You are on page 1of 9

JID: PROCI

ARTICLE IN PRESS [m;July 10, 2018;23:58]

Available online at www.sciencedirect.com

Proceedings of the Combustion Institute 000 (2018) 1–9


www.elsevier.com/locate/proci

Local supersonic and subsonic combustion mode


transition in a supersonic jet flame
Donggang Cao a,b, Guoqiang He a, Fei Qin a,∗, Dan Michaels b
a Science and Technology on Combustion, Internal Flow and Thermal-structure Laboratory, Northwestern Polytechnical
University, Xi’an, Shaanxi 710072, PR China
b Faculty of Aerospace Engineering, Technion – Israel Institute of Technology, Haifa 3200002, Israel

Received 1 December 2017; accepted 26 June 2018


Available online xxx

Abstract

An experimental and computational study has been carried out for a supersonic jet flame by using OH
chemiluminescence imaging, shadowgraph visualization, temperature measurement by TDLAS, pressure
measurement by transducers, and large eddy simulation (LES) together with a skeletal reaction mechanism
involving 13 species and 41 steps. Agreements have been found between experimental data and LES results,
which are subsequently used to analyze the flow, mixing, combustion, and heat release processes involved. A
systematic method is adopted to qualitatively as well as quantitatively investigate different combustion modes
and their contributions to heat release in the combustor. Influences of airstream temperature and pressure
on combustion mode and heat release are also discussed by comparing four different cases. Results show that
the heat is released from a combination of supersonic combustion mode and subsonic combustion mode
even when the main flow is at supersonic speed. Local mode transition occurs as the jet flame propagates and
interacts with shocks that enhance mixing because of baroclinic effects and induce subsonic combustion due
to deceleration effects. It is also observed that subsonic combustion releases more than 50% of heat at the
base of the jet flame because of recirculation zones behind the strut. Supersonic combustion mode gradually
becomes prominent in the turbulent far field with small values of heat release rate. The overall dominant
combustion mode is dependent on not only inflow conditions but also combustion intensity.
© 2018 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

Keywords: Supersonic jet flame; Combustion mode; Heat release; Large eddy simulation

1. Introduction and rocket-based combined cycle (RBCC) engines


[1–3]. In such engines the aerodynamics are quite
The interest in hypersonic flight motivates complex, permitting different flow conditions for
the development of dual-mode scramjet engines the ramjet and scramjet operation modes. These en-
gines operate in the thermal-choked ramjet-mode

at flight Mach numbers of 3–6. A shock train in
Corresponding author. the isolator slows the flow to subsonic speeds before
E-mail address: qinfei@nwpu.edu.cn (F. Qin).

https://doi.org/10.1016/j.proci.2018.06.213
1540-7489 © 2018 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213
JID: PROCI
ARTICLE IN PRESS [m;July 10, 2018;23:58]

