry. The
results show that the effect of reactants pre-heating and insulations allows the flame to propagate even in micro
channels. Furthermore, initial heat losses and wall radical quenching are the main parameters that determine the flame
propagation in such micro burners.
In order to provide a better understanding to the underlying fundamentals of the combustion of gaseous
hydrocarbon fuels like propane (C3H8) and methane (CH4) in micro-burners, numerical simulations using commercial
software were conducted. For instance, FLUENT software has been used to investigate the effect of wall conductivity,
burner dimensions and external heat losses on combustion characteristics in meso-scale burners, which modeled as two
parallel plates (Norton and Vlachos, 2003, 2004). In their work, combustion chemistry with one-step irreversible
reaction mechanism was employed. This chemistry is deemed sufficient to demonstrate flame dynamics behavior. It
was highlighted that propane-air micro-flames are more stable than methane-air flames. The lower ignition temperature
of propane fuel contributes to this flame stability enhancement. A similar reaction mechanism is used by Feng et al.
(2010) to numerically solve methane-air combustion in a circular tube combustor. Factors affecting the flame
temperature like the combustor size, inlet velocity and equivalence ratio were studied. Their results indicate that the
equivalence ratio of the mixture and the burner dimension can be considered as important parameters to the flame
temperature. Next, Li et al. (2009) presented their findings on methane-air combustion using detailed combustion
chemistry with 25-reversible reaction mechanism in two-dimensional (2-D) cylindrical tubes and parallel plates. The
partial oxidation that involves in the reaction process can be well represented by the detailed mechanism. Their studies
show that proper material selections can be performed if the flame temperature in the combustor is known.
Apart from hydrocarbon fuels, the prediction of premixed flame temperature and flame dynamics in a narrow
cylindrical tube with hydrogen as a fuel has been reported by several researchers (Brambilla, et al., 2013; Jejurkar and
Mishra, 2011). The effects of various heat transfer conditions on hydrogen combustion have been well discussed by
Hua et al. (2005). A detailed reaction mechanism with 19-reversible reaction mechanism was employed to study the
combustion stability with and without the presence of wall conduction. Their studies suggest that lower wall thermal
conductivity can reduce the heat loss, which results to the enhancement of combustion stability. Nevertheless, the low
wall thermal conductivity is likely to cause hot spot on the combustor wall that can decrease the material lifespan.
Numerical simulation was also performed on a more complex combustor geometry like Swiss-roll combustor in
order to reduce the occurrence of extinction. Chen and Buckmaster (2004) demonstrated the combustion of the
premixed fuel-lean propane–air mixture in a Swiss-roll combustor. The idea of using a flame holder in meso-scale
cylindrical tube combustors was first proposed by Mikami et al. (2013). In their experimental work, a stainless steel
wire mesh is placed between the burned and unburned gas region. The experimental results show that the flame can be
stabilized for both gas and liquid fuel. In fact, it is also demonstrated that flame could be even stabilized in a tube with
a diameter below than the classical quenching distance.
In this study, a numerical simulation of propane-air combustion in meso-scale cylindrical tubes with concentric
rings is presented. A set of rings is used as a flame holder. The effect of wall thermal conductivities of the unburned and
burned regions on flame, inner and outer wall surface temperature distribution was examined. The blowout limits and
combustion efficiency with different combination of wall thermal conductivities were also investigated. The surface
reaction is excluded in the analysis since materials that can resist radical quenching are now available.

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

2
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

2. Numerical Simulation Setup

A two-dimensional (2-D) steady state numerical simulation was performed using ANSYS Release 14.0 with
Fluent 6.3 (ANSYS Inc, 2012). The inner diameter of the combustor is 3.5 mm and the wall is 0.7 mm thick. The wall
thickness is modeled as a solid phase where only the energy equation is solved. The total length of the combustor is set
to 40 mm. The combustor is divided into two parts namely the unburned and burned gas region by concentric rings.
The concentric rings are modeled as square rectangles with 0.14 mm in length and placed side by side where the
location is set to be 30 mm at which is measured from the center of the combustor inlet to the left side of the rings. This
value is chosen to ensure that the flow will be fully developed before passing the rings. The wall of the rings is
modeled as a thermally thin-wall and its material is stainless steel that has a wall thermal conductivity of 20 W/m/K.
One-dimensional (1-D) conductive heat transfer is assumed at the wall of the rings and to model this 1-D heat transfer,
a sufficiently small wire thickness is defined to the wall of the rings, which in this case is 0.01 mm. Hence, with this
assumption, defining the mesh element inside the rings is not necessary and the total number of mesh element in the
computational domain can be reduced. The effect of this thermal conduction is minimal as compared to the thermal
conduction via the combustor wall. The main function of the rings is to stabilize the flame before any variation of other

