Applied Thermal Engineering 81 (2015) 140e153

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Research paper

Investigations of asymmetric non-premixed meso-scale vortex

combustion
Mostafa Khaleghi*, Seyed Ehsan Hosseini, Mazlan Abdul Wahid
High-Speed Reacting Flow Laboratory, Faculty of Mechanical Engineering, Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia

h i g h l i g h t s

In this paper non-premixed meso-scale asymmetric vortex combustion is investigated.

The ratio of heat loss to heat generation is maximum, in the air mass flow rate 40 mg/s.

For all air mass flow rates, blow-off occurs when the equivalent ratio is less than 0.2.

The lean side of the vortex flame is more stable.

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

Article history: A new design of asymmetric non-premixed meso-scale vortex combustor is introduced in this paper. The
Received 17 November 2014 flame stability, heat loss from the combustor wall as well as thermal efficiency and pollutant formation
Accepted 10 February 2015 are compared in various circumstances such as air/fuel inlet velocity and equivalence ratios. Further-
Available online 18 February 2015
more, direct photography method is used to capture visible flame structures at a wide range of equiv-
alence ratios in order to emphasize the exceptional stability of such flames. An essential model for the
Keywords:
stability of non-premixed flames in meso-scale combustion spaces is provided in this research. The
Vortex
temperature of the combustor wall is one of the most important factors that influence the temperature of
Meso-scale combustion
Non-premixed
the reactants (preheating phenomena) by heat conduction through the body. The results show that in the
Flame stability stoichiometric circumstance, when air mass flow rate is at the lowest rate (40 mg/s), the ratio of heat loss
to heat generation reaches the largest value (around 55%). The average temperature of the combustor
wall increases with the flow velocity for the stable flame mode and remains mostly uniform and well
distributed for the vortex flame in toroidal shape. At a constant airflow rates, the exhaust temperature
increases monotonously with the decrease in equivalence ratio until the flame blows off. This implies
that the maximum thermal efficiency of a meso-scale combustor occurs in its lean conditions.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction thermal diffusivity, which are the features of laminar flame.

Recently, combustion phenomena in meso-scale combustion
space have attracted attentions and its importance has been re-
ported in numerous review papers [1e5]. Meso-scale combustion is
defined with respect to the relative dimensions of the combustion
system to the representative quenching distance and flame length.
The dimensions of the combustor in meso-scale combustion
(which exceeds 1 mm) affect the flame thickness and effective
* Corresponding author. No.801, U8C, Perdana, Universiti Teknologi Malaysia
(UTM), Skudai, Johor 81310, Malaysia. Tel.: þ60 1112600959.
E-mail address: mostafa26_k@yahoo.com (M. Khaleghi).
Therefore, the concept of combustion phenomena in meso-scale
chambers should necessarily be taken into consideration for
developing laminar flame theories in small-scale combustors.
Flame quenching and flame stabilization are the two main barriers
of small-scale combustors' development [6]. Premixed flames are
usually used to prevent flame quenching when the combustor's size
decreases, and likewise diffusion flames (Non-premixed flames) are
employed to increase the flame stability. Due to the large surface-
to-volume ratio in meso-scale combustors, some new
configurations should be proposed.
As the micro-electronic mechanical systems (MEMS) developed,
the importance of small energy carriers has been highlighted. Chia

