American Institute of Aeronautics and Astronautics

1

Design Study of Hypersonic Components for Precooled Turbojet Engine

Takayuki Kojima1, Hiroaki Kobayashi1, Hideyuki Taguchi2, Katsuyoshi Fukiba1, Kazuhisa Fujita1, Hiroshi Hatta3, Ken Goto3 and Takuya Aoki1

Japan Aerospace Exploration Agency, 7-44-1, Jindaijihigashimachi, Chofu, Tokyo, 182-8522, JAPAN

and

Tetsuya Sato4 Waseda University

Recent studies about variable nozzles, that are a rectangular type nozzle and an axisymmetric type nozzle, of the precooled turbojet engine (S-engine) which are developed for the demonstration of the key technologies for the propulsion system of the hypersonic airplane and the first stage propulsion of the TSTO space plane are described in this paper. For the rectangular nozzle, three types of C-shaped carbon/carbon composite cowls which includes subscale model of the precooled turbojet engine are fabricated and the fine attachment to the ramp is confirmed. For the firing of the S-engine, stainless steel cowl with concrete heat insulator are fabricated and tested for 20 sec. Axisymmetric variable plug nozzle which is made of C/C material is fabricated and pressurized by the cold flow test. The axisymmetric plug nozzle can be operative up to 0.57 MPa of nozzle inlet pressure.

Nomenclature Ac = Intake Capture Area (=14100 [mm2]) Ath_n = Nozzle throat Area ER = Equivalence Ratio Tc = Afterburner combustion gas temperature NPR = Nozzle Pressure Ratio

I. Introduction precooled turbojet cycle is one of the most efficient propulsion system for the hypersonic airplane and the

first stage propulsion system of the TSTO space plane. As described in the JAXA’s long-term vision, the development of a Mach 5 demonstrator to verify the technologies on the Mach 5 vehicle is now carried out1). For the flight test of the Mach 5 demonstrator, the small scale precooled turbojet engine “S-engine” is now under development. S-engine has 224 mm x 225 mm of rectangular cross section, 2.6m in overall length and about 140 kg in weight as shown in Fig. 1. Firing tests of the precooled turbojet engine were conducted in March 20072) and October 2007 respectively. Performances of the engine components are verified by these firing tests. Start-up sequences of the core engine and the precooler are also established successfully. Air and fuel flow of the S-engine is shown in Fig. 2. The S-engine consists of hypersonic variable intake, air precooler, core engine, afterburner, and variable nozzle. LH2 which is used as fuel and coolant flows through the precooler which is a shell and tube type heat exchanger. By the air precooling effect, the core engine and other rotational components can be operative up to Mach 5 flight condition. In order to adopt the core engine to the

1 Aerospace Research and Development Directorate, AIAA Member. 2 Aviation Program Group, AIAA Member. 3 Institute of Space and Astronautical Science, AIAA Member. 4 Department of Mechanical Science and Aeronautics, Faculty of Engineering, AIAA Member.

A

15th AIAA International Space Planes and Hypersonic Systems and Technologies Conference28 April - 1 May 2008, Dayton, Ohio

AIAA 2008-2504

Copyright © 2008 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

American Institute of Aeronautics and Astronautics

2

NozzleCombustorCompressorPrecoolerIntake

Fig. 1 Firing test model of the Precooled turbojet engine (S-engine)

incoming hypersonic air flow condition, the variable systems are applied to the hypersonic intake. Variable nozzle is also one of the most important components to control the air flow rate, thrust and pressure of the engine. The afterburner fuel equivalence ratio and the combustion gas temperature of the afterburner are shown in Fig. 3. The precooler, whose cross section is 23 cm x 23 cm, is very small. Therefore, the heat exchange area of the precooler is limited and the engine equivalence ratio are higher than 1.4 to make enough heat exchange at the precooler. The regenerative cooling of the afterburner wall is applied to utilize the gaseous hydrogen which flows from the precooler. Nozzle area and pressure ratio for the S-engine are shown in Fig. 4. The throat area Ath_n changes from 23% to 47% of the intake capture area. Nozzle pressure ratio NPR also changes from 2 to 200. For the nozzle of the hypersonic air breathing engine like S-engine, control of the throat area Ath_n is required, in parallel with making large thrust for each NPR.