Fig. 1 Schematic of the computational domain (not to scale)

parameters is made. It is widely known that the flame holder prolongs the residence time in the reaction zone that leads
to the stable flame. The gap between the rings is set to 0.10 mm to enable the fluid flows and passes through this gap.
The fuel type used is propane (C3H8)-air mixture. The schematic of the computational domain for this simulation is
illustrated in Fig. 1.
In this work, Dufour, gas radiation effects and the work done by pressure and viscous force are assumed to be
negligible. Norton and Vlachos (2003) have shown that the gas phase radiation has minimal effect on the combustion
reactions. Ideal gas law is assumed for the gas density and the specific heat for all the species is calculated using a
piecewise-polynomial fit of temperature. The thermodynamics and transport data are obtained from the Fluent internal
database (ANSYS Inc, 2012) and Myer (1998).
Meanwhile, reduced one-step propane combustion with five species as suggested by Westbrook and Dryer (1981)
is employed as the combustion chemistry,
(1)
The species involved are C3H8, O2, N2, CO2 and H2O.
It should be noted that a detailed kinetics mechanism is compulsory if simulations are performed to evaluate
flame or emission characteristics. However, since the focus of this study is to examine the trend pattern of the effect of
wall thermal conductivity on the combustion stability, the single step mechanism is sufficient. A non-uniform mesh
element size is utilized where high mesh concentration is applied in the area surrounding the rings due to the small gap
between these rings. The total number of elements is 19482 and employed for all cases. A grid dependent test was
performed and revealed that greater number of elements yields no significant advantage in terms of results accuracy.
Boundary conditions are specified as proposed by Norton and Vlachos (2004) and Li et al. (2009) No-slip
boundary type condition is applied at the interface between the fluid and the solid wall. The heat flux at this interface is
calculated using Fourier’s law. Heat transfer by means of convection at the outer surface of the wall is given by:

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

23
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

(2)
where is the convective heat transfer coefficient, is the outer wall temperature and is the ambient
temperature, which is set to 300 K for all cases. The convective heat transfer coefficient (h) is fixed to be at 5 W/m2K,
which represents a weak natural convection (Li, et al., 2009).
A thermal insulation (zero heat flux boundary) is applied at both left and right wall edges of the combustor. The
wall thermal conductivity for both the unburned gas region (ku) and the burned gas region (kb) of the solid wall is varied
from 1.0 W/m/K to 1000 W/m/K. The solid zone of ku and kb can be established by defining two separate cell zones of
the solid wall in the ANSYS-Fluent. In a practical application, this region can be separated by using two composite
tubes for the unburned and burned gas region. These tubes can be adhered together using special ceramic adhesive as
demonstrated by Mikami et al. (2013). For the outlet boundary condition, a fixed pressure is used while a far-field
condition is applied for the species normal to the outlet. This means that a zero diffusive flux of species is assumed at
the outlet. An axis boundary condition is established at the centerline so that the calculation can be performed only in
the half of the domain. Table 1 below summarizes the numerical model setup while Fig.2 shows the results of grid
dependent test with different number of elements.
Table 1 Summary of the model setup
Parameters Model setup
Solver Pressure-based, steady-state and axisymmetric
Species model Laminar finite-rate with stiff chemistry solver
Pressure-velocity coupling SIMPLE
Spatial discretization First-order upwind scheme
Density Ideal gas law
Mixture specific heat Mixing law
Species specific heat Piecewise polynomial
Convergence criteria for the 1 x 10-3 for continuity, 1 x 10-3 for velocity
scaled residuals 1 x 10-6 for energy and 1 x 10-3 for species
equation

Fig. 2 Centerline gas temperature distribution along the axial displacement with different number of elements with ku=1
W/m/K and kb= 1 W/m/K for U 0.30 m/s and ɸ=1.0
A laminar flow of U with a flat velocity profile is applied at the inlet while the inlet feed temperature is
maintained at 300 K. At first, the value of U is set to 0.20 m/s and the equivalence ratio (ɸ) of the mixture is initialized
to 1. A technique that is called as “cold flow” is employed where the momentum and continuity equation is first solved.