http://dx.doi.org/10.1016/j.applthermaleng.2015.02.022
1359-4311/© 2015 Elsevier Ltd. All rights reserved.
 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153
Fig. 1. Experimental setup.
et al. [7] and Chou et al. [8] suggested the application of fossil fuels
due to their great energy density. As a practical method of using
fossil fuel, various types of small-scale combustors have been
proposed. The suggested small-scale combustors have been
employed as thrusting devices, heat sources, fuel reformers, or
photovoltaic sources [9e12]. The most important barrier to develop
small-scale combustors is the quenching limits. Many experimental
and numerical investigations have been carried out to improve
quenching limits and protect flame against flash-back so that the
safety of small-scale combustors can be ensured [13,14]. The pro-
portional relation of quenching distance and flame thickness as
well as the converse proportional relation of the flame thickness
and the burning velocity are known [15]. Therefore, in the first step,
flame thickness should decrease so that scale of the combustor
reduces. In order to fulfill these objectives, several techniques such
as enhancement of the pressure for reducing the molecular dis-
tance, application of special types of fuel or oxidants for intensi-
fying burning velocity, and the application of catalytic reactions for
preventing termination of the chemical chain reaction have been
employed [16,17]. Nevertheless, these techniques are not directly
concerned with the combustor configuration. They can be used for
most combustors after determination of the best design of these
combustors. Hence, the aim of the present study is to investigate
vortex configuration of meso-scale combustor by employing heat
recirculation technique to overcome the ordinary quenching limits.
Table 1
The details of boundary condition.
Viscos model K-epsilon (Eq. (2)) RNG, with swirl dominated
flow option
Radiation model Discrete ordinate (DO)
Reaction model Volumetric species transport reaction with Eddy
dissipation of turbulence chemistry model
Boundary Air inlet Velocity: various
condition Temperature: 300 K
Concentration: O2 ¼ 23%, N2 ¼ 77%
Fuel inlet Velocity: various
Temperature: 300 K
Concentration: CH4 ¼ 97%
Wall Material: steel
Thermal: mixed heat transfer, h ¼ 20 w/m2 K
Ambient temperature: 300 K
141
When the combustor size reduces, the surface-to-volume ratio
rises and the combustion phenomena depends conspicuously on
the combustor configuration. Thus, an investigation about the in-
fluences of practical design parameters on the performance of
small-scale combustion is essential for further development mini-
aturized power generator.
The concept of vortex flames in meso-scale combustor was
developed by Wu et al. [18] and Wang [19] in 2007 and 2006,
respectively. The reaction phenomena as well as flow dynamics on
such scales are fundamentally different from the macro-scale whirl
combustion reported by Gabler et al. [20] and the vortex flame
reported by Saqr et al. [21] and Khaleghi et al. [22].
In this paper, vortex flows as a means to stabilize non-premixed
gaseous flames in meso-scale combustor are investigated. A mini-
aturized combustor with chamber radius as small as 5 mm were
fabricated in high speed flow rate laboratory (HiREF) in Universiti
Teknologi Malaysia (UTM) and were characterized with hydrocar-
bon fuel (methane). The system is based on the concept of asym-
metric whirl combustion [23], which has illustrated unusual
stability characteristics for macro-scale combustors at very lean
conditions. Detailed computational study of the three-dimensional
(3D) flow fields as well as experimental analysis was performed to
investigate the combustion characteristics in the meso-scale
circumstance. Results further verify the favorable temperature
distribution and flow pattern of the asymmetric whirl combustion
concept at meso-scales.
2. Experimental setup
The experimental stage was settled to allow direct photography
from one plane and intrusive access of gas analyzer for emission
and temperature measurements. A schematic of the experimental
platform is shown in Fig. 1. Air is delivered at 2 bars from a gush
tank connected to a single stage reciprocating compressor, equip-
ped with a pressure gauge and regulator. Fuel (Natural-gas) is
supplied at 1 bar from a pressurized cylinder. Two digital precision
flow meters with flow-regulating screw from Cole-Palmer are used
to measure the flow rate of fuel and air. Stretchy plastic plumbing is
employed to connect the combustor chamber to the fuel-air supply
system. A Sony 300 FPS digital camera is installed perpendicular to
the outlet plane of the combustor to execute direct photographs for
142 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153

Fig. 2. The design of the vortex combustor.

the flame. A TELEGAN Tempest-100 gas analyzer with a stainless 3. Numerical model
steel probe is used to measure the emissions and temperature in-
side the combustor. A traverse system enables the axial movement 3.1. Governing equations
of the probe. An S-type thermocouple probe with ceramic insu-
lation measured the product gas temperature. TC-08 thermocouple The 3D steady-state Favre-averaged governing equations for
data logger measures temperature both fast and accurately. Tem- mass, momentum, species mass fraction and energy in Cartesian
perature data acquisition with the TC-08 (DAQ) is simply plugged coordinates are given as [24]:
into a USB port on the computer, connects the thermocouples, and
measures temperatures.

Table 2
Dimensions of the meso-scale vortex combustors (all dimensions
are mm).

L1 40
R (mm) (chamber radius) 5
L2 30
L3 32
df (mm) (fuel inlet diameter) 1
da (mm) (air inlet diameter) 1.5
Lf 5
L air 6
a (mm) 2
b (mm) 1
c (mm) 10
D'f 7
D0 air 9
L'f 5
L'air 10
Fig. 3. Numerical grid independent test.
 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153 143

2 0 13
f   f
e
vrui Yn v 6 mt BvYn C7
¼ 4 rDn þ @ A5 þ u_ n (3)
vxj vxj SCt vxj

" #
vru ~
~j H v vT~ m vh~ X N ~n
vY XN
¼ l þ t þ rDn hn  hfn u_ n (4)
vxj vxj vxj Prt vxj n¼1 vxj n¼1

where r is the density, ui/j is the velocity vector components, P is

the pressure, l is the thermal conductivity, Prt is the turbulent
Prandtl number, Dn is the mass diffusivity of species (n) which was
assumed to be constant for each species, Yn is the mass fraction of
_ n is the chemical reaction rate, m is the dynamic vis-
species (n), u
cosity and mt The ~ denotes the Favre averaging [25].