2 D. Cao et al. / Proceedings of the Combustion Institute 000 (2018) 1–9

the fuel is burned in the combustor. When the flight 2. Experimental setup and numerical method
Mach number is over 6, the pre-combustion shock
train becomes weak until the flow remains super- The present study is based on a model RBCC
sonic in the combustor and scramjet-mode opera- combustor shown in Fig. 1. The flow path consists
tion is consequently achieved [4]. of a rectangular combustor, an isolator, and a strut
As the vehicle accelerates with increasing flight in which a rocket is placed. The strut is designed
Mach number, the well-known process of the with a convergent–divergent profile and it serves as
ramjet-to-scramjet transition would occur, which a Laval nozzle to accelerate the pressurized gas to a
is characterized by the weakening of the pre- supersonic flow. The airstream is expanded through
combustion shock train [5]. If the heat addition the Laval nozzle and flows into the combustor at
to the supersonic airstream exceeds a critical value Mach number of 2.2. The rocket acts as a gas gen-
during the flight, the flow would be thermally erator and produces fuel-rich hot exhaust that is
choked. As a consequence, the reverse scramjet- injected into the combustor at Mach number of
to-ramjet transition would be triggered by the 2.1. Thus a supersonic jet flame is formed in the
pressure-rise due to significant heat release [6]. Un- combustor as the rocket exhaust gases mix and re-
derstanding the complex transition process is im- act with the surrounding air. Various experimen-
portant for the design of practical engines, and has tal methods such as OH chemiluminescence imag-
been the focus of several experimental [7] and nu- ing, shadowgraph visualization, temperature mea-
merical [8] investigations. surement by TDLAS (Tunable Diode Laser Ab-
In the ramjet-mode subsonic combustion takes sorption Spectroscopy) at 6 different locations, and
place as the flow remains subsonic in the cham- static pressure measurement by transducers along
ber. In the scramjet-mode, however, flame stabiliza- the channel length have been used to characterize
tion is typically achieved by using struts and cav- the reacting flow in the model RBCC combustor
ities [9]. Thus heat release can occur in both the shown in Fig. 2. Table 1 lists the inflow conditions
subsonic and supersonic flows. Moreover, vigor- of the airstream and rocket jet.
ous chemical reactions in the combustor would re- In LES all variables are decomposed into re-
lease a large amount of heat leading to situations solved and unresolved (subgrid) components by fil-
in which the Mach number locally falls below 1 [8]. tering reactive Navier–Stokes equations [11]. The
Recently a study on the hydrogen-fueled M12-02 one-eddy-equation model is used to calculate the
engine developed by JAXA has demonstrated that subgrid stress tensor [12–14]. An optimized 13-
local subsonic combustion still exists in the com- species 41-steps chemical mechanism is employed
bustor at flight Mach number of 12 [10]. Previous for CO/H2 /Air reaction, which has been tested
studies have shown that the true heat release picture against by a wide range of experimental data
of the scramjet-mode engine consists of not only [15]. The PaSR model is adopted to close the
the supersonic combustion mode but also the sub- LES governing equations for reactive flows [16–
sonic combustion mode. However, the local transi- 17]. The LES model equations are solved using a
tion between supersonic and subsonic combustion, density-based compressible code based on Open-
which is an underlying physics for the overall mode FOAM (rhoReactingFoam) with the second-order
transition of the engine, has received limited atten- Total Variation Diminishing (TVD) scheme and
tion in previous studies. The contributions to the the Crank–Nicholson scheme. The CFL number
heat release by different combustion modes haven’t is less than 0.3, corresponding to a physical time
been fully understood. Therefore, it is important to step at the order of 1 × 10−8 s. All variables are pre-
find out the distributions of different combustion scribed at inflow boundaries. Neumann boundary
modes and their contributions to heat release in the conditions are used at the outflow boundary where
combustor not only qualitatively but also quantita- all variables are extrapolated from the interior ones.
tively, by which the dynamic process of mode tran- The nonslip conditions are used for the strut walls
sition in the engine can be parameterized. while slip conditions are applied for the chamber
In the present study, a supersonic jet flame in a walls. Three grids with 6 million (Grid_1), 12 mil-
model RBCC combustor is investigated experimen- lion (Grid_2), and 20 million (Grid_3) cells are used
tally and numerically using static pressure distribu- for the simulation.
tions, shadow-graph visualization, OH chemilumi-
nescence images, temperature measurements, and
Large Eddy Simulations. Comparison is made be- 3. Results and discussion
tween measurements and calculations to validate
the accuracy of the LES results. LES is further 3.1. Validation with experimental data
used to identify the flame structure and combustion
characteristics over a wide range of air inlet condi- The calculated pressure profiles with different
tions. The focus of the study is on qualitative and grids are presented in Fig. 3(a) to assess the over-
quantitative assessment of combustion modes and all simulation accuracy. The simulation results from
heat release by employing filter functions. all the grids show reasonable agreements with the

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213
JID: PROCI
ARTICLE IN PRESS [m;July 10, 2018;23:58]

D. Cao et al. / Proceedings of the Combustion Institute 000 (2018) 1–9 3

Fig. 1. Computational configuration of the model RBCC combustor (mm).