Fig. 2 Flow diagram of the S-engine

American Institute of Aeronautics and Astronautics

3

1.5

2

2.5

3

3.5

4

1800

1900

2000

2100

2200

2300

0 1 2 3 4 5 6

ER Tc

Fuel Equivalence Ratio

ER

Combustion Ga

s Te

mperatur

e Tc

Mach Fig. 3 Afterburner fuel flow rate and combustion gas temperature

0.2

0.25

0.3

0.35

0.4

0.45

0.5

1

10

100

1000

0 1 2 3 4 5 6

An_th/Ac NPR

Noz

zle

Thro

at A

rea

An_

th/A

c Nozzle P

ressure Ratio NPR

Mach Fig. 4 The nozzle throat area and pressure ratio

Nozzles for jet engines and rocket engines are defined in Fig. 53-5). The simplest and the most reliable nozzle systems do not have variable geometry. However, these nozzles can not control both Ath_n and NPR. Nozzles which apply the axisymmetric configuration are lighter than rectangular nozzles from the viewpoint of nozzle structure because the axisymmetric nozzle does not have corner where the stress concentration occur. Therefore, the nozzle for the jet engine uses the axisymmetric configuration with radius direction variable system. But for the S-engine, if the radius direction variable system is introduced, the actuation system becomes too large and too heavy because the NPR of the S-engine is much higher than that of conventional jet engine. Therefore axis direction variable geometry is used for the nozzle of the S-engine. Rectangular type nozzle can make lighter variable system, because the pressure of the moving plug (or spike) can be optimized by back pressure easily. Therefore, the rectangular nozzle which has radius direction variable system and the axisymmetric nozzle which has axis direction variable system studied respectively. The present status of both variable nozzles is described from the next section.

American Institute of Aeronautics and Astronautics

4

Radius Direction Axis Direction

Axisymmetric

(Light Structuire)

Rocket Engine Jet Engine S-engine

Rectangular(Light Variable

System)

Linear Aerospike S-engine N/A

(Too Heavy)

Variable GeometryFixed

LE-7A (H-IIA)

XRS-2200 (X-33)

F100-PW-100 (F-15)

Fig. 5 Definition of the Variable Nozzle

II. Rectangular Variable Nozzle Drawing and picture of the rectangular nozzle is shown in Fig. 6 and 7. Its geometry is 23 cm in width and 88 cm

in total length including transition duct from round to rectangular, afterburner injector, variable plug and external ramp. Throat area Ath_n is controlled by moving the pear-shape variable plug. The throat height changes from 10.5mm to 28 mm, and it depends on the design plug angle and motor stroke so that Ath_n can be changed easily. A heat generated by the combustion at the afterburner to the driving motor under the ramp is shielded by the regenerative ramp. In the regenerative cooling ramp, the gaseous hydrogen flows through 20 tubes which have 5.5 mm x 2.2 mm in cross section. Furthermore, the motor itself is cooled by helium gas which controls back pressure of the variable plug to reduce the driving force of the plug.

200

880

Afterburner

Regenerative Cooling Ramp Variable Plug

CowlTransit ion Duct Injector

Sidewall

Plug Motor

Fig. 6 Drawing of rectangular variable nozzle

Fig. 7 Picture of the rectangular variable nozzle

American Institute of Aeronautics and Astronautics

5

Fabrication of Sidewall and cowl which is made of Carbon/Carbon composite material is conducted. Ramp and cowl structure of the rectangular nozzle is shown in Fig. 8. The temperature of the ramp and motor does not become over 600 K because the ramp is cooled regeneratively. Therefore, the C/C structure has “C” shape. The sidewalls of the C/C material are fixed by the metal bolts. Drawing of the C-shaped C/C cowl is shown in Fig. 9.

Fig. 8 Heat resistance structure of the rectangular nozzle (the nozzle cross section)

632

192

224

200

Fig. 9 Drawing of the C/C cowl

For the fabrication of the C-shape cowl, making the two sidewalls in parallel and rigid corner are the key

technologies. The C/C cowl is fabricated by the PY method of the across corporation6) as shown in Fig. 10. Carbon fiber, which is attached to the graphite block, is along the cowl and the sidewall, so that that the shear stresses along the cowl and sidewalls are critical against fracture. By FEM analysis the rigidity at the corner is estimated as shown in Fig. 11. The highest stresses along the cowl and sidewall surface for each cowl thickness t are shown in Fig. 12. For this type of 2-D composite material, the shear stress is critical. We confirmed that the shear stress τxy becomes lower than 10 MPa, when the thickness of the cowl is 12 mm.