3.2. Computational procedure

Fig. 4. The stability region of vortex flames in meso-scale combustor. A three dimensional (3D), finite volume solver has been used to
discretize the flow domain through a second-order upwind
scheme. Various triangular grids have been generated for ensuring
vruej that the solution is grid-independent. The SIMPLE algorithm has
¼0 (1) been employed to obtain the mass conservation between the ve-
vxj
locity terms and pressure in the discretized momentum equation.
2 0 13 Chemical reaction has considered volumetric and Eddy-Dissipation
(ED) algorithm has been selected for turbulenceechemistry in-
vruei u
~j vP v 6 v
B ie
u v e
uj C7
¼ þ 4ðm þ mt Þ@ þ A5 (2) teractions. The ED reaction model ignores chemical kinetics (i.e.,
vxj vxi vxj vxj vxi the Arrhenius rate) and uses only the parameters in reaction flow
[26]. The operating pressure and temperature were set 1.01 bars
and 300 K, respectively. The partial-equilibrium model is utilized to

Fig. 5. Predicted tangential velocity profiles at different axial distances with mass flow rate a ¼ 40 mg/s, b ¼ 80 mg/s, c ¼ 120 mg/s and d ¼ 170 mg/s respectively.
144 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153

Fig. 6. The contours of velocity on cross sections z ¼ 1.5, 5, 10, 15, 20 and 25 in the chamber with different mass flow rate at 4 ¼ 1.

predict the oxygen (O) radical concentration required for thermal 4. Results and discussion
NOx prediction. A steady state pressure based solver was applied to
solve the governing equations by CFD code ANSYS Fluent 14 in all 4.1. Flame stability
the simulations.
The solution is considered to be converged when the residuals of The diagram of stable regions for vortex flame was drawn based
each governing equation at consecutive iterations became less than on various ranges of air mass flow rate and different equivalence
1  104 except energy equation and chemical reactions equation ratios (4). The blow-off region was identified with the limitation of
which are converged at the quantities less than 1  106. At such specific values of mass flow rate according to the working param-
condition, the flow field variables reached to stable local values eters of the experimental platform. Fig. 4 depicts the stable and
with respect to any number of iterations. Also, monitor for pa-
rameters of NOx and temperature converged separately as well.
This convergence criterion was applied on reacting flow cases.
Table 1 demonstrates the details of boundary condition of simu-
lated asymmetric vortex flame.
The design of the asymmetric vortex combustor has been
illustrated in Fig. 2 and the related dimensions are given Table 2.
A grid independence test was performed to evaluate the effects
of grid sizes on the results as shown in Fig. 3. Five sets of meshes
were generated using triangular elements with M1 ¼ 63,650
nodes, M2 ¼ 127,340 nodes, M3 ¼ 254,675 nodes, M4 ¼ 509,373
nodes and M5 ¼ 1,018,752 nodes. Laminar flow with counter-flow
configuration was considered for this test where the temperatures
of inlet air as well as the inlet fuel were set 300 K. It was found
that the 509,373 nodes and 1,018,752 nodes produce almost
identical results along the chamber with a percentage error of
1.2%. Therefore, domain with 509,373 nodes was selected to
reduce the computing time. The computational domain of vortex
combustor included of both solid domain and fluid domain. A
computer with core i7 CPU and 16 GB of RAM was used to Fig. 7. Axial evolution of swirl number in the chamber with viscous head end at
perform these computations. p ¼ 1 atm.
 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153 145

blow-off region of vortex flame in meso-scale vortex combustor.
The flame stability limit was empirically obtained. Stable and un-
stable regions are separately illustrated in the graph. The unstable
region indicates that the flame will blow off at specific equivalent
ratio and mass flow rate. Blow off occurs for all mass flow rates
when the equivalent ratio is below 0.2. The blow-off phenomena
occurs when the Reynolds number is high and the equivalence ratio
is relatively small. However, at mass flow rate values, which are
different in the order of magnitude, the flame should have different
stability criterion. No other forms of instability, except for blow-off,
were possible to identify in the working conditions presented in
Fig. 4.
The extinction phenomena depends on two parameters. First
item is heat loss through the combustor's wall and the second item