Table 1
Inflow conditions of airstream and rocket exhaust.
Inlet T (K) p (MPa) m (g/s) YCO YCO2 YH2 YH2 O YO2 YN2
Air 300 0.99 780 0 0 0 0 0.233 0.767
Rocket 2070 0.71 14.9 0.355 0.053 0.018 0.091 0 0.483

airstream. Both experimental measurements and


LES suggest that a strong exothermic region is dis-
tributed in the middle of the duct. Experimental
and numerical shadowgraphs of the flow field are
presented in Fig. 3(c). As indicated before, at the en-
trance of the combustor (x = 240.4 mm) the Mach
numbers of airstream and rocket exhaust are 2.2
and 2.1, respectively. A shear layer is formed at
the interface between the two streams. Expansion
fans as well as following shocks are observed at
Fig. 2. The model RBCC combustor and TDLAS mea- the end of the strut due to the sudden change of
surement locations. flow area. The flow field becomes complex with
multiple shocks that interact with the flame. In-
stantaneous as well as time-averaged temperature
experimental data. With the increasing number contours are shown in Fig. 3(d–e). The structure
of the grid points, the simulation accuracy is im- of the supersonic jet flame is quite complex, con-
proved. Thus following discussions are based on sisting of a leading flame base, an intensive mix-
the calculation results with 20 million grid cells ing region with combustion, and a turbulent far
(Grid_3). A significant pressure drop is observed field with weaker combustion. In the intensive mix-
when the pressurized air passes through the Laval ing region, the flame surface is corrugated as the
nozzle. Further pressure decrease is found in the re- shear layer becomes distorted due to the Kelvin–
gion where expansion fans appear. Pressure rises Helmholtz [18] and Richtmyer–Meshkov [19] insta-
in the combustor due to heat release from chemi- bilities. Significant amount of heat is released in
cal reactions. Because of the superimposed shock this region as indicated by OH distributions in Fig.
train, pressure presents a wave-like pattern along 3(b). In the far field turbulent combustion becomes
the channel. weak. The jet flame spreads slowly in vertical di-
Figure 3(b) presents the experimental and nu- rections due to the compressibility of supersonic
merical distributions of OH, which is an important flows. Figure 3(f) shows time-averaged temperature
flame marker since high concentration of OH indi- contours at x = 243.4, 270.4, 282.4, 300.4, 330.4,
cates vigorous combustion and significant heat re- 350.4 mm with the laser beams involved in the TD-
lease. When the rocket exhaust is injected into the LAS measurements. The variation of line-averaged
combustor, it mixes and reacts with the incoming temperature during the experiment is presented in

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213
JID: PROCI
ARTICLE IN PRESS [m;July 10, 2018;23:58]

4 D. Cao et al. / Proceedings of the Combustion Institute 000 (2018) 1–9

Fig. 3. (a) Pressure contours and experimental as well as numerical wall pressure profiles; (b) Experimental and numer-
ical OH distributions; (c) Experimental shadowgraph and numerical grad(ρ) contours; (d) Instantaneous temperature
contours; (e) Time-averaged temperature contours; (f) Time-averaged temperature contours with laser beams used for
TDLAS; (g) Measured temperature along with time; (h) Measured and calculated temperature.

Fig. 3(g). The measured and calculated values of 3.2. Combustion mode and heat release
temperature at 6 different locations are presented
in Fig. 3(h). It shows that the averaged LES re- The instantaneous contours of Mach number
sults follow the trend of experimental data. At with sonic lines (white) and heat release rate (HRR)
x = 243.4 mm, the calculated value is a little higher using the logarithm function are shown in Fig. 4(a
than the experimental data with a relative error of and b). It is obvious that local subsonic pockets do
9.7% because there are very large temperature gra- exit in the jet flame and heat is released by super-
dients between rocket exhaust and airstream, which sonic combustion mode as well as subsonic com-
leads to lower measured temperature. With increas- bustion mode.
ing downstream distance, the flow field becomes In order to evaluate the heat release rates of su-
more uniform and good agreement between the cal- personic and subsonic combustion modes, HRRsup
culations and the measurements is found. Over- and HRRsub are defined as:
all, these comparisons indicate that qualitative and ⎧
⎪ 1 + sign(Ma − 1 )
quantitative agreements have been found between ⎨H RRsup = H RR
experimental data and LES results, which could be 2 (1)
further used to study the supersonic jet flame. ⎩H RR = 1 − sign(Ma − 1 ) H RR

sub
2

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213
JID: PROCI
ARTICLE IN PRESS [m;July 10, 2018;23:58]

D. Cao et al. / Proceedings of the Combustion Institute 000 (2018) 1–9 5

Fig. 4. Instantaneous contours of (a) Mach number with sonic lines (white); (b) HRR using the logarithm function;
(c) the logarithmic HRRsup ; (d) the logarithmic HRRsub ; (e)∇p × ∇ρ; and (f) magnitude of vorticity.