Ramp (Regenerative Cooling)

Combustion Gas (0.4MPa, 2000K)

Motor

200

Cowl (C/C Material)

Sidewall

American Institute of Aeronautics and Astronautics

6

Graphite

C/C Sheet

Fig. 10 Fabrication of the C/C cowl Fig.11 Shear stress around the corner of the cowl

0

10

20

30

40

6 7 8 9 10 11 12

Thickness of the cowl t [mm]

Str

ess

[M

Pa]

σx

τxy

R

t

Ｍｉｒｒｏｒ

Ｍｉｒｒｏｒ

Pressure

Stress Evaluation

R=t

Fig. 12 Highest stress of the cowl

Three types of the C-shape cowl is fabricated as shown in Fig. 13, 14 and 15 that is, 1) full scale model (t=6mm) 2) 1/2 scale model (t=8mm) 3) full scale model (t=12mm)7). Presently, after these problems are cleared by fabricating 1) and 2) models, full scale model 3) is fabricated and confirmed fine attachment to the S-engine. Weight of the full scale cowl 3) is 7.5 kg.

Fig. 13 t=6mm full-scale C/C cowl model

American Institute of Aeronautics and Astronautics

7

Fig. 14 1/2 Scale C/C cowl model

Fig. 15 Full scale C/C cowl model

The metal cowl with concrete heat insulator is used as the cowl of the S-engine’s nozzle for the firing test of the

S-engine, because the duration of the firing test is limited to few seconds. LC-17 (Asahi light caster corporation) is used as the insulator. Thickness of the insulator and material wall are 5 mm and 4 mm respectively. Cowl wall temperature prediction after an ignition of the afterburner is shown in Fig. 16. Initial temperature is ambient and afterburner combustion temperature is 2000 K. The temperature of the cowl will be lower than the metal melting temperature for 10~ 15 sec. The metal cowl with the insulator is attached to the S-engine as shown in Fig. 17. Weight of the metal cowl with insulator is 12.9 kg.

American Institute of Aeronautics and Astronautics

8

200

400

600

800

1000

1200

1400

1600

1800

0 10 20 30 40 50

Tw_inTw_out

Tem

pera

ture

[K]

Time [sec]

5mm

4mm SUS

Insulating Material

Adiabatic Wall

Combustion Gas (2000K)

Tw_out

Tw_in

Fig. 16 Temperature History of the nozzle cowl

Fig. 17 Fabrication of the metal cowl (View from downstream of the cowl exit)

Second firing test of the S-engine is conducted at Noshiro Test Center of JAXA in October 2007. Picture of the firing test is shown in Fig. 18. Because the fuel equivalence ratio is higher than stoichiometric condition, large frame can be seen from the exhaust of the nozzle. A rotational speed history of the S-engine firing test is shown in Fig. 19. The regenerative cooling is done for 20 seconds after the engine rotational speed become stable (90 ~ 110 sec). The nozzle ramp, plug and metal cowl with insulator are not damaged after the 2 times of the afterburner operation. Heat transfer ratio of the regenerative cooling wall is shown in Fig. 20. Design value of the heat transfer ratio of the ramp surface is 471 W/m2K, which agreed well with the test result.

Ramp

Cowl Insulator

American Institute of Aeronautics and Astronautics

9

Fig. 18 Firing test of the Precooled turbojet engine

0

1 104

2 104

3 104

4 104

5 104

-50 0 50 100 150

Rot

atio

nal S

peed

N [r

pm]

Time [sec]

Regenerative cooling ON(90~110 sec)

Fig. 19 Engine rotational speed

-200

0

200

400

600

800

0

10

20

30

40

50

80 90 100 110 120

α m_H2

Hea

t tra

nsfe

r coe

ffici

ent α

[W/m

3 K]Regenerativ

e cooling fl

ow rate

[g/s]

Time [sec]

Regenerative cooling ON(90~110 sec)

Fig. 20 Heat transfer coefficient of the nozzle ramp wall

American Institute of Aeronautics and Astronautics

10

III. Axisymmetric Variable Plug Nozzle

Isometric view of the axisymmetric plug nozzle is shown in Fig. 21. Size of the cowl is 220 mm in outside diameter and 170 mm in inside diameter. The plug nozzle consists of 4 parts that are a fixed cowl, variable cowl, strut and plug. The variable cowl can slide in axial direction. The fixed cowl is attached to the wind tunnel attachment by the cowl support panel that consists of capper. The variable cowl which is supported by the cowl support can move in axis direction. The plug is supported by the eight struts. FEM analysis result is shown in Fig. 22. Even though stress becomes high at the tip of the variable cowl where the throat of the nozzle is located, there are enough safety margins in the structure strength.

