Propulsion and Power Research 2022;11(4):444e456

http://ppr.buaa.edu.cn/

Propulsion and Power Research

w w w. s c i e n c e d i r e c t . c o m

ORIGINAL ARTICLE

Overspeeding characteristics of turbomachinery

for gas generator cycle air turbo ramjet engine
Ryojiro Minatoa,*, Daisuke Nakatab, Masaharu Uchiumib,
Ryoji Imaia, Kazuyuki Higashinob

a
Department of Engineering, Muroran Institute of Technology Muroran, Hokkaido, 050-8585, Japan
b
Aerospace Plane Research Center, Muroran Institute of Technology Muroran, Hokkaido, 050-8585, Japan

Received 30 September 2021; accepted 23 October 2022

Available online 6 December 2022

KEYWORDS Abstract The main interest in the current study focuses on the possibility of overspeeding for
the gas-generator cycle air turbo ramjet (GG-ATR) engine. The authors developed the air turbo
Air turbo ramjet engine;
Compressor map; ramjet engine and investigated its compressor performance. Based on those data, the authors
Overspeeding; developed the analytical code for the air turbo ramjet engine, which calculates the performances
Combustion efficiency of turbomachinery, gas-generator, and ram combustor. The previous study described that the
rotor overspeeding would not occur in the air turbo rocket engine. However, the current results
show that degraded ram combustion can decrease the compressor pressure ratio and the
compressor power. This reduced compressor power can cause overspeeding for the air turbo
ramjet engine. The experimental results of compressor power and turbine inlet pressure support
those analytical results.
ª 2022 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).

*Corresponding author.
E-mail addresses: [email protected] (Ryojiro Minato), 1. Introduction
[email protected] (Daisuke Nakata), uchiumi@mmm.
muroran-it.ac.jp (Masaharu Uchiumi), [email protected]
An air turborocket or air turbo ramjet (ATR) engine is a
(Ryoji Imai), [email protected] (Kazuyuki Higashino).
type of turbine-based combined cycle (TBCC) engine. It can
Peer review under responsibility of Propulsion and Power Research.
be a candidate for a propulsion engine of a hypersonic vehicle
or a spaceplane [1e8]. The ATR engine has a greater specific
thrust than turbojet engines and a higher specific impulse (Isp)
Production and Hosting by Elsevier on behalf of KeAi
than rocket engines. Kobayashi et al. consider the ATR engine

https://doi.org/10.1016/j.jppr.2022.10.001
2212-540X/ª 2022 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Overspeeding characteristics of turbomachinery 445

Nomenclature Subscript
2 air airflow
A cross section area (unit: m )
cho choke condition
C) characteristic velocity as represent in Eq. (4) (unit:
comp compressor
m/s)
cor corrected value
C0 spouting velocity at the turbine nozzle
de on design condition
CP heat capacity at constant pressure (unit: J/(kg$K))
fuel fuel
d1,k, d2,k, d3,k polynomial coefficients for turbine module in Eq.
GG gas generator
(10)
i index for the referential corrected rotation speed
M Mach number
ideal ideal condition
M_ cor corrected mass flow rate as function of pcomp,i (unit:
in inlet condition
kg/s)
k degree of polynomial
m mass flow rate (unit: kg/s)
nozzle turbine nozzle
N rotational speed (unit: rpm)
out outlet condition
P pressure (unit: Pa)
ram ram combustor
R gas constant (unit: J/(kg$K))
ref referential condition
T temperature (unit: K)
s static condition
U rotational speed of turbine blade (unit: m/s)
spout gas spouting at turbine nozzle
V velocity (unit: m/s)
sur surge condition
w specific power of compressor or turbine (unit: W)
T stagnation condition
turb turbine
Greek letters 0 ambient condition
h efficiency 1 air intake inlet
Hs adiabatic compression efficiency as function of mcor 2 compressor inlet
k specific heat capacity ratio 3 compressor outlet
Ps total-to-static compressor pressure ratio as function of 4 turbine inlet
m_ cor;i 5 turbine outlet
p pressure ratio at engine component 6 ram combustor inlet
r density (unit: kg/m3) 7 ram combustor outlet
t temperature ratio at engine component 8 nozzle throat
9 exit of nozzle

for the reusable launch vehicle due to those advantages [9]. As

well as Ref. [9], the authors also have developed the ATR
engine for the supersonic unmanned aerial vehicle [10],
including the propellant feeding system [11,12]. Unlike a
turbojet engine, the propellant gas of the ATR engine drives its
turbine. For a gas-generator cycle air turbo ramjet (GG-ATR)
engine, fuel and oxidizer are burned at the gas generator (GG)
in fuel-rich conditions. The GG combustion provides the
enthalpy for turbine power. The fuel-rich turbine effluent gas
burns with air from the compressor in the downstream ram
combustor. Previous studies have investigated various liquid
propellants [6e8] and solid propellants [13e15] for the GG-
ATR engine. Figures 1 and 2 show the ATR engine developed
at Muroran Institute of Technology and its conceptual sche-
matic, respectively. The authors conducted the GG-ATR en- Figure 1 GG-ATR engine on the test bed.
gine turbomachinery tests, which employed nitrogen gas
(GN2) or helium gas as turbine driving gas instead of GG The turbomachinery characteristics of the GG-ATR en-
combustion gas. In an actual GG-ATR engine operation, a gine differ from a turbojet engine. The source of turbine
nozzle throat plays a role in controlling the ram combustor power for the GG-ATR engine comes from GG combustion
pressure and the gas flow rate. However, the authors do not gas. The compressor power at the high altitude will be
still develop the ram combustor. Thus, the authors install the lower than the sea-level static conditions because the air
orifice at the compressor exit to control the airflow rate in the density is lower at the high altitude. Thus, propellant flow
tests. Replacing the orifices changes the compressor operating rates should be strictly controlled to keep the rotor speed
condition to obtain the compressor characteristic map data. constant. Otherwise, overspeeding will occur. The
446 Ryojiro Minato et al.

