You are on page 1of 5

AIAA JOURNAL

Vol. 52, No. 2, February 2014

Technical Notes
Experimental Investigation of Hysteresis coupled hysteresis associated with the interaction among the multiple
inlet flow patterns is reported and analyzed and indicates that the
Phenomenon for Scramjet Engine control of the bifurcation system thus depends not only on the instan-
taneous value of the input but also on the history of its operation.
Thus far, the hysteresis phenomenon in supersonic/hypersonic
Juntao Chang,∗ Lei Wang,† Wen Bao,‡ Qinchun Yang,§ and inlets is seldom studied, although it has frequently been observed
Jiang Qin¶ in aerodynamic experiments. Moreover, the reported studies of
Harbin Institute of Technology, 150001 Harbin, hysteresis phenomena in hypersonic inlets focus on the Mach number
People’s Republic of China and the angle of attack [13]. Although some efforts [14,15]
investigated the unstart mechanism of high backpressure, the
Downloaded by NORTH CAROLINA STATE UNIVERSITY on May 10, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.J052505

DOI: 10.2514/1.J052505 hysteresis phenomenon is not involved and the interaction between
isolator flow and combustion is neglected. Thus, the current works
aim to investigate experimentally the hysteresis phenomenon caused
I. Introduction by excessive heat release from combustion in a direct-connected
scramjet experiment system, of which the heater would introduce
S CRAMJET and ramjet propulsion are developing technologies
for a future generation of hypersonic atmospheric and space
vehicles. Unfortunately, scramjet engines are susceptible to a highly
combustion vitiation into the scramjet engine. Haw et al. [16] and
Tatman et al. [17] studied the vitiation effects on scramjet mode
transition [16] and flameholding in a cavity flameholder [17]; the
detrimental operation mode called unstart, which would cause an results show that vitiation has an effect on mode transition, shock
abrupt drop of both engine thrust and specific impulse and unsteady train length, and pressure rise, but it hardly influences flameholding.
aerodynamic and thermal loads. The increasing interest in faster and The focus of this Note is to present the hysteresis phenomenon and
more efficient means of atmospheric transportation has led to many analyze briefly its effect on restarting control, and so the vitiation
research efforts that aim to understand the unstart mechanism. effect is not considered.
Studies [1,2] of inlet unstart have shown that boundary-layer
separation plays an important role in the onset of unstart. Chang and
Fan investigated numerically the effects of boundary-layer bleeding II. Experimental Setup
[3] and wall cooling [4] on performance of hypersonic inlets.
The experiments were completed using the direct-connected
Although the unstart phenomenon can be restrained by an elaborate scramjet experiment system at the Harbin Institute of Technology,
design, it is impossible to completely avoid its occurrence in practice People’s Republic of China. The direct-connected experiment system
because many factors (including freestream conditions and operating runs in a blowdown mode. The high-pressure air was supplied by a
situation) can cause it. Thus, unstart detection and unstart control are big enough chamber, which can ensure that the effective run time
necessary to ensure the proper operation of scramjet engines. Many reaches about 1 min. An air heating system, in which compressed air
techniques [5,6] based on the change in oscillation intensity of was heated by the combustion of alcohol, was located upstream of the
pressure were developed to detect and predict unstart, and some supersonic nozzle. In the air heating system, additional oxygen was
active controls such as the use of vortex generator jets [7] and plasma supplied to keep the oxygen content equal to that in the atmosphere.
control [8] were proposed. In the current experiments, the total pressure and total temperature of
The hysteresis phenomenon, meaning that once the inlet is the heated air were 2.1 MPa and 1516 K, respectively, which simulate
propelled to transit from inlet start to unstart it cannot be propelled to the stagnation parameters as flying in the freestream conditions of
restart again along the same control route, however, would make the Mach number 6.0. A two-dimensional Laval nozzle with a exit area of
unstart detection and control more complex. Taking the CIAM/ 56 × 100 mm was used to accelerate the heated air to a Mach number
NASA Mach 6.5 scramjet flight test as an example of unstart of 3.0. The detailed test conditions are listed in Table 1.
detection, the scramjet restarts but the detection system shows it A simplified and two-dimensional dual-mode scramjet engine was
unstart, which led to a failure of the flight test. Many studies [9–11] used in the current experiments, and its schematic is shown in Fig. 1.
have offered important insights that the hysteresis phenomenon is an The two-dimensional model with a constant width of 100 mm
important factor in the flow mechanisms on the error detection of inlet includes an inlet, an isolator, and a combustor. A converging duct
start/unstart. At the mechanism of the inlet mode transition aspect, with a length of 192 mm and an exit area of 40 × 100 mm2 located
Cui et al. [12] analytically describe the multistability and bifurcation immediately downstream of the supersonic nozzle was used to
phenomena of supersonic inlet start/unstart under the assumption of simulate an inlet. Its contraction ratio is about 1.4. Cui et al. [18]
inviscid flow from a viewpoint of dynamical system theory. A loops- analyzed carefully the feedback mechanism, and the result shows that
the stability of the self-feedback system is determined by the profile
Presented as Paper 2013-3972 at the 49th AIAA/ASME/SAE/ASEE Joint
of the cross-sectional area. Supersonic/hypersonic inlets converge
Propulsion Conference, San Jose, CA, 14–17 July 2013; received 1 January along the flow direction, and thus the use of a converging duct to
2013; revision received 12 May 2013; accepted for publication 8 July 2013; replace an inlet for stability investigations in the current experiments
published online 16 January 2014. Copyright © 2013 by the American is rational. Immediately after the inlet throat lies a constant cross-
Institute of Aeronautics and Astronautics, Inc. All rights reserved. Copies of sectional isolator (450 mm in length and 40 mm in height).
this paper may be made for personal or internal use, on condition that the Downstream of the isolator is a simplified dump combustor,
copier pay the $10.00 per-copy fee to the Copyright Clearance Center, Inc., including an expansion section with a length of 200 mm and an exit
222 Rosewood Drive, Danvers, MA 01923; include the code 1533-385X/14 height of 70 mm and a 470 mm long constant area section. A short
and $10.00 in correspondence with the CCC. diverging (100 mm in length and 95 mm in exit height) duct is located
*Associate Professor, Academy of Fundamental and Interdisciplinary
Sciences; changjuntao@hit.edu.cn (Corresponding Author). downstream of the combustor.