The contours of logarithmic HRRsup and combustion to some extent as they pass through the
HRRsub are presented in Fig. 4(c and d). Appar- flame.
ently, the supersonic combustion mode is domi- Figure 5(a) presents the time-averaged scat-
nant but subsonic combustion also occurs locally ter plots of HRR versus Mach number at
in the combustor. The subsonic combustion zones different cross sections, which is colored by
take the forms of thin ribbons and pockets. Specif- temperature. Near the entrance of the combustor
ically, a subsonic flame is observed in the recir- (x = 250 mm), HRR is small due to limited mixing.
culation bubble at the strut base and it spreads Both supersonic combustion and subsonic com-
downstream within the shear layer. As intensive bustion take place with weak intensity. HRR rises
mixing with combustion gradually occurs between significantly at x = 300 mm and it peaks at Mach
the airstream and fuel, a large amount of heat is number of 1.5 due to intensive mixing and com-
released in the duct, which may slow down the bustion. It indicates that supersonic combustion
supersonic flow speed. Figure 4(e) plots contours releases more heat than subsonic combustion does
of ∇p × ∇ρ that is associated with the baro- at this cross section. Although supersonic com-
clinic torque effect and Fig. 4(f) presents vortic- bustion still dominates at x = 350 mm, subsonic
ity magnitude. It shows that shocks are closely combustion also plays an important role to HRR
related to baroclinic effects. The misalignment be- peaking around Mach number of 1. Downstream
tween the pressure gradient of the shock and the of the intensive mixing region, combustion be-
local density gradient across the shear layer re- comes weaker and HRR decreases correspondingly.
sults in the baroclinic vorticity source that ampli- At the cross section of x = 450 mm, HRR occurs
fies perturbations and enhances mixing. Moreover, only in the supersonic flow with very small values.
shock waves would lead to deceleration of the flow The latter shows that supersonic mixing between
speed and increase of temperature due to compres- combustion products and surrounding airstream
sion effects. Thus shocks would induce subsonic dominates the downstream section of the channel.

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213
JID: PROCI
ARTICLE IN PRESS [m;July 10, 2018;23:58]

6 D. Cao et al. / Proceedings of the Combustion Institute 000 (2018) 1–9

Fig. 5. Time-averaged (a) scatter plots of HRR (colored by temperature) versus Mach number; (b) profiles of Qsup and
Qsub ; (g) profiles of ηsup and ηsub .

In order to quantitatively assess the combustion promoted due to compression and deceleration
mode and heat release, HRR flux (Q) and contribu- effects associated with shocks. In the third re-
tion coefficient (η) of different combustion modes gion (406.7 mm < x < 550 mm) HRR flux is very
at a certain cross section are defined: small because of the consumption of reactants
  and low temperature of airstream. Figure 5(c)
Qsup =  H RRsup dA
(2) demonstrates that the small amount of heat is
Qsub = H RRsub dA
released by weak supersonic combustion. Thus
⎧ 
⎨ηsup = QQsup = HRRsup dA we can draw the conclusions that both supersonic
HRRdA combustion and subsonic combustion occur in
 (3)
⎩η = Qsub = HRRsub dA the combustor due to recirculation zones, shock
sub Q HRRdA
waves, and local thermal choking. Local mode
The time-averaged profiles of Q and η for transition between supersonic and subsonic com-
different combustion modes are plotted in bustion takes place as the jet flame propagates
Fig. 5(b and c) and three developing regions downstream. For the whole combustor, the overall
are observed accordingly. In the initial region contribution coefficient to heat release by the
(243.4 mm < x < 247.8 mm), chemical reactions supersonic combustion mode can be calculated as
are quite weak and more than 50% of heat is sup = ∫HHRsup dV/∫HRRdV = 89%, and similarly
released from subsonic combustion mode due sub = ∫HHRsub dV/∫HRRdV = 11%. Thus the
to the recirculation zone. In the second region supersonic combustion mode is more prominent
(247.8 mm < x < 406.7 mm) where intensive mix- in terms of contributions to heat release in the
ing and combustion take place, a large amount combustor.
of heat is released. Although both Qsup and Qsub
increase first and then decrease gradually, Qsup is 3.3. Influences of airstream temperature and
much larger than Qsub . From Fig. 5(c), it can be pressure
seen that more than 50% of heat is released by
supersonic combustion mode in the second region. In this section, four cases are investigated to
It is notable that subsonic combustion is enhanced identify the influences of airstream temperature
in the regions of 332.4 mm < x < 352.7 mm and pressure on combustion mode and heat release
and 388.4 mm < x < 411.9 mm. Comparing in the model RBCC combustor. In all cases, the in-
Fig. 5(c) with Fig. 3(c), one can see that these flow conditions of the rocket are the same, which
regions are located behind the shock/flame in- are listed in Table 1. As for the air inlet, the mass
teraction zones and subsonic combustion is flow rate and components of the airstream keep