2. GG-ATR engine and experimental facility

2.1. Specification of GG-ATR engine

Table 1 lists the specifications of the GG-ATR engine at

Muroran Institute of Technology. This GG-ATR engine
uses liquefied oxygen (LOX) and ethanol as an oxidizer
and a fuel, respectively. LOX and ethanol are injected into
the gas generator and burned to generate the turbine-
Figure 2 Conceptual schematic of GG-ATR engine [6,7]. driving gas. Figures 3 and 4 show the rotor element and
the mixed flow compressor configuration of the GG-ATR
engine, respectively. The rotor impeller of this mixed
propellant flow rate control is subject to the GG-ATR en-
flow compressor is made of titanium alloy (Ti-6Al-4V).
gine characteristics, such as a compressor, turbine, and gas
It has seven main blades and seven splitters in the rotor.
generator. Many researchers have investigated the GG-ATR
The diffuser has eight stator blades. The tip diameters of
engine component characteristics so far. In contrast, Lilly
the mixed flow compressor impeller are 150 mm and
et al. carried out the engine component integration and the
167.2 mm at the inlet and outlet, respectively. The outer
GG-ATR tests [16]. For engine component studies, Refs.
and inner diameters at the diffuser exit are 210 mm and
[13,14] discussed the gas generator and ram combustor
144 mm, respectively. In the on-design condition, the
performances. Liu et al. proposed the matching model of an
rotational speed is 58,000 rpm, which corresponds to
intake-compressor interaction for the GG-ATR engine and
507 m/s of the circumferential tip speed. The airflow rate of
numerically assessed its viability [17]. Bussi et al. investi-
this mixed flow compressor is 3.47 kg/s, and the total-to-
gated the off-design performance of the GG-ATR
static pressure ratio is 2.27.
compressor [18,19]. Bossard and Thomas focused on the
The turbine of the GG-ATR engine is a two-staged axial
surge/stall phenomena of the GG-ATR engine compressor
one, as shown in Figure 3. The turbine blisks are made of
[20]. They concluded that the GG-ATR engine did not have
nickel-based alloy (INCONEL 713). The turbine inlet
the possibility of overspeeding because the gas generator
pressure is 1.35 MPa, and the total-to-static turbine expan-
fully controls the shaft power to the compressor. However,
sion ratio is about 6. The turbine inlet temperature is strictly
the choke flow at the nozzle throat depends on the mass
limited to 1100 K due to the thermal limit of a turbine blade.
flow rate, total pressure, and total temperature. Thus, it can
The pitch diameter of the turbine blade is 95 mm. Thus, the
affect the compressor pressure ratio/airflow rate character-
rotational speed of the turbine blade is up to 288.5 m/s at the
istics. The overspeeding discussion in Ref. [20] did not
on-design condition.
include the ram combustor performance, i.e., combustion
efficiency. Low ram combustion efficiency decreases the
2.2. Experimental facility
total temperature at the nozzle throat, which affect the
compressor operation condition. This situation can vary the
The authors carried out the GG-ATR engine turboma-
compressor power and might increase the rotational speed
chinery tests using nitrogen (GN2) or helium gas (GHe) as
of the rotor.
turbine driving gas. The authors utilized three gas cylinder
The turbine power of the GG-ATR engine is independent
of its compressor operation. It means the GG-ATR engine
has the potential to overspeeding. Thus, the actual GG-ATR
Table 1 GG-ATR engine specification.
engine operation necessitates the appropriate model to
predict its overspeeding. This model should consider the Rotational speed at design point 58,000 rpm
Compressor inlet diameter 150 mm
interaction between the turbomachinery and the ram
Maximum diameter 231 mm
combustor. Therefore, the objective of the current study is Pressure ratio of compressor (total-to-static) 2.27
to develop the GG-ATR engine analytical model and Adiabatic compression efficiency (total-to-static) 70.2%
evaluate the overspeeding risk of the GG-ATR engine. The Air flow rate 3.47 kg/s
authors conducted the GG-ATR engine test and its CFD GG combustion temperature 1100 K
GG combustion pressure 1.35 MPa
analysis. Based on those results, the new analytical code is
Mass flow rate of GG combustion gas 0.7 kg/s
developed for the GG-ATR engine performance prediction. Expansion ratio at turbine (total-to-static) 6
This code analyzes all engine components, such as turbo- Adiabatic turbine efficiency (total-to-static) 65%
machinery, gas-generator, and ram combustor. Its code Ram combustor temperature 2380 K
validation is carried out by using the GG-ATR engine test Thrust at sea level static condition 3804 N
Isp at sea level static condition 555.4 s
data.
Overspeeding characteristics of turbomachinery 447

discharge coefficients for both orifice plate flow meters are

cited in Ref. [21]. The authors carried out the pressure and
temperature measurements at each engine component. Those
data evaluate the compressor pressure ratio, turbine expansion
ratio, and adiabatic compressor and turbine efficiencies.