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

24
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

Then, by patching a high temperature around an area that can be defined as patch zone, the energy and species
equations are then solved. This patch zone is located at 2 mm from the outlet and should not be placed too close to the
outlet since it can result to a reverse gas flow, which negatively affects the ignition process. A sufficiently high
temperature of 1700 K is applied on the patch zone to ensure that the fuel-air mixture can be ignited. Once ignited, the
flame propagated to the upstream and eventually stabilized near the rings. To obtain the blowout limit, the value of ɸ is
first fixed and the value of U is gradually stepped up with 0.01 m/s interval. The corresponding U value that causes the
flame to propagate away from the rings and subsequently blown off the tube is considered as the blowout limit. The
whole process starting from the cold flow step is repeated to obtain the blowout limits for different values of ɸ.

3. Results and Discussion

3.1 Flame propagation and stabilization
Flame propagation in a narrow channel is a broad subject matter. Investigation of flame propagation
characteristics in a narrow channel tube combustor is indeed a formidable task. In order to numerically study the
phenomenon of flame propagation inside a meso-scale tube combustor, a detailed kinetics reaction mechanism is
required that can lead to a higher computational cost. As mentioned in Section 2, this investigation is mainly on the
effect of wall thermal conductivities on the combustion of this type of combustor. Therefore, results shown in Fig. 3 are
only limited to demonstrate the capability of the concentric rings.
(a) (b)

(c) (d)

Fig. 3 Heat of reaction contours with different values of iteration number for ku=1 W/m/K and kb=1 W/m/K, ɸ =1.0 and
U=0.20 m/s, (not to scale; for easier visualization, reflection of axis is used in all figures (a) iteration number=185; (b)
iteration number= 385; (c) iteration number=755; (d) solution converged
The ignition was attained by applying a sufficiently high temperature on the patch zone. This ignition method
replicates the experimental condition reported by Mikami et al. (2013). The inlet velocity was set to U=0.20 m/s while
the equivalence ratio (ɸ) was fixed to 1.0. Since the flame burning velocity is faster than the inlet velocity, the flame
propagated towards the inlet once it was ignited. As shown in Fig. 3, the flame stopped to propagate when it reached
the rings and stabilized there until a converged solution was achieved. This condition is defined as a stable flame. The
concentric rings act as a flame holder by which the flame can be stabilized at a fixed location.

3.2 Flame, gas and wall surface temperature profile

In micro combustors, flame, inner and outer wall temperatures are interdependent to each other. These
characteristics are essentially important in a micro power generation research. Yang et al. (2012) suggest that a good
combustor for micro power generation application should have a sufficiently high and uniform outer wall surface
temperature. As for this simulation study, the wall thermal conductivity values for both the unburned (ku) and burned
(kb) gas region were varied at 1 W/m/K, 10 W/m/K, 100 W/m/K and 1000 W/m/K. The effect of wall thermal
conductivity on the temperature contours of the combustor is shown in Fig.4 while Fig. 5 illustrates the corresponding
gas centerline temperature distribution. The flame temperature can be defined as the highest gas temperature recorded
along the axial displacement. Figure 4 suggests that the variation of the wall thermal conductivity has a minimal effect
on the flame temperature. As the value of ku and kb is increased to 10 W/m/K, 100 W/m/K and 1000 W/m/K
respectively, there is notably better uniformity of gas and wall temperature distributions in the unburned gas region.