Master Student, School of Energy Science and Engineering. It can be seen from the three-dimensional view of the strut as

Professor, School of Energy Science and Engineering. shown in Fig. 2 that there are a total of 48 fuel nozzles with a diameter
§
School of Energy Science and Engineering. of 0.5 mm spreading across both sides of the strut. The strut is

Lecturer, Academy of Fundamental and Interdisciplinary Sciences. installed at the center of the isolator with the leading edge 280 mm to
447
448 AIAA JOURNAL, VOL. 52, NO. 2: TECHNICAL NOTES

Table 1 The flow conditions at leads to a decrease in pressure of all measurements due to the
the inlet entrance injection effect of supersonic flow. At the same time, fuel is given and
Parameter Inlet entrance the equivalence ratio rises to 0.375. At about t  9 s, the pressure at
T12 first increases to a high value due to the shock induced by the
Total pressure, MPa 2.1
Total temperature, K 1516 strut, which indicates that a supersonic flowfield is built. The ignition
Mach number 3.0 occurs at time period of 10.0–10.5 s and a stable combustion region is
Static pressure, kPa 55 formed, which results in an increase of combustor pressure, such as
Static temperature, K 541 the pressure at T16. Then the intermittent increase of fuel as shown in
Fig. 4 leads to an upstream propagation of the high backpressure. At
t  11.3704 s, an increase of pressure at T13 indicates that the shock
train leading edge arrives upstream of T13. The shock train leading
the isolator entrance. Kerosene fuel at room temperature is injected
from the strut in the direction normal to the airflow direction. To edge arrives upstream of T10 at t  12.3480 s. It is worth noting that
protect the strut from thermal load, the kerosene injectant serves as when the equivalence ratio steps from 1.000 to 1.250 the pressures at
the active cooling agent and is later injected into combustion, which T1–T5 all rise sharply. It is well known to us that the shock motion
ensures that the heat losses from the main flow to the strut are can be stable in a diverging channel, whereas it is unstable in a
recovered. Ignition is accomplished by an arc plasma jet (PJ) torch converging channel [18]. Hence, the pressures ranging from T1 to T5
modified from a plasma cutting machine. This PJ torch uses N2 as change consistently because a shock is unstable in a converging duct.
Downloaded by NORTH CAROLINA STATE UNIVERSITY on May 10, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.J052505

feedstock, and the maximum power is 2 kW with a voltage of 20 V This sudden increase of inlet pressure indicates that the shock train is
and a current of 100 A. It is installed at the center of the combustor ejected from the isolator, which is a flag of the onset of unstart.
upper wall next to the downstream of the strut. The plasma torch is To describe the shock train much better, Fig. 5 shows the average
turned off as long as the ignition is successful, and so it works for less pressure distributions at different equivalence ratios. The pressure
than 1 s in the ignition period of every experiment. distribution of t  10 s, when the scramjet is started without
The measurement of the pressure of the scramjet engine side wall combustion, is flat, but pressure at T12 is higher due to the induced
was made by a row of 27 fast-response pressure transducers, whose shock by the strut. With an increase of equivalence ratio, the shock
natural response frequency is 50 kHz. The transducer locations are train leading edge moves upstream, e.g., arrives at T13, T10, and T7