3. CFD analysis for turbomachinery

The authors conducted CFD analysis to evaluate the

turbomachinery performance of the GG-ATR engine in the
Figure 3 Rotor element of GG-ATR engine. present study. The authors employ the commercial software
Fine TURBO@ by Numeca INTERNATIONAL Inc. for
CFD analysis. The governing equation is the three-
dimensional turbulent Navier Stokes equation with the
Spalart-Allmaras 1-equation turbulence model. The rota-
tional speed is ranged from 50% to 110% of the rated speed.
The air in the CFD model is treated as calorically perfect gas
because the temperature in the numerical solution does not
exceed 450 K. On the other hand, the turbine analysis treats
the GG combustion gas as the thermally perfect gas. To
generate the numerical grid, the authors employ Auto
Grid@, presented by Numeca INTERNATIONAL Inc. The
average yþ on the wall boundary is about 1.0. The authors
use the mixing plane condition to model the interface be-
tween stator and rotor because of its simplicity.
The total temperature and pressure at the inlet boundary
for the compressor analysis are given as 288.15 K and
101.3 kPa, respectively. The total pressure at the turbine
inlet depends on the compressor power, as described in the
next section. The total temperature at there is determined
Figure 4 Mixed flow compressor of GG-ATR engine. under constant O/F condition and has the pressure de-
pendency. The total temperature and the total pressure at the
bundles (1 bundle has 20 gas cylinders for N2 gas and 25 for turbine inlet boundary for the rated rotational speed are
helium gas) and connected them to this test facility. GN2 is 1100 K and 1350 kPaA, respectively. In the off-design
supplied from those bundles to the turbine manifold of the condition, those ranges from 1024 K to 1111 K and from
GG-ATR engine and drives the turbine. The temperature of 400 kPaA to 1600 kPaA for total temperature and pressure,
GN2 in gas cylinder bundles is equal to room temperature, respectively.
which is much lower than GG combustion gas. Thus, the Four numerical grids are tested in the grid independence
turbine power for GN2 is lower than that for GG combustion tests for the compressor and turbine analyses. The grid
gas, and the maximum rotational speed of the turbine is point numbers are 1.22, 2.14, 4.08, and 5.92 million for the
limited to 42,000 rpm for the turbine driving gas of GN2. It compressor analysis, and those are 1.08, 2.13, 4.21, and
is about 70% of the rated rotational speed (58,000 rpm). 6.36 million for turbine ones. Figure 5(a) indicates the
Therefore, helium gas must be used as the turbine driving result of the grid independence test for the compressor
gas at the rated rotational speed. The maximum GN2 flow analysis, which shows the compressor characteristic map at
rate is 2.0 kg/s with total pressure at the turbine manifold of 70% of the rated rotational speed. The numerical results of
1.6 MPa. That of GHe is 0.58 kg/s with total pressure at the those four grids are almost identical to each other. How-
turbine manifold of 0.85 MPa. ever, only a 1.22 million grid has a slight higher pressure
In the GN2 or GHe supplying tube of this facility, the au- ratio than other three grids near the surge condition.
thors employed the orifice plate flow meter to measure the Therefore, 2.14 million grid is employed in the present
mass flow rate of GN2 or GHe. The thermocouple and the study.
pressure transducer measure the temperature and static pres- Figure 5(b) shows the grid dependence test for the tur-
sure upstream of the orifice, respectively. Moreover, the dif- bine analysis. The maximum difference in adiabatic turbine
ferential pressure across the orifice also is acquired by the efficiency is 0.9% among 2.12, 4.21, and 6.36 million grids.
pressure transducer. The bell mouth is installed on the However, the difference between 1.08 million and 6.36
compressor inlet to rectify the airflow rate. The orifice plate is million grids is nearly equal to 2.0%. Therefore, 2.12
installed in the bell mouth to measure the airflow rate. The million grid is employed for the turbine analysis.
448 Ryojiro Minato et al.

Figure 6 Total pressure distribution of the present mixed flow

compressor in on-design condition.

code for the detailed GG-ATR engine analysis. The GG-

ATR engine operation has two operating constraints, A)
compressor-turbine power balance and B) nozzle choke
condition. Equation (1) represents constraint A).


m_ air CP TT 2 k1
pcomp  1 Z hturb m_ GG CP;GG TGG
k
hcomp

 kGG 1 ! ð1Þ
kGG
PT 5
1
PGG

Equation (2a) corresponds to the constraint B) in choked

flow condition.
Figure 5 Grid dependence tests. (a) Compressor analysis, (b) tur- sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

kkram þ1
bine analysis. kram 2 ram 1
m_ air þ m_ GG Z PT 7 A8 ð2aÞ
Rram TT 7 kram þ 1
Figure 6 shows one example of the present CFD results, On the other hand, Eq. (2b) indicates the constraint B) in
which is the total pressure distribution of the present mixed unchoked flow condition.
flow compressor in the on-design condition.

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
u
u
 2
kram þ 1 !
k kram
m_ air þ m_ GG Z PT 7 A8 t
2k ram P T 7 PT 7
ram

ðkram  1ÞRram TT 7 P0 P0
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u ð2bÞ
u
u
2ðkram  1Þ
kram  1 !
kram kram
ZP0 A8 t
2k ram PT 7 PT 7

ðkram  1ÞRram TT 7 P0 P0

4. GG-ATR engine cycle analysis where PT5 and PT7 are the total pressure at the turbine and
the ram combustor outlet, respectively.
4.1. Engine system and operational constraints In Eq. (1), the compressor operating state is described by
air flow rate, mair, compressor pressure ratio, pcomp (total-to-
The prediction of a GG-ATR engine operating state is total), and adiabatic compression efficiency, hcomp (total-to-
indispensable for its development and requires its cycle total). On the other hand, GG gas flow rate, mGG, GG
analysis. The cycle analysis of a GG-ATR engine is quite combustion temperature, TGG, and GG combustion pressure,
different from that of a turbojet engine. Thus, the authors PGG, are variables concerned with GG combustion. The
developed the GG-ATR engine cycle analysis (GATRECA) chemical equilibrium calculation can determine TGG under a
Overspeeding characteristics of turbomachinery 449

given PGG and oxidizer-to-fuel (O/F) ratio. GG combustion point are the input data in the compressor module. Ncor is
gas flow is always choked at the turbine nozzle. Thus, the assumed to be between the ith and (iþ1)th referential rota-
following Eq. (3) can hold for the mGG-PGG relationship. tional speeds (donated as Nref,i and Nref,iþ1, respectively).
pcomp,i and pcomp,iþ1 are the total-to-total compressor pressure
PGG Anozzle
m_ GG Z 
ð3Þ ratios at Nref,i and Nref,iþ1, respectively. Those correlations are
CGG described in Eq. (5) by Ref. [23].
where Anozzle is the cross-section area at the turbine nozzle.
2
!k1
k

) Nref ;i k1
CGG is the characteristic velocity at the GG and is repre- pcomp;i Z 1 þ pcomp
k 1
sented in Eq. (4). Ncor
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð5Þ
u
2
!k1
k

u
kGG þ1 Nref ;iþ1 k1

) tRGG TGG kGG þ 1 kGG 1 pcomp;iþ1 Z 1 þ pcomp
k 1
CGG Z ð4Þ Ncor
kGG 2
Figure 7 indicates the correlation of Eq. (5). On the
The variables related to the compressor and GG com-
constant Nref,i operating line, specifying compressor pres-
bustion must satisfy Eqs. (1) and (2) concurrently. Thus, the
sure ratio can determine the corrected airflow rate. The
GATRECA code treats pcomp and PGG as the independent
corrected airflow rate, mcor,i, is given as a function of pcomp,i
variables and determines them by solving Eqs. (1) and (2).
in Eq. (6).
However, those equations have non-linearity, and some
 