Fig. 8. Temperature contours of the reacting flow inside the vortex combustor for different mass flow rate. (The first category is at plane (A) and the second category is at plane (B)).
146 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153
Fig. 9. Heat loss of meso-scale chamber at different mass flow rates.
is blow-off extinction which emerges when there is insufficient
residence time compared to chemical reaction time. Fig. 4 illus-
trates that the blow-off limit increases with the increase of the
mass flow rate, because in lean side of fuel/air mixtures, the
amount of heat release and chemical reaction rate increase with the
equivalence ratio [27]. Consequently, the temperature of mixture
flow remains at a higher level for larger equivalence ratios.
4.2. Velocity and swirl number
The first important aspect of the vortex structure in the asym-
metric vortex combustor is the tangential velocity profile. In Fig. 5,
the development of the tangential velocity profiles at five different
axial locations in the combustors is depicted. The figure shows that
the vortex structure reaches to a fully developed configuration by
increasing the inlet airflow. The figure also demonstrates the
asymmetry of the vortex structure. The steady-state trend of the
tangential velocity profiles demonstrate that it constitutes a forced
vortex field, rather than a Rankine vortex as in swirling flows.
Fig. 10. Isocontours of the statistical average of (a) pressure and (b) azimuthal velocity.
Considering the tangential velocity profile it can be perceived
that due to the effect of centrifugal force the velocity increases from
minimum at the center of the combustor to the maximum at a
radius which is more than half of the chamber radius. However, due
to wall friction, tangential velocity lessens at a distance close to the
combustor's wall. Due to the swirl decay the peak tangential ve-
locity decreases as the axial distances from the injection location.
The location of maximum tangential velocity moves toward the
center of combustor when (z/L) increases. This means that the core
region shrinks as swirl decays, which was mentioned by Ito et al.
[28]. The distribution of velocity on the different cross sections
from bottom of chamber to top of the chamber is shown in Fig. 6.
Due to the fully tangential inlet air, the maximum velocity magni-
tude is in the vicinity of chamber at inlet air and it decreases slightly
along the chamber length.
As it can be seen in this figure, no more changes in the velocity
contours can be detected after z ¼ 10 due to fully developed tur-
bulent flow.
4.3. Swirl number
The central recirculation zone is strongly dependent on the swirl
number, which is a non-dimensional number representing the axial
flux of swirl momentum divided by the axial flux of the axial mo-
mentum, times the equivalent nozzle radius [29].
Z
ðrux uq ÞrdA
A
S¼Z (5)
ru2x dA  D=2
A
The swirl number was calculated at various axial locations along
the chamber at p ¼ 1 atm which is plotted in Fig. 7. This was per-
formed by integrations of Eq (5) on different surfaces in the XY
plane. In these computations, there is no assumption of axisym-
metric distribution of the azimuthal velocity. The predicted swirl
number using Eq (5) is based on the local value of axial and
azimuthal velocity component at every grid cell, on each specific
plane. From Fig. 7, it can be construed that an augmentation in
 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153
Fig 11. Digital photos (a) two stoichiometric fuel jets without the air vortex and (b)
stoichiometric vortex flame.
injection velocity can lead to an increase in the swirl level. High
swirl favors the generation of a central recirculation zone.
Regions of high swirling flow are important to increase mixing
of reactants and flame stabilization. To ensure efficient combustion,
fast mixing and long residence time are required. The present flow
enhances flame stabilization through generation of a relatively
quiescent core flow.
Swirling flow is an effective way to stabilize flames in combustor
by establishing a high adverse pressure gradient which causes a
flow reversal and determines a central recirculation zone.
The central recirculation zone is formed when the swirl number
is more than 0.6 which was obtained experimentally [30]. The swirl
number increases with the inlet air flow velocity. Strong swirling
flows in the meso-scale combustor, on the axial distance of about 1/
3 of the combustor length.
4.4. Temperature pattern
Temperature patterns of the vortex flame in question are of
significant importance to investigate. This importance is driven
147
from the demand to establish a theory for the aerodynamic stability
mechanism for such flames. Fig. 8 illustrates 2D temperature pro-
files of vortex flames in stoichiometric Natural-gas/air flame at
atmospheric conditions. The temperature patterns are shown at
two different planes to indicate the asymmetry of the reacting flow
field. Comparing the temperature profiles taken at two different
planes, A and B, the asymmetric of the flow can be easily depicted.
In the first group of Fig. 8, where the temperature field is plotted on
a coplanar to the asymmetry region, the temperature in the region
downstream the asymmetry is found to be relatively low, while by
increasing the mass flow rate, the temperature in the region
downstream the asymmetry is found to be relatively high.
The average temperature in the combustion zone is lower than
that of the adiabatic flame temperature (adiabatic flame tempera-
ture of natural gas is 2233 K) due to heat loss from the combustor's
wall. The temperature peak located near flame front reaches to
2145 K.
The hot flow occupies larger volumes in the central region of the
combustor with the increase of mass flow rate in stoichiometric
condition due to higher swirling flow velocity and simultaneously
presence of stronger central recirculation zone.
4.5. Heat loss through the combustor wall
For the case with stoichiometric equivalence ratio, wall tem-
peratures at four mass flow rates are calculated to compare the
combustion heat release and heat loss to the environment. The
calculated combustion heat and heat loss are presented in Fig. 9. It
demonstrates that with the increase of mass flow rate, wall heat
loss as well as combustion heat increase. When equivalence ratio is
close to unity, in lowest mass flow rate (40 mg/s), they reach the
largest values and the ratio of heat loss to combustion heat reaches
the peak value, 55%. This means that the wall heat loss at meso-
scale chamber takes a large amount of the combustion heat,
which is tens of times of a traditional combustor. Indeed, convec-
tion and radiation are the major contribution to heat loss and takes
all portion of the total heat loss. Compared to the stoichiometric
condition, combustion temperature is lower for lean conditions.
Furthermore, the exhaust temperature rises with the airflow rate.
This implies that the heat loss to the surrounding abates as the
equivalence ratio mitigates. Hence, it can be concluded that a lean
non-premixed vortex combustion which acts like premixed com-
bustion, is the most thermally efficient condition for meso-scale
combustors.
The thermal conductivity of steel is 16.27 (W/m. K) while the
emissivity of steel is high and taken 0.85 related to high tempera-
ture. Total heat loss is dominated by radiation and hence, convec-
tive heat loss is very small. When the mass flow rate of air is 170
(mg/s), the radiation heat loss is 71% of the total heat loss.
4.6. Vortex structure
The flow in the asymmetric vortex combustor was found to
exhibit recirculation phenomena. A central recirculation zone was
found to exist in the core of the combustor, and two other recir-
culation zones were found to exist in the in the asymmetric region.
The central recirculation zone was formed due to the strong
azimuthal velocity component, which is created due to the negative
pressure. Fig. 10 depicts the statistically averaged values of
azimuthal velocity and pressure to show the recirculation and
negative pressure zones, respectively.
Fig. 11 displays the stoichiometric vortex flame digital photos of
in daylight settings. Two stoichiometric fuel jets have been shown
with and without the vortex flow field. Fig. 11(a) provides image of
the flame shape and color (in the web version) of the jet flame is
148 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153