Wind Tunnel AttachmentWind Tunnel Attachment PlugPlug

Variable Cowl Variable Cowl (Pre(Pre--Fixed)Fixed)

Fixed CowlFixed Cowl

Cowl SupportCowl Support

StrutStrut

Wind Tunnel AttachmentWind Tunnel Attachment PlugPlug

Variable Cowl Variable Cowl (Pre(Pre--Fixed)Fixed)

Fixed CowlFixed Cowl

Cowl SupportCowl Support

StrutStrut

Fig. 21 Isometric view of the C/C plug nozzle

Cowl

Plug

0 100µS Fig. 22 Circumferential strain of the C/C plug nozzle

In order to confirm pressure resistant characteristics of the plug nozzle, cold flow test of the nozzle is

conducted. Photograph of the C/C plug nozzle on the University of Tokyo Kashiwa Hypersonic and High Temperature Wind Tunnel test section is shown in Fig. 23. In this test facility, up to 1 kg/s of high temperature air can be supplied to the engine. Pressure, temperature and air flow rate history are shown in Fig. 24. Approximately 0.55 kg/s of air is supplied to the nozzle for approximately 190 seconds. During these seconds, the nozzle pressure was 0.57 MPa. Design of the actuation mechanism for the axisymmetric plug is further issues for the realization of this plug nozzle.

American Institute of Aeronautics and Astronautics

11

Fig. 23 C/C plug nozzle test model

0

50

100

150

200

250

300

350

400

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 40 80 120 160 200

T0Tc_in

Tc_out WairP0

Noz

zle

Inle

t Tem

pera

ture

T0

Noz

zle

Cow

l Ins

ide

Tem

pera

ture

Tc_

inN

ozzl

e C

owl O

utsi

de T

empe

ratu

re T

c_ou

t

Air Flow R

ate Wair [kg/s]

Nozzle Inlet P

ressure P0 [MP

a]

Time1 [sec] Fig. 24 Pressure and temperature of the C/C plug nozzle

IV. Conclusion Two types of variable nozzle for S-engine are fabricated and tested in this study. As results, followings are

obtained. - The area of the nozzle throat and NPR changes widely. Therefore, conventional nozzle configuration is not adequate. Variable mechanism is necessary. Rectangular Nozzle - Radius direction variable rectangular type nozzle and axis direction variable axisymmetric plug nozzle are appropriate for the S-engine - C-shaped Carbon/Carbon composite cowl are fabricated successfully. Fine attachment of the cowl to the ramp is confirmed. - If the firing duration is limited within 20 seconds. The material cowl which has a concrete insulator is adequate. Axisymmetric Nozzle - The axisymmetric plug nozzle can be operative up to 0.57 MPa of nozzle inlet pressure.

American Institute of Aeronautics and Astronautics

12

The authors acknowledge the contribution of Dr. Imamura (University of Tokyo), Mr. Hongoh (JAXA) and all members of the S-engine testing for the preparation and testing, Mr. Suzuki (JAXA) for the FEM analysis and Mrs. Yamashita (Across corp.) for the fabrication of the C/C nozzle.

References

1 JAXA Long Term Vision, March 2005, http://www.jaxa.jp/about/2025/pdf/2025_02.pdf 2 Sato, T.: Development Study of a Precooled Turbojet Engine for Flight Demonstration, AJCPP 2008. 3 Przemieniecki, J, S.: Aircraft Engine Design Second Edition, 2002, pp494.

4 Harmon, Tim. : Systems engineering and integration strategy on the X-33 linear aerospike engine, AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Seattle Washington, July 6-9 1997, AIAA 97-3317. 5 John S. Orme. : Initial Flight Test Evaluation of the F-15 Active Axisymmetric Vectoring Nozzle Performance, AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Cleveland Ohio, July 13-15, AIAA 98-3871. 6 www.across-cc.co.jp 7 Kojima, T.: Development Study on C/C cowl for Precooled Turbojet Engine, FY17 Space Transportation Symposium, Sagamihara Japan, 2006 (in Japanese).