labeled T1–T27 in the schematic of Fig. 1. All of them have ranges of when the equivalence ratio is 0.625, 0.875, and 1.000, respectively.
0–1 MPa. The transducers were calibrated first to a temperature shift According to the pressure distribution at t  14 s, it can be seen that
below 0.05% full scale FS∕°C, a nonlinear error of 0.1% FS, a the inlet pressure is high when the inlet is unstart. The shock train
repeatability error of 0.1% FS, and a comprehensive accuracy of motion is a phenomenon of the scramjet, which does not exist in the
0.1% FS. A row of 28 test holes with a diameter of 0.5 mm were ramjet engine.
drilled in a row near the centerline of the side wall. To protect After about 13.9 s, the equivalence ratio starts decreasing as shown
transducers from combustion, transducers were connected to test in Fig. 4. The decreasing of the equivalence ratio leads to a gradual
points by hollow metal ducts, with a diameter of 2.5 mm and a length reduction of all pressures. The equivalence ratio steps from 0.875 to
of about 300 mm, which will decay the cutoff frequency of pressure 0.750 at t  14.8750 s; however, the pressures of both the inlet (such
measurements. To decrease this decay as much as possible, the as T1 and T5) and isolator (such as T7 and T10) drop sharply and then
hollow metal ducts were filled with kerosene. According to a the inlet pressure recovers slowly. This sharp drop must be caused by
conservative estimation, the cutoff frequency of the system is higher an abrupt change of flowfield, namely, the shock structure in the
than 1.1 kHz, taking the speed of sound in kerosene as 1324 m∕s. In converging inlet. Hence, the sudden drop of pressures in the range
the current experiments, the pressures at the side wall were sampled at T1–T5 is regarded as the flag of restart onset. With a further reduction
5 kHz during the whole test process. of equivalence ratio, the downstream pressure recovers gradually. In
The detailed operation sequence of the system is shown in Fig. 3. addition, the pressure at T27 keeps a relatively low value (close to the
First, the pressure data acquisition starts to run at t0  0. After a short atmosphere pressure) as shown in Fig. 5 because it locates near the
period, the heated airflow enters the scramjet through both a heater combustor exit.
and a two-dimensional nozzle. Once a steady flow is established, the Through the preceding analysis, the sudden rise and drop in
fuel injectors are turned on. Ignition then occurs in the combustor at pressure of the converging inlet are the onset flags of unstart and
t3, and the processes of ignition lasts 0.5 s and ends at t4 . Finally, a restart, respectively. We now focus on the inlet unstart and restart
sustaining combustion heat release is formed in the combustor. processes. The histories of pressures at T5 and T8 and equivalence
ratio are plotted in Fig. 6. At t  13.3876 s (point A), the equivalence
ratio begins increasing and reaches 1.250 at t  13.5028 s (point B).
III. Analysis of Experimental Results The unstart starts at t  13.5876 s (point C); however, it does not
A. Run 1 occur immediately after fuel rate rises but lags behind 85 ms. Here the
Some typical pressure signals with the history of the equivalence reasons causing lag are analyzed briefly. The shock train leading edge
ratio are plotted in Fig. 4. At about 8.5 s, the wind tunnel starts, which has already been upstream of T10 when the equivalence ratio is