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

25
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

(a) (b)

(c) (d)

Fig. 4 (a)-(d) Temperature contours with different values of ku and kb for U=0.30 m/s and ɸ=1.0 (not to scale; for easier
visualization, a reflection of the axis is used in all figures)
Meanwhile, Fig. 5 shows that there is a direct relationship between ku values and the gas temperature in the
unburned gas region particularly from x=3 mm onwards. This change is due to the role of the combustor wall as a heat
(a) (b)

(c) (d)

Fig. 5 Temperature distribution of gas at the centerline along the axial displacement with different values of ku and kb for
U=0.30 m/s and ɸ=1.0: (a) ku=1 W/m/K; (b) ku=10 W/m/K; (c) ku=100 W/m/K; (d) ku=1000 W/m/K
conductor. The main method of transferring the heat from the hot burned gas to the unburned reactants in this type of
combustor is by the wall conduction. The process of heating up the unburned gas prior to the combustion starts from
the inner wall. Thus, the intensity of this process depends on the value of ku. Figure 6 shows the temperature
distribution of inner wall and gas inside the combustor with two different values of ku and a fixed kb value. For ku=10
W/m/K, higher inner wall temperature is obtained in the unburned gas region up to x=29.0 mm as more heat energy is

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

26
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

distributed to the unburned gas region. Generally, this trend is repeated with a further increment of the ku value.
Nonetheless, using ku=1000 W/m/K is no longer effective in boosting up the unburned gas temperature. The next
important criterion for meso and micro scale combustion is the outer wall temperature distribution, which is presented
in Fig. 7. Obviously, it is shown in the figure that the combination of low wall thermal conductivities in the unburned
and burned gas region leads to steeper temperature gradients.

Fig. 6 Temperature distribution of the inner wall and gas at the centerline along the axial displacement from the inlet with
ku=1 W/m/K and ku=10 W/m/K for kb=10 W/m/K, U=0.30 m/s and ɸ=1.0

(a) (b)

(c) (d)

Fig. 7 Outer wall surface temperature distribution along the axial displacement from the inlet with different values of ku
and kb for U=0.30 m/and ɸ = 1: (a) ku=1 W/m/K; (b) ku=10 W/m/K; (c) ku=100 W/m/K; (d) ku=1000 W/m/K
According to Norton & Vlachos (2003), such large temperature gradients can significantly reduce the material
lifespan. In severe cases, melting of the combustor material might occur. Therefore, to avoid such a problem, a uniform
outer wall temperature distribution is desired. This uniform temperature can be achieved by using a high wall thermal
conductivity in the unburned (ku) and burned gas region (kb). However, it is noted that by utilizing a high value of ku
and kb, the maximum wall temperature is dramatically reduced. For instance, with ku=1000 W/m/K and kb=1000
W/m/K, the maximum wall temperature is only at 694.7 K. This reduction is mainly due to the heat energy being

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

27
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

distributed to a wider range of area. A lower wall temperature means a reduction of the combustor efficiency in
converting heat to electrical energy. As such, a proper consideration is necessary in order to balance between the need
of a high performance combustor and a durable combustor.

3.3 Effects of the inlet velocity and equivalence ratio

According to Park et al. (2012), the efficiency of a micro combustor in converting the heat produced from
combustion to electrical energy is significantly affected by the inlet velocity. This fact is due to the temperature
variation with respect to the inlet velocity. With the elevation of the inlet velocity, more reactant mixture is delivered to
the burned region that eventually escalates the reaction rate. Consequently, there is a surge of flame temperature that
results to a greater outer wall surface temperature distribution.
Figure 8 depicts the outer wall temperature distribution along the axial displacement with different values of inlet
velocity (U) and equivalence ratio (ɸ). As shown in Fig. 8 (a), the increment of the inlet velocity escalates the wall
(a) (b)

Fig. 8 Outer wall surface temperature distribution along the axial displacement from the inlet with different U and ɸ for
ku=1 W/m/K and kb=10 W/m/K: (a) for different values of U with constant ɸ=1.0; (b) for different values of ɸ with
constant U=0.20 m/s
temperature particularly in the burned gas region. This pattern is also valid for larger values of ku where the increase of
wall temperature is more noticeable in the unburned gas region. It is important to note that the process of obtaining a
higher outer wall temperature distribution by applying a higher inlet velocity should be prudently performed as blowout
might occur.
The other way to elevate the wall temperature without adjusting the inlet velocity is by increasing the equivalence
ratio (ɸ). Figure 8(b) clearly shows that a richer fuel-air ratio can result to an increase in the wall temperature due to
greater reaction rate. Nevertheless, considering the environment where micro power generators are utilized, combustion
of fuel mixture with below the stoichiometric ratio is preferred.