root-finding algorithms are necessary, such as Newton- m_ cor;i Z M_ cor pcomp;i ð6Þ
Raphson method. PT5, PT7, and mair are the dependent
variables of pcomp. On the other hand, mGG and TGG are
those of PGG. ps;comp;i Z Ps ðm_ cor;i Þ ð7Þ
The GATRECA code contains some engine component
modules and evaluates each component’s performance, such hs;comp;i Z Hs ðm_ cor;i Þ ð8Þ
as turbomachinery, gas-generator, and ram combustor.
Subsections 4.2 to 4.5 describe the analytical model for The total-to-static pressure ratio of the compressor,
those engine components, and subsection 4.6 treats the en- ps,comp,i, and the adiabatic compression efficiency, hs,comp,i,
gine cycle analysis procedure. are given as functions of mcor,i in Eqs. (7) and (8), respec-
tively. Functions from Eqs. (6)e(8) algebraically approxi-
4.2. Compressor analysis mate the relationships among mcor,i, ps,comp,i, hs,comp,i, and
pcomp. Those equations are established along the constant
The compressor operation analysis requires a mathe- Nref,i line.
matical model for a compressor map. Reference [22] de- Finally, the following linear interpolations give the cor-
scribes the simplified exponential model for compressor rected airflow rate, mcor, compressor pressure ratio (total-to-
characteristics. This model employs the exponential static), ps,comp, and the adiabatic compression efficiency
approximation and requires defining three compressor (total-to-static), hs,comp, at the operating point.
operating states, i.e., the stall-surge line, the operating line,
m_ cor Zm_ cor;i þ N cor ðm_ cor;iþ1  m_ cor;i Þ
and the choke condition. The choke condition means the  
low-pressure side of the pressure ratio-airflow rate curve in ps;comp Zps;comp;i þ N cor ps;comp;iþ1  ps;comp;i
the compressor map. The pressure ratio in the choke con-  
hs;comp Zhs;comp;i þ N cor hs;comp;iþ1  hs;comp;i ð9Þ
dition does not have a particular physical meaning. How-
ever, it is necessary to set up the present model. Ncor  Nref ;i
N cor Z
Unfortunately, this simplified exponential model is not Nref ;iþ1  Nref ;i
necessarily accurate in describing the compressor behaviors
because it identifies only three points to draw one pressure mcor, ps,comp, and hs,comp can determine the total temperature
ratio-airflow rate curve. at the compressor exit, TT3, the total-to-total adiabatic
The authors present the following approximation model compression efficiency, hcomp, and the specific compression
for a more accurate prediction of the compressor operation. power, wcomp. Figure 8 indicates the validation of the pre-
This model assigns several referential rotational speeds in sent methods in the GATRECA code with CFD analytical
advance. Those referential rotational speeds are from 70% to results. The analytical results in the GATRECA code agree
110% of the rated speed with a 5% interval. Besides those, well with the CFD results of compressor performance.
50% and 60% of the rated speed are added to the referential
speeds. The authors conducted CFD analysis for those 4.3. Turbine analysis
referential speeds to evaluate the compressor performance.
The corrected rotational speed, Ncor, and the Total-to-total The present subsection describes the analytical procedure
compressor pressure ratio, pcomp, at the arbitrary operating in the turbine module. The input data for the turbine module
450 Ryojiro Minato et al.

spouting Mach number, Mspout, and its sonic speed. Mspout is

approximated as the polynomial of the pressure ratio of
pturb. The second formula represents this relationship be-
tween Mspout and pturb. Finally, the third formula is the
polynomial of the adiabatic turbine efficiency, hturb, as a
function of velocity ratio, U/C0. With those three formulas,
hturb can be calculated from ps,turb. The authors utilize the
CFD results to determine those polynomial coefficients in
Eq. (10).
The total temperature at the turbine exit, TT5, can be
calculated from TGG, pturb, and hturb. Those results can
determine the specific turbine power, wturb. The authors
consider that the turbine expansion is a chemical frozen
process because the turbine power in the chemical frozen
Figure 7 Concepts of Eq. (5) on compressor map.
condition is identical to that in the chemical equilibrium
condition [25]. In the turbine module, hturb, wturb, PT5, and
are the rotational speed, N, static pressure at the turbine exit,
TT5, are its output data.
Ps5, GG combustion pressure, PGG, and temperature, TGG.
Ps5 is equal to the static pressure at the compressor exit, Ps3.
4.4. Gas generator analysis
PGG and TGG correspond to turbine inlet pressure and
temperature, respectively. The pressure ratio of PGG/Ps5 is
The objective of the GG module is to specify the
defined as the total-to-static turbine expansion ratio, ps,turb.
oxidizer-to-fuel ratio for a given GG combustion condition.
The present turbine analysis employs the following three
The GG module assigns the combustion temperature, pres-
polynomial correlations.
sure, and type of fuel/oxidizer as the input data. The O/F
ratio is the output data of the GG module.
X
4
k In the gas generator (GG) module, the GG combustion
pturb Z d1;k ðps;turb Þ temperature and pressure are specified in the On-design
kZ0
condition in advance. The O/F ratio is determined in the GG
X4
Mspout Z d2;k ðpturb Þk ð10Þ module to achieve a given combustion temperature. This O/
kZ0 F ratio is always constant, even if the engine operating states
X
4
k
change. The GG module contains the chemical equilibrium
hturb Z d3;k ðU =C0 Þ calculation to evaluate the combustion temperature and the
kZ0 gas species. The authors developed the computation code
for the chemical equilibrium calculation and thoroughly
The first formula in Eq. (10) is the polynomial of the
verified its accuracy compared to the computation of
total-to-total turbine expansion ratio, pturb as a function of
NASA-CEA [26].
ps,turb. The total pressure at the turbine outlet, PT5, is equal
to the product of PGG and pturb. Reference [24] describes
4.5. Ram combustor and nozzle analyses
that the adiabatic turbine efficiency is given as a function of
the ratio of the pitch-line rotor velocity, U, to theoretical gas
This subsection describes the analytical procedure in the
spouting velocity, C0. U is linearly proportional to the
ram combustor and the nozzle modules. A ram combustor is
rotational speed. C0 is represented as the product of the gas
an element where the fuel-rich GG combustion gas burns
with the air discharged from the compressor. Rayleigh flow
theory treats a flow in a duct with heating or cooling, such as
the ram combustor flow, and can evaluate the pressure re-
covery there [27]. Rayleigh’s flow theory builds on mass
conservation and momentum conservation laws.