Supersonic nozzle Inlet Isolator Combustor


Plasma jet
T26
T1 …… T5 T6 …… T13 T14 …… T17 T18 ……
70

95

T25 T27

Strut
100

T1 …… T5 T6 …… T13 T14 …… T17 T18 …… T27

0 192 642 842 1222 1322 X/mm


Fig. 1 The schematic of the scramjet engine.
AIAA JOURNAL, VOL. 52, NO. 2: TECHNICAL NOTES 449

700
= 1.000 = 0.625 t =10 s
= 0.875 = 0.375
600
= 0.750 t =14 s
500

400

300

200

100

0
bottom wall
Fig. 2 The three-dimensional view of the strut. 0 200 400 600 800 1000 1200 1400
X/mm
Fig. 5 The pressure distribution at different equivalence ratios.
n
isit
io low io n ction
cqu s ir f ect n off nje
da nj no el i off
t a a r t ate tarts el i on i tio nition Fu
Da sta e Fu n
Downloaded by NORTH CAROLINA STATE UNIVERSITY on May 10, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.J052505

H s Ig Ig
600
Equivalence ratio
500 T5
t0=0 t1 t2 t3 t4 t5 T10

Static pressure/kPa
T13 D
Sustaining combustion 400
Fig. 3 Schematic of the operation sequence for the test facility. B 1.250
300 E
1.0 1.0
0.875 0.875
200 0.750 A 0.750 F
0.625
0.750, so the pressure at T13 can be forced to rise immediately by an
100
increase of backpressure. When the equivalence ratio increases to C
0.875, the pressure at T13 increases at about 12.34 s, whereas the 0
12 12.5 13 13.5 14 14.5 15 15.5 16
pressure at T10 rises at about 12.35 s, and so the shock propagation Time/s
speed is about 15 m∕s, which is close to the result obtained by Fig. 6 The histories of some typical pressures and equivalence ratios.
Wagner et al. [14]. Because the onset instant cannot be obtained
accurately, the shock train propagation speed is a rough estimate.
Thus, the main factor of delay is not shock propagation, and it may be B. Run 2
caused by the combustion and measurement system. To determine the equivalence ratios of the inlet mode transition, an
The pressure distributions of both point C and D are plotted in experiment was conducted under a continuous change in equivalence
Fig. 7a. It can be seen from this plot that the backpressure is very high ratio (0.25 per second), and some typical pressure histories and the
and the shock train leading edge has arrived upstream of T6. This variation of the equivalence ratio are plotted in Fig. 8a. The pressure
station is very critical because once a disturbance affects the flowfield at T13 is first to rise at about 15.6742 s, then the shock train leading
inlet unstart occurs. Thus, this special location is used to predict inlet edge propagates to T8 at about 16.3742 s. A sudden rise occurs in all
unstart [19]. In this experiment, inlet unstart occurs due to a further pressure histories at t  16.9147 s (φ1  1.065), meaning the onset
increase of backpressure. For the restarting process, the pressure of the unstart process. For the decreasing process of fuel, a sudden
distributions of both t  14.8750 s (point E) and t  15.3750 s drop in these four pressure histories occurs at t  19.5189 s
(point F) are plotted in Fig. 7b. It can be seen from the picture that the
pressures of both the converging inlet and isolator reduce remarkably,
600
which is the critical state for the restarting process. Therefore, there C
exists an obvious hysteresis effect in the mode transition of the 500 D
t=10 s
Static pressure/kPa

