You are on page 1of 6

Acta Astronautica 57 (2005) 384 – 389

Fundamental study of supersonic combustion in pure air flow with

use of shock tunnel
Shigeru Aso∗ , Arif Nur Hakim, Shingo Miyamoto, Kei Inoue, Yasuhiro Tani
Department of Aeronautics and Astronautics, Kyushu University, Japan

An experimental study using reflected-type of shock tunnel has been conducted to investigate the phenomena of supersonic
combustion. In the experiment, test air is compressed by reflected shock wave up to stagnation temperature of 2800 K and
stagnation pressure of 0.35 MPa. Heated air is used as a reservoir gas of supersonic nozzle. Hydrogen is injected transversely
through circular hole into freestream of Mach 2. Flow duration is 300 s. Schlieren method and CCD UV camera are used
to obtain information on the shock structures and the region of combustion. The effects of total pressure of injection gas to
the fuel penetration and the region of combustion have been obtained.
© 2005 Elsevier Ltd. All rights reserved.

1. Introduction fast mixing and fast combustion with minimum total

pressure loses is required.
Supersonic combustion ramjet (SCRAM-jet) en- There are many experimental studies of this field in
gine is currently one of the promising propulsion ground-based facilities with the use of vitiation heated
systems for achieving hypersonic airbreathing propul- tunnel to generate a supersonic hot flow [1]. This
sion. In the engine, the flow in the combustor leads facility is relatively simple and easy to operate and
to a short residence time of air and fuel, thus injec- has a long test time of the order of seconds or longer.
tion method, mixing and combustion process become However, the high-temperature test air delivered to
important problems for the success of airbreathing the combustion chamber contains a large concentra-
hypersonic vehicles. To achieve efficient combustion, tion of H2 O as well as OH radicals. And the existence
of those impurities increases the uncertainty of the
supersonic combustion experiments. To solve such a
∗ Corresponding author.
problem, impulse facility can be considered as an al-
E-mail addresses: ternative method. Although the test time is very short,
(S. Aso), (A.N. Hakim),
impulse facilities are capable of providing the required (S. Miyamoto), (K. Inoue), high total temperature and total pressure without hav-
(Yasuhiro Tani). ing problem of test air with H2 O and OH radicals

0094-5765/$ - see front matter © 2005 Elsevier Ltd. All rights reserved.
S. Aso et al. / Acta Astronautica 57 (2005) 384 – 389 385

which causes a condition different from the real flight two-dimensional supersonic nozzle of Mach number
combustion condition. of 2.
There are some types of impulse facility used in the The schematic of experimental setup is shown in
supersonic combustion experiments, i.e. shock tube Fig. 1. The test section, whose cross section is rect-
[2], expansion tube [3,4] and reflected shock tunnel angular, is 44 mm wide and 33 mm high. Hydrogen
[5]. However, the reflected shock tunnel offers longer or Helium was injected transversely through circu-
test time than the shock tube and the expansion tube. lar sonic nozzle of 2 mm diameter into freestream by
In the reflected tunnel, high pressure and high temper- using a solenoid valve operated by time control sys-
ature reservoir gas is made by compressing test air by tem. The primary parameter used in the experiment
incident shock wave and reflected shock wave. is the total pressure ratio, p0j /p0 (p0j : total pressure
The requirement of efficient injection method and of injected gas, p0 : total pressure of free stream). The
flame holding has motivated the study of mixing [6,7], pressure ratio is set to 0.5, 1.0 and 1.5, as shown in
as well as combustion [6], of transverse injection into Table 1.
supersonic crossflow. The method is expected to gener-
ate recirculation upstream of injection point as flame- 2.2. Measurements of test conditions
holding. When fuel is injected into supersonic cross-
flow through nozzle, the fuel flow will expand and To determine the experimental conditions, measure-
interact with it. As a result, a three-dimensional bow ments of total pressure and total temperature of reser-
shock is formed upstream of the injection point and voir gas were conducted. The total pressure of test air
interacts with the boundary layer, resulting in a separa- was obtained by measuring stagnation pressure using
tion region. The second separated zone which appears Kistler pressure transducer located in driven section,
downstream of injection point, between the injection 10 mm before nozzle throat. On the other hand, mea-
point and the boundary layer reattachment point is also surement of the stagnation heat flux of blunt body has
expected to operate as flame holding. been conducted in no-injection case to obtain the total
In addition to recirculation zone, the total pressure temperature of test air. The measurement was made
of injection has an important role in controlling the using a semi-sphere cylinder platinum with thin film
combustion, due to different environments of chamber heat transfer gauge located at the center of test section
during the flight. Change of total pressure of injection on the same position with injection point. Assuming
will affect the penetration of fuel [7], thus the condi- an isentropic expansion of test air from the driven sec-
tion of combustion. tion to the test section and using combined relation-
In the present study, transverse injection has been ship of stagnation pressure of reservoir and the heat
conducted through a circular nozzle into supersonic flux in the test section, the total temperature of test air
crossflow of Mach 2 generated by reflected shock tun- in the test section can be obtained.
nel. The objective is to obtain a better understanding Fig. 2 shows time history of wall temperature of
of the behaviors of supersonic combustion when total semi-sphere cylinder model. The temperature be-
pressure of injection is changed. gins to increase when starting shock wave arrived to
body surface. Figs. 3 and 4 show the time history of
stagnation heat flux calculated from measured wall
2. Facility and experiment description temperature data and the calculated total tempera-
ture, respectively. The total temperature of test air
2.1. Experimental facility begin to increase rapidly just as starting shock wave
arrives to the surface of gauge and get relatively con-
The experiments have been performed in the re- stant at about 2800 K. In the supersonic combustion,
flected shock tunnel at Kyushu University, Japan. In if the static temperature sufficiently high for fuel
the shock tunnel, test air is compressed by reflected self-ignition, no additional heating devices is needed
shock wave up to stagnation pressure 0.35 MPa and to be provided. In the present investigation, the air
total temperature of 2800 K. Heated air is used as a static temperature at the location of injection is about
reservoir gas and provided into test section through 1550 K, which is quite enough for the self-ignition.
386 S. Aso et al. / Acta Astronautica 57 (2005) 384 – 389