m_ ram Z m_ air þ m_ GG ð11Þ

     
Ps7 A7 1 þ kram M72 Z Ps3 A3 1 þ kM32 þ Ps5 A5 1 þ kGG M52
ð12Þ

The mass flow rate at the ram combustor in Eq. (11) can
be represented by using static pressure, Ps7, and Mach
Figure 8 Comparison between CFD and GATRECA code. number, M7.
Overspeeding characteristics of turbomachinery 451

m_ ram Z r7 V7 A7 One of the most significant factors in determining the

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

 GG-ATR engine operating state is to estimate the cross-
kram kram  1 2 ð13Þ
ZPs7 A7 M7 1þ M7 sectional area at the nozzle throat, A8, in Eq. (2). As indi-
Rram TT 7 2 cated in Eq. (2), A8 affects mairþmGG, PT7, and TT7. TT7
strongly depends on the combustion efficiency at the ram
Eq. (14) can be obtained from Eq. (13). combustor. Thus, the actual design of the nozzle throat re-
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi quires an accurate estimation of the combustion efficiency.
m_ ram Rram TT 7 kram  1 2 The combustion performance of a rocket engine is usually
Z M7 1 þ M7 ð14Þ
Ps7 A7 kram 2 evaluated by C * efficiency, which is the ratio of the actual
characteristic velocity to the ideal one. C * efficiency is
The static pressure at the compressor exit is equal to the valid for the ram combustor efficiency evaluation in a
turbine exit static pressure, which is the same as that at the choked nozzle flow. However, it is not for an unchoked
ram combustor inlet. The compressor-turbine power balance nozzle flow. Therefore, Eqs. (2a) and (2b) derive the
in Eq. (1) can determine the mixture ratio of the GG com- following Eq. (18). The term on the left-hand side is pro-
bustion gas to the air. The chemical equilibrium calculation portional to the square root of TT7, whether the nozzle flow
can determine the theoretical total temperature at the ram is choked or not.
combustor, TT7,ideal, under a given mixture ratio of mGG/mair
and the pressure of PT3. The effect of the pressure difference PT 7 A8 pffiffiffiffiffiffiffi
f TT 7 ð18Þ
between PT3 and PT7 on TT7 is negligible. Considering the m_ air þ m_ GG
ram combustor efficiency, the total temperature at the ram
combustor, TT7, is given in Eq. (15). The ram combustion efficiency is defined as the square
root of temperature ratio of TT7 and TT7,ideal.
TT 7 Z TT 7;ideal h2ram ð15Þ sffiffiffiffiffiffiffiffiffiffiffiffiffiffi
TT 7
The definition of ram combustion efficiency is described hram Z ð19Þ
TT 7;ideal
later. The momentum conservation law in Eq. (12) gives the
static pressure ratio, Ps7/Ps3.
where TT7,ideal is the theoretical ram combustor temperature.
    As well as TGG, the chemical equilibrium calculation can
Ps7 A3 1 þ kM32 þ A5 1 þ kGG M52
Z   ð16Þ evaluate TT7,ideal under a given PT7 and mGG/mair ratio. The
Ps3 A7 1 þ kram M72
authors define the referential ram combustion efficiency,
hram, to determine the cross-sectional area, A8, from Eq. (2).
M3 and M5 are Mach numbers at the exit of the
In the present study, the referential hram is set to 100%,
compressor and the turbine, respectively, and are known
which means that A8 is determined under the ram combustor
quantities. The unknown variables in Eqs. (14) and (16) are
temperature of TT7,ideal.
the Mach number at the ram combustor exit, M7, and Ps7.
The combustion process in the GG-ATR engine ram
Equations (14) and (16) are the dual simultaneous equation.
combustor is more like a turbojet engine afterburner instead
Using Eq. (16) and eliminating the variable Ps7A7 from Eq.
of a rocket engine. The combustion efficiency at an after-
(14), it can be solved as a quadric equation of M27. M7 and
burner is defined as the ratio of the actual temperature rise to
Ps7 can determine the total pressure at the ram combustor
the ideal one [19]. In the present study, hram is ranged from
exit, PT7.
82% to 100%, which means the minimum TT7 is identical to
The GATRECA code in the nozzle module evaluates the
two-thirds of TT7,ideal. hram of 82% in Eq. (19) corresponds
nozzle performance based on the compressible gas dy-
to 58% of combustion efficiency defined in Ref. [28]. This
namics theory. Equation (17) can give the static pressure at
value of combustion efficiency is close to the minimum
the nozzle exit and the exhaust gas velocity. In some cases,
afterburner combustion efficiency. Therefore, this assump-
shock waves may be present in the divergent nozzle. The
tion gives a reasonable range for the ram combustion
actual calculation in the nozzle module considers the effect
efficiency.
of the shock wave.

 kram þ1 4.6. Sequence of analytical procedure
A9 1 2 þ ðkram  1ÞM92 2ðkram 1Þ
Z
A 8 M9 kram þ 1 The GG-ATR engine cycle analysis (GATRECA) code

 kram integrates those engine component modules and solves the
PT 9 kram  1 2 kram 1
Z 1þ M9 ð17Þ two constraints in Eqs. (1) and (2). The computational
Ps9 2 procedure of the GATREAC code is listed below.
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
kram 1 !
u 2kram P kram
V 9 Zt 1) At first, the rotational speed of the rotor, N, the ambient
s9
Rram TT 9 1 
kram  1 PT 9 pressure, P0, and temperature, T0, are specified. The
present study assumes the sea-level static conditions.
452 Ryojiro Minato et al.

Thus, the ambient pressure and temperature are

101.3 kPa and 288.15 K, respectively.
2) The total pressure, PT2, and temperature, TT2, at the
intake outlet are calculated by considering the intake
performance. The corrected air flow rate and the cor-
rected rotational speed are calculated with PT2 and TT2.
3) Set the assumed values for PGG and pcomp. Those are the
independent variables in the iterative procedure.
4) The compressor and the GG modules evaluate their per-
formances by using PGG, pcomp, the corrected air flow rate,
mcor, and the corrected rotational speed, Ncor. The turbine
module uses the static pressure at the compressor outlet,
Ps3. The GG modules provides TGG to the turbine module.
5) The ram combustor module conducts the chemical
equilibrium calculation to determine its combustion
temperature, gas species, and pressure recovery by using
the results in the compressor and the turbine modules.
6) The output data from each engine components module
can evaluate the errors in the constraint of Eqs. (1) and
(2). The modification of the assumed values of PGG
and pcomp is determined by Newton-Raphson method.
The computation goes back to 3) to repeat if those errors
do not converge less than a certain level.

Figure 9 indicates the flow chart of GATRECA compu-

tation. The exhaust gas velocity, V9, is obtained at the end of
the flow chart. V9 can determine the propulsive performance
of the GG-ATR engine.