Fig. 12. Photographs of the vortex flame at various equivalence ratios.

similar to the typical non-premixed free jet flames. On the other

hand, the abbreviated flame length of the vortex flame is quite clear
in Fig. 11(b). In comparison with the total length of the meso-
combustor (400 mm), the vortex flame has a short length. The
color of vortex flame is blue as depicted in Fig. 11(b) which is a main
characteristic of premixed flames [32].
The structures of flame at various equivalence ratios have been
captured and reported in Fig. 12. To emphasize the flame structure
by its visible color (in the web version), the related photographs
were captured in dark-room. At different equivalence ratios, the
most notable feature of vortex flame is unchanged circumferential
structure. Enhancement of the flame brightness with the increase
of equivalence ratio is another notable flame characteristic.

4.7. Meso-scale vortex flame height

It is crucial to study the flame height in order to investigate the

Fig. 13. Predicted flame height and maximum flame temperature as functions of
equivalence ratio at Re ¼ 308.
structure of the vortex flame. There are several definitions of the
flame height, but the definition that suits the current work is: “it is
the distance between the fuel inlet port and the point where flame
temperature is at maximum value” [33,34]. This definition is quite
difficult to observe experimentally, as it requires accurate local
measurements of the temperature field in the combustor. In this
 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153 149

Fig. 14. Variation of NOx concentration respect to the mass flow rate of air.