3.4 Effects of the wall thermal conductivity on the flame stability and combustion efficiency
In meso and micro-scale combustors, the wall thermal conductivity plays a significant role in determining the
flame stability. The axial heat conduction from the burned gas region promotes the pre-heating of the unburned gas that
eventually enhances the flame stability. One of the indicators of flame instability in a micro combustor is blowout. In
general, blowout occurs when the combustor exit velocity is higher than flame burning velocity (Turns, 2003). It is
important to determine the blowout limits for meso and micro-scale combustors so that the operative range can be
ascertained.
In this study, the determination of blowout limits is as follow. Firstly, a stable flame is established in the
combustor. For all cases, a stable flame can be achieved with the inlet velocity (U) of 0.20 m/s regardless of the
equivalence ratio (ɸ) value. Then, the value of U is gradually stepped up with 0.01 m/s interval and further calculation
is conducted until a converged solution is achieved. For instance, if increasing the value of U from 0.25 m/s to 0.26 m/s

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

28
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

causes the flame to propagate away from the rings toward the outlet, then U=0.25 m/s is considered as the blowout
limit. This step is repeated for each case of ɸ and both wall thermal conductivities (ku and kb). Figure 9 shows the flame
image with a blowout condition while Fig. 10 summarizes the blowout limits with respect to all variable parameters. As
depicted in Fig.9, with the value of U increased to 0.42 m/s, the flame is blown out to the tube end resulting to the

(a) (b)

(c)

Fig. 9 Heat of reaction contours in near blowout condition with different number of iteration for ku=100 W/m/K, kb=100
W/m/K, ɸ=1.0. This phenomenon of blowout occurs when U was increased from 0.41 m/s to 0.42 m/s: (a) iteration
number=100; (b) iteration number=7000; (c) iteration number=9500(not to scale, for easier visualization, reflection of
axis is used)
(a) (b)

(c) (d)

Fig. 10 Blowout limits with different values of ku, kb and ɸ (a) kb=1 W/m/K; (b) kb=10 W/m/K; (c) kb=100 W/m/K;
(d) kb=1000 W/m/K

flame extinction. Meanwhile, Figure 10 indicates that as kb is increased from 1 W/m/K to 100 W/m/K, the flame
stability is enhanced. In overall, the combustion of mixture that is rich in fuel significantly improves the blowout limits
as Mikami et al. (2013) experimentally showed for meso-scale quartz tube combustors with wire mesh. However, there
is only a slight improvement in the limit with the use of kb=1000 W/m/K. It is also shown that there is an inverse
relationship between the value of ku and the blowout limits except for ɸ=0.90 as indicated in Fig. 10 (b) to (d). This
reduction of limit is probably due to the variation of pre-heating effect that occurs in the unburned gas region.
Figure 11 presents the temperature distribution along the inner wall with different values of ku and kb. It can be
seen from the figure that the inner wall temperature at x=29.5 mm for ku=1 W/m/K is the highest. This maximum inner

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

29
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

(a) (b)

(c) (d)

Fig. 11 Inner wall temperature with different values of ku for U=0.30 m/s and ɸ=1.0; (a) kb=1 W/m/K; (b) kb=10
W/m/K; (c) kb=100 W/m/K; (d) kb=1000 W/m/K

wall temperature is still within the unburned gas region. Figure 12(a) depicts the numerical value of inner wall
temperature at x=29.5 mm while Fig. 12(b) shows the blowout limits for ɸ=1.0. The highest inner wall temperature is
recorded to be 848.7 K while the maximum blow out limits is 0.52 m/s. Clearly, Fig. 12 suggests that the blowout
limits is affected by the inner wall temperature, particularly in the small area just before the ring position. The
increased inner wall temperature at such a position conceivably increases the local burning velocity at the flame
attachment region. The use of a bigger ku value can result to a lower temperature in this area. Consequently, the
blowout limit is decreased. However, it is observed from the results that for the combustor with kb=1 W/m/K, the
blowout limit does not change dramatically with the increase of ku. It is important to note that at near blowout
conditions, the flame is slightly displaced towards the downstream. For the combustor with ku=1 W/m/K and kb=1
W/m/K, the flame is located at 32.2 mm from the inlet when U is increased to 0.32 m/s (the blowout limit is 0.33 m/s).
There is possibility that the elevation of the inner wall temperature has no effect on the flame stability at this near
blowout condition.
(a) (b)