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213
JID: PROCI
ARTICLE IN PRESS [m;July 10, 2018;23:58]

D. Cao et al. / Proceedings of the Combustion Institute 000 (2018) 1–9 7

Fig. 6. Instantaneous contours of temperature, pressure, Mach number, and logarithmic HRR in different cases.

Table 2 thermal choking. Although the jet flame is formed


Combustion mode and heat release in different cases. by two supersonic streams, both supersonic com-
Case 1 (Test) 2 3 4 bustion mode and subsonic combustion mode re-
lease heat simultaneously in all cases.
T air inlet (K) 300 600 900 1200 The time-averaged heat release rate flux (Q) and
p air inlet (MPa) 0.99 1.41 1.74 2.02
m (g/s) 780 780 780 780
contribution coefficient (η) for supersonic as well as
∫HHRdV (kJ/s) 2.5 73.6 87.1 87.5 subsonic combustion modes are plotted in Fig. 7.
∫HHRsup dV/∫HRRdV 89% 28% 55% 68% Table 2 lists the integral of HRR in the combustor
∫HHRsub dV/∫HRRdV 11% 72% 45% 32% and overall contribution coefficients of supersonic
and subsonic combustion modes in different cases.
However, the dominant mode depends on combus-
tion intensity and inflow conditions. In case 1 al-
constant values as presented in Table 1, whereas though ηsup is 89%, it should be stressed that only
temperature rises from 300 K to 1200 K with steps a very small amount of heat is released and com-
of 300 K as shown in Table 2. The airstream pres- bustion is very weak. Combustion becomes inten-
sure also increases to obtain the same mass flow sive in case 2, which leads to strong heat release and
rate. Besides, at the entrance of the combustor the significant pressure-rise. A distinct subsonic zone
airstream is 2.2 Ma and the rocket jet is 2.1 Ma for is found in the intensive mixing and combustion
all cases. region as the incoming airstream pressure is rela-
Figure 6 shows the instantaneous contours of tively low. Through the integral operation, it sug-
temperature, pressure, Mach number, and logarith- gests 72% of heat is released by subsonic combus-
mic HRR. In all cases one can observe a lead- tion, which becomes the dominant mode in case 2.
ing flame base, an intensive mixing region with The jet flame becomes longer and thinner when the
combustion, and a turbulent far field with weaker airstream’s temperature and pressure are further in-
combustion. With increasing temperature of the creased in case 3 and case 4. The supersonic com-
airstream, more and more heat is released in the bustion mode dominates in these two cases with
combustor since the chemical reaction rate is en- contribution coefficients of 55% and 68%, respec-
hanced. Meanwhile, heat release results in pressure- tively. It is also clear from Fig. 7 that subsonic com-
rise in the combustion chamber and leads to local bustion mode is always prominent at the base of the

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213
JID: PROCI
ARTICLE IN PRESS [m;July 10, 2018;23:58]

8 D. Cao et al. / Proceedings of the Combustion Institute 000 (2018) 1–9

Fig. 7. Time-averaged plots of Q and η of supersonic and subsonic combustion modes in different cases.

jet flame because of recirculating flows in all cases. supersonic combustion dominates but releases only
In the turbulent far field with weaker combustion, a small amount of heat. When intensive combus-
supersonic combustion mode becomes dominant tion occurs with high-temperature (Tairinlet = 600,
with small values of HRR. Another notable phe- 900, 1200 K), the dominant mode varies from sub-
nomenon is that subsonic combustion pockets are sonic combustion to supersonic combustion be-
located behind the shock/flame interaction zones, cause of interaction between the shock train and
suggesting that these shocks induce subsonic com- the flame.
bustion regions. Therefore it is not the fact that
supersonic combustion is always dominant in the
combustor when the main flow is at supersonic Acknowledgments
speed because local mode transition between sub-
sonic and supersonic combustion would take place This work is supported by the National Natu-
as the jet-flame propagates downstream. ral Science Foundation of China through grants
51376152 and in part by a Technion fellowship.