Fig. 1. Schematic of experimental setup.

Table 1 1.5E+07
Experimental conditions

Gas Freestream Injectant

Air He, H2 1.0E+07

qw [W/m2]
Total pressure 0.350 MPa 0.175 MPa
0.350 MPa
0.525 MPa
Total temperature 2800 K 290 K 5.0E+06
Mach number 2 1

800 600 700 800 900 1000 1100
time [microsec]
Fig. 3. Time history of heat flux.
Twall [K]

500 To visualize the shock structure of the supersonic

flowfield, Schlieren method has been conducted. The
400 instantaneous light source of less than 1 s of ex-
posure period is used. Helium and Hydrogen were
300 used as injectant gas in this experiment to investi-
gate shock structure in both non-reacting and reacting
200 cases. CCD-UV camera was also used to observe the
600 700 800 900 1000 1100 OH self-luminescence generated in hydrogen combus-
time [microsec] tion to investigate the combustion processes in the su-
personic flowfield.
Fig. 2. Time history of wall temperature of platinum thin film heat
transfer gauge.

3. Results and discussions

Test time in this experiment, as shown in Fig. 4,
is defined as the test time of about 300 s, in which To investigate the condition of the combustion phe-
the total temperature remains constant. Table 1 sum- nomena with relation to the shock structure in the
marizes the experimental conditions for the present flowfield, both Schlieren method and CCD UV cam-
investigation. era were employed.
S. Aso et al. / Acta Astronautica 57 (2005) 384 – 389 387

test time (~300 microsec)

T0 [K]



600 700 800 900 1000 1100
time [microsec]

Fig. 4. Time history of total temperature of freestream.