Fig. 12 (a) Inner wall temperature with different values of kb at x=29.5 mm for U=0.30 m/s and ɸ=1.0 (b) Blowout limits
with different values of kb for ɸ=1.0.
Meanwhile, the combustion efficiency with respect to the wall thermal conductivity is presented in Fig. 13. The
combustion efficiency for the combustor with wall thermal conductivity kb of 1000 W/m/K is not shown in the graph as
it has the same tendency as the combustor with kb=100 W/m/K. It can be seen from the figure that there is an inverse

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

210
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

relationship between the wall thermal conductivity ku and the efficiency. Jianlong et al. (2012) stated that the
combustion efficiency for a micro combustor depends on the inlet velocity, heat losses and flame temperature. Thus,
this variation of combustion efficiency is mainly due to the transverse heat loss to the ambient air with the use of higher
wall thermal conductivity. Nevertheless, the efficiency change is relatively small.

Fig. 13 Combustion efficiency with different values of ku and kb for U=0.20 m/s and ɸ= 0.90

4. Conclusions
The numerical simulation of combustion of propane-air mixture in meso-scale tube combustors with concentric
rings has been successfully demonstrated. The function of the concentric rings is to stabilize the flame at a fixed
location. With the use of these rings, it is possible to vary the wall thermal conductivity in the two different regions.
In this type of combustor, the wall and gas temperature distribution in the unburned gas region is tremendously
affected by the value of wall thermal conductivity. However, the use of a high value of ku or kb in which for this case
k=1000 W/m/K produces no observable effect on both the wall and gas temperature profile. It is demonstrated that the
flame temperature is independent to the wall thermal conductivity. On the other hand, the inlet velocity and the
equivalence ratio have a significant impact on the flame temperature, which in turn affects the wall and gas temperature
distribution.
It is a well-known fact that the utilization of low wall thermal conductivity values leads to the high temperature
gradients that eventually contributes to hot spot related problems. Thus, by establishing the two separate regions of wall
thermal conductivity, the problems related to the high thermal stress could be minimized without compromising the
combustor efficiency.
Lastly, the blowout limits greatly depend on the wall thermal conductivity. The blowout limits are significantly
affected by the inner wall temperature particularly in the area just before the ring position. The increased inner wall
temperature at such a position conceivably increases the local burning velocity at the flame attachment region. The use
of higher wall thermal conductivity in the unburned gas region brings down the value of the inner wall temperature in
the area just before the ring position. As a result, the blowout limits are also reduced. It is also demonstrated that the
combustion efficiency is influenced by the variation of the wall thermal conductivity ku. The use of a combustor with a
low value of ku can improve the combustion efficiency. However, other critical factors such as the outer wall
temperature and the blowout limits should also be considered before selecting the appropriate wall thermal
conductivity.

Acknowledgment
The authors would like to express heartiest gratitude to Universiti Teknologi Malaysia (UTM) for providing the
software. This research was partly subsidized by JSPS KAKENHI Grant-in-Aid for Challenging Exploratory Research

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

211
 Munir and Mikami, Journal of Thermal Science and Technology, Vol.10, No.1 (2015)

(26630443). The first author would like to the Malaysian Ministry of Education (MOE) and Universiti Teknikal
Malaysia Melaka (UTeM) for the study scholarship.