section, the flame height is determined based on the maximum 4.8. Exhaust gas analysis
computed temperature from the numerical simulation. Fig. 13 il-
lustrates the flame height as predicted by the numerical simulation 4.8.1. NOx emission characteristics
and corresponding maximum flame temperature linear regression The nitrogen oxide, NOx is emitted harmful gaseous that greatly
based on the least-square method which was used to correlate the affect the environment, hence it needs to be reduced [35]. The main
results. Both flame height and maximum temperature were found reason is that the heat absorption by water vaporization causes a
to be proportional to the equivalence ratio. With a constant air flow mitigation of local adiabatic flame temperature and therefore de-
rate, the fuel is consumed at a specific rate. As the fuel flow rate creases the chemical reaction in gas phase to produce thermal NOx.
increases, the flame propagation improves as a result of the excess The gas analyzer from TELEGAN was used for measuring NOx
fuel being transported into the mainstream. The flame height can concentration of the vortex flame at different equivalence ratios
be expressed as a function of the equivalence ratio for Re ¼ 308 as: and mass flow rate of air. The measurements are taken in the center
hf ¼ b0 þ b14 Where b0 ¼ 7.3940, b1 ¼ 0.2225 and 4 is equivalence of the combustor at a distance of 3 mm from the outlet plane. Fig. 14
ratio. While the maximum temperature at the same Reynolds shows the absolute NOx concentration at different equivalence ra-
number, can be expressed as: Tf ¼ b0 þ b14 Where b0 ¼ 2004.19, tios for three values of four air mass flow rate.
b1 ¼ 0.2886 and 4 is equivalence ratio. NOx is produced primarily in the combustor whenever the
combustion temperature and the concentration of O and OH atoms
are high. These situations are close to stoichiometric conditions that
is the broad regions near the flame tip. When the mass flow rate of
Table 3
Values of least-square regression constants for different Reynolds number. air is constant at 40 (mg/s), NOx formation increases due to high
residence time. By increasing the equivalence ratio, residence time
Constant m_ ¼ 40 mg/s m_ ¼ 80 mg/s m_ ¼ 120 mg/s m_ ¼ 170 mg/s
reduces because the fuel inlet is perpendicular in this non-premixed
y0 0.977 6.704 10.386 14.022 vortex combustor design. Consequently the rate of NOx formation
y1 4.617 0.772 2.179 4.962
decreases in air mass flow rates of 120 mg/s and 170 mg/s.
150 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153
Fig. 15. Mole fraction of O2 in the exhaust gas for different mass flow rate of air.
The level of NOx is well below 10 ppm for all mass flow rates and
equivalence ratios. The least-square method was used to correlate
the NOx level with the equivalence ratio with the form
CNOx ¼ y0 þ y14, and the values of constants y0 and y1 are given in
Table 3 for the three values of Reynolds number.
Although the NOx concentration is obviously affected by the
equivalence ratio, the regression results indicates that such effect is
far less significant by the Reynolds number. This result agrees well
with the established theory of NOx formation developed by Zel-
dovich. The Zeldovich mechanism for NOX formation is very sen-
sitive to temperature. The increase of mass flow rate of air in
stoichiometric equivalence ratio dictates corresponding increase in
the adiabatic flame temperature, thus the rate of NOx increases
[36].
4.8.2. O2 and CO2
Fig. 15 shows the percentage of O2 in the flue gases as the
function of equivalence ratio at four Reynolds number ¼ 308, 590,
879 and 1179, respectively. It is observed that the conversion of fuel
is identical in all cases. The high percentage of O2 in the lower
equivalence ratio suggests that burning efficiency is less at lower
mass flow rates of fuel. That can also be supported by the lower
exhaust temperature in the combustors at lower mass flow rate of
fuel.
Thus, as the mass flow rate of air increases the fuel-air mixture
becomes leaner and the percentage of the fuel that takes part in the
combustion increases.
The effect of Reynolds number in the percentage of CO2 in the
exhaust can be observed in Fig. 16 for different equivalence ratios.
The amount of mass flow rate of air is kept constant for each case.
The study suggests that the percentage of fuel conversion decreases
with enhancement of equivalence ratio for each constant Reynolds
number. The percentage of CO2 in the exhaust is also increasing
since the amount of fuel in flow increases by rising equivalence
ratio.
For all cases in lean side, less CO2 is formed and it is approxi-
mately the same, even by increasing mass flow rate of air. The
reason is that the lean side within meso-scale vortex combustor has
fast combustion and suitable fuel/air mixing by fully vortex flow.
4.9. Exhaust temperature of natural gas combustion
The combustor and its support structure were not thermally
insulated in these experiments. Therefore, heat is allowed to escape
the combustor in convection and radiation forms. Fig. 17 depicts the
exhaust flue gases temperature at stoichiometric circumstance at
various thermal powers. The temperature increases linearly with