Fig. 5 shows Schlieren images of flowfield. In all

cases, small recirculation region can be seen just up-
stream of the injection point. However, due to the
small diameter of hole and the sensitivity of camera,
the recirculation region only appeared very small. Fig.
3a–c shows the images of non-reacting flow cases.
As reported previously [2], the total pressure of in-
jection will affect the height in crossflow. The images
indicated that when the injection total pressure is in- Fig. 5. Schlieren images of flowfield: (a) Helium injection,
creased, the bow shock formed by the injection jet be- p0j /p0 = 0.5; (b) Helium injection, p0j /p0 = 1.0; (c) Helium
injection, p0j /p0 = 1.5; (d) Hydrogen injection, p0j /p0 = 1.0.
comes stronger and the angle of shock wave increases.
Also the increase of total pressure of the injection jet
results in the increase of total pressure lose of the flow.
Additional oblique shock wave is observed in the joint
part of the wall located at 72 mm downstream of the era and the Schlieren image of the helium injection
injection point. when p0j /p0 = 1.0. The reaction started at about
Similar results are also observed in reacting flow 45 mm downstream of the injection point, and grew
(hydrogen injection). However, comparing Fig. 5b and larger as far as downstream. This distance between
d, it can be found that shock wave generated in react- injection point and location of the reaction start indi-
ing flow appeared weaker than those in non-reacting cates that the induction time of combustion is in the
flow. This difference of shock wave strength might order of 10−5 s, thus reasonable for this experimental
come frome the difference of molecular weight of He condition. However, the reaction cannot be found at
and H2 . the recirculation just upstream and downstream of
In the hydrogen combustion, the condition of the the injection point. The reason may result in the poor
combustion can investigated by observing the pres- sensitivity of the camera.
ence of OH-radical. By recording the self-luminesence The strong reaction is established after oblique
of OH radicals using CCD UV camera, location shock generated by joint of wall, about 80 mm down-
and intensity of the reaction zone were observed. stream of the injection point and suggests that to
Fig. 6 shows the intensified averaged image of self- accelerate the reaction it can be very useful using
luminescence of OH-radical taken by CCD UV cam- shock generator.
388 S. Aso et al. / Acta Astronautica 57 (2005) 384 – 389

of fuel can be achieved. Therefore, reaction zone ex-

pands to the center of flowfield.

4. Conclusions

The study of supersonic combustion of hydrogen

has been conducted by using a reflected-type shock
tunnel which generated a stable supersonic air flow
of Mach number of 2 with the total temperature of
2800 K and the total pressure of 0.35 MPa. The static
temperature in the test section is about 1550 K which
enables the supersonic combustion with self-ignition.
The Schlieren images show that the increase of in-
jection pressure generated strong bow shock, resulting
in the pressure loses.
The combustion begins almost at the same location
for all injection total pressure cases, about 45 mm
Fig. 6. UV images of the combustion (hydrogen injection, downstream of the injection point, which indicates
p0j /p0 = 1.0) and the position of shock wave generagted by joint the reasonable induction time. After the oblique shock
wall (helium injection p0j /p0 = 1.0).
generated by joint part of the wall located at 72 mm
downstream from the injection point, the strong re-
action is established. This demonstrates the shock
generator be an effective method to accelerate the
The increase of the injection total pressure raises the
penetration of fuel; thus, the reaction zone expands to
the center of flowfield.


We really appreciate Mr. Minoru Orino and Mr.

Yuzo Inokuchi of Department of Aeronautics and As-
tronautics, Kyushu University, for every support and
help in the experiments.


[1] R.C. Roger, D.P. Capriotti, R.W. Guy, Experimental supersonic

Fig. 7. UV images in case of hydrogen injection: (a) p0j /p0 =0.5; combustion research at NASA Langley, AIAA-Paper, 98-2506,
(b) p0j /p0 = 1.0; (c) p0j /p0 = 1.5. 1998.
[2] G. Smeets, C. Quenett, Shock tube investigation of H2
combustion in a high temperature supersonic air flow
(Scramjet), Proceeding of IUTAM Symposium on Combustion
Fig. 7 shows the reaction zone in different case of in Supersonic Flows, Netherlands, 1997, pp. 173–178.
injection total pressure. The images show when the [3] B.K. McMillin, J.L. Palmer, R.K. Hanson, Temporally
injection total pressure increases, the high penetration resolved, two-line fluorescence imaging of NO temperature
S. Aso et al. / Acta Astronautica 57 (2005) 384 – 389 389

in a transverse jet in a supersonic cross flow, Applied Optics [6] S. Aso, M. Tannou, S. Maekawa, S. Okuyama, A study of
32 (36) (1993) 7532–7545. mixing phenomena in three-dimensional supersonic flow with
[4] A. Ben-Yakar, M. Kamel, C. Morris, R.K. Hanson, circular nozzle, AIAA-Paper, 94-0707, 1994.
Experimental investigation of H2 transverse jet combustion in [7] M.R. Grubber, A.S. Nejad, T.H. Chen, J.C. Dutton, Mixing
hypervelocity flows, AIAA-Paper, 97-3019, 1997. and penetration studies of sonic jets in a mach 2 freestream,
[5] R.R. Boyce, A. Paull, R.J. Stalker, M. Wendt, N. Chinsei, Journal of Propulsion and Power 11 (2) (1995) 315–323.
H. Miyajima, Comparison of supersonic combustion between
impulse and vitiation-heated facilities, Journal of Propulsion
and Power 16 (4) (2000) 709–717.