current scramjet, and the equivalence ratio of start-to-unstart 400


transition φ1 is in the range 1.000–1.250, whereas the equivalence
300
ratio of unstart-to-start transition ϕ2 is in the range 0.750–0.875.
200
600 100
T1
T5 0
500
Static pressure/kPa

T7 Bottom wall
T10 0 200 400 600 800 1000 1200 1400
400 T12
T13 X/mm
300 T16 a) Pressure distributions of unstart process
T12
T10
200 T16 600
T13 T7 E
100 500 F
T5
Static pressure/kPa

t=10 s
T1 400
0
1.5 300
1.250
Equivalence ratio

1.125
1.000 1.000 200
1 0.875 0.875
0.750 0.750 100
0.625 0.625
0.5 0.375 0.375
0
Bottom wall
0 0 200 400 600 800 1000 1200 1400
8 10 12 14 16 18 20 X/mm
Time/s
Fig. 4 The histories of some typical pressure transducers and b) Pressure distributions of restart process
equivalence ratios. Fig. 7 Some typical pressure distributions.
450 AIAA JOURNAL, VOL. 52, NO. 2: TECHNICAL NOTES

600 Equivalence ratio 600


T13 Equivalence ratio
T13 T13
Static pressure/kPa 500 T8 500
T8 T8
T5

Static pressure/kPa
400 T1 T5 400 T5 T13
T1
300 T1
300 T5
200
200 T1 T8
100 1.250
0.625 0.625 100 1.125
0
0.375 t1 t2 0.0 0.625
0 0.875
15 16 17 18 19 20 21 t1 t2 t3 0.474
Time/s 11 12 13 14 15 16 17
a) Some typical pressure signals Time/s
a) The histories of both the pressure at T5 and
600 equivalence ratio
T5
T13
500 600
Static pressure/kPa

T13
T5
Downloaded by NORTH CAROLINA STATE UNIVERSITY on May 10, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.J052505