increasing thermal power for vortex chamber.
As it can be seen in Fig. 17, the exhaust temperature is directly
proportional to the thermal power. Although it should be noted
that the relationship between the Reynolds number and exhaust
temperature is not the same. The increase of mass flow rate shows
significant effect on the exhaust temperature similar to increase of
the thermal power. This characteristic feature of vortex flames has
been also reported in Saqr et al. [21], while it's a bit different for
current study due to the size of chamber.
It can be interpreted that when the thermal power is greater,
more combustion heat is released which leads to a higher
 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153
Fig. 16. Mole fraction of CO2 in the exhaust gas for different mass flow rate of air.
temperature in the combustor. As a result, the exhaust gas tem-
perature also raises. Moreover, the exhaust gas temperature in-
creases first and then decreases moderately with the increase of
inlet fuel mass flow rate. This means that the exhaust gas tem-
perature reaches its maximum at stoichiometric condition.
4.10. Thermal efficiency
According to the technique of employment generated thermal
energy, various definitions of thermal efficiency are expressed
[34,37]. For example, in thermal-photovoltaic power generator and
heater, the beneficial thermal energy is related to a portion that is
transmitted through the wall, however this is considered as heat
“loss” for applications where high sensible energy is required in the
exhaust gas. In the present study, the thermal efficiency is defined
as the ratio of the sensible energy in the combustion products to the
total chemical power input. Such a definition is more suited for
power plants or micro propulsion systems. It can be written
mathematically in the following form:
P Experimental and numerical investigations were implemented
_
k mk C p;k ðTc  Ti Þ
ht ¼ P  100%
_
i mi hc;i
where m_ i and hc,i are the mass flow rate and heat of combustion of
reactant species i, respectively, and m_ k is the mass flow rate of
species k in the product. The constant-pressure heat capacity of
species k, C p;k , is evaluated at the measured temperature, Tc, and the
inlet temperature Ti is set at 300 K. Thermal efficiency for the
chamber with diameter of 10 mm were calculated 42%, 55%, 62%
and 69% for Re ¼ 308, 590, 879 and 1179 respectively. All of these
151
cases are considered in stoichiometry equivalence ratio. As
mentioned above, the measured temperatures were not necessarily
the combustor average temperatures. Furthermore, attempts con-
trol thermal losses were not considered, and consequently, thermal
efficiencies were low while the chemical efficiencies were high.
By increasing the Reynolds number in the stoichiometric
equivalence ratio, the thermal efficiency increases (which is same
as variations of the exhaust temperature showed in Fig. 17). At
stoichiometric condition when Reynolds number is 308, thermal
efficiency is low because a considerable amount of thermal energy
is lost in the environment. In the low Reynolds number the flame
propagates close to the wall. Due to high surface to volume ratio of
meso-scale combustor, heat loss from the wall is around 60% of heat
release. Moreover, when the Reynolds number is low, the residence
time is longer which provides longer time for heat loss from the
wall.
5. Conclusion
for reacting flow fields of asymmetric meso-scale turbulent vortex
(6)
flames. Vortex flames are stabilized in vicinity of wall and bottom of
chamber wherever is fully tangential velocity of forced vortex flow
fields. Blow-off occurs for all mass flow rates when the equivalent
ratio is less than 0.2. The lean side of the vortex flame is more
stable. The temperature of the combustor wall is one of the most
important factors that influences on the reactants temperature by
conducting the heat through the body. The results show that in the
stoichiometric circumstance, when air mass flow rate is at the
lowest rate (40 mg/s), the ratio of heat loss to heat generation
152 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153
Fig. 17. Exhaust temperature of natural-gas combustion at varying mass flow rate a ¼ 40 mg/s, b ¼ 80 mg/s, c ¼ 120 mg/s and d ¼ 170 mg/s.
reaches the largest value (around 55%). Despite the fact that the
current study of meso-scale combustor deals with the non-
premixed vortex flames, the features of premixed flames such as
color and temperature were also observed. The maximum tem-
perature reaches to 2145 K in the area closed to flame front. Ther-
mal efficiency for meso-scale combustor were calculated 42%, 55%,
62% and 69% for Re ¼ 308, 590, 879 and 1179 respectively. In the
lower rates of air mass flow (40 mg/s and 80 mg/s); when equiv-
alence ratio increases the rate of NOx formation raises. However, in
higher air mass flow rates (120 mg/s and 170 mg/s) when equiva-
lence ratio increases, the rate of NOx formation decreases.
References
[1] K. Maruta, Micro and mesoscale combustion, Proc. Combust. Inst. 33 (1)
(2011) 125e150.
[2] M. Lee, N. Kim, The stabilization of a methaneeair edge flame within a mixing