5. Results and discussion

5.1. Experimental and analytical compressor

performance

The authors have investigated the mixed-flow

compressor performance for the GG-ATR engine in the
Figure 9 Flow chart of GATRECA computation.
turbomachinery tests. Figures 10 and 11 indicate the char-
acteristic map for this compressor. The compressor pressure
ratio in Figures 10 and 11 are evaluated as total-to-static and
turbine driving gas, and GHe is done for tests in higher Ncor
total-to-total ones, respectively. The authors experimentally
conditions. The GG-ATR engine turbomachinery tests
measured the static pressure, Ps3, and total temperature at
require a large amount of helium gas to drive at the rated
the compressor exit, TT3, to evaluate the compressor per-
rotational speed (Ncor Z 100%). Unfortunately, helium gas
formance. With Eq. (20), the authors calculated the airflow
is very costly, and the opportunities are strictly limited to
Mach number at the compressor exit, M3, and determined
testing in higher Ncor conditions. A limited number of tests
the total pressure at the compressor exit, PT3.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi are possible to use GHe as turbine driving gas. However,

ffi CFD results agree well with the experimental data in both
k k1 2
m_ air Z Ps3 A3 M3 1þ M3 ð20Þ Figures 10 and 11.
RTT 3 2
It should be noted that the actual compressor operation
where A3 is the cross-section area at the compressor exit. might encounter not only surge/stall but also other unsteady
The total pressure at the compressor exit, PT3. The total-to- phenomena, such as supersonic unstalled flutter [29]. The
total pressure ratio, pcomp, can be evaluated by using PT3. flutter prediction requires the aeroelastic analysis of the
The solid lines in Figures 10 and 11 indicate the CFD compressor blades. Fortunately, the present mixed flow
results, and the dots do the experimental data. The error bar compressor tests did not observe such unsteady phenomena.
indicates the measurement error in those figures. Up to 70% Thus, the engine cycle analysis in the following section does
of the corrected rotational speed, Ncor, GN2 is used as not deal with unsteady phenomena. Figure 12 shows the
Overspeeding characteristics of turbomachinery 453

Figure 10 Compressor characteristic map for mixed-flow

Figure 12 Adiabatic compression efficiency for mixed-flow
compressor (total-to-static).
compressor (total-to-static).

curves are the referential rotational speeds. This figure also

shows the operating lines of constant ram combustion effi-
ciency, hram. hram in Figure 13 is equal to 85%, 90%, 95%,
and 100%.
Those operating lines in Figure 13 indicate the quasi-
steady throttling behaviors, where the compressor power
strictly equals the turbine power. The actual GG-ATR en-
gine has dynamic throttling behaviors, where the turbine
power is no longer equal to the compressor one. The dy-
namic throttling behaviors of a gas turbine engine are
described in Refs. [30,31]. On the other hand, the dynamic
throttling analysis necessitates the non-steady behaviors of a
Figure 11 Compressor characteristic map for mixed-flow compressor and a turbine. The authors treat the transient
compressor (total-to-total). responses in Figure 13 as quasi-steady states, and those
responses have tiny throttling deviations. Figure 13 shows
that a low hram reduces the compressor pressure ratio.
experimental and numerical results of the adiabatic Lowering hram decreases the ram combustor temperature,
compression efficiency (total-to-static), hs,comp. The error TT7. As indicated in Eq. (18), the reduction of TT7 leads to
bars in Figure 12 also indicate measurement errors. The the increase in the airflow rate, mair, at the nozzle throat and
analytical results of hs,comp also have good agreements with the decrease in the ram combustor pressure, PT7. This sit-
the experimental data. From the results in Figures 10e12, uation will also reduce the compressor pressure ratio pcomp.
the present CFD analysis can predict this mixed flow Therefore, lower TT7 shifts the compressor operation to
compressor performance appropriately. The discussion choke side condition in the GG-ATR engine operation. A
about the compressor operating characteristics in the sub-
sequent section is based on those CFD results.

5.2. Effect of ram combustion efficiency on

compressor operation

The experimental data of the mixed-flow compressor

validates its numerical prediction in the preceding subsec-
tion. To analyze the GG-ATR engine performance, the au-
thors arrange those mixed flow compressor data into the
GATRECA code. The constraints of the GG-ATR engine
operation are shown in Eqs. (1) and (2). The GATRACA
code resolves those constraints based on the engine
component performances and predicts the compressor
operating state on its characteristic maps. The pressure ratio/
corrected airflow rate curves of the mixed flow compressor
are shown in Figure 13. The rotational speeds of those Figure 13 Operating line of constant hram efficiency.
454 Ryojiro Minato et al.

turbojet engine with an afterburner has a variable nozzle,

which changes the nozzle throat cross-sectional area with
the exhaust gas temperature. However, the present GG-ATR
engine uses a fixed nozzle to avoid complicated mechanical
structures. The nozzle throat cross-sectional area is set based
on hram Z 100% in the current study. If the nozzle throat
cross-sectional area is too small, the compressor might enter
surge/stall conditions. However, it is difficult to estimate the
combustion efficiency of a ram combustor. Therefore, the
design of the nozzle throat must proceed by trial-and-error.
This design process will be costly and time-consuming.
Thus, it is reasonable to design the nozzle throat based on
hram Z 100% to avoid the risk of surge/stall.
Figure 15 Ncor-hram efficiency correlation.
5.3. Throttling characteristics of GG-ATR engine
prominent for higher PGG, and it is negligible in the case of
Throttling of the GG-ATR engine requires controlling the low PGG. The numerical and experimental compressor
mass flow rates of propellants, mGG. The right-hand side of powers are shown in Figure 16 to explore the mechanism of
Eq. (1) indicates the turbine power, which drives the GG- this overspeeding. The error bar of the compressor power
ATR engine rotor. In Eq. (1), the mass flow rate of GG indicates measurement errors. The left-hand side of Eq. (1)
propellants can directly contribute to the turbine power. corresponds to the compressor power. The experimental
Moreover, as shown in Eq. (3), mGG is proportional to the compressor power is evaluated by using the airflow rate and
GG combustion pressure, PGG. Thus, PGG can be the indi- the temperature rise on the compressor as Eq. (21).
cator to show the throttling level.
Figure 14 shows the constant PGG lines on the mixed-
k1 !
flow compressor characteristic map. The higher Ncor re- m_ air CP TT 2
quires high PGG in general. Figure 15 indicates the de- wcomp Z k
pcomp  1
hcomp ð21Þ
pendency of Ncor on hram with constant PGG. Along the
constant PGG line, the compressor operating state shifts to Zm_ air CP ðTT 3  TT 2 Þ
low pcomp condition as hram decreases. This behavior in-
dicates the same tendencies in Figure 14. In the actual GG- Figure 17 indicates the experimental PGG in the turbo-
ATR engine throttling, Ncor is controlled by monitoring machinery tests using GN2 as the turbine driving gas.
PGG. The rotational speed becomes higher than expected for Those data are expressed as functions of mcor. In Figure
lower ram combustion efficiency. In Figure 15, on the line 16, the right side of each compressor power curve corre-
of PGG Z 1.3 MPa, Ncor is nearly equal to 100% at sponds to the choke side condition. In the CFD results, the
hram Z 100%, and Ncor accelerates to 107.7% when hram compressor powers decrease on the choked side in high
decreases to 82%. However, Ncor is 69.7% at hram Z 100% Ncor conditions. However, the compressor power curves
on the line of PGG Z 0.6 MPa, while Ncor slightly increases for Ncor from 50% to 70% are flat to mcor. Low Ncor
to 72.3% at hram Z 82%. That means that the lower ram generally leads to low pcomp. Different from higher Ncor,
combustion efficiency can cause overspeeding for the GG- the reduction of pcomp due to degraded hram is not so
ATR engine. The overspeeding by degraded hram is significant in lower Ncor conditions. Therefore, the varia-
tion of the compressor power is relatively small. Unfor-
tunately, the helium gas is very costly, although it is