References
ANSYS Inc., ANSYS Fluent Tutorial Guide, ANSYS Inc., Canonsburg PA, (2012).
Brambilla, A., Frouzakis, C.E., Mantzaras, J., Bombach, R., Boulouchos, K., Flame dynamics in lean premixed CO/H 2
/air combustion in a mesocale channel, Combustion and Flame, (2013), In press.
Chen, M., Buckmaster, J., Modelling of combustion and heat transfer in ‘Swiss roll’ micro-scale combustors,
Combustion Theory and Modeling, Vol. 8, (2004), pp. 701-720.
Feng, L., Liu, Z., Li, Y., Numerical study of methane and air combustion inside a small tube with an axial temperature
gradient at the wall, Applied Thermal Engineering, Vol. 30, (2010), pp. 2804-2807.
Hua, J., Wu, M., Kumar, K., Numerical simulation of the combustion of hydrogen–air mixture in micro-scaled
chambers Part II: CFD analysis for a micro-combustor, Chemical Engineering Science, Vol. 60, (2005),
pp.3507-3515.
Jejurkar, S.Y., Mishra, D.P., Effects of wall thermal conductivity on entropy generation and exergy losses in a H 2-air
premixed flame microcombustor, International Journal of Hydrogen Energy, Vol. 36, (2011), pp. 15851-15859.
Jianlong, W., Aiwu, F., Maruta, K., Hong, Y., Wei, L., Experimental and numerical investigation on combustion
characteristics of premixed hydrogen/air flame in a micro-combustor with a bluff body, International Journal of
Hydrogen Energy, Vol. 37, (2012), pp. 19190-19197.
Kaisare, N.S., Vlachos, D.G., A review on microcombustion: Fundamentals, devices and applications, Progress in
Energy and Combustion Science, Vol. 38, (2012), pp. 321-359.
Kurdyumov, V.N., Pizza, G., Frouzakis, C.E., Mantzaras, J., Dynamics of premixed flames in a narrow channel with a
step-wise wall temperature, Combustion and Flame, Vol. 156, (2009), pp. 2190-2200.
Li, J., Chou, S.K., Yang, W.M., Li, Z.W., A numerical study on premixed micro-combustion of CH4–air mixture:
Effects of combustor size, geometry and boundary conditions on flame temperature, Chemical Engineering
Journal, Vol. 150, (2009), pp. 213-222.
Maruta, K., Micro and mesoscale combustion, Proceedings of the Combustion Institute, Vol. 33, (2011), pp. 125-150.
Miesse, C., Masel, R., Short, M., Shannon, M., Experimental observations of methane–oxygen diffusion flame
structure in a sub-millimetre microburner, Combustion Theory and Modeling, Vol. 9, (2005), pp. 77-92.
Mikami, M., Maeda, Y., Matsui, K., Seo, T., Yuliati, L., Combustion of gaseous and liquid fuels in meso-scale tubes
with wire mesh, Proceedings of the Combustion Institute, Vol. 34, (2013), pp. 3387-3394.
Myer K., Mechanical Engineers Handbook Second Edition (1998), John Wiley & Sons.
Norton, D.G., Vlachos, D.G., Combustion characteristics and flame stability at the microscale: a CFD study of
premixed methane/air mixtures, Chemical Engineering Science, Vol. 58, (2003), 4871-4882.
Norton, D.G., Vlachos, D.G., A CFD study of propane/air microflame stability, Combustion and Flame, Vol. 138,
(2004), pp. 97-107.
Park, J.H, Lee, S.I., Wu, H., Kwon, O.C.,Thermophotovoltaic power conversion from a heat-recirculating
micro-emitter, International Journal of Heat and Mass Transfer, Vol. 55 (2012), 4878–4885
Raimondeau, S., Norton, D., Vlachos, D.G., Masel, R.I., Modeling of high temperature microburners, Proceedings of
the Combustion Institute, Vol. 29, (2002), pp. 901-907.
Suenaga, Y., Kitano, M., Yanaoka, H., Development of Ultra-Micro Combustor Using Cylindrical Flames, Journal of
Thermal Science and Technology, Vol. 6, (2011), pp. 406-415.
Turns, S., R., An Introduction to Combustion Concepts and Applications 2nd Edition (2003), p.298, McGraw-Hill.
Westbrook, C.K., Dryer, F.L., Simplified Reaction Mechanism for the Oxidation of Hydrocarbon Fuels in Flames,
Combustion Science and Technology, Vol. 27, (1981), pp. 31-43.
Yang, W.M., Jiang, D.Y., Chou, S.K., Chua, K.J., Karthikeyan, K.H., Experimental study on micro modular combustor
for micro-thermophotovoltaic system application, International Journal of Hydrogen Energy, Vol. 37, (2012), pp.
9576-9583.

[DOI: 10.1299/jtst.2015jtst0008] © 2015 The Japan Society of Mechanical Engineers

212