layer in a narrow channel, Combust. Flame 157 (1) (2010) 201e203.
[3] M.J. Lee, S.M. Cho, B Il Choi, N Il Kim, Scale and material effects on flame
characteristics in small heat recirculation combustors of a counter-current
channel type, Appl. Therm. Eng. 30 (2010) 2227e2235.
[4] M. Khaleghi, M.A. Wahid, M.M. Seis, A. Saat, Investigation of vortex reacting
flows in asymmetric meso scale combustor, Appl. Mech. Mater. 388 (2013)
246e250.
[5] V. Shirsat, A.K. Gupta, A review of progress in heat recirculating meso-scale
combustors, Appl. Energy 88 (2011) 4294e4309.
[6] G. Bagheri, S.E. Hosseini, M.A. Wahid, Effects of bluff body shape on the flame
stability in premixed micro-combustion of hydrogeneair mixture, Appl.
Therm. Eng. 67 (2014) 266e272.
[7] L.C. Chia, B. Feng, The development of a micropower (micro-thermophoto-
voltaic) device, J. Power Sources 165 (2007) 455e480.
[8] S.K. Chou, W.M. Yang, K.J. Chua, J. Li, K.L. Zhang, Development of micro power
generators e a review, Appl. Energy 88 (2011) 1e16.
[9] W. Yang, S. Chou, C. Shu, Z. Li, H. Xue, Combustion in micro-cylindrical
combustors with and without a backward facing step, Appl. Therm. Eng. 22
(2002) 1777e1787.
[10] K.B. Kim, O.C. Kwon, Studies on a two-staged micro-combustor for a micro-
reformer integrated with a micro-evaporator, J. Power Sources 182 (2008)
609e615.
[11] J.F. Pan, J. Huang, D.T. Li, W.M. Yang, W.X. Tang, H. Xue, Effects of major pa-
rameters on micro-combustion for thermophotovoltaic energy conversion,
Appl. Therm. Eng. 27 (2007) 1089e1095.
[12] W. Yang, S. Chou, K. Chua, H. An, K. Karthikeyan, X. Zhao, An advanced micro
modular combustor-radiator with heat recuperation for micro-TPV system
application, Appl. Energy 97 (2012) 749e753.
[13] V. Vijayan, A.K. Gupta, Combustion and heat transfer at meso-scale with
thermal recuperation, Appl. Energy 87 (2010) 2628e2639.
[14] B. Lewis, G. von Elbe, Combustion, Flames and Explosions of Gases, Elsevier,
2012.
[15] N Il Kim, T. Kataoka, S. Maruyama, K. Maruta, Flammability limits of stationary
flames in tubes at low pressure, Combust. Flame 141 (2005) 78e88.
[16] T. Okamasa, G.-G. Lee, Y. Suzuki, N. Kasagi, S. Matsuda, Development of a
micro catalytic combustor using high-precision ceramic tape casting,
J. Micromech. Microeng. 16 (2006) S198eS205.
[17] J. Ahn, C. Eastwood, L. Sitzki, P.D. Ronney, Gas-phase and catalytic combustion
in heat-recirculating burners, Proc. Combust. Inst. 30 (2005) 2463e2472.
[18] M. Wu, Y. Wang, V. Yang, R.A. Yetter, Combustion in meso-scale vortex
chambers, Proc. Combust. Inst. 31 (2007) 3235e3242.
[19] Y. Wang, A Numerical Study of Combustion in Meso-scale Vortex Chambers,
The Pennsylvania State University, 2006.
[20] H. Gabler, An Experimental and Numerical Investigation of Asymmetrically-
Fueled Whirl Flames, Diss Princet Univ, 1998.
[21] K.M. Saqr, H.S. Aly, M.M. Sies, M.A. Wahid, Computational and experimental
investigations of turbulent asymmetric vortex flames, Int. Commun. Heat.
Mass Transf. 38 (2011) 353e362.
[22] M. Khaleghi, S. Ehsan Hosseini, M. Abdul Wahid, Emission and combustion
characteristics of hydrogen in vortex flame, J. Teknol. 66 (2014).
 M. Khaleghi et al. / Applied Thermal Engineering 81 (2015) 140e153
[23] H.C. Gabler, R.A. Yetter, I. Glassman, Asymmetric Whirl Combustion: a New
Approach for Non-premixed Low NOx Gas Turbine Combustor Design, 1998,
pp. 1e11.
[24] A. Obieglo, J. Gass, D. Poulikakos, Comparative study of modeling a hydrogen
nonpremixed turbulent flame, Combust. Flame 122 (2000) 176e194.
[25] K. Saqr, M. Sies, M. Wahid, Numerical investigation of the turbulence-
combustion interaction in nonpremixed CH4/air flames, Int. J. Appl. Math.
Mech. 5 (8) (2009) 69e79.
[26] T.D. Canonsburg, ANSYS FLUENT User ’ S Guide, vol. 15317, 2011, pp.
724e746.
[27] S. Turns, An Introduction to Combustion, 1996.
[28] M. Ito, H. Kimura, Boiling heat transfer and pressure drop in internal spiral-
grooved tubes, Bull. JSME 22 (1979) 1251e1257.
[29] Y.A. Eldrainy, K.M. Saqr, H.S. Aly, M.N.M. Jaafar, CFD insight of the flow dy-
namics in a novel swirler for gas turbine combustors, Int. Commun. Heat.
Mass Transf. 36 (2009) 936e941.
153
[30] A.K. Gupta, D.G. Lilley, N. Syred, Swirl Flows. Tunbridge Wells, Abacus Press,
Kent, England, 1984, p. 488.
[32] R. Marvodineanu, H. Boiteux, R. Mavrodineanu, Flame Spectroscopy, 1965.
[33] A. Mukhopadhyay, I. Puri, Numerical Simulation of Methane-air Nozzle
Burners for Aluminum Remelt Furnaces, vol. 369, ASME-PUBLICATIONS-HTD,
2001, pp. 65e72.
[34] J.L. Consalvi, Y. Pizzo, B. Porterie, J.L. Torero, On the flame height definition for
upward flame spread, Fire Saf. J. 42 (2007) 384e392.
[35] S.E. Hosseini, M. a. Wahid, A.A.A. Abuelnuor, Pollutant reduction and energy
saving in industrial sectors by applying high temperature air combustion
method, Int. Rev. Mech. Eng. 6 (2012) 1667e1672.
[36] M.C. Drake, R.J. Blint, Relative importance of nitric oxide formation mechanisms
in laminar opposed-flow diffusion flames, Combust. Flame 83 (1991) 185e203.
[37] C.M. Spadaccini, A. Mehra, J. Lee, X. Zhang, S. Lukachko, I.A. Waitz, High power
density silicon combustion systems for micro gas turbine engines, J. Eng. Gas.
Turbines Power Trans. Asme 125 (2002) 709e719.