400 500 T13

Static pressure/kPa
300 400 T13

200 T5 300
100 T5
200
3 2 1
0
0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 100
Equivalence ratio φ2 φ1
0
b) The hysteresis loops of both T5 and T13 0.6 0.7 0.8 0.9 1 1.1 1.2
Equivalence ratio
Fig. 8 The results of run 2.
b) The hysteresis loops of both T5 and T13 in Run 3
Fig. 9 The results of run 3.
(φ2  0.815), indicating the onset of the restarting process. To
present the hysteresis effect better, the relationships between the
pressure at T5 and T13 and the equivalence ratio are plotted in Fig. 8b, results are shown in Fig. 9. Similar to previous runs, the pressure at
which shows two obvious hysteresis loops. The pressure of the T13 rises at φ  0.625. All pressures have a sudden rise and a sudden
hysteresis loop at T5 and T13 changes suddenly at φ1 and φ2 , drop at t1 and t2 , respectively, as shown in Fig. 9a.
respectively. It can be seen from the hysteresis loop of T13 shown in Fig. 9b that
The hysteresis effect has a great effect on the unstart control; thus, there are two sudden rises, including one caused by the successful
it should be discussed. On the one hand, fuel supply should be ignition when the equivalence ratio stays at 0.625 and one denoting
decreased to far less than that of the start-to-unstart critical state due to the unstart onset, and a sudden drop meaning the onset of the
the existence of the hysteresis effect, namely, φ2 < φ1 . In the current restarting process. The pressure at T5 has a sudden rise and a sharp
facility, φ2 is lower than φ1 by about 0.25. On the other hand, the drop. It is worth noting that the equivalence ratio of unstart transition
practical unstart control, which is generally based on the unstart φ1 is 1.107, slightly higher than that of run 2 (1.065). The restart
margin [20], demands a further decrease of fuel. It is known to us that transition equivalence ratio is about 0.800, slightly smaller than that
the backpressure (the pressure at the isolator exit) can be used to of run 2 (0.815). The width of the hysteresis loop is nearly equal to
define unstart margin ε, which is defined by a quotient of current that of run 2. At the same time, the pressures at T13 and T5 at the inlet
backpressure Pb and pressure of the critical station Pcritical , namely, mode transition equivalence ratios are close to those of run 2. It
ε  Pcritical − Pb ∕Pcritical − Pinitial  [21]. A positive ε indicates a should be mentioned that the results of other runs that are not
started mode, and the smaller its value the more dangerously an inlet described in the current Note have similar results. Therefore, the fuel
operates, whereas a negative ε means an unstarted mode and demands supply rate has little effect on the hysteresis effect.
a reduction in fuel supply for the restarting inlet. In the current In addition, it is worth noticing that the pressure is very high
scramjet, the pressure at T13 can denote the backpressure, and it has although the inlet has already began the restart process. If the pressure
two critical stations. Generally, the pressure of the critical station of of the converging inlet is used to detect the operation mode and set a
the start-to-unstart transition is chosen as Pcritical . It can be seen from threshold, for example, 80 kPa in the current run, the pressure at T5
Fig. 8b that Pcritical and Pinitial equal about 400 and 100 kPa, does not recover until the equivalence ratio decreases to about 0.625
respectively. Then, the unstart margin ε  400 − Pb ∕400 − 100 (namely, t3 as shown in Fig. 9a). The inlet, however, will restart as
is reasonable for the start-to-unstart transition, but it has a great effect long as the equivalence ratio is lower than ϕ2 . This can be proved by
on the unstart-to-start transition. The backpressure (T13) at φ2 is the result of run 1 as shown in Fig. 6. The pressure at T5 recovers
about 500 kPa, so unstart margin ε  −1∕3. That is to say, although gradually as long as the onset of the restarting process occurs, even if
the inlet restarting process begins at equivalence ratio φ2 , the unstart the equivalence ratio keeps a constant of 0.750. Thus, a prior
control system based on backpressure still denotes an unstart mode, knowledge of hysteresis effects benefits unstart control.
which demands a further reduction in fuel for restarting. For the
current scramjet facility, the unstart control based unstart margin does IV. Conclusions
not indicate inlet start until the equivalence ratio decreases to φ3 as Unstart is detrimental to the performances of hypersonic vehicles,
shown in Fig. 8b. which should be controlled for a normal operation in practice. In the
current work, the inlet mode transition between start and unstart was
C. Run 3 investigated experimentally in a simplified scramjet engine. The
In practice, if the control system detects the inlet unstart, results indicate that there exists an obviously hysteresis effect in the
decreasing the fuel rate for a reduction of combustor pressure may be inlet mode transition. The hysteresis effect has a great effect on
the simplest method to restart inlet. To study the effect of the fuel unstart controls. On the one hand, to restart the inlet, the fuel supply
supplying rate on the hysteresis phenomenon, the equivalence ratio should be decreased to far less than that of the critical state in the start-
changes at 0.125 per second from 0.875 to 0.375 in this run. The to-unstart transition due to the existence of the hysteresis effect,
AIAA JOURNAL, VOL. 52, NO. 2: TECHNICAL NOTES 451