Figure 14 Constant PGG lines on compressor map. Figure 16 Numerical and experimental compressor power.
Overspeeding characteristics of turbomachinery 455

Figure 17 Experimental PGG in turbomachinery test for GN2. Figure 18 Correlation between PGG and compressor power in
turbomachinery test.
necessary for tests higher than Ncor Z 70%. Thus, the
experimental data of more than Ncor Z 70% are limited. results in Figures 17 and 18 corroborate the analytical re-
The experimental compressor power data for Ncor Z 50% sults of overspeeding due to low hram in Figures 13 and 14.
and 60% agree with the CFD results well. Those for The ram combustor characteristics were not included in the
Ncor Z 70% agree with the CFD results well in the surge previous study by Bossard and Thomas [11], and the po-
side condition. However, some differences between the tential for overspeeding was not recognized in the GG-ATR
experimental and numerical results appear in the choke engine. However, the current GG-ATR engine cycle anal-
side condition. The experimental compressor power ysis, which takes into account the characteristics of the
for Ncor Z 70% shows a reduction in the choke side compressor, the turbine, and the ram combustor, indicates
condition. This decrease corresponds to those of the nu- that overspeeding may occur. The current result suggests
merical results in higher Ncor conditions. If the compressor that the nozzle throat design requires estimating the accurate
power becomes less in the choke side condition, the hram to keep the GG-ATR engine from overspeeding.
required turbine power also decreases. Those results
can explain the overspeeding mechanism in lower hram 6. Conclusion
conditions, as shown in Figures 14 and 15. As discussed in
Subsection 5.2, Eq. (18) indicates that lower TT7 The authors conducted the numerical and experimental
causes mcor to increase and PT7 to decrease. This means investigation of the mixed-flow compressor for the GG-ATR
that lower hram results in the compressor operation engine and discussed the possibility of rotor overspeeding.
under choked conditions. The compressor power The conclusions of the present study are summarized below.
becomes lower in this situation. Therefore, in lower hram
conditions, overspeeding will occur if PGG keeps at the 1. The CFD results agree well with the experimental results
same level. for the pressure ratio/the corrected airflow rate charac-
The experimental PGG in Figure 17 corroborates the be- teristics and the adiabatic compression efficiency. Using
haviors of compressor power in Figure 16. The pressure those CFD results, the authors developed the GG-ATR
level of GN2 in Figure 17 is quite different from those of GG engine cycle analysis (GATRECA) code, including the
combustion gas because the TGG of GN2 is much lower than turbomachinery, the gas generator, and the ram
that of GG combustion gas. PGG is proportional to the mass combustor performances. The analytical results of the
flow rate of GN2 as described in Eq. (3). Thus, PGG can be GATRECA code can obtain a good agreement with CFD
the indicator to show the throttling level because the turbine analysis for the mixed flow compressor performance.
power depends on the mass flow rate of the turbine driving 2. The GATRECA analysis indicates that the compressor
gas. In Figure 17, the behaviors of PGG for Ncor Z 50% and pressure ratio, pcomp, and compression power decrease
60% are nearly flat to mcor. However, PGG for Ncor Z 70% for a fixed nozzle if the ram combustion efficiency
tends to decrease with mcor. This decrease in PGG corre- decrease. The low ram combustion efficiency results in
sponds to the decline of the compressor power in Figure 16. the low ram combustion temperature, which increases the
Finally, Figure 18 shows the correlation between PGG airflow rate and reduces the compressor pressure ratio,
and the compressor power in the turbomachinery tests using pcomp. Low pcomp reduces the compressor power, and the
GN2. The compressor power is linearly proportional to PGG. required turbine power decreases in turn. The turbine
Both the compressor power and PGG become less on the inlet pressure does not have to be so high when the
choke side, as shown in Figures 16 and 17. The turbine compressor operates on the choke side. As a result of the
power largely depends on the mass flow rate of GN2 and is lower compressor power, this scenario leads to the
equal to the compressor power. PGG is also linearly pro- overspeeding of the rotor. To avoid stall/surge conditions
portional to the mass flow rate of GN2. The experimental or overspeeding, an accurate model to predict the ram
456 Ryojiro Minato et al.