4. Conclusion
Supplementary materials
In the current study, a jet flame formed by the
2.2 Ma airstream and 2.1 Ma rocket exhaust, has Supplementary material associated with this ar-
been experimentally and numerically investigated ticle can be found, in the online version, at doi:
in a model RBCC combustor. A filter method is 10.1016/j.proci.2018.06.213.
introduced to quantitatively analyze subsonic and
supersonic combustion mode and their contribu- References
tions to heat release. Results show that although
the two streams are supersonic, both supersonic [1] R.J. Yentsch, D.V. Gaitonde, J. Propul. Power 30 (2)
combustion and subsonic combustion occur and (2014) 474–489.
release heat in the combustor. Local mode tran- [2] J.E. Quinn, AIAA Paper 2003-5235.
sition takes place as the flame propagates down- [3] L. Shi, G. He, P. Liu, et al., Acta Astronaut. 128
(2016) 350–362.
stream. The subsonic combustion mode is always
[4] D.J. Micka, J.F. Driscoll, Pro. Combust. Inst. 32
prominent at the base of the jet flame because of (2009) 2397–2404.
recirculation bubble, and supersonic combustion [5] G.A. Sullins, J. Propul. Power 9 (4) (1993) 515–520.
mode becomes dominant in the turbulent down- [6] E.T. Curran, W.H. Heiser, D.T. Pratt, Annu. Rev.
stream field with small heat release rate. In the pres- Fluid Mech. 28 (1996) 323–360.
ence of shock waves, mixing and combustion are [7] C. Aguilera, K.H. Yu, Pro. Combust. Inst. 36 (2017)
enhanced by inducing vorticity because of baro- 2911–2918.
clinic effects. Subsonic combustion pockets are ob- [8] J. Larsson, S. Laurence, I. Bermejo-Moreno, et al.,
served closely behind the shocks due to the as- Combust. Flame 162 (2015) 907–920.
[9] H. Wang, Z. Wang, M. Sun, et al., Pro. Combust.
sociated deceleration effects. The dominant mode
Inst. 34 (2013) 2073–2082.
in the model RBCC combustor is dependent on [10] S. Zhang, J. Li, F. Qin, et al., Acta Astronaut. 129
temperature and pressure of the airstream when (2016) 357–366.
the rocket’s conditions and air mass flow rate are [11] Z. Wang, M. Sun, Supersonic Turbulent Flow, Com-
fixed. When the chemical reactions are weak due bustion Modeling and Large Eddy Simulation, Science
to low-temperature (Tairinlet = 300 K) airstream, Press, Beijing, China, 2013.

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213
JID: PROCI
ARTICLE IN PRESS [m;July 10, 2018;23:58]

D. Cao et al. / Proceedings of the Combustion Institute 000 (2018) 1–9 9

[12] A. Yoshizawa, K. Horiuti, J. Phys. Soc. Jpn. 54 (8) [17] F. Qin, Z. Huang, G. He, et al., Int. J. Hydrogen En-
(1985) 2834–2839. ergy 42 (2017) 21360–21370.
[13] S. Menon, P.K. Yeung, W.W. Kim, Comput. Fluids 25 [18] S.R. Choudhury, J. Plasma Phys. 35 (1986) 375–392.
(2) (1996) 165–180. [19] M. Brouillette, Annu. Rev. Fluid Mech. 34 (2002)
[14] A. Nakayama, S. Vengadesan, Int. J. Numer. Meth- 445–468.
ods Fluids 38 (3) (2002) 227–253.
[15] S.G. Davis, A.V. Joshi, H. Wang, et al., Proc. Com-
bust. Inst. 30 (2005) 1283–1292.
[16] A.H. Kadar, Modelling Turbulent Non-Premixed
Combustion in Industrial Furnaces Using the Open
Source Toolbox OpenFOAM Thesis, Delft University
of Technology, 2015.

Please cite this article as: D. Cao et al., Local supersonic and subsonic combustion mode
transition in a supersonic jet flame, Proceedings of the Combustion Institute (2018),
https://doi.org/10.1016/j.proci.2018.06.213

You might also like