namely, φ2 < φ1 . On the other hand, the practical unstart control, [8] Im, S.-K., Do, H., Mungal, M. G., and Cappelli, M. A., “Experimental
which is generally based on the unstart margin defined by Study and Plasma Control of an Unstarting Supersonic Flow,” AIAA
backpressure, will demand a further decrease of fuel. The Paper 2012-2809, 2012.
backpressure at restart onset is still far higher than the critical [9] McClinton, C., Roudakov, A., Semenov, V., and Kopehenov, V.,
“Comparative Flow Path Analysis and Design Assessment of an
pressure of the start-to-unstart transition, which leads to a negative Axisymmetric Hydrogen Fueled Scramjet Flight Test Engine at a Mach
unstart margin. The unstart control based on the unstart margin Number of 6.5,” AIAA Paper 1996-4571, 1996.
demands a further reduction in the fuel supply. Therefore, the [10] Rodriguez, C. G., “Computational Fluid Dynamics Analysis of the
hysteresis phenomenon severely affects unstart control. Central Institute of Aviation Motors/NASA Scramjet,” Journal of
Based on experiments conducted under different changes in the Propulsion and Power, Vol. 19, No. 4, 2003, pp. 547–555.
equivalence ratio, the results show that the inlet will restart as long as doi:10.2514/2.6165
the equivalence ratio is lower than a constant, which indicates that the [11] Cui, T., He, X. G., Yu, D. R., and Tang, S. L., “Multistability and Loops-
adjustment rate of fuel has little effect on the hysteresis loop. Thus, a Coupled Hysteresis: Flight-Test Analysis on Error Detection of Inlet
prior knowledge of hysteresis effects will benefit the restart control, Start/Unstart,” Journal of Propulsion and Power, Vol. 28, No. 3, 2012,
pp. 496–503.
and it can avoid the undue reduction of fuel rate due to hysteresis and doi:10.2514/1.58149
the slow recovering of measured pressure. [12] Cui, T., Lv, Z., and Yu, D. R., “Multistability and Complex Routes of
It is worth mentioning that when a scramjet engine undergoes inlet Supersonic Inlet Start/Unstart,” Journal of Propulsion and Power,
unstart total airflow rate through the combustor should be reduced Vol. 27, No. 6, 2011, pp. 1204–1211.
Downloaded by NORTH CAROLINA STATE UNIVERSITY on May 10, 2014 | http://arc.aiaa.org | DOI: 10.2514/1.J052505