combustion efficiency is necessary for the design of the considering a heat input, Trans. JSASS Aerospace Technol. Jpn. 19
ram combustor nozzle throat. (2) (2021) 217e223.
[13] J. Li, K. Liu, Y. Gao, S. Liu, W. Wang, Y. Liu, Combustion characteristics
3. In the current experiment, the compressor power tends to experimental study of solid hydrocarbon propellant for air turbo rocket, J.
be lower in the choke condition. The turbine inlet pres- Propul. Power 34 (5) (2018) 1198e1205.
sure is proportional to the compressor power. Those be- [14] S. Liu, J. Li, G. Zhu, W. Wang, Y. Liu, Mixing and combustion
haviors can explain that the low ram combustion enhancement of turbocharged solid propellant ramjet, Acta Astronaut.
efficiency could cause overspeeding for the GG-ATR 143 (2018) 193e202.
[15] J. Li, K. Liu, Y. Liu, S. Liu, Concept and performance study of
engine. turbocharged solid propellant ramjet, Acta Astronaut. 147 (2018)
431e444.
The author’s GG-ATR engine is being repaired to [16] J.S. Lilly, S.E. Hecht, B.G. Kirkham, C.A. Eadon, Experimental
improve the reliability of its rotor-bearing system at the evaluation of an air turbo ramjet, in: Proceeding of the 30th
moment. In the future study, the authors will conduct the AIAA/ASME/SAE/ASEE joint propulsion conference, Indianapolis,
IN, June 27-29, 1994, AIAA Paper 94-3386.
combustion test of this GG-ATR engine, where the GG
[17] Y. Liu, Y. Gao, X. Pu, J. Li, G.Q. He, Operation matching model and
combustion gas is employed to drive the turbine. The cur- analysis between an air inlet and a compressor in an air turbo rocket,
rent results can be a valuable guideline for the GG-ATR Aero. Sci. Technol. 81 (2018) 306e315.
engine operation. The authors will verify the present [18] G. Bussi, G. Colasurdo, D. Pastrone, Analysis of air-turborocket
analytical result in the GG-ATR engine combustion test. performance, J. Propul. Power 11 (5) (1995) 950e954.
[19] M.E. Thomas, Monorotor turbomachinery for air-turbo-rocket pro-
pulsion, in: Proceedings of the 31th AIAA/ASME/SAE/ASEE Joint
References Propulsion Conference and Exhibit San Diego CA, July 10-12, 1995,
AIAA Paper 95-2804.
[1] K. Christensen, Air turborocket/vehicle performance comparison, J. [20] J.A. Bossard, M.E. Thomas, The influence of turbomachinery
Propul. Power 15 (5) (1999) 706e712. characteristics on air turbo rocket engine operation, in: Proceedings
[2] C.A. Snyder, A Parametric Study of a Gas-Generator Airturbo Rocket of the 36th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and
(ATR), 1986, NASA TM 88808. Exhibit Huntsvill AL, July 16-19, 2000, AIAA Paper 2000-3308.
[3] W.H. Heiser, D.T. Pratt, Hypersonic airbreathing propulsion, in: [21] ISO/Technical Committees 30, Measurement of fluid flow by means
AIAA Educational Series, AIAA Inc., Washington DC, 1994, of pressure differential devices inserted in circulat cross-section con-
pp. 457e464. duits running full d Part 2: Orifice plates, first ed., ISO 5167-2, 2003.
[4] H. Pan, P. Zhou, Performace Analysis of Liquid Air Turborocket, in: [22] G.C. Oates, Aerothermodynamics of gas turbine and rocket propul-
Proceeding of the 46th AIAA Aerospace Sciences Meeting and sion, in: AIAA Educational Series, AIAA Inc., Reston, Virginia, 1997,
Exhibit, Reno, NV, January 7-10, 2008, AIAA Paper 2008-70. pp. 313e326.
[5] V.F. Villace, G. Paniagua, J. Steelant, Installed performance evaluation [23] J.D. Mattingly, Element of propulsion, in: AIAA Educational Series,
of an air turbo-rocket expander engine, Aero. Sci. Technol. 35 (2014) AIAA Inc., Reston, Virginia, 2006, p. 459.
63e79. [24] Liquid Rocket Engine Turbine, 1974, NASA SP-8110.
[6] R. Minato, Advantage of ethanol fuel for gas generator cycle air turbo [25] K. Christensen, Comparison of methods for calculating turbine work
ramjet engine, Aero. Sci. Technol. 50 (2016) 161e172. in the air turborocket, J. Propul. Power 17 (2) (2001) 256e261.
[7] R. Minato, Alcohol and alkane fuel performance for gas generator [26] S. Gordon, B.J. McBride, Computer Program for Calculation of
cycle air turbo ramjet engine, Aero. Sci. Technol. 91 (2019) 337e356. Complex Chemical Equilibrium Compositions and Applications,
[8] R. Minato, Low toxic nitromethane based monopropellant for gas 1994, NASA RP-1311.
generator cycle air turbo ramjet engine, Propulsion and Power [27] A.H. Shapiro, The Dynamics and the Thermodynamics of
Research 11 (3) (2022) 311e324. Compressible Fluid Flow, The Ronald Press Company, New York,
[9] H. Kobayashi, Y. Maru, M.P. Richardson, K. Kinefuchi, T. Sato, 1953, pp. 190e203.
Conceptual design study of a vertical takeoff and landing airbreather, [28] J.E. McAulay, M. Abdelwahab, Experimental Evaluation of a TF30-P-
J. Spacecraft Rockets 58 (5) (2021) 1279e1292. 3 Turbofan Engine in an Altitude Facility: Afterburner Performance
[10] K. Mizobata, R. Minato, K. Higuchi, M. Ueba, S. Takagi, D. Nakata, and Engine-Afterburner Operating Limits, July, 1972, NASA TN-D-
K. Higashino, N. Tanatsugu, Development of a small-scale supersonic 6839.
flight experiment vehicle as a flying test bed for future space transportation [29] L.E. Snyder, G.L. Commerford, Supersonic unstalled flutter in fan
research, in: ISTS Special Issue: transactions of JSASS, Aerospace rotors; analytical and experimental results, J. Eng. Power Trans.
Technology Japan, 12, No. ISTS 29, 2014, pp. Po 3 1ePo3 10. ASME Ser. A 96 (4) (1974) 379e386.
[11] Y. Sato, R. Imai, D. Nakata, R. Minato, M. Uchiumi, Study on pro- [30] B. Maclsaac, R. Langton, Gas Turbine Propulsion System, John Wiley
pellant supply system for small-scale supersonic flight experiment & Sons, 2011, pp. 295e306.
vehicle (development of design technology for LOX supply system), [31] L.C. Jaw, J.D. Mattingliy, Aircraft engine controls design, system
Int. J. Microgravity Sci. Appl. 37 (1) (2020) 370104.1e370104.7. analysis, and health monitoring, in: AIAA Educational Series, AIAA
[12] K. Arimatsu, D. Nakata, R. Minato, I. Yoshikawa, H. Yagihashi, Inc., Reston VA, 2009, pp. 37e93.
M. Uchiumi, On the pressure loss characteristics of liquid oxygen