due to the spillage of the inlet. However, with the current direct- doi:10.2514/1.56384
connect configuration, the airflow rate is fixed. Therefore, the [13] Chang, J. T., Yu, D. R., and Bao, W., “Characteristic Analysis of Unstart/
hysteresis phenomenon observed in current works could be different Restart of Hypersonic Inlets Caused by Variations of Attack Angle of
from the actual phenomenon occurring in flight. The purpose of this Freestream,” Journal of Aerospace Power, Vol. 23, No. 5, 2008,
work, however, is to present the hysteresis effect in a converging inlet pp. 816–821 (in Chinese).
[14] Wagner, J. L., Yuceil, K. B., Valdivia, A., Clemens, N. T., and Dolling,
and discuss its effects on unstart control, which is an important issue D. S., “Experimental Investigation of Unstart in an Inlet/Isolator Model
for ramjet/scramjet engines because supersonic/hypersonic inlets are in Mach 5 Flow,” AIAA Journal, Vol. 47, No. 6, 2009, pp. 1528–1542.
also converging types. doi:10.2514/1.40966
[15] Chang, J. T., Wang, L., Bao, W., Qin, J., Niu, J., and Xue, W., “Novel
Acknowledgments Oscillatory Patterns of Hypersonic Inlet Buzz,” Journal of Propulsion
and Power, Vol. 28, No. 6, 2012, pp. 1214–1221.
This work was supported by China National Natural Science doi:10.2514/1.60461
Foundation (grant numbers 91216105 and 11372092). [16] Haw, W. L., Goyne, C. P., Rockwell, R. D., Krauss, R. H., and
McDaniel, J. C., “Experimental Study of Vitiation Effects on Scramjet
Mode Transition,” Journal of Propulsion and Power, Vol. 27, No. 2,
References 2011, pp. 506–508.
[1] Neaves, M. D., McRae, S., and Edwards, J. R., “High-Speed Inlet doi:10.2514/1.49090
Unstart Calculations Using an Implicit Solution Adaptive Mesh [17] Tatman, B. J., Rockwell, R. D., Goyne, C. P., McDaniel, J. C., and
Algorithm,” AIAA Paper 2001-0825, 2001. Donohue, J. M., “Experimental Study of Vitiation Effects on
[2] McDaniel, K. S., and Edwards, J. R., “Three-Dimensional Simulation of Flameholding in a Cavity Flameholder,” Journal of Propulsion and
Thermal Choking in a Model Scramjet Combustor,” AIAA Paper 2001- Power, Vol. 29, No. 2, 2013, pp. 417–423.
0382, 2001. doi:10.2514/1.B34687
[3] Chang, J. T., and Fan, Y., “Effects of Boundary-Layers Bleeding on [18] Cui, T., Yu, D. R., and Bao, W., “Analysis of the Equilibrium Manifold
Performance Parameters of Hypersonic Inlets,” Aircraft Engineering Linearization Model for Normal Shock Position Control,” AIAA Paper
and Aerospace Technology, Vol. 81, No. 3, 2009, pp. 204–211. 2006-4454, 2006.
doi:10.1108/00022660910954709 [19] Chang, J. T., Wang, L., Yu, D. R., and Bao, W., “Real-Time Unstart
[4] Chang, J. T., Bao, W., Yu, D. R., Fan, Y., and Shen, Y., “Effects of Wall Prediction and Detection of Hypersonic Inlet Based on Recursive
Cooling on Performance Parameters of Hypersonic Inlets,” Acta Fourier Transform,” AIAA Paper 2012-4150, 2012.
Astronautica, Vol. 65, Nos. 3–4, 2009, pp. 467–476. [20] Chang, J. T., Fan, Y., Bao, W., Yu, D. R., and Shen, Y., “Unstart Margin
[5] Trapier, S., Deck, S., Duveau, P., and Sagaut, P., “Time-Frequency Control of Hypersonic Inlets,” Acta Astronautica, Vol. 66, Nos. 1–2,
Analysis and Detection of Supersonic Inlet Buzz,” AIAA Journal, 2010, pp. 78–87.
Vol. 45, No. 9, 2007, pp. 2273–2284. doi:10.1016/j.actaastro.2009.05.021
doi:10.2514/1.29196 [21] Chandra, K. P. B., Gupta, N. K., Ananthkrishnan, N., Park, I. S., and
[6] Srikant, S., Wagner, J. L., Valdivia, A., Akella, M. R., and Clemens, N., Yoon, H. G., “Modeling, Simulation, and Controller Design for an Air-
“Unstart Detection in a Simplified-Geometry Hypersonic Inlet-Isolator Breathing Combustion System,” Journal of Propulsion and Power,
Flow,” Journal of Propulsion and Power, Vol. 26, No. 5, 2010, Vol. 26, No. 3, 2010, pp. 562–574.
pp. 1059–1071. doi:10.2514/1.42368
doi:10.2514/1.46937
[7] Valdivia, A., Yuceil, K. B., Wagner, J. L., Clemens, N. T., and Dolling, G. Candler
D. S., “Active Control of Supersonic Inlet Unstart Using Vortex Associate Editor
Generator Jets,” AIAA Paper 2009-4022, 2009.

You might also like